Theory of Asset Pricing

Theory of Asset Pricing
George Pennacchi
Part I
Single-period Portfolio Choice and Asset Pricing
Chapter 1
Expected Utility and Risk

Aversion
Asset prices are determined by investors risk preferences and by the distributions of assets risky future payments.
Economists refer to these two bases
of prices as investor "tastes" and the economys "technologies" for generating

asset returns. A satisfactory theory of asset valuation must consider how individuals allocate their wealth among assets having dierent future payments.
This chapter explores the development of expected utility theory, the standard
approach for modeling investor choices over risky assets. We first analyze the
conditions that an individuals preferences must satisfy to be consistent with an
expected utility function. We then consider the link between utility and riskaversion, and how risk-aversion leads to risk premia for particular assets. Our
final topic examines how risk-aversion aects an individuals choice between a
risky and a risk-free asset.
Modeling investor choices with expected utility functions is widely-used.
However, significant empirical and experimental evidence has indicated that
3
CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION
individuals sometimes behave in ways inconsistent with standard forms of expected utility.
These findings have motivated a search for improved models
of investor preferences.
Theoretical innovations both within and outside the
expected utility paradigm are being developed, and examples of such advances
are presented in later chapters of this book.
1.1
Preferences when Returns are Uncertain
Economists typically analyze the price of a good or service by modeling the

nature of its supply and demand. A similar approach can be taken to price an
asset. As a starting point, let us consider the modeling of an investors demand
for an asset. In contrast to a good or service, an asset does not provide a current
consumption benefit to an individual. Rather, an asset is a vehicle for saving. It
is a component of an investors financial wealth representing a claim on future
consumption or purchasing power.
The main distinction between assets is
the dierence in their future payos. With the exception of assets that pay a
risk-free return, assets payos are random. Thus, a theory of the demand for
assets needs to specify investors preferences over dierent, uncertain payos.
In other words, we need to model how investors choose between assets that
have dierent probability distributions of returns. In this chapter we assume
an environment where an individual chooses among assets that have random
payos at a single future date.
Later chapters will generalize the situation
to consider an individuals choices over multiple periods among assets paying

returns at multiple future dates.
Let us begin by considering potentially relevant criteria that individuals
might use to rank their preferences for dierent risky assets.
One possible
measure of the attractiveness of an asset is the average or expected value of

its payo.
Suppose an asset oers a single random payo at a particular
1.1. PREFERENCES WHEN RETURNS ARE UNCERTAIN
future date, and this payo has a discrete distribution with n possible outcomes,
n
P
pi = 1 and
(x1 , ..., xn ), and corresponding probabilities (p1 , ..., pn ), where
i=1
pi 0.1 Then the expected value of the payo (or, more simply, the expected
n
P
pi xi .
payo) is x
E [e
x] =
i=1
Is it logical to think that individuals value risky assets based solely on the
assets expected payos?
This valuation concept was the prevailing wisdom
until 1713 when Nicholas Bernoulli pointed out a major weakness. He showed
that an assets expected payo was unlikely to be the only criterion that individuals use for valuation.
He did it by posing the following problem that
became known as the St. Petersberg Paradox:
Peter tosses a coin and continues to do so until it should land "heads"

when it comes to the ground. He agrees to give Paul one ducat if
he gets "heads" on the very first throw, two ducats if he gets it on
the second, four if on the third, eight if on the fourth, and so on, so
that on each additional throw the number of ducats he must pay is
doubled.2 Suppose we seek to determine Pauls expectation (of the
payo that he will receive).
Interpreting Pauls prize from this coin flipping game as the payo of a risky
asset, how much would he be willing to pay for this asset if he valued it based
on its expected value? If the number of coin flips taken to first arrive at a heads
i
is i, then pi = 12 and xi = 2i1 so that the expected payo equals
1 As is the case in the following example, n, the number of possible outcomes, may be
infinite.
2 A ducat was a 3.5 gram gold coin used throughout Europe.
x =
=
=
X
pi xi = 12 1 + 14 2 + 18 4 +
i=1
1
2 (1
1
2 (1
1
16 8
+ ...
(1.1)
+ 12 2 + 14 4 + 18 8 + ...
+ 1 + 1 + 1 + ... =
The "paradox" is that the expected value of this asset is infinite, but, intuitively, most individuals would pay only a moderate, not infinite, amount to play
this game. In a paper published in 1738, Daniel Bernoulli, a cousin of Nicholas,
provided an explanation for the St. Petersberg Paradox by introducing the concept of expected utility.3 His insight was that an individuals utility or "felicity"
from receiving a payo could dier from the size of the payo and that people
cared about the expected utility of an assets payos, not the expected value of
Pn
its payos. Instead of valuing an asset as x = i=1 pi xi , its value, V , would
be
V E [U (e
x)] =
Pn
i=1 pi Ui
(1.2)
where Ui is the utility associated with payo xi . Moreover, he hypothesized

that the "utility resulting from any small increase in wealth will be inversely
proportionate to the quantity of goods previously possessed." In other words,
the greater an individuals wealth, the smaller is the added (or marginal) utility
received from an additional increase in wealth. In the St. Petersberg Paradox,
prizes, xi , go up at the same rate that the probabilities decline. To obtain a
finite valuation, the trick is to allow the utility of prizes, Ui , to increase slower
3 An English translation of Daniel Bernoullis original Latin paper is printed in Econometrica (Bernoulli 1954). Another Swiss mathematician, Gabriel Cramer, oered a similar
solution in 1728.
than the rate that probabilities decline. Hence, Daniel Bernoulli introduced the
principle of a diminishing marginal utility of wealth (as expressed in his quote
above) to resolve this paradox.
The first complete axiomatic development of expected utility is due to John
von Neumann and Oskar Morgenstern (von Neumann and Morgenstern 1944).
Von Neumann, a renowned physicist and mathematician, initiated the field of
game theory, which analyzes strategic decision making. Morgenstern, an economist, recognized the fields economic applications and, together, they provided
a rigorous basis for individual decision-making under uncertainty. We now outline one aspect of their work, namely, to provide conditions that an individuals
preferences must satisfy for these preferences to be consistent with an expected
utility function.
Define a lottery as an asset that has a risky payo and consider an individuals optimal choice of a lottery (risky asset) from a given set of dierent lotteries. All lotteries have possible payos that are contained in the set {x1 , ..., xn }.
In general, the elements of this set can be viewed as dierent, uncertain outcomes. For example, they could be interpreted as particular consumption levels
(bundles of consumption goods) that the individual obtains in dierent states
of nature or, more simply, dierent monetary payments received in dierent
states of the world. A given lottery can be characterized as an ordered set
n
P
pi = 1 and pi 0. A
of probabilities P = {p1 , ..., pn }, where, of course,
i=1
dierent lottery is characterized by another set of probabilities, for example,
P = {p1 , ..., pn }. Let , , and denote preference and indierence between

lotteries.4 We will show that if an individuals preferences satisfy the following
conditions (axioms), then these preferences can be represented by a real-valued
4 Specifically, if an individual prefers lottery P to lottery P , this can be denoted as P P
or P P . When the individual is indierent between the two lotteries, this is written as
P P . If an individual prefers lottery P to lottery P or she is indierent between lotteries
P and P , this is written as P P or P P .
utility function defined over a given lotterys probabilities, that is, an expected
utility function V (p1 , ..., pn ).
Axioms:
1) Completeness
For any two lotteries P and P , either P P , or P P , or P P .
2) Transitivity
If P P and P P , then P P .
3) Continuity
If P P P , there exists some [0, 1] such that P P +(1)P ,
where P + (1 )P denotes a compound lottery, namely with probability
one receives the lottery P and with probability (1 ) one receives the
lottery P .
These three axioms are analogous to those used to establish the existence
of a real-valued utility function in standard consumer choice theory.5
The
fourth axiom is unique to expected utility theory and, as we later discuss, has
important implications for the theorys predictions.
4) Independence
For any two lotteries P and P , P P if for all (0,1] and all P :
P + (1 )P P + (1 )P
Moreover, for any two lotteries P and P , P P if for all (0,1] and all
P :
5A
primary area of microeconomics analyzes a consumers optimal choice of multiple goods

(and services) based on their prices and the consumers budget contraint. In that context,
utility is a function of the quantities of multiple goods consumed. References on this topic
include (Kreps 1990), (Mas-Colell, Whinston, and Green 1995), and (Varian 1992). In contrast, the analysis of this chapter expresses utility as a function of the individuals wealth. In
future chapters, we introduce multi-period utility functions where utility becomes a function
of the individuals overall consumption at multiple future dates. Financial economics typically bypasses the individuals problem of choosing among dierent consumption goods and
focuses on how the individual chooses a total quantity of consumption at dierent points in
time and dierent states of nature.
P + (1 )P P + (1 )P
To better understand the meaning of the independence axiom, note that P
is preferred to P by assumption. Now the choice between P + (1 )P
and P + (1 )P is equivalent to a toss of a coin that has a probability
(1 ) of landing tails, in which case both compound lotteries are equivalent
to P , and a probability of landing heads, in which case the first compound
lottery is equivalent to the single lottery P and the second compound lottery
is equivalent to the single lottery P . Thus, the choice between P + (1 )P
and P + (1 )P is equivalent to being asked, prior to the coin toss, if one
would prefer P to P in the event the coin lands heads.
It would seem reasonable that should the coin land heads, we would go
ahead with our original preference in choosing P over P . The independence
axiom assumes that preferences over the two lotteries are independent of the
way in which we obtain them.6 For this reason, the independence axiom is
also known as the no regret axiom.
However, experimental evidence finds
some systematic violations of this independence axiom, making it a questionable

assumption for a theory of investor preferences. For example, the Allais Paradox is a well-known choice of lotteries that, when oered to individuals, leads
most to violate the independence axiom.7
Machina (Machina 1987) summa-
rizes violations of the independence axiom and reviews alternative approaches

to modeling risk preferences.
In spite of these deficiencies, von Neumann -
Morgenstern expected utility theory continues to be a useful and common ap6 In the context of standard consumer choice theory, would be interpreted as the amount
(rather than probability) of a particular good or bundle of goods consumed (say C) and
(1 ) as the amount of another good or bundle of goods consumed (say C ). In this case,
it would not be reasonable to assume that the choice of these dierent bundles is independent.
This is due to some goods being substitutes or complements with other goods. Hence, the
validity of the independence axiom is linked to outcomes being uncertain (risky), that is, the
interpretation of as a probability rather than a deterministic amount.
7 A similar example is given as an exercise at the end of this chapter.
10
proach to modeling investor preferences, though research exploring alternative

paradigms is growing.8
The final axiom is similar to the independence and completeness axioms.
5) Dominance
Let P 1 be the compound lottery 1 P +(11 )P and P 2 be the compound
lottery 2 P + (1 2 )P . If P P , then P 1 P 2 if and only if 1 > 2 .
Given preferences characterized by the above axioms, we now show that the
choice between any two (or more) arbitrary lotteries is that which has the higher
(highest) expected utility.
The completeness axioms ordering on lotteries naturally induces an ordering on the set of outcomes. To see this, define an "elementary" or "primitive"
lottery, ei , which returns outcome xi with probability 1 and all other outcomes
with probability zero, that is, ei = {p1 , ...,pi1 ,pi ,pi+1, ...,pn } = {0, ..., 0, 1, 0, ...0}
where pi = 1 and pj = 0 j 6= i. Without loss of generality, suppose that the
outcomes are ordered such that en en1 ... e1 . This follows from the
completeness axiom for this case of n elementary lotteries. Note that this ordering of the elementary lotteries may not necessarily coincide with a ranking
of the elements of x strictly by the size of their monetary payos, as the state
of nature for which xi is the outcome may dier from the state of nature for
which xj is the outcome, and these states of nature may have dierent eects
on how an individual values the same monetary outcome. For example, xi may
be received in a state of nature when the economy is depressed, and monetary
payos may be highly valued in this state of nature. In contrast, xj may be
received in a state of nature characterized by high economic expansion, and
monetary payments may not be as highly valued.
Therefore, it may be that
ei ej even if the monetary payment corresponding to xi was less than that

8 This research includes "behavioral finance," a field that encompasses alternatives to both
expected utility theory and market eciency. An example of how a behavioral finance - type
utility specification can impact asset prices will be presented in Chapter 15.
11
corresponding to xj .
From the continuity axiom, we know that for each ei , there exists a Ui [0, 1]
such that
ei Ui en + (1 Ui )e1
(1.3)
and for i = 1, this implies U1 = 0 and for i = n, this implies Un = 1. The values
of the Ui weight the most and least preferred outcomes such that the individual
is just indierent between a combination of these polar payos and the payo of
xi . The Ui can adjust for both dierences in monetary payos and dierences
in the states of nature during which the outcomes are received.
Now consider a given arbitrary lottery, P = {p1 , ..., pn }. This can be considered a compound lottery over the n elementary lotteries, where elementary
lottery ei is obtained with probability pi . By the independence axiom, and using
equation (1.3), the individual is indierent between the compound lottery, P ,
and the following lottery given on the right-hand-side of the equation below:
p1 e1 + ... + pn en
p1 e1 + ... + pi1 ei1 + pi [Ui en + (1 Ui )e1 ]

+pi+1 ei+1 + ... + pn en
(1.4)
where we have used the indierence relation in equation (1.3) to substitute for
ei on the right hand side of (1.4). By repeating this substitution for all i,
i = 1, ..., n, we see that the individual will be indierent between P , given by
the left hand side of (1.4), and
n
!
!
n
X
X
pi Ui en + 1
pi Ui e1
p1 e1 + ... + pn en
i=1
i=1
(1.5)
12

Now define
n
P
pi Ui . Thus, we see that lottery P is equivalent to a
i=1
compound lottery consisting of a probability of obtaining elementary lottery

en and a (1 ) probability of obtaining elementary lottery e1 . In a similar
manner, we can show that any other arbitrary lottery P = {p1 , ..., pn } is equivalent to a compound lottery consisting of a probability of obtaining en and
n
P
pi Ui .
a (1 ) probability of obtaining e1 , where
i=1
Thus, we know from the dominance axiom that P P if and only if > ,
n
n
P
P
pi Ui >
pi Ui . So defining an expected utility function as
which implies
i=1
i=1
V (p1 , ..., pn ) =
n
X
pi Ui
(1.6)
i=1
will imply that P P if and only if V (p1 , ..., pn ) > V (p1 , ..., pn ).
The function given in equation (1.6) is known as von Neumann - Morgenstern
expected utility. Note that it is linear in the probabilities and is unique up to
a linear monotonic transformation.9 This implies that the utility function has
cardinal properties, meaning that it does not preserve preference orderings
for all strictly increasing transformations.10 For example, if Ui = U (xi ), an
individuals choice over lotteries will be the same under the transformation
aU (xi ) + b, but not a non-linear transformation that changes the shape of
U (xi ).
The von Neumann-Morgenstern expected utility framework may only partially explain the phenomenon illustrated by the St. Petersberg Paradox. Suppose an individuals utility is given by the square root of a monetary payo, that
is, Ui = U (xi ) = xi . This is a monotonically increasing, concave function of

9 The intuition for why expected utility is unique up to a linear transformation can be
traced to equation (1.3). The derivation chose to compare elementary lottery i in terms of
the least and most preferred elementary lotteries. However, other bases for ranking a given
lottery are possible.
10 An "ordinal" utility function preserves preference orderings for any strictly increasing
transformation, not just linear ones. The utility functions defined over multiple goods and
used in standard consumer theory are ordinal measures.
13
x, which, here, is assumed to be simply a monetary amount (in units of ducats).

Then the individuals expected utility of the St. Petersberg payo is
X
X
1 i1 X 2i
pi Ui =
2
=
2
2i
i=1
i=1
i=2
2
(1.7)
= 2 2 + 2 2 + ...
i

X
1
1
1

=
1 =
= 1.707
2
2
2
2
i=0
which is finite. This individual would get the same expected utility from receiving a certain payment of 1.7072
= 2.914 ducats since V =
expected (and actual) utility of 1.707.
2.914also gives
Hence we can conclude that the St.
Petersberg gamble would be worth 2.914 ducats to this square-root utility maximizer.
However, the reason that this is not a complete resolution of the paradox
is that one can always construct a super St. Petersberg paradox where even
expected utility is infinite. Note that in the regular St. Petersberg paradox,
the probability of winning declines at rate 2i while the winning payo increases
at rate 2i . In a super St. Petersberg paradox, we can make the winning payo
increase at a rate xi = U 1 (2i1 ) and expected utility would no longer be finite.
If we take the example of square-root utility, let the winning payo be xi = 22i2 ,
that is, x1 = 1, x2 = 4, x3 = 16, etc. In this case, the expected utility of the
super St. Petersberg payo by a square-root expected utility maximizer is
V =
X
X
1 2i2
pi Ui =
2
=
2i
i=1
i=1
(1.8)
Should we be concerned by the fact that if we let the prizes grow quickly
enough, we can get infinite expected utility (and valuations) for any chosen form
of expected utility function? Maybe not. One could argue that St. Petersberg
14
games are unrealistic, particularly ones where the payos are assumed to grow
rapidly. The reason is that any person oering this asset has finite wealth (even
Bill Gates). This would set an upper bound on the amount of prizes that could
feasibly be paid, making expected utility, and even the expected value of the
payo, finite.
The von Neumann-Morgenstern expected utility approach can be generalized to the case of a continuum of outcomes and lotteries having continuous
probability distributions. For example, if outcomes are a possibly infinite number of purely monetary payos or consumption levels denoted by the variable
x, a subset of the real numbers, then a generalized version of equation (1.6) is
V (F ) = E [U (e
x)] =
U (x) dF (x)
(1.9)
where F (x) is a given lotterys cumulative distribution function over the payos,
x.11 Hence, the generalized lottery represented by the distribution function F
is analogous to our previous lottery represented by the discrete probabilities
P = {p1 , ..., pn }.
Thus far, our discussion of expected utility theory has said little regarding
an appropriate specification for the utility function, U (x). We now turn to a
discussion of how the form of this function aects individuals risk preferences.
1.2
Risk Aversion and Risk Premia
As mentioned in the previous section, Daniel Bernoulli proposed that utility

functions should display diminishing marginal utility, that is, U (x) should be
an increasing but concave function of wealth. He recognized that this concavity
implies that an individual will be risk averse.
By risk averse we mean that
11 When the random payo, x

h, is absolutely continuous, then Uexpected utility can be written
in terms of the probability density function, f (x), as V (f ) = U (x) f (x) dx.
1.2. RISK AVERSION AND RISK PREMIA
15
the individual would not accept a fair lottery (asset), where a fair or pure
risk lottery is defined as one that has an expected value of zero. To see the
relationship between fair lotteries and concave utility, consider the following
example. Let there be a lottery that has a random payo, e
, where
1 with probability p
e
=
with probability 1 p
(1.10)
The requirement that it be a fair lottery restricts its expected value to equal
zero:
E [e
] = p1 + (1 p)2 = 0
(1.11)
which implies 1 /2 = (1 p) /p, or, solving for p, p = 2 / (1 2 ).
Of
course since 0 < p < 1, 1 and 2 are of opposite signs.

Now suppose a von Neumann-Morgenstern expected utility maximizer whose
current wealth equals W is oered the above lottery. Would this individual
accept it, that is, would she place a positive value on this lottery?
If the lottery is accepted, expected utility is given by E [U (W + e
)]. Instead,
if it is not accepted, expected utility is given by E [U (W )] = U (W ). Thus, an

individuals refusal to accept a fair lottery implies
U (W ) > E [U (W + e
)] = pU (W + 1 ) + (1 p)U (W + 2 )
(1.12)
To show that this is equivalent to having a concave utility function, note that
U (W ) can be re-written as
16

U(W ) = U (W + p1 + (1 p)2 )
(1.13)
since p1 + (1 p)2 = 0 by the assumption that the lottery is fair. Re-writing

inequality (1.12), we have
U (W + p1 + (1 p)2 ) > pU (W + 1 ) + (1 p)U (W + 2 )
(1.14)
which is the definition of U being a concave function. A function is concave

if a line joining any two points of the function lies entirely below the function.
When U (W ) is concave, a line connecting the points U(W + 2 ) to U (W + 1 )
lies below U (W ) for all W such that W +2 < W < W +1 . As shown in Figure
1.1, pU (W + 1 ) + (1 p)U (W + 2 ) is exactly the point on this line directly
below U (W ). This is clear by substituting p = 2 /(1 2 ). Note that when
U (W ) is a continuous, second dierentiable function, concavity implies that its
second derivative, U 00 (W ), is less than zero.
To show the reverse, that concavity of utility implies the unwillingness to
accept a fair lottery, we can use a result from statistics known as Jensens
inequality. If U () is some concave function, and x
e is a random variable, then
Jensens inequality says that
E[U (
x)] < U (E[
x])
(1.15)
Therefore, substituting x = W + e
, with E[e
] = 0, we have
E [U (W + e
)] < U (E [W + e
]) = U (W )
which is the desired result.
(1.16)
Utility
17
Concave U tility Function
U(W + 1 )
U(W )
[- 2 U(W + 1 )+
1 U (W + 2 )]/( 1 - 2 )
U(W + 2 )
= p U (W + 1 ) +
(1-p) U (W + 2 )
W + 2
W + 1
W ealth
Figure 1.1: Fair Lotteries Lower Utility
We have defined risk aversion in terms of the individuals utility function.12

Let us now consider how this aversion to risk can be quantified. This is done
by defining a risk premium, the amount that an individual is willing to pay to
avoid a risk.
Let denote the individuals risk premium for a particular lottery, e
. It
can be likened to the maximum insurance payment an individual would pay to
avoid a particular risk. John W. Pratt (Pratt 1964) defined the risk premium
for lottery (asset) e
as
U (W ) = E [U (W + e
)]
(1.17)
12 Based on the same analysis, it is straightforward to show that if an individual strictly

preferred a fair lottery, his utility function must be convex in wealth. Such an individual
is said to be risk-loving. Similarly, an individual that is indierent between accepting or
refusing a fair lottery is said to be risk-neutral and must have utility that is a linear function
of wealth.
18
W is defined as the certainty equivalent level of wealth associated with the

lottery, e
. Since utility is an increasing, concave function of wealth, Jensens
inequality ensures that must be positive when e
is fair, that is, the individual
would accept a level of wealth lower than her expected level of wealth following
the lottery, E [W + e
], if the lottery could be avoided.
To analyze this Pratt (1964) risk premium, we continue to assume the indi-
vidual is an expected utility maximizer and that e

is a fair lottery, that is, its
expected value equals zero. Further, let us consider the case of e

being small,
so that we can study its eects by taking a Taylor series approximation of equa-
tion (1.17) around the point e

= 0 and = 0.13 Expanding the left hand side
of (1.17) around = 0 gives
U (W )
= U(W ) U 0 (W )
(1.18)
and expanding the right hand side of (1.17) around e

= 0 (and taking a three
term expansion since E [e

] = 0 implies that a third term is necessary for a
limiting approximation) gives
h
i
E [U(W + e
)]
U 0 (W ) + 12 e
2 U 00 (W )
= E U (W ) + e
(1.19)
= U (W ) + 12 2 U 00 (W )
h i
2 is the lotterys variance. Equating the results in (1.18) and
where 2 E e
(1.19), we have
= 12 2
U 00 (W )
12 2 R(W )
U 0 (W )
(1.20)
13 By describing the random variable h

as small we mean that its probability density is
concentrated around its mean of 0.
19
where R(W ) U 00 (W )/U 0 (W ) is the Pratt (1964) - Arrow (1971) measure

of absolute risk aversion. Note that the risk premium, , depends on the uncertainty of the risky asset, 2 , and on the individuals coecient of absolute
risk aversion. Since 2 and U 0 (W ) are both greater than zero, concavity of the
utility function ensures that must be positive.
From (1.20) we see that the concavity of the utility function, U 00 (W ), is
insucient to quantify the risk premium an individual is willing to pay, even
though it is necessary and sucient to indicate whether the individual is riskaverse. In order to determine the risk premium, we also need the first derivative,
U 0 (W ), which tells us the marginal utility of wealth. An individual may be very
risk averse (U 00 (W ) is large), but he may be unwilling to pay a large risk
premium if he is poor since his marginal utility is high (U 0 (W ) is large).
To illustrate this point, consider the following negative exponential utility
function:
U (W ) = ebW , b > 0
(1.21)
Note that U 0 (W ) = bebW > 0 and U 00 (W ) = b2 ebW < 0. Consider the

behavior of a very wealthy individual, that is, one whose wealth approaches
infinity:
lim U 0 (W ) = lim U 00 (W ) = 0
(1.22)
As W , the utility function is a flat line. Concavity disappears, which

might imply that this very rich individual would be willing to pay very little
for insurance against a random event, e
, certainly less than a poor person with
the same utility function. However, this is not true because the marginal utility
of wealth is also very small. This neutralizes the eect of smaller concavity.
20
Indeed:
R(W ) =
b2 ebW
=b
bebW
(1.23)
which is a constant. Thus, we can see why this utility function is sometimes
referred to as a constant absolute risk aversion utility function.
If we want to assume that absolute risk aversion is declining in wealth, a
necessary, though not sucient, condition for this is that the utility function
have a positive third derivative, since
U 000 (W )U 0 (W ) [U 00 (W )]2
R(W )
=
W
[U 0 (W )]2
(1.24)
Also, it can be shown that the coecient of risk aversion contains all relevant
information about the individuals risk preferences. To see this, note that
R(W ) =
(ln [U 0 (W )])
U 00 (W )
=
U 0 (W )
W
(1.25)
Integrating both sides of (1.25), we have
R(W )dW = ln[U 0 (W )] + c1
(1.26)
where c1 is an arbitrary constant. Taking the exponential function of (1.26),

one obtains
R(W )dW
= U 0 (W )ec1
(1.27)
Integrating once again gives

Z
R(W )dW
dW = ec1 U(W ) + c2
(1.28)

where c2 is another arbitrary constant.
21
Because expected utility functions
are unique up to a linear transformation, ec1 U(W ) + c1 reflects the same risk
preferences as U (W ). Hence, this shows one can recover the risk-preferences of
U (W ) from the function R (W ).
Relative risk aversion is another frequently used measure of risk aversion and
is defined simply as
Rr (W ) = W R(W )
(1.29)
In many applications in financial economics, an individual is assumed to have

relative risk aversion that is constant for dierent levels of wealth. Note that this
assumption implies that the individuals absolute risk aversion, R (W ), declines
in direct proportion to increases in his wealth. While later chapters will discuss
the widely varied empirical evidence on the size of individuals relative risk
aversions, one recent study based on individuals answers to survey questions
finds a median relative risk aversion of approximately 7.14
Let us now examine the coecients of risk aversion for some utility functions
that are frequently used in models of portfolio choice and asset pricing. Power
utility can be written as
U (W ) = 1 W , < 1
(1.30)
14 The mean estimate was lower, indicating a skewed distribution. Robert Barsky, Thomas
Juster, Miles Kimball, and Matthew Shapiro (Barsky, Juster, Kimball, and Shapiro 1997)
computed these estimates of relative risk aversion from a survey that asked a series of questions regarding whether the respondent would switch to a new job that had a 50-50 chance
of doubling their lifetime income or decreasing their lifetime income by a proportion . By
varying in the questions, they estimated the point where an individual would be indierent
between keeping their current job or switching. Essentially, they attempted to find such
that 12 U (2W ) + 12 U ( W ) = U (W ). Assuming utility displays constant relative risk aversion of the form U (W ) = W /, then the coecient of relative risk aversion, 1 satisfies
2 + = 2. The authors warn that their estimates of risk aversion may be biased upward if
individuals attach non-pecuniary benefits to maintaining their current occupation. Interestingly, they confirmed that estimates of relative risk aversion tended to be lower for individuals
who smoked, drank, were uninsured, held riskier jobs, and invested in riskier assets.
22

2
implying that R(W ) = (1)W

W 1
(1)
W
and, therefore, Rr (W ) = 1
. Hence, this form of utility is also known as constant relative risk aversion.
Logarithmic utility is a limiting case of power utility.
power utility function as 1 W 1 =
W 1 15
.
To see this, write the
Next take the limit of this utility
function as 0. Note that the numerator and denominator both go to zero,

so that the limit is not obvious. However, we can re-write the numerator in
terms of an exponential and natural log function and apply LHpitals rule to
obtain:
W 1
e ln(W ) 1
ln(W )W
= lim
= lim
= ln(W )
0
0
0
(1.31)
lim
Thus, logarithmic utility is equivalent to power utility with = 0, or a coecient

of relative risk aversion of unity:
2
R(W ) = W
W 1 =
1
W
and Rr (W ) = 1.
Quadratic utility takes the form
U (W ) = W 2b W 2 , b > 0
(1.32)
Note that the marginal utility of wealth is U 0 (W ) = 1 bW and is positive only

when b <
1
W
. Thus, this utility function makes sense (in that more wealth is
preferred to less) only when W < 1b . The point of maximum utility, 1b , is known
as the bliss point. We have R(W ) =
b
1bW
and Rr (W ) =
bW
1bW
Hyperbolic absolute risk aversion (HARA) utility is a generalization of all of

the aforementioned utility functions. It can be written as
1
U (W ) =
W
+
1
(1.33)
15 Recall that we can do this because utility functions are unique up to a linear transformation.
23
W
subject to the restrictions 6= 1, > 0, 1
+ > 0, and = 1 if = .
1
W
+
. Since R(W ) must be > 0, it implies > 0 when
Thus, R(W ) = 1
1
W
> 1. Rr (W ) = W 1
+
. HARA utility nests constant absolute risk
aversion ( = , = 1), constant relative risk aversion ( < 1, = 0), and

quadratic ( = 2) utility functions. Thus, depending on the parameters, it is
able to display constant absolute risk aversion or relative risk aversion that is
increasing, decreasing, or constant.
We will re-visit HARA utility in future
chapters as it can be an analytically convenient assumption for utility when

deriving an individuals intertemporal consumption and portfolio choices.
Pratts definition of a risk premium in (1.17) is commonly used in the insurance literature because it can be interpreted as the payment that an individual
is willing to make to insure against a particular risk. However, in the field of
financial economics, a somewhat dierent definition is often employed. Financial economists seek to understand how the risk of an assets payo determines
the assets rate of return. In this context, an assets risk premium is defined as
its expected rate of return in excess of the risk-free rate of return. This alternative concept of a risk premium was used by Kenneth Arrow (Arrow 1971) who
independently derived a coecient of risk aversion that is identical to Pratts
measure. Let us now outline Arrows approach. Suppose that an asset (lottery), e
, has the following payos and probabilities (this could be generalized to
other types of fair payos):
where
0.
+ with probability
e
=
with probability
1
2
(1.34)
1
2
Note that, as before, E [e

] = 0. Now consider the following
question. By how much should we change the expected value (return) of the
asset, by changing the probability of winning, in order to make the individual
24
indierent between taking and not taking the risk? If p is the probability of
winning, we can define the risk premium as
= prob (e
= + ) prob (e
= ) = p (1 p) = 2p 1
(1.35)
Therefore, from (1.35) we have

prob (e
= + ) p = 12 (1 + )
prob (e
= ) 1 p = 12 (1 )
(1.36)
These new probabilities of winning and losing are equal to the old probabilities,
1
2,
plus half of the increment, . Thus, the premium, , that makes the individual
indierent between accepting and refusing the asset is
U (W ) =
1
1
(1 + )U (W + ) + (1 )U(W )
2
2
Taking a Taylor series approximation around
U (W ) =
= 0, gives
1
(1 + ) U (W ) + U 0 (W ) + 12 2 U 00 (W )
2
1
+ (1 ) U (W ) U 0 (W ) + 12 2 U 00 (W )
2
= U (W ) + U 0 (W ) +
(1.37)
(1.38)
1 2 00
2 U (W )
Re-arranging (1.38) implies
1
2
R(W )
(1.39)
which, as before, is a function of the coecient of absolute risk aversion. Note

that the Arrow premium, , is in terms of a probability, while the Pratt measure,
, is in units of a monetary payment. If we multiply by the monetary payment
1.3. RISK AVERSION AND PORTFOLIO CHOICE
25
received, , then equation (1.39) becomes
=
Since
1 2
2 R(W )
(1.40)
is the variance of the random payo, e

, equation (1.40) shows that the
Pratt and Arrow measures of risk premia are equivalent. Both were obtained
as a linearization of the true function around e
= 0.
The results of this section showed how risk aversion depends on the shape of
an individuals utility function. Moreover, it demonstrated that a risk premium,

equal to either the payment an individual would make to avoid a risk or the
individuals required excess rate of return on a risky asset, is proportional to
the individuals Pratt-Arrow coecient of absolute risk aversion.
1.3
Risk Aversion and Portfolio Choice
Having developed the concepts of risk aversion and risk premiums, we now
consider the relation between risk aversion and an individuals portfolio choice
in a single period context. While the portfolio choice problem that we analyze
is very simple, many of its insights extend to the more complex environments
that will be covered in later chapters of this book. We shall demonstrate that
absolute and relative risk aversion play important roles in determining how
portfolio choices vary with an individuals level of wealth. Moreover, we show
that when given a choice between a risk-free asset and a risky asset, a risk-averse
individual always chooses at least some positive investment in the risky asset if
it pays a positive risk premium.
The models assumptions are as follows. Assume there is a riskless security
that pays a rate of return equal to rf . In addition, for simplicity suppose there
is just one risky security that pays a stochastic rate of return equal to re. Also,
26
let W0 be the individuals initial wealth, and let A be the dollar amount that
the individual invests in the risky asset at the beginning of the period. Thus,
W0 A is the initial investment in the riskless security. Denoting the individuals
, it satisfies:
end-of-period wealth as W
= (W0 A)(1 + rf ) + A(1 + r)
(1.41)
= W0 (1 + rf ) + A(
r rf )
Note that in the second line of equation (1.41), the first term is the individuals
return on wealth when the entire portfolio is invested in the risk-free asset, while
the second term is the dierence in return gained by investing A dollars in the
risky asset.
We assume that the individual cares only about consumption at the end of
this single period. Therefore, maximizing end-of-period consumption is equivalent to maximizing end-of-period wealth. Assuming that the individual is a von
Neumann-Morgenstern expected utility maximizer, she chooses her portfolio by
maximizing the expected utility of end-of-period wealth:
)] = maxE [U (W0 (1 + rf ) + A(
r rf ))]
maxE[U (W
A
(1.42)
The solution to the individuals problem in (1.42) must satisfy the following
first order condition with respect to A:
i
h
(
r rf ) = 0
E U0 W
(1.43)
This condition determines the amount, A, that the individual invests in the

risky asset.16
27
Consider the special case in which the expected rate of re-
turn on the risky asset equals the risk-free rate. In that case A = 0 satisfies the first order condition.
To see this, note that when A = 0, then

= U 0 (W0 (1 + rf )) are non-stochastic.
= W0 (1 + rf ) and, therefore, U 0 W
W
i
h
(
r rf ) = U 0 (W0 (1 + rf )) E[
r rf ] = 0. This result is
Hence, E U 0 W
reminiscent of our earlier finding that a risk-averse individual would not choose
to accept a fair lottery. Here, the fair lottery is interpreted as a risky asset that
has an expected rate of return just equal to the risk-free rate.
Next, consider the case in which E[
r]rf > 0. Clearly, A = 0 would not sati
h
(
r rf ) = U 0 (W0 (1 + rf )) E[
r
isfy the first order condition because E U 0 W
rf ] > 0 when A = 0. Rather, when E[
r] rf > 0 condition (1.43) is satisfied
only when A > 0. To see this, let rh denote a realization of r such that it exceeds
. Also let rl denote a realization
rf , and let W h be the corresponding level of W
.
of r such that it is lower than rf , and let W l be the corresponding level of W

(
r rf )
Obviously, U 0 (W h )(rh rf ) > 0 and U 0 (W l )(rl rf ) < 0. For U 0 W
to average to zero for all realizations of r, it must be the case that W h > W l

so that U 0 W h < U 0 W l due to the concavity of the utility function. This is
because since E[
r]rf > 0, the average realization of rh is farther above rf than

(
r rf )
the average realization of rl is below rf . Therefore, to make U 0 W
average to zero, the positive (rh rf ) terms need to be given weights, U 0 W h ,

that are smaller than the weights, U 0 (W l ), that multiply the negative (rl rf )
realizations. This can occur only if A > 0 so that W h > W l . The implication
is that an individual will always hold at least some positive amount of the risky
asset if its expected rate of return exceeds the risk-free rate.17
k

2 l
r rf
second order condition for a maximum, E U 00 W
0, is satisfied be
0 due to the assumed concavity of the utility function.
cause U 00 W
17 Related to this is the notion that a risk-averse expected utility maximizer should accept
a small lottery with a positive expected return. In other words, such an individual should
be close to risk-neutral for small-scale bets. However, Matthew Rabin and Richard Thaler
(Rabin and Thaler 2001) claim that individuals frequently reject lotteries (gambles) that are
16 The
28

Now, we can go further and explore the relationship between A and the
individuals initial wealth, W0 . Using the envelope theorem, we can dierentiate

the first order condition to obtain18
i
h
i
h
)(
)(
E U 00 (W
r rf )(1 + rf ) dW0 + E U 00 (W
r rf )2 dA = 0
(1.44)
or
h
i
)(
r rf )
(1 + rf )E U 00 (W
dA
i
h
=
dW0
)(
E U 00 (W
r rf )2
(1.45)
The denominator of (1.45) is positive because concavity of the utility function

) is negative. Therefore, the sign of the expression depends
ensures that U 00 (W
on the numerator, which can be of either sign because realizations of (
r rf )
can turn out to be both positive and negative.
To characterize situations in which the sign of (1.45) can be determined, let
us first consider the case where the individual has absolute risk aversion that
is decreasing in wealth. As before, let rh denote a realization of r such that it
. Then for A > 0, we
exceeds rf , and let W h be the corresponding level of W
have W h > W0 (1 + rf ). If absolute risk aversion is decreasing in wealth, this
modest in size yet have positive expected returns. From this they argue that concave expected
utility is not a plausible model for predicting an individuals choice of small-scale risks.
18 The envelope theorem is used to analyze how the maximized value of the objective function
and the control variable
change
l when one of the models parameters changes. In our context,
k
i
so that v (W0 ) = maxf (A, W0 ) is the maximized value of the
define f (A, W0 ) E U W
A
objective function when the control variable, A, is optimally chosen. Also define A (W0 ) as
the value of A that maximizes f for a given value of W0 . Then applying the chain rule, we
dv(W )
f (A,W0 ) dA(W0 )
f (A(W0 ),W0 )
f(A,W0 )
have dW 0 =
+
. But since
= 0, from the first
A
dW
W
A
0
dv(W0 )
0 ),W0 )
= f(A(W
. Again applying the chain rule
dW0
W0
(f(A(W0 ),W0 )/A)
2 f (A(W0 ),W0 ) dA(W0 )
=0=
+
to the first order condition, one obtains
W0
dW0
A2
2 f (A(W0 ),W0 )
dA(W0 )
2 f (A(W0 ),W0 ) 2 f(A(W0 ),W0 )
. Re-arranging gives us dW
=
/
, which is
AW0
AW0
A2
0
order condition, this simplifies to just
equation (1.45).
29
implies
R W h 6 R (W0 (1 + rf ))
(1.46)
where, as before, R(W ) = U 00 (W )/U 0 (W ). Multiplying both terms of (1.46)

by U 0 (W h )(rh rf ), which is a negative quantity, the inequality sign changes:
U 00 (W h )(rh rf ) > U 0 (W h )(rh rf )R (W0 (1 + rf ))
(1.47)
Next, we again let rl denote a realization of r that is lower than rf and define W l
. Then for A > 0, we have W l 6 W0 (1 + rf ).
to be the corresponding level of W
If absolute risk aversion is decreasing in wealth, this implies
R(W l ) > R (W0 (1 + rf ))
(1.48)
Multiplying (1.48) by U 0 (W l )(rl rf ), which is positive, so that the sign

of (1.48) remains the same, we obtain
U 00 (W l )(rl rf ) > U 0 (W l )(rl rf )R (W0 (1 + rf ))
(1.49)
Notice that inequalities (1.47) and (1.49) are of the same form. The inequality
holds whether the realization is r = rh or r = rl . Therefore, if we take expectations over all realizations, where r can be either higher than or lower than rf ,
we obtain
h
i
h
i
)(
)(
E U 00 (W
r rf ) > E U 0 (W
r rf ) R (W0 (1 + rf ))
(1.50)
Since the first term on the right-hand-side is just the first order condition,
inequality (1.50) reduces to
30
i
h
)(
r rf ) > 0
E U 00 (W
(1.51)
Thus, the first conclusion that can be drawn is that declining absolute risk aversion implies dA/dW0 > 0, that is, the individual invests an increasing amount of
wealth in the risky asset for larger amounts of initial wealth. For two individuals
with the same utility function but dierent initial wealths, the more wealthy one
invests a greater dollar amount in the risky asset if utility is characterized by
decreasing absolute risk aversion. While not shown here, the opposite is true,
namely, that the more wealthy individual invests a smaller dollar amount in the
risky asset if utility is characterized by increasing absolute risk aversion.
Thus far, we have not said anything about the proportion of initial wealth
invested in the risky asset. To analyze this issue, we need the concept of relative
risk aversion. Define
dA W0
dW0 A
(1.52)
which is the elasticity measuring the proportional increase in the risky asset for
an increase in initial wealth. Adding 1 A
A to the right hand side of (1.52) gives
=1+
(dA/dW0 )W0 A
A
(1.53)
Substituting the expression dA/dW0 from equation (1.45), we have

i
h
i
h
)(
)(
r rf ) + AE U 00 (W
r rf )2
W0 (1 + rf )E U 00 (W
i
h
=1+
)(
AE U 00 (W
r rf )2
)(
r rf ), this can be re-written as
Collecting terms in U 00 (W
(1.54)
i
h
)(
r rf ){W0 (1 + rf ) + A(
r rf )}
E U 00 (W
i
h
= 1+
)(
AE U 00 (W
r rf )2
h
i
)(
E U 00 (W
r rf )W
i
h
= 1+
)(
AE U 00 (W
r rf )2
31
(1.55)
The denominator is always positive. Therefore, we see that the elasticity, , is
greater than one, so that the individual invests proportionally more in the risky
i
h
)(
> 0. Can we relate
r rf )W
asset with an increase in wealth, if E U 00 (W
this to the individuals risk aversion? The answer is yes and the derivation is
almost exactly the same as that just given.
Consider the case where the individual has relative risk aversion that is
decreasing in wealth. Let rh denote a realization of r such that it exceeds
. Then for A > 0, we have

rf , and let W h be the corresponding level of W
W h > W0 (1 + rf ). If relative risk aversion, Rr (W ) W R(W ), is decreasing in
wealth, this implies
W h R(W h ) 6 W0 (1 + rf )R (W0 (1 + rf ))
(1.56)
Multiplying both terms of (1.56) by U 0 (W h )(rh rf ), which is a negative

quantity, the inequality sign changes:
W h U 00 (W h )(rh rf ) > U 0 (W h )(rh rf )W0 (1 + rf )R (W0 (1 + rf )) (1.57)

Next, let rl denote a realization of r such that it is lower than rf , and let W l
. Then for A > 0, we have W l 6 W0 (1 + rf ). If
be the corresponding level of W
relative risk aversion is decreasing in wealth, this implies
32
W l R(W l ) > W0 (1 + rf )R (W0 (1 + rf ))
(1.58)
Multiplying (1.58) by U 0 (W l )(rl rf ), which is positive, so that the sign of

(1.58) remains the same, we obtain
W l U 00 (W l )(rl rf ) > U 0 (W l )(rl rf )W0 (1 + rf )R (W0 (1 + rf ))
(1.59)
Notice that inequalities (1.57) and (1.59) are of the same form. The inequality
holds whether the realization is r = rh or r = rl . Therefore, if we take expectations over all realizations, where r can be either higher than or lower than rf ,
we obtain
h
i
h
i
U 00 (W
)(
)(
E W
r rf ) > E U 0 (W
r rf ) W0 (1+rf )R(W0 (1+rf )) (1.60)
Since the first term on the right-hand-side is just the first order condition,
inequality (1.60) reduces to
i
h
)(
U 00 (W
r rf ) > 0
E W
(1.61)
Thus, we see that an individual with decreasing relative risk aversion has > 1
and invests proportionally more in the risky asset as wealth increases. The
opposite is true for increasing relative risk aversion: < 1 so that this individual
invests proportionally less in the risky asset as wealth increases. The following
table provides another way of writing this sections main results.
1.4. SUMMARY
33
Risk Aversion
Investment Behavior
Decreasing Absolute
A
W0
>0
Constant Absolute
A
W0
=0
Increasing Absolute
A
W0
<0
Decreasing Relative
A
W0
>
A
W0
Constant Relative
A
W0
A
W0
Increasing Relative
A
W0
<
A
W0
A point worth emphasizing is that absolute risk aversion indicates how the
investors dollar amount in the risky asset changes with changes in initial wealth
while relative risk aversion indicates how the investors portfolio proportion
(or portfolio weight) in the risky asset, A/W0 , changes with changes in initial
wealth.
1.4
Summary
This chapter is a first step toward understanding how an individuals preferences

toward risk aects his portfolio behavior. It was shown that if an individuals
risk preferences satisfied specific plausible conditions, then her behavior could
be represented by a von Neumann-Morgenstern expected utility function. In
turn, the shape of the individuals utility function determines a measure of risk
aversion that is linked to two concepts of a risk premium.
The first one is
the monetary payment that the individual is willing to pay to avoid a risk, an
example being a premium paid to insure against a property/casualty loss. The
second is the rate of return in excess of a riskless rate that the individual requires
to hold a risky asset, which is the common definition of a security risk premium
used in the finance literature. Finally, it was shown how an individuals absolute
and relative risk aversion aects his choice between a risky and risk-free asset. In
34
particular, individuals with decreasing (increasing) relative risk aversion invest

proportionally more (less) in the risky asset as their wealth increases. Though
based on a simple single-period, two asset portfolio choice model, this insight
generalizes to the more complex portfolio choice problems that will be studied
in later chapters.
1.5
Exercises
1. Suppose there are two lotteries P = {p1 , ..., pn } and P = {p1 , ..., pn }. Let
n
P
pi Ui be an individuals expected utility function defined
V (p1 , ..., pn ) =
i=1
over these lotteries. Let W (p1 , ..., pn ) =
n
P
pi Qi where Qi = a + bUi and
i=1
a and b are constants. If P P , so that V (p1 , ..., pn ) > V (p1 , ..., pn ),

must it be the case that W (p1 , ..., pn ) > W (p1 , ..., pn )? In other words, is
W also a valid expected utility function for the individual? Are there any
restrictions needed on a and b for this to be the case?
2. (Allais Paradox) Asset A pays $1,500 with certainty, while asset B pays
$2,000 with probability 0.8 or $100 with probability 0.2.
If oered the
choice between asset A or B, a particular individual would choose asset

A. Suppose, instead, the individual is oered the choice between asset
C and asset D. Asset C pays $1,500 with probability 0.25 or $100 with
probability 0.75 while asset D pays $2,000 with probability 0.2 or $100
with probability 0.8.
If asset D is chosen, show that the individuals
preferences violate the independence axiom.

3. Verify that the HARA utility function in 1.33 becomes the constant absolute risk aversion utility function when = 1 and = . Hint: recall
x
that ea = lim 1 + xa .
x
4. Consider the individuals portfolio choice problem given in 1.42.
As-
1.5. EXERCISES
35
sumeU (W ) = ln (W ) and the rate of return on the risky asset equals
4rf with probability 1

2
. Solve for the individuals proportion
re =
rf with probability 1
2
of initial wealth invested in the risky asset, A/W0 .
5. An expected utility maximizing individual has constant relative risk-aversion

utility, U (W ) = W / with relative risk-aversion coecient of = 1.
The individual currently owns a product that has a probability p of failing,
an event that would result in a loss of wealth that has a present value equal
to L. With probability 1-p the product will not fail and no loss will result.
The individual is considering whether to purchase an extended warranty
on this product. The warranty costs C and would insure the individual
against loss if the product fails. Assuming that the cost of the warranty
exceeds the expected loss from the products failure, determine the individuals level of wealth at which she would be just indierent between
purchasing or not purchasing the warranty.
6. In the context of portfolio choice problem (1.42), show that an individual
with increasing relative risk aversion invests proportionally less in the risky
asset as her initial wealth increases.
7. Consider the following four assets whose payos are as follows:
Asset A =
Asset C =
X with probability px
0 with probability 1 px
X with probability px
0 with probability 1 px
Asset B =
Y with probability py
0 with probability 1 py
Asset D =
Y with probability py
0 with probability 1 py
36
where 0 < X < Y , py < px , px X < py Y , and (0, 1).

7.a When given the choice of asset C versus asset D, an individual chooses asset C. Could this individuals preferences be consistent with von Neumann
- Morgenstern expected utility theory? Explain why or why not.
7.b When given the choice of asset A versus asset B, an individual chooses
asset A. This same individual, when given the choice between asset C and
asset D, chooses asset D. Could this individuals preferences be consistent
with von Neumann - Morgenstern expected utility theory? Explain why
or why not.
8. An individual has expected utility of the form

i
h i
h
i
f = E ebW
E U W
where b > 0. The individuals wealth is normally distributed as N W , 2W .
What is this individuals certainty equivalent level of wealth?
Chapter 2
Mean-Variance Analysis
The preceding chapter studied an investors choice between a risk-free asset
and a single risky asset.
This chapter adds realism by giving the investor
the opportunity to choose among multiple risky assets.
As a University of
Chicago graduate student, Harry Markowitz, wrote a path-breaking article on

this topic (Markowitz 1952).1
Markowitzs insight was to recognize that, in
allocating wealth among various risky assets, a risk-averse investor should focus
on the expectation and the risk of her combined portfolios return, a return
that is aected by the individual assets diversification possibilities. Because of
diversification, the attractiveness of a particular asset when held in a portfolio
can dier from its appeal when it is the sole asset held by an investor.
Markowitz proxied the risk of a portfolios return by the variance of its return. Of course, the variance of an investors total portfolio return depends on
the return variances of the individual assets included in the portfolio. But portfolio return variance also depends on the covariances of the individual assets
1 His work on portfolio theory, of which this article was the beginning, won him a share of
the Nobel prize in economics in 1990. Initially, the importance of his work was not widelyrecognized. Milton Friedman, a member of Markowitzs doctoral dissertation committee and
who also became a Nobel laureate, questioned whether the work met the requirements for an
economics Ph.D. See Bernstein (Bernstein 1992).
37
38
CHAPTER 2. MEAN-VARIANCE ANALYSIS
returns.
Hence, in selecting an optimal portfolio, the investor needs to con-
sider how the co-movement of individual assets returns aects diversification

possibilities.
A rational investor would want to choose a portfolio of assets that eciently
trades o higher expected return for lower variance of return.
Interestingly,
not all portfolios that an investor can create are ecient in this sense. Given
the expected returns and covariances of returns on individual assets, Markowitz
solved the investors problem of constructing an ecient portfolio. His work has
had an enormous impact on the theory and practice of portfolio management
and asset pricing.
Intuitively, it makes sense that investors would want their wealth to earn
a high average return with as little variance as possible.
However, in gen-
eral, an individual who maximizes expected utility may care about moments
of the distribution of wealth in addition to its mean and variance.2
Though
Markowitzs mean - variance analysis fails to consider the eects of these other
moments, in later chapters of this book we will see that his models insights can
be generalized to more complicated settings.
The next section outlines the assumptions on investor preferences and the
distribution of asset returns that would allow us to simplify the investors portfolio choice problem to one that considers only the mean and variance of portfolio
returns. We then analyze a risk-averse investors preferences by showing that
he has indierence curves that imply a trade-o of expected return for variance.
2 For
example, expected utility can depend on the skewness (the third moment) of the return
on wealth. The observation that some people purchase lottery tickets, even though these
investments have a negative expected rate of return, suggests that their utility is enhanced
by positive skewness. Alan Kraus and Robert Litzenberger (Kraus and Litzenberger 1976)
developed a single-period portfolio selection and asset pricing model that extends Markowitzs
analysis to consider investors who have a preference for skewness. Their results generalize
Markowitzs model, but his fundamental insights are unchanged. For simplicity, this chapter
focuses on the orginal Markowitz framework. Recent empirical work by Campbell Harvey
and Akhtar Siddique (Harvey and Siddique 2000) examines the eect of skewness on asset
pricing.
2.1. ASSUMPTIONS ON PREFERENCES AND ASSET RETURNS
39
Subsequently, we show how a portfolio can be allocated among a given set of

risky assets in a mean-variance ecient manner. We solve for the ecient frontier, defined as the set of portfolios that maximize expected returns for a given
variance of returns, and show that any two frontier portfolios can be combined
to create a third. In addition, we show that a fundamental simplification to the
investors portfolio choice problem results when one of the assets included in the
investors choice set is a risk-free asset. The final section of this chapter applies
mean-variance analysis to a problem of selecting securities to hedge the risk of
commodity prices.
This application is an example of how modern portfolio
analysis has influenced the practice of risk management.
2.1
Assumptions on Preferences and Asset Returns
Suppose an expected utility maximizing individual invests her beginning-ofep be the random
period wealth, W0 , in a particular portfolio of assets. Let R
=
return on this portfolio, so that the individuals end-of-period wealth is W
ep .3 Denote this individuals end-of-period utility by U (W

). Given W0 , for
W0 R
ep ) , because W
ep as just U (R
) = U W0 R
notational simplicity we write U (W
ep .
is completely determined by R
ep ) by expanding it in a Taylor series around the mean of

Let us express U(R
ep ]. Let U 0 (), U 00 (), and U (n) () denote the first, second,

ep , denoted as E[R
R
and nth derivatives of the utility function:
3 Thus,
hp is defined as one plus the rate of return on the portfolio.

R
40
ep ]
ep ] + R
ep E[R
ep ] U 0 E[R
ep ) = U E[R
U (R
ep E[R
ep ] + ...
ep ] U 00 E[R
+ 12 R
1
ep E[R
ep ] + ...
ep ] U (n) E[R
R
+ n!
(2.1)
Now let us investigate the conditions that would make this individuals expected utility depend only on the mean and variance of the portfolio return.
We first analyze the restrictions on the form of utility, and then the restrictions
on the distribution of asset returns, that would produce this result.
Note that if the utility function is quadratic, so that all derivatives of order 3
and higher are equal to zero (U (n) = 0, n 3), then the individuals expected
utility is
h
00
ep E[R
ep ]
ep ] + 1 E R
ep ]
ep ) = U E[R
E[
R
E U (R
U
2
ep ]
ep ]U 00 E[R
ep ] + 1 V [R
= U E[R
2
(2.2)
ep ] is the variance of the return on the portfolio.4 Therefore, for any

where V [R
ep , quadratic utility leads to

probability distribution of the portfolio return, R
ep .
expected utility that depends only on the mean and variance of R
Next, suppose that utility is not quadratic, but any general increasing, con-
cave form. Are there particular probability distributions for portfolio returns
that make expected utility, again, depend only on the portfolio returns mean
4 The
expected value of the second term in the Taylor series, E

equals zero.
k

l

hp E[R
hp ] U 0 E[R
hp ] ,
R
2.1. ASSUMPTIONS ON PREFERENCES AND ASSET RETURNS
41
and variance? Such distributions would need to be fully determined by their

means and variances, that is, they must be two-parameter distributions whereby
higher order moments could be expressed in terms of the first two moments
(mean and variance).
Many distributions, such as the gamma, normal, and
lognormal, satisfy this criterion. But in the context of an investors portfolio

selection problem, such distributions need to satisfy another condition. Since
an individual is able to choose which assets to combine into a portfolio, all portfolios created from a combination of individual assets or other portfolios must
have distributions that continue to be determined by their means and variances.
In other words, we need a distribution where if the individual assets return distributions depend on just mean and variance, then the return on a combination
(portfolio) of these assets has a distribution that depends on just mean and
variance. The only distributions that satisfy this "additivity" restriction is the
stable family of distributions, and, among this family, the only distribution that
has finite variance is the normal (Gaussian) distribution.
A portfolio (sum)
of assets whose returns are multivariate normally distributed also has a return
that is normally distributed.
To verify that expected utility depends only on the portfolio returns mean
and variance when this return is normally distributed, note that the third,
fourth, and all higher central moments of the normal distribution are either
n i
h
ep ]
ep E[R
= 0, for n odd, and
zero or a function of the variance: E R
h
n i
n/2
n!
1
ep ]
e
ep E[R
= (n/2)!
E R
, for n even. Therefore, for this case
2 V [Rp ]
the individuals expected utility equals
h
i
ep ] + 0 + 1 V [ R
ep ]
ep ] U 0000 E[R
ep ) = U E[R
ep ] + 1 V [R
ep ]U 00 E[R
E U(R
2
8
n/2
1 e
1
ep ] + ...
V [Rp ]
U (n) E[R
(2.3)
+0 + ... +
(n/2)! 2
42
which depends only on the mean and variance of the portfolio return.
In summary, restricting utility to be quadratic or restricting the distribution
h
i
ep ) as a function of only
of asset returns to be normal allows us to write E U (R
ep ], of the portfolio return. Are either

ep ], and the variance, V [R
the mean, E[R
of these assumptions realistic? If not, it may be unjustified to suppose that
only the first two moments of the portfolio return distribution matter to the
individual investor.
The assumption of quadratic utility clearly is problematic. As mentioned

earlier, quadratic utility displays negative marginal utility for levels of wealth
greater than the bliss point, and it has the unattractive characteristic of increasing absolute risk aversion. There are also diculties with the assumption
of normally distributed asset returns. When asset returns measured over any
finite time period are normally distributed, there exists the possibility that their
end of period values could be negative since realizations from the normal distribution have no lower (or upper) bound. This is an unrealistic description of
returns for many assets such as stocks and bonds because, being limited-liability
assets, their minimum value is non-negative.5
As will be demonstrated in Chapter 12, the assumption of normal returns
can be modified if we generalize the model to have multiple periods and assume
that asset rates of return follow continuous-time stochastic processes. In that
context, one can assume that assets rates of return are instantaneously normally
distributed, which implies that if their means and variances are constant over
infinitesimal intervals, then over any finite interval asset values are lognormally
distributed. This turns out to be a better way of modeling limited liability assets
because the lognormal distribution bounds these assets values to be no less than
zero.
When we later study continuous-time, multi-period models, we shall
5 A related problem is that many standard utility functions, such as constant relative risk
aversion, are not defined for negative values of portfolio wealth.
2.2. INVESTOR INDIFFERENCE RELATIONS
43
see that the results derived here assuming a single-period, discrete-time model
continue to hold, under particular conditions, in the more realistic multi-period
context.
Moreover, in more complex multiperiod models that permit assets
to have time-varying return distributions, we will show that optimal portfolio

choices are straightforward generalizations of the mean-variance results derived
in this chapter.
2.2
Investor Indierence Relations
Therefore, let us proceed by assuming that the individuals utility function, U , is

a general concave utility function and that individual asset returns are normally
p that is normally
distributed. Hence, a portfolio of these assets have a return R
p , 2p ), where we use the
distributed with probability density function f (R; R
p ] and 2p V [R
p ]. In this section we analyze
p E[R
short-hand notation R
an investors "tastes," that is, the investors risk - expected return preferences
when utility depends on the mean (expected return) and variance (risk) of the
return on wealth.
The following section analyzes investment "technologies"
represented by the combinations of portfolio risk and expected return that can
be created from dierent portfolios of individual assets.
Historically, mean
- variance analysis has been illustrated graphically, and we will follow that
convention while also providing analytic results.
Note that an investors expected utility can then be written
h i Z
ep =
E U R
p , 2p )dR
U (R)f (R; R
(2.4)
To gain insight regarding this investors preferences over portfolio risk and expected return, we wish to determine the characteristics of this individuals indifference curves in portfolio mean-variance space. An indierence curve represents
44
the combinations of portfolio mean and variance that would give the individual
the same level of expected utility.6 To understand this relation, let us begin by
defining x
e
p
p R
R
p .
Then
h i Z
ep =
E U R
p + x p )n(x)dx
U(R
(2.5)
where n(x) f (x; 0, 1) is the standardized normal probability density function,

that is, the normal density having a zero mean and unit variance. Now consider
how expected utility varies with changes in the mean and variance of the return
p:
on wealth. Taking the partial derivative with respect to R
h i
Z
ep
E U R
=
U 0 n(x)dx > 0
p
R
(2.6)
since U 0 is always greater than zero. Next, take the partial derivative of (2.5)
with respect to 2p :
h i
ep
E U R
2p
h i
Z
ep
E
U R
1
1
=
=
U 0 xn(x)dx
2p
p
2p
(2.7)
While U 0 is always positive, x ranges between and +. Because x has a

standard normal distribution, which is symmetric, for each positive realization
there is a corresponding negative realization with the same probability density. For example, take the positive and negative pair +xi and xi . Then
n(+xi ) = n(xi ). Comparing the integrand of equation (2.7) for equal absolute
realizations of x, we can show
6 Indierence curves are used in microeconomics to analyze an individuals choice of consuming dierent quantities of multiple goods. For example, if utility, u (x, y), derives from
consuming two goods, with x being the quantity of good X consumed and y being the quantity of good Y consumed, then an indierence curve is the locus of points in X, Y space that
gives a constant level of utilitiy. That is, combinations of goods X and Y for which the
individual would be indierent between consuming. Mathematically, these combinations are
represented as the points (x, y) such that u (x, y) = U, a constant. In this section, we employ
a similar concept but where expected utility depends on the mean and variance of the return
on wealth.
2.2. INVESTOR INDIFFERENCE RELATIONS
p + xi p )xi n(xi ) + U 0 (R
p xi p )(xi )n(xi )
U 0 (R
45
(2.8)
p + xi p )xi n(xi ) U 0 (R
p xi p )xi n(xi )
= U 0 (R
p + xi p ) U 0 (R
p xi p ) < 0
= xi n(xi ) U 0 (R
because
+ xi p ) < U 0 (R
xi p )
U 0 (R
(2.9)
due to the assumed concavity of U, that is, the individual is risk averse so that
U 00 < 0. Thus, comparing U 0 xi n(xi ) for each positive and negative pair, we
conclude that
h i
ep
E U R
2p
1
2 p
U 0 xn(x)dx < 0
(2.10)
which is the intuitive result that higher portfolio variance, without higher portfolio expected return, reduces a risk-averse individuals expected utility.
Finally, an indierence curve is the combinations of portfolio mean and variance that leaves expected utility unchanged. In other words, it is combinations
h i
ep = U , a constant. Higher lev p , 2 that satisfy the equation E U R

of R
p
els of U denote dierent indierence curves providing a greater level of utility.

If we totally dierentiate this equation, we obtain:
h i
h i
ep
ep
h i E U R
E U R
2
p = 0
ep =
d
+
dR
dE U R
p
p
2p
R
(2.11)
which, based on our previous results, tells us that each indierence curve is
p , 2 space:
positively sloped in R
p
46
Rp
Increasing Utility
Figure 2.1: Indierence Curves

h i
h i
ep
ep
E U R
E U R
p
dR
=
/
>0
p
d2p
2p
R
(2.12)
Thus, the indierence curves slope in (2.12) quantifies the extent to which
the individual requires a higher portfolio mean for accepting a higher portfolio
variance.
Indierence curves are typically drawn in mean - standard deviation space,
rather than mean - variance space, because standard deviations of returns are in
the same unit of measurement as returns or interest rates (rather than squared
returns).
Figure 2.1 illustrates such a graph, where the arrow indicates an
increase in the utility level, U.7 It is left as an end-of-chapter exercise to show

that the curves are convex due to the assumed concavity of the utility function.
Having analyzed an investors preferences over dierent combinations of
7 Clearly, these indierence curves cannot "cross" (intersect) because we showed that utility is always increasing in expected portfolio return for a given level of portfolio standard
deviation.
2.3. THE EFFICIENT FRONTIER
47
portfolio means and standard deviations (or variances), let us consider next
what portfolio means and standard deviations are possible given the available
distributions of returns for individual assets.
2.3
The Ecient Frontier
The individuals optimal choice of portfolio mean and variance is determined by

the point where one of these indierence curves is tangent to the set of means
and standard deviations for all feasible portfolios, what we might describe as the
risk versus expected return investment opportunity set. This set represents
all possible ways of combining various individual assets to generate alternative
combinations of portfolio mean and variance (or standard deviation). This set
includes inecient portfolios (those in the interior of the opportunity set) as well
as ecient portfolios (those on the frontier of the set). Ecient portfolios are
those that make best use of the benefits of diversification.
As we shall later
prove, ecient portfolios have the attractive characteristic that any two ecient
portfolios can be used to create any other ecient portfolio.
2.3.1
A Simple Example
To illustrate the eects of diversification, consider the following simple example.

A and
Suppose there are two assets, assets A and B, that have expected returns R
B and variances of 2 and 2 , respectively. Further the correlation between
R
A
B
B but 2 < 2 . Now
A < R
their returns is given by . Let us assume that R
A
B
form a portfolio with a proportion invested in asset A and a proportion 1
invested in asset B.8 The expected return on this portfolio is
8 It is assumed that can be any real number. A < 0 indicates a short position in asset
A while > 1 indicates a short position in asset B.
48
A + (1 )R
B
Rp = R
(2.13)
The expected return of a portfolio is a simple weighted average of the expected

returns of the individual financial assets. Expected returns are not fundamentally transformed by combining individual assets into a portfolio. The standard
deviation of the return on the portfolio is
p = 2 2A + 2(1 )A B + (1 )2 2B 2
(2.14)
In general, portfolio risk, as measured by the portfolios return standard deviation, is a nonlinear function of the individual assets variances and covariances.
Thus, risk is altered in a relatively complex way when individual assets are
combined in a portfolio.
Let us consider some special cases regarding the correlation between the two
assets. Suppose = 1, so that the two assets are perfectly positively correlated.
Then the portfolio standard deviation equals
1
2 2
A + 2(1 ) A B + (1 )2 2B 2
(2.15)
= | A + (1 )B |
which is a simple weighted average of the individual assets standard deviations.

Solving (2.15) for asset As portfolio proportion gives = (B p )/ ( B A ).
Then, by substituting for in (2.13), we obtain
p
R
B + p B (R
A)
B R
= R
B A
B R
A A R
B
A
B R
R
=
p
B A
B A
49
(2.16)
Thus, the relationship between portfolio risk and expected return are two straight
A AR
B / (B A )
p space. They have the same intercept of B R
lines in p , R
and have slopes of the same magnitude but opposite signs. The positively sloped
A ) and (B , R
B ) when = 1 and = 0,
line goes through the points (A , R
respectively. When = B / (B A ) > 1, indicating a short position in asset
B, we see from (2.15) that all portfolio risk is eliminated (p = 0). Figure 2.2
provides a graphical illustration of these relationships.
Next, suppose = 1, so that the assets are perfectly negatively correlated.
Then
( A (1 ) B )2 2
(2.17)
= | A (1 )B |
In a manner similar to the previous case, we can show that
p = A RB + B RA RB RA p
R
A + B
A + B
(2.18)
p space. The intercept at

which, again, represents two straight lines in p , R
p = 0 is given by = B / (A + B ), so that all portfolio risk is eliminated
with positive amounts invested in each asset. Furthermore, the negatively sloped
A ) when = 1, while the positively sloped
line goes through the point (A , R
B ) when = 0. Figure 2.2 summarizes these
line goes through the point (B , R
50
=1
Rp
RB
-1 < < 1
= -1
RA
= -1
=1
Figure 2.2: Ecient Frontier with Two Risky Assets
risk - expected return constraints.
For either the = 1 or = 1 case, an investor would always choose a

portfolio represented by the positively sloped lines because they give the highest average portfolio return for any given level of portfolio risk.
These lines
represent the so-called ecient portfolio frontier. The exact portfolio chosen by
the individual would be where her indierence curve is tangent to the frontier.
When correlation between the assets is imperfect (1 < < 1), the relationship between portfolio expected return and standard deviation is not linear, but,
as illustrated in Figure 2.2, is hyperbolic. In this case, it is no longer possible
to create a riskless portfolio, so that the portfolio having minimum standard
deviation is one where p > 0. We now set out to prove these assertions for
the general case of n assets.
2.3.2
51
Mathematics of the Ecient Frontier
Robert C. Merton (Merton 1972) provided an analytical solution to the following portfolio choice problem: Given the expected returns and the matrix of
covariances of returns for n individual assets, find the set of portfolio weights
that minimizes the variance of the portfolio for each feasible portfolio expected
return. The locus of these points in portfolio expected return - variance space is
the portfolio frontier. This section presents the derivation of Mertons solution.
We begin by specifying the problems notation and assumptions.
2 ... R
n )0 be an n 1 vector of the expected returns of the
= (R
1 R
Let R
n assets.
Also let V be the n n covariance matrix of the returns on the n
assets. V is assumed to be of full rank.9 Since it is a covariance matrix, it is

also symmetric and positive definite. Next, let = ( 1 2 ... n )0 be an n 1
vector of portfolio proportions, such that i is the proportion of total portfolio
wealth invested in the ith asset.
It follows that the expected return on the
portfolio is given by
p = 0 R
(2.19)
and the variance of the portfolio return is given by
2p = 0 V
(2.20)
The constraint that the portfolio proportions must sum to 1 can be written as
0 e = 1 where e is defined to be an n 1 vector of ones.
The problem of finding the portfolio frontier now can be stated as a quadratic
optimization exercise: minimize the portfolios variance subject to the con9 This implies that there are no redundant assets among the n assets. An asset would be
redundant if its return was an exact linear combination of the the returns on other assets.
If such an asset exists, it can be ignored, since its availability does not aect the ecient
portfolio frontier.
52
straints that the portfolios expected return equals Rp and the portfolios weights
sum to one.10
+ [1 0 e]
min 21 0 V + Rp 0 R
(2.21)
The first order conditions with respect to and the two Lagrange multipliers,
and , are:
e = 0
V R
(2.22)
=0
Rp 0 R
(2.23)
1 0 e = 0
(2.24)
Solving (2.22), the optimal portfolio weights satisfy
+ V 1 e
= V 1 R
(2.25)
0 , we have
Pre-multiplying equation (2.25) by R
0 = R
0 V 1 R
+ R
0 V 1 e
Rp = R
(2.26)
Pre-multiplying equation (2.25) by e0 we have
+ e0 V 1 e
1 = e0 = e0 V 1 R
(2.27)
Equations (2.26) and (2.27) are two linear equations in two unknowns, and .
10 In (2.21), the problem actually minimizes one-half the portfolio variance to avoid carrying
an extra "2" in the first order condition (2.22). The solution is the same as minimizing the
total variance and only changes the scale of the Lagrange multipliers.
53
The solution is
Rp
2
(2.28)
Rp
2
(2.29)
R
0 V 1 R,
and e0 V 1 e are scalars. Note
0 V 1 e = e0 V 1 R,
where R
that the denominators of and , given by 2 , are guaranteed to be positive
when V is of full rank.11 Substituting for and in equation (2.25), we have
Rp 1 Rp 1
V R+
V e
2
2
(2.30)
p gives
Collecting terms in R
= a + bRp
where a
(2.31)
V 1 e
V 1 R
V 1 e V 1 R
and
b
.
2
2
Equation (2.31) is both a necessary and sucient condition for a frontier

portfolio. Given Rp , a portfolio must have weights satisfying (2.31) to minimize
its return variance.
Having found the optimal portfolio weights for a given Rp , the variance of
the frontier portfolio is
11
1
To see
this, note thatsince V is positive definite,
so is V . Therefore the quadratic form
R e V 1 R e = 2 22 + 2 = 2 is positive. But since RV 1 R

is a positive quadratic form, then 2 must also be positive.
54
p2
2
mv
Rmv
Rp
Figure 2.3: Frontier Portfolios
= 0 V = (a + bRp )0 V (a + bRp )
2p
=
=
(2.32)
2
Rp
2Rp +
2
2
1 Rp
+
where the second line in equation (2.32) results from substituting in the definitions of a and b and simplifying the resulting expression.
Equation (2.32)
is a parabola in 2p , Rp space and is graphed in Figure 2.3.
From the third
line in equation (2.32), it is obvious that the unique minimum is at the point
Rp = Rmv
which corresponds to a global minimum variance of 2mv 1 .
Substituting Rp =
into 2.30 shows that this minimum variance portfolio has
weights mv = 1 V 1 e.
Each point on the parabola in Figure 2.3 represents an investors lowest possible portfolio variance given some target level of expected return, Rp . However,
55
an investor whose utility is increasing in expected portfolio return and is decreasing in portfolio variance would never choose a portfolio having Rp < Rmv ,
that is, points on the parabola to the left of Rmv . This is because the frontier
portfolios variance actually increases as the target expected return falls when
Rp < Rmv , making this target expected return region irrelevant to an optimizing investor. Hence, the ecient portfolio frontier is represented only by the
region Rp Rmv .
Traditionally, portfolios satisfying (2.32) are graphed in p , Rp space. Taking the square root of both sides of equation (2.32), p becomes a hyperbolic
function of Rp . When this is graphed as in Figure 2.4 with Rp on the vertical
axis and p on the horizontal one, only the upper arc of the hyperbola is relevant because, as stated above, investors would not choose target levels of Rp <
Rmv . Dierentiating (2.32), we can also see that the hyperbolas slope equals
Rp
2
p
=
p
Rp
The upper arc asymptotes to the straight line Rp = Rmv +
(2.33)
q
2
p,
while the
lower arc, representing inecient frontier portfolios, asymptotes to the straight

q
2
p .12
line Rp = Rmv
2.3.3
Portfolio Separation
We now state and prove a fundamental result:

Every portfolio on the mean-variance frontier can be replicated
by a combination of any two frontier portfolios; and an individual
will be indierent between choosing among the n financial assets, or
s
see that the slope of the hyperbola asymptotes to a magnitude of ( 2 ) /, use
t
s

in (2.33) to obtain Rp /p = ( 2 )/ 1/ 2p .
(2.32) to substitute for Rp
Taking the limit of this expression as p gives the desired result.

12 To
56
Rp
Asymptote of upper arc
R p = Rmv +
2
p
Efficient Frontier
Rmv
R p = Rmv
R p2 2 R p +
2
2
p
mv
Figure 2.4: Ecient Frontier

choosing a combination of just two frontier portfolios.
This remarkable finding has an immediate practical implication. If all investors have the same beliefs regarding the distribution of asset returns, namely,
returns are distributed N R, V and, therefore, the frontier is (2.32), then they
can form their individually preferred frontier portfolios by trading in as little as
two frontier portfolios. For example, if a security market oered two mutual
funds, each invested in a dierent frontier portfolio, any mean-variance investor
could replicate his optimal portfolio by appropriately dividing his wealth between only these two mutual funds.13
1p and R
2p be the
The proof of this separation result is as follows. Let R
3p be the expected
expected returns on any two distinct frontier portfolios. Let R
13 To form his preferred frontier portfolio, an investor may require a short position in one
of the frontier mutual funds. Since short positions are not possible with typical open ended
mutual funds, the better analogy would be that these funds are exchange-traded funds (ETFs)
which do permit short positions.
57
return on a third frontier portfolio. Now consider investing a proportion of

wealth, x, in the first frontier portfolio and the remainder, (1 x), in the
second frontier portfolio. Clearly, a value for x can be found that makes the
expected return on this composite portfolio equal to that of the third frontier
portfolio.14
= xR
1p + (1 x)R
2p
R
3p
(2.34)
In addition, because portfolios 1 and 2 are frontier portfolios, we can write their
portfolio proportions as a linear function of their expected returns. Specifically,
1p and 2 = a +bR
2p where i is the n 1 vector of optimal
we have 1 = a +bR
portfolio weights for frontier portfolio i. Now create a new portfolio with an n1
vector of portfolio weights given by x 1 + (1 x)2 . The portfolio proportions
of this new portfolio can be written as
x1 + (1 x) 2
1p ) + (1 x)(a + bR
2p )
= x(a + bR
(2.35)
1p + (1 x)R
2p )
= a + b(xR
3p = 3
= a + bR
where, in the last line of (2.35) we have substituted in equation (2.34). Based
on the portfolio weights of the composite portfolio, x 1 + (1 x) 2 ,equalling
3p , which is the portfolio weights of the third frontier portfolio, 3 , this
a + bR
composite portfolio equals the third frontier portfolio. Hence, any given ecient
portfolio can be replicated by two frontier portfolios.
Portfolios on the mean-variance frontier have an additional property that will
prove useful to the next sections analysis of portfolio choice when a riskless asset
14 x
may be any positive or negative number.
58
exists and also to understanding equilibrium asset pricing in Chapter 3. Except

for the global minimum variance portfolio, mv , for each frontier portfolio one
can find another frontier portfolio with which its returns have zero covariance.
That is, one can find pairs of frontier portfolios whose returns are orthogonal.
To show this, note that the covariance between two frontier portfolios, w1 and
w2 , is
1V 2
= (a + bR1p )0 V (a + bR2p )
+
=
R
1p
2p
2
(2.36)
Setting this equal to zero and solving for R2p , the expected return on the portfolio that has zero covariance with portfolio 1 is
R2p
R1p
= Rmv
(2.37)
R1p Rmv
2
Note that if R1p Rmv > 0 so that frontier portfolio 1 is ecient, then
equation (2.37) indicates that R2p < Rmv , implying that frontier portfolio 2
must be inecient.
We can determine the relative locations of these zero
covariance portfolios by noting that in p , Rp space, a line tangent to the
frontier at the point 1p , R1p is of the form
Rp = R0 +
where
Rp
p
Rp
p
p p =1p
(2.38)
p =1p denotes the slope of the hyperbola at point 1p , R1p and
R0 denotes the tangent lines intercept at p = 0. Using (2.33) and (2.32), we
59
Tangent to Portfolio 1
Rp
R1 p
R mv
2
R0 = R 2 p
mv 2 p
1p
Figure 2.5: Frontier Portfolios with Zero Covariance
can solve for R0 by evaluating (2.38) at the point 1p , R1p :

R0
Rp
2
1p 1p
p =1p 1p = R1p
p
R1p
"
2 #
1 R1p
2
+
= R1p
2
R1p
= R1p
(2.39)
R1p
= R2p
Hence, as shown in Figure 2.5, the intercept of the line tangent to frontier portfolio 1 equals the expected return of its zero covariance counterpart, frontier
portfolio 2 .
60
2.4
The Ecient Frontier with a Riskless Asset
Thus far, we have assumed that investors can hold only risky assets. An implication of our analysis was that while all investors would choose ecient portfolios
of risky assets, these portfolios would dier based on the particular investors
level of risk aversion. However, as we shall now see, introducing a riskless asset
can simplify the investors portfolio choice problem. This augmented portfolio
choice problem, whose solution was first derived by James Tobin (Tobin 1958),
is one that we now consider.15
Assume that there is a riskless asset with return Rf . Let continue to
be the n 1 vector of portfolio proportions invested in the risky assets. Now,
however, the constraint 0 e = 1 does not apply, because 1 0 e is the portfolio
proportion invested in the riskless asset. We can impose the restriction that the
portfolio weights for all n + 1 assets sum to one by writing the expected return
on the portfolio as
Rf e)
p = Rf + 0 (R
R
(2.40)
The variance of the return on the portfolio continues to be given by 0 V . Thus,

the individuals optimization problem is changed to:
Rf e)
min 21 0 V + Rp Rf + 0 (R
(2.41)
In a manner similar to the previous derivation, the first order conditions lead
to the solution
Rf e)
= V 1 (R
(2.42)
15 Tobins work on portfolio selection was one of his contributions cited by the selection
committee that awarded him the Nobel prize in economics in 1981.
2.4. THE EFFICIENT FRONTIER WITH A RISKLESS ASSET
61
Rp Rf
Rp Rf
1
where
is posi=
0
2 . Since V
1
2R
+
R
f
R Rf e V (R Rf e)
f
tive definite, is non-negative when Rp Rf , the region of the ecient frontier
where investors expected portfolio return is at least as great as the risk-free
return. Given (2.42), the amount optimally invested in the riskless asset is determined by 1 e0 w . Note that since is linear in Rp , so is , similar to the
previous case of no riskless asset. The variance of the portfolio now takes the
form
2p = 0 V =
(Rp Rf )2
2Rf + Rf2
(2.43)
Taking the square root of each side of (2.43) and re-arranging

1
Rp = Rf 2Rf + R2f 2 p
(2.44)
which indicates that the frontier is linear in p , Rp space. Corresponding to

the hyperbola for the no riskless asset case, the frontier when a riskless asset
is included becomes two straight lines, each with an intercept of Rf but one
12
having a positive slope of 2Rf + Rf2 , the other having a negative

12
slope of 2Rf + Rf2 . Of course, only the positively sloped line is the
ecient portion of the frontier.
Since is linear in Rp , the previous sections separation result continues

to hold: any portfolio on the frontier can be replicated by two other frontier
portfolios. However, when Rf 6= Rmv
principle obtains.16
holds, an even stronger separation
In this case, any portfolio on the linear ecient frontier
can be replicated by two particular portfolios: one portfolio that is located on

the "risky asset only" frontier and another portfolio that holds only the riskless
16 We continue to let R
mv denote the expected return on the minimum variance portfolio
that holds only risky assets. Of course with a riskless asset, the minimum variance portfolio
would be one that holds only the riskless asset.
62
Efficient Frontier
Rp
Rp = R f +
RA
RA R f
Rmv
Rf
Frontier with only

Risky Assets
Frontier
mv A
Figure 2.6: Ecient Frontier with a Riskless Asset
asset.
Let us start by proving this result for the situation where Rf < Rmv . We as 12
sert that the ecient frontier given by the line Rp = Rf + 2Rf + R2f p
can be replicated by a portfolio consisting of only the riskless asset and a portfo-
lio on the risky asset only frontier that is determined by a straight line tangent
to this frontier whose intercept is Rf . This situation is illustrated in Figure 2.6
where A denotes the portfolio of risky assets determined by the tangent line
having intercept Rf . If we can show that the slope of this tangent line equals
12
2Rf + Rf2 , then our assertion is proved.17 Let RA and A be the
expected return and standard deviation of return, respectively, on this tangency
portfolio. Then the results of (2.37) and (2.39) allow us to write the slope of
17 Note that if a proportion x is invested in any risky asset portfolio having expected return
and standard deviation of RA and A , respectively, and a proportion 1 x is invested in the
riskless asset having certain return Rf , then the combined portfolio has an expected return

and standard deviation of Rp = Rf + x RA Rf and p = x A , respectively. When
graphed in Rp , p space, we can substitute for x to show that these combination portfolios
are represented by the straight line Rp = Rf +
RA Rf
A
p whose intercept is Rf .
63
the tangent as
"
#
2
Rf /A
2
=
Rf
"
#
2Rf Rf2
=
/ A
Rf
RA Rf
A
(2.45)
Furthermore, we can use (2.32) and (2.37) to write
2A
=
=
=
2
1 RA
+
2
2
1
+ 3
Rf 2
(2.46)
Rf2 2Rf +
2
2 Rf
Substituting the square root of (2.46) into (2.45) gives18
RA Rf
A
"
2Rf Rf2
Rf
2
1
Rf 2Rf + 2
Rf
12
R2f 2Rf +
(2.47)
which is the desired result.

This result is an important simplification.
If all investors agree on the
distribution of asset returns (returns are distributed N R, V ), then they all

12
consider the linear ecient frontier to be Rp = Rf + 2Rf + Rf2 p

and all will choose to hold risky assets in the same relative proportions given
by the tangency portfolio A . Investors dier only in the amount of wealth

18 Because it is assumed that R <
f
order for A to be positive.
the square root of (2.46) has an opposite sign in
64
they choose to allocate to this portfolio of risky assets versus the risk-free asset.
Along the ecient frontier depicted in Figure 2.7, the proportion of an investors
total wealth held in the tangency portfolio, e , increases as one moves to the
right. At point p , Rp = (0, Rf ), e = 0 and all wealth is invested in the
risk-free asset. In between points (0, Rf ) and A , RA , which would be the
case if, say, investor 1 had an indierence curve tangent to the ecient frontier
at point 1 , Rp1 , then 0 < e < 1 and positive proportions of wealth are
invested in the risk-free asset and the tangency portfolio of risky assets. At
point A , RA , e = 1 and all wealth is invested in risky assets and none in

the risk-free asset. Finally, to the right of this point, which would be the case
if, say, investor 2 had an indierence curve tangent to the ecient frontier at
point 2 , Rp2 , then e > 1. This implies a negative proportion of wealth in
the risk-free asset. The interpretation is that investor 2 borrows at the risk-free
rate to invest more than 100 percent of her wealth in the tangency portfolio of
risky assets.
In practical terms, such an investor could be viewed as buying
risky assets "on margin," that is leveraging her asset purchases with borrowed
money.
It will later be argued that Rf < Rmv , the situation depicted in Figures
2.6 and 2.7, is required for asset market equilibrium.
However, we briefly
describe the implications of other parametric cases. When Rf > Rmv , the
12
ecient frontier of Rp = Rf + 2Rf + Rf2 p is always above the riskyasset-only frontier.
Along this ecient frontier, the investor short-sells the
tangency portfolio of risky assets.
This portfolio is located on the inecient
portion of the risky-asset-only frontier at the point where the line Rp = Rf

12
2Rf + Rf2 p becomes tangent. The proceeds from this short selling
are then wholly invested in the risk-free asset.
Lastly, when Rf = Rmv the
portfolio frontier is given by the asymptotes illustrated in Figure 2.4.
It is
65
Efficient Frontier
Rp
Indifference Curve
Investor 2
Rp2
RA
Indifference Curve
Investor 1
R p1
Rf
Figure 2.7: Investor Portfolio Choice

straightforward to show that e = 0 for this case, so that total wealth is
invested in the risk-free asset.
However, the investor also holds a risky, but
zero net wealth, position in risky assets.
In other words, the proceeds from
short-selling particular risky assets are used to finance long positions in other
risky assets.
2.4.1
An Example with Negative Exponential Utility
To illustrate our results, let us specify a form for an individuals utility function.
This enables us to determine the individuals preferred ecient portfolio, that
is, the point of tangency between the individuals highest indierence curve and
the ecient frontier. Given a specific utility function and normally distributed
asset returns, we show how the individuals optimal portfolio weights can be
derived directly by maximizing expected utility.
be the individuals end-of-period wealth and assume that
As before, let W
66
she maximizes expected negative exponential utility.
) = ebW
U(W
(2.48)
where b is the individuals coecient of absolute risk aversion. Now define

br bW0 , which is the individuals coecient of relative risk aversion at initial
wealth W0 . Equation (2.48) can be re-written:
) = ebr W /W0 = ebr Rp

U(W
(2.49)
p is the total return (one plus the rate of return) on the portfolio.
where R
In this problem, we assume that initial wealth can be invested in a riskless
asset and n risky assets. As before, denote the return on the riskless asset as
Also as before,
Rf and the returns on the n risky assets as the n 1 vector R.
let = (1 ... n )0 be the vector of portfolio weights for the n risky assets. The
risky assets returns are assumed to have a joint normal distribution where R
is the n 1 vector of expected returns on the n risky assets and V is the n n
covariance matrix of returns. Thus, the expected return on the portfolio can be
Rf e) and the variance of the return on the portfolio
p Rf + 0 (R
written R
is 2p 0 V .
Now recall the properties of the lognormal distribution. If x
is a normally
distributed random variable, for example, x
N(, 2 ), then z = exh is lognor-
mally distributed. The expected value of ze is

1
E[
z ] = e+ 2
(2.50)
Rf e) is normally distributed, then

p = Rf + 0 (R
From (2.49), we see that if R

is lognormally distributed. Using equation (2.50), we have
U W
h i
1 2 0
f e)]+ 2 br V
f = ebr [Rf +0 (RR
E U W
67
(2.51)
The individual chooses portfolio weights by maximizing expected utility:

h i
1 2 0
f e)]+ 2 br V
f = max ebr [Rf +0 (RR
maxE U W
(2.52)
Because the expected utility function is monotonic in its exponent, the maximization problem in (2.52) is equivalent to
Rf e) 1 br 0 V
max 0 (R
2
(2.53)
The n first order conditions are
Rf e br V = 0
R
(2.54)
Solving for , we obtain
1 1
V (R Rf e)
br
(2.55)
Thus, we see that the individuals optimal portfolio choice depends on br ,

her coecient of relative risk aversion, and the expected returns and covariances
of the assets. Comparing (2.55) to (2.42), note that
Rp Rf
1
=
0
Rf e)
br
R Rf e V 1 (R
(2.56)
so that the greater the investors relative risk aversion, br , the smaller is her
target mean portfolio return, Rp , and the smaller is the proportion of wealth
invested in the risky assets. In fact, multiplying both sides of (2.55) by W0 , we
see that the absolute amount of wealth invested in the risky assets is
68
1
Rf e)
W0 = V 1 (R
b
(2.57)
Therefore, the individual with constant absolute risk aversion, b, invests a fixed
dollar amount in the risky assets, independent of her initial wealth. As wealth
increases, each additional dollar is invested in the risk-free asset. Recall that
this same result was derived at the end of Chapter 1 for the special case of a
single risky asset.
As in this example, constant absolute risk aversions property of making risky
asset choice independent of wealth often allows for simple solutions to portfolio
choice problems when asset returns are assumed to be normally distributed.
However, the unrealistic implication that both wealthy and poor investors invest
the same dollar amount in risky assets limits the empirical applications of using
this form of utility. As we shall see in later chapters of this book, models where
utility displays constant relative risk aversion is more typical.
2.5
An Application to Cross-Hedging
The following application of mean-variance analysis is based on Anderson and

Danthine (Anderson and Danthine 1981). Consider a one-period model of an
individual or institution that is required to buy or sell a commodity in the
future and would like to hedge the risk of such a transaction by taking positions
in futures (or other financial securities) markets. Assume that this financial
operator is committed at the beginning of the period, date 0, to buy y units of
a risky commodity at the end of the period, date 1, at the then-prevailing spot
price p1 . For example, a commitment to buy could arise if the commodity is
a necessary input in the operators production process.19
19 An
Conversely, y < 0
example of this case would be a utility that generates electricity from oil.
2.5. AN APPLICATION TO CROSS-HEDGING
69
represents a commitment to sell y units of a commodity, which could be due to

the operator producing a commodity that is non-storable.20 What is important
is that, as of date 0, y is deterministic while p1 is stochastic.
There are n financial securities (for example, futures contracts) in the economy. Denote the date 0 price of the ith financial security as psi0 . Its date 1
price is psi1 , which is uncertain as of date 0. Let si denote the amount of the
ith security purchased at date 0. Thus, si < 0 indicates a short position in the
security.
Define the n 1 quantity and price vectors s [s1 ... sn ]0 , ps0 [ps10 ... psn0 ]0 ,
and ps1 [ps11 ... psn1 ]0 . Also define ps ps1 ps0 as the n1 vector of security price
changes. This is the profit at date 1 from having taken unit long positions in
each of the securities (futures contracts) at date 0, so that the operators profit
from its security position is ps0 s. Also define the first and second moments of
the date 1 prices of the spot commodity and the financial securities:
E[p1 ] = p1 , V ar[p1 ] = 00 , E [ps1 ] = ps1 , E[ps ] = ps , Cov[psi1 , psj1 ] = ij ,
Cov[p1 , psi1 ] = 0i , and the (n + 1) (n + 1) covariance matrix of the spot
commodity and financial securities is
00
=
001
01
11
(2.58)
where 11 is an n n matrix whose i, j th element is ij , and 01 is an 1 n

vector whose ith element is 0i .
For simplicity, let us assume that y is fixed and, therefore, is not a decision variable at date 0. Then the end-of-period profit (wealth) of the financial
operator, W , is given by
20 For example, the operator could be a producer of an agricultural good, such as corn,
wheat, or soybeans.
70
W = ps0 s p1 y
(2.59)
What the operator must decide is the date 0 positions in the financial securities. We assume that the operator chooses s in order to maximize the following
objective function that depends linearly on the mean and variance of profit:
maxE[W ] 12 V ar[W ]
s
(2.60)
As was shown in the previous sections (2.53), this objective function results
from maximizing expected utility of wealth when portfolio returns are normally
distributed and utility displays constant absolute risk aversion.21 Substituting
in for the operators profit, we have
max ps0 s p1 y 12 y 2 00 + s0 11 s 2y01 s

s
(2.61)
The first order conditions are
ps [11 s y001 ] = 0
(2.62)
Thus, the optimal positions in financial securities are
s =
=
1 1 s
0
p + y1
11 01
11
1 1 s
0
(p ps0 ) + y1
11 01
11 1
(2.63)
21 Similar to the previous derivation, the objective function (2.60) can be derived from an
expected utility function of the form E [U (W )] = exp [W ] where is the operators
coecient of absolute risk aversion. Unlike the previous example, here the objective function
is written in terms of total profit (wealth) not portfolio returns per unit wealth. Also, risky
asset holdings, s, are in terms of absolute amounts purchased, not portfolio proportions.
Hence, is the coecient of absolute risk aversion, not relative risk aversion.
2.5. AN APPLICATION TO CROSS-HEDGING
71
Let us first consider the case of y = 0. This can be viewed as the situation
faced by a pure speculator, by which we mean a trader who has no requirement to
hedge. If n = 1 and ps1 > ps0 , the speculator takes a long position in (purchases)
the security, while if ps1 < ps0 , the speculator takes a short position in (sells)
the security. The magnitude of the position is tempered by the volatility of the
security (1
11 = 1/ 11 ), and the speculators level of risk aversion, . However,
for the general case of n > 1, an expected price decline or rise is not sucient
to determine whether a speculator takes a long or short position in a particular
security. All of the elements in 1
11 need to be considered, since a position in
a given security may have particular diversification benefits.
For the general case of y 6= 0, the situation faced by a hedger, the demand for
financial securities is similar to that of a pure speculator in that it also depends
on price expectations. In addition, there are hedging components to the demand
for financial assets, call them sh :
0
sh y1
11 01
(2.64)
This is the solution to the problem min V ar(W ). Thus, even for a hedger,
s
it is never optimal to minimize volatility (risk) unless risk aversion is infinitely

large. Even a risk-averse, expected utility maximizing hedger should behave
somewhat like a speculator in that securities expected returns matter. From
definition (2.64), note that when n = 1 the pure hedging demand per unit of
the commodity purchased, sh /y, simplifies to22
Cov(p1 , ps1 )
sh
=
y
V ar(ps1 )
(2.65)
22 Note that if the correlation between the commodity price and the financial security return
were equal to 1, so that a perfect hedge exists, then (2.65) beomces sh /y = 00 / 11 , that
is, the hedge ratio equals the ratio of the commodity prices standard deviation to that of the
security price.
72
0
For the general case, n > 1, the elements of the vector 1
11 01 equal the coe-
cients 1 , ..., n in the multiple regression model
p1 = 0 + 1 ps1 + 2 ps2 + ... + n psn +
(2.66)
where p1 p1 p0 , psi ps1i ps0i , and is a mean-zero error term. An

implication of (2.66) is that an operator might estimate the hedge ratios, sh /y,
by performing a statistical regression using an historical times series of the n 1
vector of security price changes. In fact, this is a standard way that practitioners
use to calculate hedge ratios.
2.6
Summary
When the returns on individual assets are multivariate normally distributed,

a risk-averse investor optimally chooses among a set of mean-variance ecient
portfolios. Such portfolios make best use of the benefits of diversification by
providing the highest mean portfolio return for a given portfolio variance. The
particular ecient portfolio chosen by a given investor depends on her level of
risk-aversion.
However, the ability to trade in only two ecient portfolios is
sucient to satisfy all investors, because any ecient portfolio can be created
from any other two.
When a riskless asset exists, the set of ecient portfo-
lios has the characteristic that the portfolios mean returns are linear in their
portfolio variances. In such a case, a more risk averse investor optimally holds
a positive amount of the riskless asset and a positive amount of a particular
risky asset portfolio, while a less risk averse investor optimally borrows at the
riskless rate to purchase the same risky asset portfolio in an amount exceeding
his wealth.
This chapter provided insights on how individuals should optimally allocate
2.7. EXERCISES
their wealth among various assets.
73
Taking the distribution of returns for all
available assets as given, we determined any individuals portfolio demands for

these assets. Having now derived a theory of investor asset demands, the next
chapter will consider the equilibrium asset pricing implications of this investor
behavior.
2.7
Exercises
1. Prove that the indierence curves graphed in Figure 2.1 are convex if
the utility function is concave. Hint: Suppose there are two portfolios,
portfolios 1 and 2, that lie on the same indierence curve, where this
indierence curve has expected utility of U.
Let the mean returns on
2p , respectively, and let the standard

1p and R
portfolios 1 and 2 be R
deviations of returns on portfolios 1 and 2 be 1p and 2p , respectively.
Consider a third portfolio located in (Rp , p ) space that happens to be
on a straight line between portfolios 1 and 2, that is, a portfolio having
1p + (1 x) R
2p and
3p = xR
a mean and standard deviation satisfying R
3p = x 1p + (1 x) 2p where 0 < x < 1. Prove that the indierence
curve is convex by showing that the expected utility of portfolio 3 exceeds
U . Do this by showing that the utility of portfolio 3 exceeds the convex
combination of utilities for portfolios 1 and 2 for each standardized normal
realization. Then, integrate over all realizations to show this inequality
holds for expected utilities.
2. Show that the covariance between the return on the minimum variance
portfolio and the return on any other portfolio equals the variance of the
return on the minimum variance portfolio. Hint: Write down the variance
of a portfolio that consists of a proportion x invested in the minimum
74

variance portfolio and a proportion (1 x) invested in any other portfolio.
Then minimize the variance of this composite portfolio with respect to x.
3. Show how to derive the solution for the optimal portfolio weights for a
frontier portfolio when there exists a riskless asset, that is, equation (2.42)
Rp Rf
Rf e) where
given by = V 1 (R
=
Rf e 0 V 1 (R
Rf e)
R
Rp Rf
. The derivation is similar to the case with no riskless
2Rf + Rf2
asset.
4. Show that when Rf = Rmv , the optimal portfolio involves e = 0.
5. Consider the mean-variance analysis covered in class where there are n

risky assets whose returns are jointly normally distributed. Assume that
investors dier with regard to their (concave) utility functions and their
initial wealths. Also assume that investors can lend at the risk-free rate,
Rf < Rmv , but investors are restricted from risk-free borrowing, that is,
no risk-free borrowing is permitted.
a. Given this risk-free borrowing restriction, graphically show the ecient frontier for these investors in expected portfolio return - standard
deviation space Rp , p .
b. Explain why only three portfolios are needed to construct this ecient
frontier, and locate these three portfolios on your graph. (Note these
portfolios may not be unique.)
c. At least one of these portfolios will sometimes need to be sold short
to generate the entire ecient frontier. Which portfolio(s) is it (label it
on the graph) and in what range(s) of the ecient frontier will it be sold
short? Explain.
2.7. EXERCISES
75
6. Suppose there are n risky assets whose returns are multi-variate normally
distributed. Denote their n 1 vector of expected returns as R and their
nn covariance matrix as V . Let there also be a riskless asset with return
Rf . Let portfolio a be on the mean-variance ecient frontier and have an
expected return and standard deviation of Ra and a , respectively. Let
portfolio b be any other (not necessarily ecient) portfolio having expected
return and standard deviation Rb and b , respectively.
Show that the
correlation between portfolios a and b equals portfolio bs Sharpe ratio divided by portfolio as Sharpe ratio, where portfolio is Sharpe ratio equals
i Rf /i . (Hint: write the correlation as Cov (Ra , Rb ) / ( a b ), and

R
derive this covariance using the properties of portfolio eciency.)
7. A corn grower has utility of wealth given by U (W ) = eaW where a >

0. This farmers wealth depends on the total revenue from the sale of corn
at harvest time. This total revenue is a random variable se = qepe, where
qe is the number of bushels of corn harvested and pe is the spot price, net
of harvesting costs, of a bushel of corn at harvest time. The farmer can
enter into a corn futures contract having a current price of f0 and a random
price at harvest time of fe. If k is the number of short positions in this
futures contract taken by the farmer, then the farmers wealth at harvest
f = se k fe f0 . If seN s, 2s , feN f, 2 , and

time is given by W
f

Cov se, fe = s f , then solve for the optimal number of futures contract
short positions, k, that the farmer should take.
8. Consider the standard Markowitz mean-variance portfolio choice problem

where there are n risky assets and a riskfree asset. The risky assets n 1
e have a multi-variate normal distribution N R, V ,

vector of returns, R,
where R is the assets n 1 vector of expected returns and V is a non-
singular n n covariance matrix. The riskfree assets return is given by
76

Rf > 0. As usual, assume no labor income so that the individuals end-ofep ,
f = W0 R
period wealth depends only on her portfolio return, that is, W
e Rf e where w is an nx1
ep = Rf + w0 R
where the portfolio return is R
vector of portfolio weights for the risky assets and e is an n 1 vector of

1s. Recall that we solved for the optimal portfolio weights, w for the
case of an individual with expected utility displaying constant absolute
i
h i
h
i
f = E ebW
. Now, in this problem, consider
risk aversion, E U W
the dierent case of an individual with expected utility displaying constant
i
h i
h
f = E 1W
f where < 1. What is w
relative risk aversion, E U W
for this constant relative risk aversion case?
Hint: Recall the ecient
frontier and consider the range of the probability distribution of the tangency portfolio. Also consider what would be the individuals marginal
utility should end-of-period wealth be non-postive. This marginal utility
will restrict the individuals optimal portfolio choice.
Chapter 3
CAPM, Arbitrage, and

Linear Factor Models
In this chapter, we analyze the asset pricing implications of the previous chapters mean - variance portfolio analysis. From one perspective, the Markowitz
- Tobin portfolio selection rules is a normative theory instructing how an individual investor can best allocate wealth among various assets.
However, it
also could be interpreted as a positive or descriptive theory of how an investor

actually behaves. If this latter view is taken, then a logical extension of portfolio selection theory is to consider the equilibrium asset pricing consequences
of investors individually rational actions. The portfolio choices of individual
investors represent their particular demands for assets. By aggregating these
investor demands and equating them to asset supplies, equilibrium asset prices
can be determined. In this way, portfolio choice theory can provide a foundation for an asset pricing model. Indeed, such a model, the Capital Asset Pricing
Model (CAPM), was derived at about the same time by four individuals: Jack
77
78 CHAPTER 3. CAPM, ARBITRAGE, AND LINEAR FACTOR MODELS

Treynor, William Sharpe, John Lintner, and Jan Mossin.1
CAPM has influ-
enced financial practice in highly diverse ways. It has provided foundations for
capital budgeting rules, for the regulation of utilities rates of return, for performance evaluation of money managers, and for the creation of indexed mutual
funds.
This chapter starts by deriving the CAPM and studying its consequences
for assets rates of return. The notion that investors might require higher rates
of return for some types of risks, but not others, is an important insight of
CAPM and extends to other asset pricing models. CAPM predicts that assets
risk premia result from a single risk factor, the returns on the market portfolio
of all risky assets which, in equilibrium, is a mean-variance ecient portfolio.
However, it is not hard to imagine that a weakening of CAPMs restrictive
assumptions could generate risk premia deriving from multiple factors. Hence,
we then consider how assets risk premia may be related when multiple riskfactors generate assets returns.
We derive this relationship, not based on a
model of investor preferences as was done in deriving CAPM, but based on

the concept that competitive and ecient securities markets should not permit
arbitrage.
As a prelude to considering a multi-factor asset pricing model, we define
and give examples of arbitrage. Arbitrage pricing is the primary technique for
valuing one asset in terms of another. It is the basis of so-called relative pricing
models, contingent claims models, or derivative pricing models.
We look at
some simple applications of arbitrage pricing and then study the multi-factor
Arbitrage Pricing Theory (APT) developed by Stephen Ross (Ross 1976). APT
is the basis of the most popular empirical multi-factor models of asset pricing.
1 William Sharpe, a student of Harry Markowitz, shared the 1990 Nobel prize with
Markowitz and Merton Miller. See (Treynor 1961), (Sharpe 1964), (Lintner 1965), and
(Mossin 1966).
3.1. THE CAPITAL ASSET PRICING MODEL
3.1
79
The Capital Asset Pricing Model
We proved that if investors maximize expected utility that depends only on

the expected return and variance of end-of-period wealth, then, no matter what
their particular levels of risk-aversion, they would be interested only in portfolios
on the ecient frontier. This mean-variance ecient frontier was the solution to
the problem of computing portfolio weights that would maximize a portfolios
expected return for a given portfolio standard deviation or, alternatively, minimizing a portfolios standard deviation for a given expected portfolio return.
The point on this ecient frontier ultimately selected by a given investor was
that combination of expected portfolio return and portfolio standard deviation
that maximized the particular investors expected utility.
For the case of n
risky assets and a riskfree asset, the optimal portfolio weights for the n risky
assets were shown to be
Rf e
= V 1 R
(3.1)
Rp Rf
R
0 V 1 R,
and
0 V 1 e = e0 V 1 R,
, R
2Rf + R2f
e0 V 1 e. The amount invested in the riskfree asset is then 1 e0 . Since
where
is a scalar quantity that is linear in Rp , which is the individual investors

equilibrium portfolio expected return, the weights in (3.1) are also linear in Rp .
Rp is determined by where the particular investors indierence curve is tangent
to the ecient frontier. Thus, because dierences in Rp just aect the scalar,
, we see that all investors, no matter what their degree of risk aversion, choose
to hold the risky assets in the same relative proportions.
Mathematically, we showed that the ecient frontier is given by
Rp Rf
p =
12
2Rf + R2f
(3.2)
Efficient Frontier
Rp
Rm
Rp = R f +
Rm R f
Rmv
Frontier with only

Risky Assets
Rf
mv m
Figure 3.1: Capital Market Equilibrium
which, as illustrated in Figure 3.1, is linear when plotted in p , Rp space.
3.1.1
Characteristics of the Tangency Portfolio
The ecient frontier, given by the line through Rf and m , implies that investors optimally choose to hold combinations of the riskfree asset and the
ecient frontier portfolio of risky assets having portfolio weights wm . We can
easily solve for this unique tangency portfolio of risky assets since it is the
point where an investor would have a zero position in the riskfree asset, that is,
0 . Pre-multiplying (3.1) by e0 , setting the result to 1,
e0 = 1, or Rp = R
and solving for , we obtain = m [ Rf ]1 , so that
Rf e)
m = mV 1 (R
(3.3)
81
Let us now investigate the relationship between this tangency portfolio and
individual assets. Consider the covariance between the tangency portfolio and
the individual risky assets. Define M as the n 1 vector of covariances of the
tangency portfolio with each of the n risky assets. Then using (3.3) we see that
Rf e)
M = V wm = m(R
(3.4)
Note that the variance of the tangency portfolio is simply 2m = m0 V m . So

that if we then pre-multiply equation (3.4) by m0 we obtain
2m
Rf e)
= m0 M = m m0 (R
(3.5)
m Rf )
= m(R
m m0 R
is the expected return on the tangency portfolio.2 Rewhere R
arranging (3.4) and substituting in for m from (3.5), we have
Rf e) =
(R
where
M
2m
1
M
M = 2 (R
m Rf ) = (Rm Rf )
m
m
is the n 1 vector whose ith element is
m ,R
hi )
Cov(R
m ) .
V ar(R
(3.6)
Equation
(3.6) shows that a simple relationship links the excess expected return (expected
m Rf ), to
return in excess of the risk-free rate) on the tangency portfolio, (R
Rf e).
the excess expected returns on the individual risky assets, (R
2 Note that the elements of m sum to 1 since the tangency portfolio has zero weight in
the riskfree asset.
3.1.2
Market Equilibrium
Now suppose that individual investors, each taking the set of individual assets
expected returns and covariances as fixed (exogenous), all choose mean-variance
ecient portfolios. Thus, each investor decides to allocate his or her wealth
between the riskfree asset and the unique tangency portfolio. Because individual
investors demand the risky assets in the same relative proportions, we know
that the aggregate demands for the risky assets will have the same relative
proportions, namely those of the tangency portfolio. Recall that our derivation
of this result does not assume a representative investor in the sense of requiring
all investors to have identical utility functions or beginning-of-period wealth.
It does assume that investors have identical beliefs regarding the probability
distribution of asset returns, that all risky assets can be traded, that there are
no indivisibilities in asset holdings, and that there are no limits on borrowing
or lending at the riskfree rate.
We can now define an equilibrium as a situation where asset returns are such
that the investors demands for the assets equal the assets supplies.
What
determines the assets supplies? One way to model asset supplies is to assume
they are fixed. For example, the economy could be characterized by a fixed
quantity of physical assets that produce random output at the end of the period.
Such an economy is often referred to as an endowment economy, and we detail a
model of this type in Chapter 6. In this case, equilibrium occurs by adjustment
of the date 0 assets prices so that investors demands conform to the inelastic
assets supplies. The change in the assets date 0 prices eectively adjusts the
assets return distributions to those which make the tangency portfolio and the
net demand for the riskfree asset equal to the fixed supplies of these assets.
An alternative way to model asset supplies is to assume that the economys
asset return distributions as fixed but allow the quantities of these assets to be
83
elastically supplied. This type of economy is known as a production economy,

and a model of it is presented in Chapter 13. Such a model assumes that there
are n risky, constant-returns-to-scale technologies. These technologies require
date 0 investments of physical capital and produce end-of-period physical in and covariance matrix of
vestment returns having a distribution with mean R
V at the end of the period. Also, there could be a riskfree technology that
generates a one-period return on physical capital of Rf . In this case of a fixed
return distribution, supplies of the assets adjust to the demands for the tangency portfolio and the riskfree asset determined by the technological return
distribution.
As it turns out, how one models asset supplies does not aect the results
that we now derive regarding the equilibrium relationship between asset returns.
We simply note that the tangency portfolio having weights m must be the
equilibrium portfolio of risky assets supplied in the market. Thus, equation (3.6)
can be interpreted as an equilibrium relationship between the excess expected
return on any asset and the excess expected return on the market portfolio. In
other words, in equilibrium, the tangency portfolio chosen by all investors must
be the market portfolio of all risky assets. Moreover, as mentioned earlier, the
only case for which investors have a long position in the tangency portfolio is
Rf < Rmv . Hence, for asset markets to clear, that is, for the outstanding stocks
of assets to be owned by investors, the situation depicted in Figure 3.1 can be
the only equilibrium ecient frontier.3
The Capital Asset Pricing Models prediction that the market portfolio is
mean - variance ecient is an important solution to the practical problem of
identifying a mean-variance ecient portfolio. As a theory, CAPM justifies the
3 This presumes that the tangency portfolio is composed of long positions in the individual
risky assets, that is, m
i > 0 for i = 1, ..., n. While our derivation has not restricted the sign
of these portfolio weights, since assets must have non-negative supplies, equilibrium market
clearing implies that assets prices or individuals choice of technologies must adjust (eectively
changing R and/or V ) to make the portfolio demands for individual assets to be non-negative.

practice of investing in a broad market portfolio of stocks and bonds.
This
insight has led to the growth of "indexed" mutual funds and exchange traded
funds (ETFs) that hold market-weighted portfolios of stocks and bonds.
Lets now consider some additional implications of CAPM when we consider
realized, rather than expected, asset returns. Note that asset is realized return,
i , can be defined as R
i +
R
i where i is the unexpected component of the assets
m , can be defined
return. Similarly the realized return on the market portfolio, R
m + m , where m is the unexpected part of the market portfolios return.
as R
Substituting these into (3.6) we have
i
R
m m Rf ) + i
= Rf + i (R
(3.7)
m Rf ) + i i m
= Rf + i (R
m Rf ) + e
= Rf + i (R
i
where e
i i i m . Note that
m, e
m , i ) i Cov(R
m , m )
Cov(R
i ) = Cov(R
(3.8)
m, R
i ) i Cov(R
m, R
m)
= Cov(R
m ) i V ar(R
m) = 0
= i V ar(R
which, along with (3.7), implies that the total variance of risky asset i, 2i , has
two components
2i = 2i 2m + 2i
(3.9)
where 2i 2m is proportional to the return variance of the market portfolios and

i , and it is orthogonal to the market portfolios return.
2i is the variance of e
85
Since (3.8) shows that e

i is the part of the return on risky asset i that is un-
correlated with the return on the market portfolio, this implies that equation
(3.7) represents a regression equation.
In other words, an unbiased estimate
of i can be obtained by running an Ordinary Least Squares regression of asset is excess return on the market portfolios excess return. The orthogonal,
mean zero residual, e
i , is sometimes referred to as idiosyncratic, unsystematic,
or diversifiable risk. This is the particular assets risk that is eliminated or di-
versified away when the asset is held in the market portfolio. Since this portion
of the assets risk can be eliminated by the individual who invests optimally,
there is no price or risk premium attached to it in the sense that the assets
equilibrium expected return is not altered by it.
To make clear what risk is priced, let us denote the covariance between
the return on the ith asset and the return on the market portfolio as Mi =
ei ), which is the ith element of M . Also let im be the correlation
m, R
cov(R
between the return on the ith asset and the return on the market portfolio.
Then equation (3.6) can be re-written as
i Rf
R
m Rf )
Mi (R
m
m
m Rf )
(R
= im i
m
(3.10)
= im i Se
where Se
m Rf )
(R
m
is the equilibrium excess return on the market portfolio
per unit of market risk and known as the market Sharpe ratio, named after
William Sharpe, one of the developers of the CAPM. Se can be interpreted as
the market price of systematic or non-diversifiable risk. It is also referred to as
the slope of the capital market line, where the capital market line is defined as

the ecient frontier which connects the points Rf and m in Figure 3.1. Now
th
if we define m
i as the weight of asset i in the market portfolio and Vi as the i
row of covariance matrix V , then
N
1 2m
1 mV m
1
1 X m
m
m
=
=
=
2V
=
ij
i
m
2 m m
2 m m
2m
m j=1 j
i
i
i
m =
where ij is the i, j th element of V . Since R
n
P
(3.11)

m
j Rj , then Cov(Ri , Rm ) =
j=1
n
n
i , P mR
j ) = P m
Cov(R
j
j ij . Hence, (3.11) can be re-written
j=1
j=1
1
m
i, R
m ) = im i
=
Cov(R
m
i
m
(3.12)
Thus, im i can be interpreted as the marginal increase in market risk,

m , from a marginal increase of asset i in the market portfolio. In this sense,
im i is the quantity of asset is systematic or non-diversifiable risk. Equation
(3.10) shows that this quantity of systematic risk, multiplied by the price of
systematic risk, Se , determines the assets required excess expected return or
risk premium.
If a riskless asset does not exist so that all assets are risky, Fischer Black
(Black 1972) showed that a similar asset pricing relationship exists. Here, we
outline his zero-beta CAPM. Note that an implication of the portfolio separation
result of section 2.3.3 is that since every frontier portfolio can be written as
p , then a linear combination of these frontier portfolios is also a
= a + bR
frontier portfolio.
Let Wi be the proportion of the economys total wealth
owned by investor i, and let i be this investors desired frontier portfolio so

ip . If there are a total of I investors, then the weights of the
that i = a + bR
market portfolio are given by
I
X
Wi i =
i=1
= a
I
X
i=1
I
X
i=1
Wi + b
ip
Wi a + bR
I
X
87
(3.13)
ip = a + bR
m
Wi R
i=1
m PI Wi R
ip and where the last equality of (3.13) uses the fact that
where R
i=1
the sum of the proportions of total wealth must equal 1. Equation (3.13) shows
that the market portfolio, the aggregation of all individual investors portfolios,
is a frontier portfolio.
Its expected return, Rm , is a weighted average of the
expected returns of the individual investors portfolios. Because each individual

investor optimally chooses a portfolio on the ecient portion of the frontier (the
upper arc in Figure 2.4), then the market portfolio, being a weighted average,
is also on the ecient frontier.
Now, let us compute the covariance between the market portfolio and any
arbitrary portfolio of risky assets, not necessarily a frontier portfolio. Let this
e0p , and an
arbitrary risky asset portfolio have weights 0 , a random return of R
expected return of R0p . Then
em , R
e0p
m 0 V 0
Cov R
= m0 V 0 = a + bR
(3.14)
1
0
V 1 R
V 1 e
V e V 1 R
m V 0
+
R
=
2
2
0 1
0
0
1
0
V V
e V V R
=
2

mR
0 V 1 V 0 R
m e0 V 1 V 0
R
+
2

m
R0p + Rm R0p R
=
2

Rearranging 3.14 gives
R0p =
2
m
R
em , R
e0p
+
Cov
R
m
m
R
R
(3.15)
Re-writing the first term on the right-hand side of equation (3.15) and multiplying and dividing the second term by the definition of a frontier portfolios
variance given in equation (2.32), equation (3.15) becomes
R0p
e0p
em , R
Cov R
1
2 !
Rm
2
+
m
2m
2
R
!
e0p
em , R
Cov R
2
2
2
Rm + 2
2m
Rm
Rm
(3.16)
2
+
2
Rm
From equation (2.39), we recognize that the first two terms on the right-hand
side of (3.16) equal the expected return on the portfolio that has zero-covariance
with the market portfolio, call it Rzm . Thus, equation (3.16) can be written as
R0p
e0p
em , R
Cov R
= Rzm +
Rm Rzm
2m
= Rzm + 0 Rm Rzm
(3.17)
Since the portfolio having weights 0 can be any risky-asset portfolio, it includes
a portfolio that invests solely in a single asset.4 In this light, 0 becomes the
covariance of the individual assets return with that of the market portfolio,
and the relationship equation (3.17) is identical to the previous CAPM result in
equation (3.10) except that Rzm replaces Rf . Hence, when a riskless asset does
not exist, we measure an assets excess returns relative to Rzm , the expected
return on a portfolio that has a zero beta.
4 One
of the elements of 0 would equal 1 while the rest would be zero.
3.2. ARBITRAGE
89
Because the CAPM relationship in equations (3.10) or (3.17) implies that

assets expected returns dier only due to dierences in their betas, it is considered a single "factor" model, this risk factor being the return on the market
portfolio. Stephen Ross (Ross 1976) derived a similar multi-factor relationship,
but starting from a dierent set of assumptions and using a derivation based
on the arbitrage principle. Frequently during this book, we will see that asset
pricing implications can often be derived based on investor risk-preferences, as
was done in the CAPM when we assumed investors cared only about the mean
and variance of their portfolios return. However, another powerful technique
for asset pricing is to rule out the existence of arbitrage. We now turn to this
topic, first by discussing the nature of arbitrage.
3.2
Arbitrage
The notion of arbitrage is simple. It involves the possibility of getting something

for nothing while having no possibility of loss. Specifically, consider constructing
a portfolio involving both long and short positions in assets such that no initial
wealth is required to form the portfolio.5 If this zero-net-investment portfolio
can sometimes produce a positive return, but can never produce a negative
return, then it represents an arbitrage: starting from zero wealth, a profit can
sometimes be made but a loss can never occur. A special case of arbitrage
is when this zero-net-investment portfolio produces a riskless return. If this
certain return is positive (negative), an arbitrage is to buy (sell) the portfolio
and reap a riskless profit or free lunch. Only if the return is zero would there
be no arbitrage.
An arbitrage opportunity can also be defined in a slightly dierent context.
5 Proceeds from short sales (or borrowing) are used to purchase (take long positions in)
other assets.

If a portfolio that requires a non-zero initial net investment is created such that
it earns a certain rate of return, then this rate of return must equal the current
(competitive market) risk free interest rate. Otherwise, there would also be an
arbitrage opportunity. For example, if the portfolio required a positive initial
investment but earned less than the risk free rate, an arbitrage would be to
(short) sell the portfolio and invest the proceeds at the risk free rate, thereby
earning a riskless profit equal to the dierence between the risk free rate and
the portfolios certain (lower) rate of return.6
In ecient, competitive, asset markets where arbitrage trades are feasible, it
is reasonable to think that arbitrage opportunities are rare and fleeting. Should
arbitrage temporarily exist, then trading by investors to earn this profit will tend
to move asset prices in a direction that eliminates the arbitrage opportunity. For
example, if a zero-net-investment portfolio produces a riskless positive return,
as investors create (buy) this portfolio, the prices of the assets in the portfolio
will be bid up. The cost of creating the portfolio will then exceed zero. The
portfolios cost will rise until it equals the present value of the portfolios riskless
return, thereby eliminating the arbitrage opportunity. Hence, for competitive
asset markets where it is also feasible to execute arbitrage trades, it may be
reasonable to assume that equilibrium asset prices reflect an absence of arbitrage
opportunities. As will be shown, this assumption leads to a law of one price: if
dierent assets produce exactly the same future payos, then the current prices
of these assets must be the same. This simple result has powerful asset pricing
implications.
6 Arbitrage defined in this context is really equivalent to the previous definition of arbitrage.
For example, if a portfolio requiring a positive initial investment produces a certain rate of
return in excess of the riskless rate, then an investor should be able to borrow the initial funds
needed to create this portfolio and pay an interest rate on this loan that equals the risk-free
interest rate. That the investor should be able to borrow at the riskless interest rate can be
seen from the fact that the portfolio produces a return that is always sucient to repay the
loan in full, making the borrowing risk-free. Hence, combining this initial borrowing with
the non-zero portfolio investment results in an arbitrage opportunity that requires zero initial
wealth.
3.2. ARBITRAGE
91
However, as a word of caution, not all asset markets meet the conditions
required to justify arbitrage pricing. For some markets, It may be impossible
to execute pure arbitrage trades due significant transactions costs and/or restrictions on short-selling or borrowing. In such cases of limited arbitrage, the
law of one price can fail.7 Alternative methods, such as those based on a model
of investor preferences, are required to price assets.
3.2.1
Examples of Arbitrage Pricing
An early use of the arbitrage principle is the covered interest parity condition
which links spot and forward foreign exchange markets to foreign and domestic
money markets. To illustrate, let F0 be the current date 0 forward price for
exchanging one unit of a foreign currency periods in the future. This forward
price represents the dollar price to be paid periods in the future for delivery
of one unit of foreign currency periods in the future. In contrast, let S0 be
the spot price of foreign exchange, that is, the current date 0 dollar price of
one unit of foreign currency to be delivered immediately. Also let Rf be the
per-period risk free (money market) return for borrowing or lending in dollars
over the period 0 to , and denote as Rf the per-period risk free return for
borrowing or lending in the foreign currency over the period 0 to .8
Now construct the following portfolio that requires zero net wealth. First,
we sell forward (take a short forward position in) one unit of foreign exchange
at price F0 .9 This contract means that we are committed to delivering one
unit of foreign exchange at date in return for receiving F0 dollars at date .
Second, let us also purchase the present value of one unit of foreign currency,
7 Andrei Shleifer and Robert Vishny (Shleifer and Vishny 1997) discuss why the conditions
needed to apply arbitrage pricing are not present in many asset markets.
8 For example, if the foreign currency is the Japanese yen, R would be the per period
f
return for a yen-denominated risk-free investment or loan.
9 Taking a long or short position in a forward contract requires zero initial wealth, as
payment and delivery all occur at the future date .

1/Rf , and invest it in a foreign bond yielding the per-period return, Rf . In
terms of the domestic currency, this purchase costs S0 /Rf , which we finance
by borrowing dollars at the per-period return Rf .
What happens at date as a result of these trades? When date arrives,
we know that our foreign currency investment yields Rf /Rf = 1 unit of the
foreign currency. This is exactly what we need to satisfy our short position in
the forward foreign exchange contract. For delivering this foreign currency, we
receive F0 dollars. But we also now owe a sum of Rf S0 /R
f due to our dollar
borrowing. Thus, our net proceeds at date are
F0 Rf S0 /Rf
(3.18)
Note that these proceeds are non-random, that is, the amount is known at date
0 since it depends only on prices and riskless rates quoted at date 0. If this
amount is positive, then we should indeed create this portfolio as it represents
an arbitrage. If, instead, this amount is negative, then an arbitrage would be
for us to sell this portfolio, that is, we reverse each trade discussed above (take
a long forward position, invest in the domestic currency financed by borrowing
in foreign currency markets). Thus, the only instance in which arbitrage would
not occur is if the net proceeds are zero, which implies
F0 = S0 Rf /R
f
(3.19)
Equation 3.19 is referred to as the covered interest parity condition.

The forward exchange rate, F0 , represents the dollar price for buying or
selling a foreign currency at date , a future date when the foreign currencys
dollar value is unknown.
Though F0 is the price of a risky cashflow, it has
been determined without knowledge of the utility functions of investors or their
3.2. ARBITRAGE
93
expectations regarding the future value of the foreign currency. The reason
for this simplification is due to the law of one price, which states that in the
absence of arbitrage, equivalent assets (or contracts) must have the same price.
A forward contract to purchase a unit of foreign currency can be replicated by
buying, at the spot exchange rate S0 , a foreign currency investment paying the
per-period risk-free return Rf and financing this by borrowing at the dollar riskfree return Rf . In the absence of arbitrage, these two methods for obtaining
foreign currency in the future must be valued the same. Given the spot exchange
rate, S0 , and the foreign and domestic money market returns, Rf and Rf ,
the forward rate is pinned down. Thus, when applicable, pricing assets or
contracts by ruling out arbitrage is attractive in that assumptions regarding
investor preferences or beliefs are not required.
To motivate how arbitrage pricing might apply to a very simple version of the
CAPM, suppose that there is a risk free asset that returns Rf and multiple risky
assets. However, assume that only a single source of (market) risk determines
all risky asset returns and that these returns can be expressed by the linear
relationship
ei = ai + bi f
R
(3.20)
ei is the return on the ith asset and f is the single risk factor generating
where R
= 0. ai is asset is expected
all asset returns, where it is assumed that E[f]
ei ] = ai . bi is the sensitivity of asset i to the risk factor and

return, that is, E[R
can be viewed as asset is beta coecient. Note that this is a highly simplified
example in that all risky assets are perfectly correlated with each other. Assets
have no idiosyncratic risk (residual component e
i ). A generalized model with
idiosyncratic risk will be presented in the next section.
Now suppose that a portfolio of two assets is constructed, where a proportion

of wealth of is invested in asset i and the remaining proportion of (1 ) is
invested in asset j. This portfolios return is given by
ep
R
= ai + (1 )aj + bi f + (1 )bj f
(3.21)
= (ai aj ) + aj + [(bi bj ) + bj ] f
If the portfolio weights are chosen such that
bj
bj bi
(3.22)
then the uncertain (random) component of the portfolios return is eliminated.

The absence of arbitrage then requires that Rp = Rf , so that
Rp = (ai aj ) + aj = Rf
(3.23)
or
bj (ai aj )
+ aj = Rf
bj bi
which implies
ai Rf
aj Rf
=
bi
bj
(3.24)
This condition states that the expected return in excess of the risk free rate,
per unit of risk, must be equal for all assets, and we define this ratio as .
is the risk premium per unit of the factor risk. The denominator, bi , can be
interpreted as asset is quantity of risk from the single risk factor, while ai Rf
can be thought of as asset is compensation or premium in terms of excess
expected return given to investors for holding asset i. Thus, this no-arbitrage
3.3. LINEAR FACTOR MODELS
95
condition is really a law of one price in that the price of risk, , which is the
risk premium divided by the quantity of risk, must be the same for all assets.
Equation (3.24) is a fundamental relationship, and similar law of one price
conditions hold for virtually all asset pricing models.
For example, we can
re-write the CAPM equation (3.10) as

m Rf )
i Rf
(R
R
=
Se
im i
m
(3.25)
i Rf , to its quantity
so that the ratio of an assets expected return premium, R
of market risk, im i , is the same for all assets and equals the slope of the
capital market line, Se . We next turn to a generalization of the CAPM that
derives from arbitrage pricing.
3.3
Linear Factor Models
The CAPM assumption that all assets can be held by all individual investors is
clearly an over-simplification. Transactions costs and other trading "frictions"
that arise from distortions such as capital controls and taxes might prevent individuals from holding a global portfolio of marketable assets. Furthermore, many
assets simply are non-marketable and cannot be traded.10 The preeminent example of a non-marketable asset is the value of an individuals future labor
income, what economists refer to as the individuals human capital.
There-
fore, in addition to the risk from returns on a global portfolio of marketable

assets, individuals are likely to face multiple sources of non-diversifiable risks.
It is then not hard to imagine that, in equilibrium, assets risk premia derive
10 Richard Roll (Roll 1977) has argued that CAPM is not a reasonable theory because a true
"market" portfolio consisting of all risky assets cannot be observed or owned by investors.
Moreover, empirical tests of CAPM are infeasible because proxies for the market portfolio
(such as the S&P500 stock index) may not be mean-variance ecient, even if the true market
portfolio is. Conversely, a proxy for the market portfolio could be mean-variance ecient
even though the true market portfolio is not.

from more than a single risk factor. Indeed, the CAPMs prediction that risk
from a market portfolio is the only source of priced risk has not received strong
empirical support.11
This is a motivation for the multi-factor Arbitrage Pricing Theory (APT)
model. APT assumes that an individual assets return is driven by multiple risk
factors and by an idiosyncratic component, though the theory is mute regarding
the sources of these multiple risk factors. APT is a relative pricing model in
the sense that it determines the risk premia on all assets relative to the risk
premium for each of the factors and each assets sensitivity to each factor.12
It does not make assumptions regarding investor preferences but uses arbitrage
pricing to restrict an assets risk premium. The main assumptions of the model
are that the returns on all assets are linearly related to a finite number of risk
factors, and that the number of assets in the economy is large relative to the
number of factors. Let us now detail the models assumptions.
Assume that there are k risk factors and n assets in the economy, where
n > k. Let biz be the sensitivity of the ith asset to the z th risk factor, where
i be the idiosyncratic
fz is the random realization of risk factor z. Also let e
risk component specific to asset i, which by definition is independent of the k
j .
risk factors, f1 ,...,fk , and the specific risk component of any other asset j, e
e
i must be independent of the risk factors or else it would aect all assets, thus
not being truly a specific source of risk to just asset i. If ai is the expected
return on asset i, then the return generating process for asset i is given by the
linear factor model
11 Ravi Jagannathan and Ellen McGrattan (Jagannathan and McGrattan 1995) review the
empirical evidence for CAPM.
12 This is not much dierent from the CAPM. CAPM determined each assets risk premium
based on the single factor market risk premium, Rm Rf , and the assets sensitivity to this
single factor, i . The only dierence is that CAPM provides somewhat more guidance as to
the identity of the risk factor, namely, the return on a market portfolio of all assets.
ei = ai +
R
97
k
X
z=1
biz fz + e
i
(3.26)
h
i
h i
i fz = 0, and E [e
ie
j ] = 0 for i 6= j. For simplicity,
where E [e
i ] = E fz = E e
i
h
we also assume that E fz fx = 0 for z 6= x, that is, the risk factors are
mutually independent. In addition, let us further assume that the risk factors
h i
are normalized to have a variance equal to one, so that E fz2 = 1. As it turns
out, these last two assumptions are not important, as a linear transformation
of correlated risk factors can allow them to be redefined as independent, unit
variance risk factors.13
A final assumption is that the idiosyncratic risk (variance) for each asset be
finite, that is
h i
E e
2i s2i < S 2
(3.27)
ei , fz =
where S 2 is some finite number. Under these assumptions, note that Cov R
Cov biz fz , fz = biz Cov fz , fz = biz . Thus, biz is the covariance between
the return on asset i and factor z.
In the simple example of the previous section, assets had no idiosyncratic

risk, and their expected returns could be determined by ruling out a simple
arbitrage. This was because a hedge portfolio, consisting of appropriate combinations of dierent assets, could be created that had a riskless return. Now,
however, when each assets return contains an idiosyncratic risk component, it is
not possible to create a hedge portfolio having a purely riskless return. Instead,
we will argue that if the number of assets is large, a portfolio can be constructed
that has "close" to a riskless return, because the idiosyncratic components of
13 For example, suppose g
h is a kx1 vector of mean-zero, correlated risk factors with kxk
covariance matrix E [h
gh
g 0 ] = . Then create a transformed kx1 vector of risk factors
k given
l
g . The covariance matrix of these transformed risk factors is E fhfh0 =

by fh = 1 h
1 E [h
gh
g 0 ] 1 = Ik where Ik is a kxk identify matrix.

assets returns are diversifiable. While ruling out pure arbitrage opportunities
is not sucient to constrain assets expected returns, we can use the notion of
asymptotic arbitrage to argue that assets expected returns will be "close" to
the relationship that would result if they had no idiosyncratic risk. So let us
now state what we mean by an asymptotic arbitrage opportunity.14
Definition: Let a portfolio containing n assets be described by the vector of
investment amounts in each of the n assets, W n [W1n W2n ... Wnn ]0 . Thus, Win
is the amount invested in asset i when there are n total assets in the economy.
Consider a sequence of these portfolios where n is increasing, n = 2, 3, ... . Let
ij be the covariance between the returns on assets i and j. Then an asymptotic
arbitrage exists if the following conditions hold:
(A) The portfolio requires zero net investment:
n
X
Win = 0
i=1
(B) The portfolio return becomes certain as n gets large:
lim
n X
n
X
Win Wjn ij = 0
i=1 j=1
(C) The portfolios expected return is always bounded above zero

n
X
i=1
Win ai > 0
We can now state the Arbitrage Pricing Theorem (APT):

Theorem: If no asymptotic arbitrage opportunities exist, then the expected
return of asset i, i = 1, ..., n, is described by the following linear relation
14 This proof of Arbitrage Pricing Theory based on the concept of asymptotic arbitrage is
due to Gur Huberman (Huberman 1982).
ai = 0 +
99
k
X
biz z + i
()
z=1
where 0 is a constant, z is the risk premium for risk factor fez , z = 1, ..., k,
and the expected return deviations, i , satisfy
n
X
i = 0
(i)
biz i = 0, z = 1, ..., k
(ii)
i=1
n
X
i=1
1X 2
i = 0
n n
i=1
lim
(iii)
Note that condition (iii) says that the average squared error (deviation) from
the pricing rule () goes to zero as n becomes large. Thus, as the number of
assets increase relative to the risk factors, expected returns will, on average,
P
become closely approximated by the relation ai = 0 + kz=1 biz z . Also note
that if the economy contains a risk free asset (implying biz = 0, z), the risk
free return will be approximated by 0 .
Proof : For a given number of assets, n > k, think of running a cross-sectional
regression of the ai s on the biz s. More precisely, project the dependent variable
vector a = [a1 a2 ... an ]0 on the k explanatory variable vectors bz = [b1z b2z ... bnz ]
, z = 1, ..., k. Define i as the regression residual for observation i, i = 1, ..., n.
Denote 0 as the regression intercept and z , z = 1, ..., k, as the estimated
coecient on explanatory variable z. The regression estimates and residuals
must then satisfy
ai = 0 +
k
X
z=1
biz z + i
(3.28)
100CHAPTER 3. CAPM, ARBITRAGE, AND LINEAR FACTOR MODELS

where by the properties of an orthogonal projection (Ordinary Least Squares
P
regression) the residuals sum to zero, ni=1 i = 0, and are orthogonal to the
P
regressors, ni=1 biz i = 0, z = 1, ..., k. Thus, we have shown that (), (i), and
(ii), can be satisfied. The last, but most important part of the proof, is to show
that (iii) must hold in the absence of asymptotic arbitrage.
Thus, let us construct a zero-net investment arbitrage portfolio with the
following investment amounts
i
Wi = pPn
(3.29)
2
i=1 i n
so that greater amounts are invested in assets having the greatest relative expected return deviation. The total arbitrage portfolio return is given by
p
R
n
X
i=1
Since
ei
Wi R
1
pPn
2
i=1 i n
Pn
i=1 biz i
(3.30)
"
n
X
i=1
ei = pP
iR
n
2
i=1 i n
" n
X
i ai +
i=1
k
X
z=1
biz fz + e
i
!#
= 0, z = 1, ..., k, this equals
p = pP
R
n
2
i=1 i n
" n
X
i=1
i (ai + e
i )
(3.31)
Let us calculate this portfolios mean and variance. Taking expectations, we

obtain
h i
p = pP 1
E R
n
2
i=1 i n
since E[e
i ] = 0. Substituting in for ai = 0 +
" n
X
i=1
Pk
i ai
z=1 biz z
(3.32)
+ i , we have
h i
p = pP 1
E R
n
2
i=1 i n
and since
Pn
i=1 i
= 0 and
h
p
E R
"
n
X
i +
i=1
Pn
101
i=1 i biz
k
X
z=1
n
X
i biz
i=1
n
X
2i
i=1
(3.33)
= 0, this simplifies to
v
u n
u1 X
2
t
= pPn
=
2
i
2
n i=1 i
i=1 i n i=1
n
X
(3.34)
To calculate the portfolios variance, start by subtracting (3.32) from (3.31)

h i
p E R
p = pP 1
R
n
2
i=1 i n
" n
X
i=1
ie
i
(3.35)
j ] = 0 for i 6= j and E[e

2i ] = s2i , the portfolio variance is
Then because E[e
ie
Pn
h i2 Pn 2 s2
2 2
S2
i i
i=1
i=1 i S
P
E Rp E Rp
<
=
= Pn
n
n
n i=1 2i
n i=1 2i
(3.36)
Thus, as n becomes large (n ), the variance of the portfolio goes to zero,

that is, the expected return on the portfolio becomes certain. This implies that
in the limit the actual return equals the expected return in (3.34)
v
n
h i u
u X
p = E R
p = t 1
lim R
2
n
n i=1 i
(3.37)
and so if there are no asymptotic arbitrage opportunities, this certain return on

the portfolio must equal zero. This is equivalent to requiring
n
1X 2
i = 0
n n
i=1
lim
(3.38)
which is condition (iii).

We see that APT, given by the relation ai = 0 +
Pk
z=1 biz z ,
can be inter-

preted as a multi-beta generalization of CAPM. However, whereas CAPM says
that its single beta should be the sensitivity of an assets return to that of the
market portfolio, APT gives no guidance as to what are the economys multiple underlying risk-factors. An empirical application of APT by Nai-Fu Chen,
Richard Roll, and Stephen Ross (1986) (Chen, Roll, and Ross 1986) assumed
that the risk factors were macroeconomic in nature, as proxied by industrial
production, expected and unexpected inflation, the spread between long- and
short-maturity interest rates, and the spread between high- and low-credit quality bonds. Other researchers, have tended to select risk factors based on those
that provide the best fit to historical asset returns.15
The well-known Eu-
gene Fama and Kenneth French (Fama and French 1993) model is an example of
this. Its risk factors are returns on three dierent portfolios: a market portfolio
of stocks (like CAPM), a portfolio that is long the stocks of small firms and
short the stocks of large firms, and a portfolio that is long the stocks having
high book-to-market ratios (value stocks) and short stocks having low book-tomarket ratios (growth stocks). The latter two portfolios capture the empirical
finding that the stocks of smaller firms and those of value firms tend to have
higher expected returns than would be predicted solely by the one-factor CAPM
model. The Fama-French model predicts that a given stocks expected return is
determined by its three betas for these three portfolios.16 It has been criticized
for lacking a theoretical foundation for its risk factors.17
However, there have been some attempts to provide a rationale for the Fama15 Gregory
Connor and Robert Korajczk (Connor and Korajczyk 1995) survey empirical
tests of the APT.
16 A popular extension of the Fama-French three-factor model is the four-factor model proposed by Mark Carhart (Carhart 1997). His model adds a proxy for stock momentum.
17 Moreover, some researchers argue that what the model interprets as risk factors may
be evidence of market ineciency. For example, the relatively low returns on growth stocks
relative to value stocks may represent market mis-pricing due to investor over-reaction to high
growth firms. Josef Lakonishok, Andrei Shleifer, and Robert Vishny (Lakonishok, Shleifer,
and Vishny 1994) find that various measures of risk cannot explain the higher average returns
of value stocks relative to growth stocks.
3.4. SUMMARY
103
French models good fit of asset returns. Heaton and Lucas (Heaton and Lucas
2000) provide a rationale for the additional Fama-French risk factors. They note
that many stockholders may dislike the risks of small firm and value stocks, the
latter often being stocks of firms in financial distress, thereby requiring higher
average returns on these stocks. They provide empirical evidence that many
stockholders are, themselves, entrepreneurs and owners of small businesses, so
that their human capital is already subject to the risks of small firms with
relatively high probabilities of failure.
Hence, these entrepreneurs wish to
avoid further exposure to these types of risks.

We will later develop another multi-beta asset pricing model, namely Robert
Mertons Intertemporal Capital Asset Pricing Model (ICAPM) (Merton 1973a),
which is derived from an intertemporal consumer-investor optimization problem.
It is a truly dynamic model that allows for changes in state variables that could
influence investment opportunities.
While the ICAPM is sometimes used to
justify the APT, the static (single-period) APT framework may not be compatible with some of the predictions of the more dynamic (multi-period) ICAPM.
In general, the ICAPM allows for changing risk-free rates and predicts that
assets expected returns should be a function of such changing investment opportunities. The model also predicts that an assets multiple betas are unlikely
to remain constant through time, which can complicate deriving estimates of
betas from historical data.18
3.4
Summary
In this chapter, we took a first step in understanding the equilibrium determinants of individual assets prices and returns. The Capital Asset Pricing Model
18 Ravi Jagannathan and Zhenyu Wang (Jagannathan and Wang 1996) find that the
CAPM better explains stock returns when stocks betas are permitted to change over
time and a proxy for the return on human capital is included in the market portfolio.

(CAPM) was shown to be a natural extension of Markowitzs mean - variance
portfolio analysis. However, in addition to deriving CAPM from investor meanvariance risk-preferences, we showed that CAPM, and its multi-factor generalization Arbitrage Pricing Theory (APT), could result from assumptions of a
linear model of asset returns and an absence of arbitrage opportunities.
Arbitrage pricing will arise frequently in subsequent chapters, especially in
the context of valuing derivative securities. Furthermore, future chapters will
build on our single-period CAPM and APT results to show how equilibrium
asset pricing is modified when multiple periods and time-varying asset return
distributions are considered.
3.5
Exercises
1. Assume that individual investor k chooses between n risky assets in order

to maximize the following utility function
max Rk
{k
i}
1
Vk
k
where the mean and variance of investor ks portfolio are Rk =
n
P
ki Ri
i=1
and Vk =
n
n P
P
ki kj ij , respectively, and where Ri is the expected re-
i=1j=1
turn on risky asset i and ij is the covariance between the returns on risky
asset i and risky asset j. ki is investor ks portfolio weight invested in
n
P
ki = 1. k is a positive constant and equals
risky asset i, so that
i=1
investor ks risk tolerance.
(a) Write down the Lagrangian for this problem and show the first order
conditions.
(b) Re-write the first order condition to show that the expected return
3.5. EXERCISES
105
on asset i is a linear function of the covariance between risky asset

is return and the return on investor ks optimal portfolio.
(c) Assume that investor k has initial wealth equal to Wk and that there
are k = 1, ..., M total investors, each with dierent initial wealth and
risk tolerance. Show that the equilibrium expected return on asset i
is of a similar form to the first order condition found in part 1.b), but
depends on the wealth-weighted risk tolerances of investors and the
covariance of the return on asset i with the market portfolio. (Hint:
begin by multiplying the first order condition in 1.b) by investor ks
wealth times risk tolerance, and then aggregate over all investors.
2. Let the U.S. dollar ($) - Swiss franc (SF) spot exchange rate be $0.68 per
SF and the one year forward exchange rate be $0.70 per SF. The one-year
interest rate for borrowing or lending dollars is 6.00 percent.
(a) What must be the one-year interest rate for borrowing or lending
Swiss francs in order for there to be no arbitrage opportunity?
(b) If the one-year interest rate for borrowing or lending Swiss francs was
less than your answer in (a), describe the arbitrage opportunity.
3. Suppose that the Arbitrage Pricing Theory holds with k = 2 risk factors,
so that asset returns are given by
ei = ai + bi1 f1 + bi2 f2 + e
i
R
where ai
= f 0 + bi1 f 1 + bi2 f 2 . Maintain all of the assumptions made in
the notes and, in addition, assume that both f 1 and f 2 are positive. Thus,
the positive risk premia imply that both of the two orthogonal risk factors are
priced sources of risk. Now define two new risk factors from the original risk

factors:
ge1 = c1 fe1 + c2 fe2
ge2 = c3 fe1 + c4 fe2
Show that there exists a c1 , c2 , c3 , and c4 such that ge1 is orthogonal to e

g2 , they
each have unit variance, and g1 > 0, but that g2 = 0, where g1 and g2 are
the risk premia associated with e

g1 and ge2 , respectively. In other words, show
that any economy with two priced sources of risk can also be described by an
economy with one priced source of risk.
Chapter 4
Consumption-Savings
Decisions and State Pricing
Previous chapters studied the portfolio choice problem of an individual who
maximizes the expected utility of his end-of-period wealth. This specification
of an individuals decision making problem may be less than satisfactory since,
traditionally, economists have presumed that individuals derive utility from consuming goods and services, not by possessing wealth per se. Taking this view,
our prior analysis can be interpreted as implicitly assuming that the individual
consumes only at the end of the single investment period, and all end-of-period
wealth is consumed.
Utility from the individual consuming some of her ini-
tial beginning-of-period wealth was not modeled, so that all initial wealth was
assumed to be saved and invested in a portfolio of assets.
In this chapter we consider the more general problem where an individual
obtains utility from consuming at both the initial and terminal dates of her
decision period and where non-traded labor income also may be received. This
allows us to model the individuals initial consumption-savings decision as well
107
108
CHAPTER 4. CONSUMPTION-SAVINGS AND STATE PRICING
as her portfolio choice decision. In doing so, we can derive relationships between
asset prices and the individuals optimal levels of consumption that extend many
of our previous results.
We introduce the concept of a stochastic discount
factor that can be used to value the returns on any asset.
This stochastic
discount factor equals each individuals marginal rate of substitution between

initial and end-of-period consumption for each state of nature, that is, each
random outcome.
After deriving this stochastic discount factor, we demonstrate that its volatility restricts the feasible excess expected returns and volatilities of all assets.
Importantly, we discuss empirical evidence that appears inconsistent this restriction for standard, time-separable utility functions, casting doubt on the
usefulness of a utility-of-consumption-based stochastic discount factor. Fortunately, however, a stochastic discount factor for pricing assets need not rely on
this consumption-based foundation. We provide an alternative derivation of a
stochastic discount factor based on the assumptions of an absence of arbitrage
and market completeness.
Markets are said to be complete when there are
a sucient number of non-redundant assets whose returns span all states of

nature.
The chapter concludes by showing how the stochastic discount factor approach can be modified to derive an asset valuation relationship based on riskneutral probabilities.
These probabilities transform the true probabilities of
each state of nature to incorporate adjustments for risk premia.
Valuation
based on risk-neutral probabilities is used extensively to price assets, and this

technique will be employed frequently in future chapters.
4.1. CONSUMPTION AND PORTFOLIO CHOICES
4.1
109
Consumption and Portfolio Choices
In this section, we introduce an initial consumption-savings decision into an

investors portfolio choice problem. This is done by permitting the individual
to derive utility from consuming at the beginning, as well as at the end, of the
investment period. The assumptions of our model are as follows.
Let W0 and C0 be the individuals initial date 0 wealth and consumption,
respectively. At date 1, the end of the period, the individual is assumed to consume all of his wealth which we denote as C1 . The individuals utility function is
defined over beginning- and end-of-period consumption and takes the following
form
h i
e1
U (C0 ) + E U C
(4.1)
where is a subjective discount factor that reflects the individuals rate of time
preference and E [] is the expectations operator conditional on information
at date 0.1
The multi-date specification of utility in (4.1) is an example of
a time-separable utility function.
Time separability means that utility at a
particular date (say 0 or 1) depends only on consumption at that same date.

Later chapters will analyze the implications of time-separability and consider
generalized multi-period utility functions that permit utility to depend on past
or expected future consumption.
Suppose that the individual can choose to invest in n dierent assets. Let
Pi be the date 0 price per share of asset i, i = 1, ..., n , and let Xi be the date
1 random payo of asset i. For example, a dividend-paying stock might have a
1
is sometimes written as 1+
where is the rate of time preference. A value of < 1 ( >
0) reflects impatience on the part of the individual, that is, a preference for consuming early.
A more general two-date utility function could be expressed as U0 (Co ) + E [U1 (C1 )] where
U0 and U1 are any dierent increasing, concave functions of consumption. Our presentation
assumes U1 (C) = U0 (C), but the qualitative results we derive also hold for the more general
specification.
1
110
ei = Pe1i + D
e 1i , where f
date 1 random payo of X
P 1i is the date 1 stock price and
e 1i is the stocks dividend paid at date 1. Alternatively, for a coupon-paying

D
e 1i would be the bonds coupon

bond, f
P 1i would be the date 1 bond price and D
paid at date 1.2
Given this definition, then we can also define Ri Xi /Pi
to be the random return on asset i.
The individual may also receive labor
income of y0 at date 0 and random labor income of y1 at date 1.3 If i is the

proportion of date 0 savings that the individual chooses to invest in asset i, then
his intertemporal budget constraint is
C1 = y1 + (W0 + y0 C0 )
n
X
i Ri
(4.2)
i=1
where (W0 + y0 C0 ) is the individuals date 0 savings. The individuals maximization problem can then be stated as
max U (C0 ) + E [U (C1 )]
(4.3)
C0 ,{i }
subject to equation (4.2) and the constraint
Pn
i=1 i
= 1. The first order
conditions are with respect to C0 and the i , i = 1, ..., n are

"
#
n
X
i Ri = 0
U (C0 ) E U (C1 )
(4.4)
E [U 0 (C1 ) Ri ] = 0,
(4.5)
i=1
i = 1, ..., n
where 0 / (W0 + y0 C0 ) and 0 is the Lagrange multiplier for the conPn

straint i=1 i = 1. The first order conditions in (4.5) describe how the investor
2 The coupon payment would be uncertain if default on the payment is possible and/or the
coupon is not fixed but floating (tied to a market interest rate).
3 There is an essential dierence between tangible wealth, W , and wage income, y. The
present value of wage income, which is referred to as "human capital," is assumed to be a
non-tradeable asset. The individual can re-balance his tangible wealth to change his holdings
of marketable assets, but his endowment of human capital (and its cashflows in the form of
wage income) is assumed to be fixed.
111
chooses between dierent assets. Substitute out for and one obtains:
E [U 0 (C1 ) Ri ] = E [U 0 (C1 ) Rj ]
(4.6)
for any two assets, i and j. Equation (4.6) tells us that the investor trades o
investing in asset i for asset j until their expected marginal utility-weighted
returns are equal. If this were not the case, the individual could raise his total
expected utility by investing more in assets whose marginal utility weighted
returns were relatively high and investing less in assets whose marginal utility
weighted returns were low.
How does the investor act to make the optimal equality of expected marginal
utility weighted returns in (4.6) come about? Note from (4.2) that C1 becomes
more positively correlated with Ri the greater is i . Thus, the greater is asset is
portfolio weight, the lower will be U 0 (C1 ) when Ri is high due to the concavity of
utility. Hence, as i becomes large, smaller marginal utility weights multiply
the high realizations of asset is return, and E [U 0 (C1 ) Ri ] falls.
Intuitively,
this occurs because the investor becomes more undiversified by holding a larger
proportion of asset i. By adjusting the portfolio weights for asset i and each
of the other n 1 assets, the investor changes the random distribution of C1 in
a way that equalizes E [U 0 (C1 ) Rk ] for all assets k = 1, ..., n, thereby attaining
the desired level of diversification.
Another result of the first order conditions involves the intertemporal allocation of resources. Substituting (4.5) into (4.4) gives
"
#
n
n
X
X
i Ri =
i E [U 0 (C1 ) Ri ]
U (C0 ) = E U (C1 )
0
n
X
=
i =
i=1
i=1
i=1
(4.7)
112
Therefore, substituting = U 0 (C0 ), the first order conditions in (4.5) can be

written as:
E [U 0 (C1 ) Ri ] = U 0 (C0 ) ,
i = 1, ..., n
(4.8)
or since Ri = Xi /Pi
Pi U 0 (C0 ) = E [U 0 (C1 ) Xi ] ,
i = 1, ..., n
(4.9)
Equation (4.9) has an intuitive meaning and, as will be shown in subsequent

chapters, generalizes to multi-period consumption and portfolio choice problems.
It says that when the investor is acting optimally, he invests in asset i until the
loss in marginal utility of giving up Pi dollars at date 0 just equals the expected
marginal utility of receiving the random payo of Xi at date 1. To see this more
clearly, suppose that one of the assets pays a risk-free return over the period.
Call it asset f so that Rf is the risk-free return (one plus the risk-free interest
rate). For the risk-free asset, equation (4.9) can be re-written
U 0 (C0 ) = Rf E [U 0 (C1 )]
(4.10)
which states that the investor trades o date 0 for date 1 consumption until the
marginal utility of giving up $1 of date 0 consumption just equals the expected
marginal utility of receiving $Rf of date 1 consumption. For example, suppose
that utility is of a constant relative risk aversion form: U (C) = C /, for < 1.
Then equation (4.10) can be re-written as
1
= E
Rf
"
C0
C1
1 #
(4.11)
Hence, when the interest rate is high, so will be the expected growth in consump-
113
tion. For the special case of there being only one risk-free asset and non-random
labor income, so that C1 is non-stochastic, equation (4.11) becomes
Rf =
C1
C0
(4.12)
Taking logs of both sides of the equation, we obtain
ln (Rf ) = ln + (1 ) ln
C1
C0
(4.13)
Since ln(Rf ) is the continuously-compounded risk-free interest rate and ln(C1 /C0 )
is the growth rate of consumption, then we can define the elasticity of intertemporal substitution, , as
1
ln (C1 /C0 )
=
ln (Rf )
1
Hence, with power (constant relative risk aversion) utility,
(4.14)
is the reciprocal
of the coecient of relative risk aversion. That is, the single parameter determines both risk aversion and the rate of intertemporal substitution.4 When
0 < < 1,
exceeds unity and a higher interest rate raises second period
consumption more than one-for-one.
This implies that if utility displays less
risk-aversion than logarithmic utility, this individual increases his savings as the
interest rate rises. Conversely, when < 0, then < 1 and a rise in the interest
rate raises second period consumption less than one-for-one, implying that such
an individual decreases her initial savings when the return to savings is higher.
For the logarithmic utility individual ( = 0 and, therefore
= 1), a change in
the interest rate has no eect on savings. These results can be interpreted as
4 An end-of-chapter exercise shows that this result extends to an environment with risky
assets. In Chapter 14, we will examine a recursive utility generalization of multi-period
power utility for which the elasticity of intertemporal substitution is permitted to dier from
the inverse of the coecient of relative risk aversion. There these two characteristics of
multi-period utility are modeled by separate parameters.
114
an individuals response to two eects from an increase in interest rates. The

first is a substitution eect that raises the return from transforming current
consumption into future consumption. This higher benefit from initial savings
provides an incentive to do more of it. The second eect is an income eect due
to the greater return that is earned on a given amount of savings. This makes
the individual better o and, ceteris paribus, would raise consumption in both
periods. Hence, initial savings could fall and still lead to greater consumption
in the second period. For
> 1, the substitution eect outweighs the income
eect, while the reverse occurs when
< 1.
When
= 1, the income and
substitution eects exactly oset each other.

A main insight of this section is that an individuals optimal portfolio of
assets is one where the assets expected marginal utility-weighted returns are
equalized.
If this were not the case, the individuals expected utility could
be raised by investing more (less) in assets whose average marginal utilityweighted returns are relatively high (low). It was also demonstrated that an
individuals optimal consumption-savings decision involves trading-o higher
current marginal utility of consuming for higher expected future marginal utility
obtainable from invested saving.
4.2
An Asset Pricing Interpretation
Until now, we have analyzed the consumption-portfolio choice problem of an

individual investor. For such an exercise, it makes sense to think of the individual taking the current prices of all assets and the distribution of their payos
as given when deciding on his optimal consumption - portfolio choice plan.
Importantly, however, the first order conditions we have derived might be reinterpreted as asset pricing relationships. They can provide insights regarding
the connection between individuals consumption behavior and the distribution
4.2. AN ASSET PRICING INTERPRETATION
115
of asset returns.
To see this, let us begin by re-writing equation (4.9) as:
Pi
U 0 (C1 )
Xi
= E
U 0 (C0 )
= E [m01 Xi ]
(4.15)
where m01 U 0 (C1 ) /U 0 (C0 ) is the marginal rate of substitution between initial and end-of-period consumption. For any individual who can trade freely in
asset i, equation (4.15) provides a condition that equilibrium asset prices must
satisfy. Condition (4.15) appears in the form of an asset pricing formula. The
current asset price, Pi , is an expected discounted value of its payos, where the
discount factor, m01 , is a random quantity because it depends on the random
level of future consumption.
Hence, m01 is also referred to as the stochastic
discount factor for valuing asset returns. In states of nature where future consumption turns out to be high (due to high asset portfolio returns or high labor
income), marginal utility, U 0 (C1 ), is low and the assets payos in these states
are not highly valued. Conversely, in states where future consumption is low,
marginal utility is high so that the assets payos in these states are much desired. This insight explains why m01 is also known as the state price deflator.
It provides a dierent discount factor (deflator) for dierent states of nature.
It should be emphasized that the stochastic discount factor, m01 , is the same
for all assets that a particular investor can hold. It prices these assets payos
only by dierentiating in which state of nature the payo is made. Since m01
provides the core or kernel for pricing all risky assets, it is also referred to as
the pricing kernel. Note that the random realization of m01 may dier across
investors because of dierences in random labor income that can cause the
random distribution of C1 to vary across investors. Nonetheless, the expected
116
product of the pricing kernel and asset is payo, E [m01 Xi ], will be the same
for all investors who can trade in asset i.
4.2.1
Real versus Nominal Returns
In writing down the individuals consumption - portfolio choice problem, we

implicitly assumed that returns are expressed in real or purchasing power terms,
that is, returns should be measured after adjustment for inflation. The reason
is that an individuals utility should depend on the real, not nominal (currency
denominated), value of consumption. Therefore, in the budget constraint (4.2),
if C1 denotes real consumption, then asset returns and prices (as well as labor
income) need to be real values. Thus if PiN and XiN are the initial price and
end-of-period payo measured in currency units (nominal terms), we need to
deflate them by a price index to convert them to real quantities. Letting CP It
denote the consumer price index at date t, the pricing relationship in (4.15)
becomes
0
U (C1 ) XiN
PiN
=E
CP I0
U 0 (C0 ) CP I1
(4.16)
or if we define Its = CP Is /CP It as one plus the inflation rate between dates t
and s, equation (4.16) can be re-written as
PiN
1 U 0 (C1 ) N
X
= E
I01 U 0 (C0 ) i
= E M01 XiN
(4.17)
where M01 (/I01 ) U 0 (C1 ) /U 0 (C0 ) is the stochastic discount factor (pricing
kernel) for discounting nominal returns. Hence, this nominal pricing kernel is
simply the real pricing kernel, m01 , discounted at the (random) rate of inflation
117
between dates 0 and 1.
4.2.2
Risk Premia and the Marginal Utility of Consumption
The relation in (4.15) can be re-written to shed light on an assets risk premium.
Dividing each side of equation (4.15) by Pi results in
= E [m01 Ri ]
(4.18)
= E [m01 ] E [Ri ] + Cov [m01 , Ri ]
Cov [m01 , Ri ]
= E [m01 ] E [Ri ] +
E [m01 ]
Recall from (4.10) that for the case of a risk-free asset, E [U 0 (C1 ) /U 0 (C0 )] =
E [m01 ] = 1/Rf . Then (4.18) can be re-written as
Rf = E [Ri ] +
Cov [m01 , Ri ]
E [m01 ]
(4.19)
or
Cov [m01 , Ri ]
E [m01 ]
Cov [U 0 (C1 ) , Ri ]
= Rf
E [U 0 (C1 )]
E [Ri ] = Rf
(4.20)
Equation (4.20) states that the risk premium for asset i equals minus the
covariance between the marginal utility of end-of-period consumption and the
asset return divided by the expected end-of-period marginal utility of consumption. If an asset pays a higher return when consumption is high, its return has
a negative covariance with the marginal utility of consumption, and therefore
118
the investor demands a positive risk premium over the risk free rate.
Conversely, if an asset pays a higher return when consumption is low, so
that its return positively covaries with the marginal utility of consumption,
then it has an expected return less than the risk-free rate. Investors will be
satisfied with this lower return because the asset is providing a hedge against
low consumption states of the world, that is, it is helping to smooth consumption
across states.
4.2.3
The Relationship to CAPM
em that is perfectly
Now suppose there exists a portfolio with a random return of R

e1 ,
negatively correlated with the marginal utility of date 1 consumption, U 0 C
implying that it is also perfectly negatively correlated with the pricing kernel,
m01 :
Then this implies
em ,
U 0 (C1 ) = R
> 0.
(4.21)
Cov[U 0 (C1 ), Rm ] = Cov[Rm , Rm ] = V ar[Rm ]
(4.22)
Cov[U 0 (C1 ), Ri ] = Cov[Rm , Ri ].
(4.23)
and
em , the risk premium relation (4.20) is

For the portfolio having return R
E[Rm ] = Rf
Cov[U 0 (C1 ), Rm ]
V ar[Rm ]
= Rf +
0
E[U (C1 )]
E[U 0 (C1 )]
(4.24)
Using (4.20) and (4.24) to substitute for E[U 0 (C1 )], and using (4.23), we obtain
V ar[Rm ]
E[Rm ] Rf
=
E[Ri ] Rf
Cov[Rm , Ri ]
(4.25)
119
and re-arranging
E[Ri ] Rf =
Cov[Rm , Ri ]
(E[Rm ] Rf )
V ar[Rm ]
(4.26)
or
E[Ri ] = Rf + i (E[Rm ] Rf ) .
(4.27)
So we obtain the CAPM if the return on the market portfolio is perfectly negatively correlated with the marginal utility of end-of-period consumption, that
is, perfectly negatively correlated with the pricing kernel.
Note that for an
arbitrary distribution of asset returns and non-random labor income, this will
always be the case if utility is quadratic because marginal utility is linear in consumption, and consumption also depends linearly on the markets return. In
addition, for the case of general utility, normally distributed asset returns, and
non-random labor income, marginal utility of end-of-period consumption is also
perfectly negatively correlated with the return on the market portfolio because
each investors optimal portfolio is simply a combination of the market portfolio
and the (non-random) risk-free asset. Thus, consistent with Chapters 2 and 3,
under the assumptions needed for mean-variance analysis to be equivalent with
expected utility maximization, asset returns satisfy the CAPM.
4.2.4
Bounds on Risk Premia
Another implication of the stochastic discount factor is that it places bounds

on the means and standard deviations of individual securities and, therefore,
determines an ecient frontier. To show this, re-write the first line in (4.20) as
E [Ri ] = Rf m01 ,Ri
m01 Ri
E [m01 ]
(4.28)
120
where m01 , Ri , and m01 ,Ri are the standard deviation of the discount factor,
the standard deviation of the return on asset i, and the correlation between the
discount factor and the return on asset i, respectively. Re-arranging (4.28) leads
to
m01
E [Ri ] Rf
= m01 ,Ri
Ri
E [m01 ]
(4.29)
The left hand side of (4.29) is the Sharpe ratio for asset i. Since 1 m01 ,Ri
1, we know that
E [Ri ] Rf
m01 = m01 Rf
E [m01 ]
Ri
(4.30)
This equation was derived by Robert Shiller (Shiller 1982), was generalized by
Lars Hansen and Ravi Jagannathan (Hansen and Jagannathan 1991), and is
known as a Hansen-Jagannathan bound.
Given an assets Sharpe ratio and
the risk-free rate, equation (4.30) sets a lower bound on the volatility of the
economys stochastic discount factor.
Conversely, given the volatility of the
discount factor, equation (4.30) sets an upper bound on the maximum Sharpe
ratio that any asset, or portfolio of assets, can attain.
If there exists an asset (or portfolio of assets) whose return is perfectly
negatively correlated with the discount factor, m01 , then the bound in (4.30)
holds with equality. As we just showed in equations (4.21) to (4.27), such
a situation implies the CAPM, so that the slope of the capital market line,
Se
E[Rm ]Rf
Rm
, equals m01 Rf . Thus, the slope of the capital market line,
which represents (ecient) portfolios that have a maximum Sharpe ratio, can
be related to the standard deviation of the discount factor.
The inequality in (4.30) has empirical implications. m01 can be estimated
if we could observe an individuals consumption stream and if we knew his or
121
her utility function. Then, according to (4.30), the Sharpe ratio of any portfolio
of traded assets should be less than or equal to m01 /E [m01 ]. For power utility,
U (C) = C /, < 1, so that m01 (C1 /C0 )1 = e(1) ln(C1 /C0 ) .
If
C1 /C0 is assumed to be lognormally distributed, with parameters c and c

then
m01
E [m01 ]
=
=
=
V ar e(1) ln(C1 /C0 )
E e(1) ln(C1 /C0 )

q
2
E e2(1) ln(C1 /C0 ) E e(1) ln(C1 /C0 )
E e(1) ln(C1 /C0 )

q
2
E e2(1) ln(C1 /C0 ) /E e(1) ln(C1 /C0 ) 1
q
p
2
2
2
e2(1)c +2(1) 2c /e2(1)c +(1) 2c 1 = e(1) 2c 1
(1 ) c
(4.31)
where in the fourth line of (4.31) the expectations are evaluated assuming C1
is lognormally distributed.5 Hence, with power utility and lognormally distributed consumption, we have
E [Ri ] Rf
(1 ) c
Ri
(4.32)
Suppose, for example, that Ri is the return on a broadly diversified portfolio

of U.S. stocks, such as the S&P500.
Over the last 75 years, this portfolios
annual real return in excess of the risk-free (U.S. Treasury bill) interest rate
has averaged 8.3 percent, suggesting E [Ri ] Rf = .083. The portfolios annual standard deviation has been approximately Ri = 0.17, implying a Sharpe
ratio of
E[Ri ]Rf
Ri
= 0.49. Assuming a representative agent and using per
5 The fifth line of (4.31) is based on taking a two-term approximation of the series ex =
2
3
1 + x + x2! + x3! + ..., which is reasonable when x is a small positive number.
122
capita U.S. consumption data to estimate the standard deviation of consumption growth, researchers have come up with annualized estimates of c between
0.01 and 0.0386.6 Thus, even if a diversified portfolio of U.S. stocks was an ecient portfolio of risky assets, so that (4.32) held with equality, it would imply a
E[Ri ]Rf
/c between -11.7 and -48.7 Since reasonable levels
value of = 1
R
i
of risk aversion estimated from other sources imply values of much smaller
in magnitude, say in the range of -1 to -5, the inequality (4.32) appears to not
hold for U.S. stock market data and standard specifications of utility.8 In other
words, consumption appears to be too smooth ( c is too low) relative to the

premium that investors demand for holding stocks. This inconsistency between
theory and empirical evidence was identified by Rajnish Mehra and Edward
Prescott (Mehra and Prescott 1985) and is referred to as the equity premium
puzzle. Attempts to explain this puzzle have involved using dierent specifications of utility and questioning whether the ex-post sample mean of U.S. stock
returns is a good estimate of the a priori expected return on U.S. stocks.9
Even if one were to accept a high degree of risk-aversion in order to fit the
historical equity premium, additional problems may arise because this high riskaversion could imply an unreasonable value for the risk-free return, Rf . Under
our maintained assumptions and using (4.10), the risk-free return satisfies
6 See John Y. Campbell (1999) (Campbell 1999) and Stephen G. Cecchetti, Pok-Sam Lam,
and Nelson C. Mark (1994)(Cecchetti, Lam, and Mark 1994).
7 If the stock portfolio was less than ecient, so that a strict inequality held in (4.32), the
magnitude of the risk-aversion coecient would need to be even higher.
8 Rajnish Mehra and Edward Prescott (Mehra and Prescott 1985) survey empirical work
finding values of of -1 or more (equivalent to coecients of relative risk aversion, 1 , of
2 or less).
9 Jeremy J. Siegel and Richard H. Thaler (Siegel and Thaler 1997) review this literature. It
should be noted that recent survey evidence from academic financial economists (Welch 2000)
finds that a consensus believes that the current equity risk-premium is significantly lower
than the historical average. Moreover, at the begining of 2006, the Federal Reserve Bank
of Philadelphias Survey of Professional Forecasters found that the median predicted annual
returns over the next decade on the S&P500 stock portfolio, the 10-year U.S. Treasury bond,
and 3-month U.S. Treasury bill are 7.00 %, 5.00 %, and 4.25 %, respectively. This implies
a much lower equity risk premium (7.00 % - 4.25 % = 2.75 %) compared to the historical
average dierence between stocks and bills of 8.3%.
1
Rf
123
= E [m01 ]
i
h
= E e(1) ln(C1 /C0 )
2
= e(1)c + 2 (1)
(4.33)
2c
and therefore
ln (Rf ) = ln () + (1 ) c
1
(1 )2 2c
2
(4.34)
If we set = 0.99, reflecting a 1% rate of time preference, and c = 0.018,

which is the historical average real growth of U.S. per capita consumption, then
a value of = 11 and c = 0.036 implies
ln (Rf ) = ln () + (1 ) c
1
(1 )2 2c
2
= 0.01 + 0.216 0.093 = 0.133
(4.35)
which is a real risk-free interest rate of 13.3%. Since short-term real interest
rates have averaged about 1% in the U.S., we end up with a risk-free rate puzzle.
The notion that assets can be priced using a stochastic discount factor, m01 ,
is attractive because the discount factor is independent of the asset being priced:
it can be used to price any asset no matter what its risk. We derived this discount factor from a consumption - portfolio choice problem and, in this context,
showed that it equaled the marginal rate of substitution between current and
end-of-period consumption. However, the usefulness of this approach is in doubt
since empirical evidence using aggregate consumption data and standard specifications of utility appears inconsistent with the discount factor equaling the
124
marginal rate of substitution.10
Fortunately, a general pricing relationship of
the form Pi = E0 [m01 Xi ] can be shown to hold without assuming that m01
represents a marginal rate of substitution.
Rather, it can be derived using
alternative assumptions. This is the subject of the next section.
4.3
Market Completeness, Arbitrage, and State

Pricing
We need not assume a consumption- portfolio choice structure to derive a stochastic discount factor pricing formula. Instead, our derivation can be based
on the assumptions of a complete market and the absence of arbitrage, an
approach pioneered by Kenneth Arrow and Gerard Debreu.11
With these al-
ternative assumptions, one can show that a law of one price holds and that a
unique stochastic discount factor exists. This new approach makes transparent
the derivation of relative pricing relationships and is an important technique for
valuing contingent claims (derivatives).
4.3.1
Complete Markets Assumptions
To illustrate, suppose once again that an individual can freely trade in n dierent
assets. Also, let us assume that there are a finite number of end-of-period states
of the nature, with state s having probability s .12 Let Xsi be the cashflow
generated by one share (unit) of asset i in state s. Also assume that there are
k states of nature and n assets. The following vector describes the payos to
10 As will be shown in Chapter 14, some specifications of time-inseparable utility can improve
the consumption-based stochastic discount factors ability to explain asset prices.
11 See Kenneth Arrow (Arrow 1953) reprinted in (Arrow 1964) and Gerard Debreu (Debreu
1959).
12 As is discussed later, this analysis can be extended to the case of an infinite number of
states.
4.3. MARKET COMPLETENESS, ARBITRAGE, AND STATE PRICING125

financial asset i:
X1i
.
.
Xi =
.
Xki
(4.36)
Thus, the per-share cashflows of the universe of all assets can be represented by
the k n matrix
X =
X11
..
.
..
.
X1n
..
.
Xk1
Xkn
(4.37)
We will assume that n = k and that X is of full rank. This implies that the
n assets span the k states of nature, an assumption that indicates a complete
market. We would still have a complete market (and, as we will show, unique
state-contingent prices) if n > k, as long as the payo matrix X has rank k. If
the number of assets exceeds the number of states, some assets are redundant,
that is, their cashflows in the k states are linear combinations of others. In such
a situation, we could reduce the number of assets to k by combining them into
k linearly independent (portfolios of) assets.
An implication of the assumption that the assets returns span the k states
of nature is that an individual can purchase amounts of the k assets so that
she can obtain target levels of end-of-period wealth in each of the states. To
show this complete markets result, let W denote an arbitrary kx1 vector of
end-of-period levels of wealth.
W1
.
.
W =
. .
Wk
(4.38)
where Ws is the level of wealth in state s. To obtain W , at the initial date
126
the individual needs to purchase shares in the k assets.
Let the vector N =
[N1 . . . Nk ]0 to be the number of shares purchased of each of the k assets. Hence,

N must satisfy
XN = W
(4.39)
Because X is a non-singular matrix of rank k, its inverse exists so that
N = X 1 W.
(4.40)
Hence, because the assets payos span the k states, arbitrary levels of wealth in
the k states can be attained if initial wealth is sucient to purchase the required
shares, N. Denoting P = [P1 . . . Pk ]0 as the kx1 vector of beginning-of-period
per share prices of the k assets, then the amount of initial wealth required to
produce the target level of wealth given in (4.38) is simply P 0 N .
4.3.2
Arbitrage and State Prices
Given our assumption of complete markets, the absence of arbitrage opportunities implies that the price of a new, redundant security or contingent claim can
be valued based on the prices of the original k securities. For example, suppose
a new asset pays a vector of end-of-period cashflows of W . In the absence of
arbitrage, its price must be P 0 N . If its price exceeded P 0 N , an arbitrage would
be to sell this new asset and purchase the original k securities in amounts N .
Since the end-of-period liability from selling the security is exactly oset by the
returns received from the k original securities, the arbitrage profit equals the
dierence between the new assets price and P 0 N . Conversely, if the new assets
price was less than P 0 N , an arbitrage would be to purchase the new asset and
sell the portfolio N of the k original securities.
Lets apply this concept of complete markets, no-arbitrage pricing to the

special case of a security that has a payo of 1 in state s and 0 in all other
states. Such a security is referred to as a primitive, elementary, or Arrow-Debreu
security. Specifically, elementary security s has the vector of cashflows
W1
..
.

es = Ws =

.
..

Wk
0
..
.
1
..
.
0
(4.41)
Let ps be the beginning of period price of elementary security s, that is, the
price of receiving 1 in state s. Then as we just showed, its price in terms of the
payos and prices of the original k assets must equal
ps = P 0 X 1 es , s = 1, ..., k
(4.42)
so that a unique set of state prices exists in a complete market.13 Furthermore,

we would expect that these elementary state prices should each be positive,
since a unit amount of wealth received in any state will have a value greater
than zero whenever individuals are assumed to be non-satiated.14
Hence the
equations in (4.42) along with the conditions ps > 0 s restrict the payos, X,
and the prices, P , of the original k securities.
We can now derive a stochastic discount factor formula by considering the
value of any other security or contingent claim in terms of these elementary state
security prices. Note that the portfolio composed of the sum of all elementary
13 If markets were incomplete, for example, if n was the rank of X and k > n, then state
prices would not be uniquely determined by the absence of arbitrage. The no-arbitrage
conditions would place only n linear restrictions on the set of k prices, implying that there
could be an infinity of possible state prices.
14 This would be the case whenever individuals marginal utilities are positive for all levels
of end-of-period consumption.
128
securities gives a cashflow of 1 unit with certainty. The price of this portfolio
defines the risk free return, Rf , by the relation:
k
X
ps =
s=1
1
.
Rf
(4.43)
In general, let there be some multi-cashflow asset, a, whose cashflow paid in

state s is Xsa . In the absence of arbitrage, its price, Pa , must equal
Pa =
k
X
ps Xsa
(4.44)
s=1
Note that the relative pricing relationships that we have derived did not require
using information on the state probabilities.
However, let us now introduce
these probabilities to see their relationship to state prices and the stochastic
discount factor.
Define ms ps /s to be the price of elementary security
s divided by the probability that state s occurs. Note that if, as was argued
earlier, a sensible equilibrium requires ps > 0 s, then ms > 0 s when there is a
positive probability of each state occurring. Using this new definition, equation
(4.44) can be written as
Pa
k
X
s=1
k
X
ps
Xsa
s
(4.45)
s ms Xsa
s=1
= E [m Xa ]
where m denotes a stochastic discount factor whose expected value is
Pk
s=1
s ms ,
and Xa is the random cashflow of the multi-cashflow asset a. Equation (4.45)

shows that the stochastic discount factor equals the prices of the elementary se-

curities normalized by their state probabilities. Hence, we have shown that in
a complete market that lacks arbitrage opportunities, a unique, positive-valued
stochastic discount factor exists. When markets are incomplete, the absence of
arbitrage, alone, cannot determine the stochastic discount factor. One would
need to impose additional conditions, such as the previous sections assumptions on the form of individuals utility, in order to determine the stochastic
discount factor.
For example, if dierent states of nature led to dierent re-
alizations of an individuals non-traded labor income, and there did not exist
assets that could span or insure against this wage income, then a unique stochastic discount factor may not exist. In this case of market incompleteness, a
utility-based derivation of stochastic discount factor may be required for asset
pricing.
While the stochastic discount factor relationship (4.45) is based on state
prices derived from assumptions of market completeness and the absence of arbitrage, it is interesting to interpret these state prices in terms of the previouslyderived consumption-based discount factor.
Note that since ps = s ms , the
price of the elementary security paying 1 in state s is higher the greater is

the likelihood of the state s occurring and the greater is the stochastic discount factor for state s.
In terms of the consumption-based model, ms =
U 0 (C1s ) /U 0 (C0 ) where C1s is the level of consumption at date 1 in state s.

Hence, the state s price, ps , is greater when C1s is low, that is, state s is a low
consumption state, such as an economic recession.
4.3.3
Risk-Neutral Probabilities
The state pricing relationship of equation (4.44) can be used to develop an

important alternative formula for pricing assets. Define
bs ps Rf as the price
of elementary security s times the risk-free return. Then
130
Pa
k
X
ps Xsa
(4.46)
s=1
k
1 X
ps Rf Xsa
Rf s=1
k
1 X
bs Xsa
Rf s=1
Now these
bs , s = 1, ..., k, have the characteristics of probabilities because
Pk
Pk
bs =
they are positive,
bs = ps / s=1 ps > 0, and they sum to 1,
s=1
Pk
Rf s=1 ps = Rf /Rf = 1. Using this insight, we can re-write equation (4.46)
as
Pa
k
1 X
bs Xsa
Rf s=1
1 b
E [Xa ]
Rf
(4.47)
b [] denotes the expectation operator evaluated using the "pseudo" probwhere E
abilities
bs rather than the true probabilities s . Since the expectation in (4.47)
b [Xa ] as the certainty

is discounted by the risk-free return, we can recognize E
equivalent expectation of the cashflow Xa . In comparison to the stochastic dis-
count factor approach, the formula works by modifying the probabilities of the
cashflows in each of the dierent states, rather than discounting the cashflows
by a dierent discount factor.
To see this, note that since ms ps /s and
bs can be written as
Rf = 1/E [m],
bs
= Rf ms s
=
ms
s
E [m]
(4.48)

so that the pseudo probability transforms the true probability by multiplying
by the ratio of the stochastic discount factor to its average value.
In states
of the world where the stochastic discount factor is greater than its average
value, the pseudo probability exceeds the true probability.
For example, if
bs exceeds s in states of the world with relatively

ms = U 0 (C1s ) /U 0 (C0 ),
low consumption where marginal utility is high.
As a special case, suppose that in each state of nature, the stochastic discount
factor equaled the risk-free discount factor, that is, ms =
1
Rf
= E [m].
This
circumstance implies that the pseudo probability equals the true probability
and Pa = E [mXa ] = E [Xa ] /Rf . Because the price equals the expected payo
discounted at the risk-free rate, the asset is priced as if investors are risk-neutral.
b []
Hence, this explains why
bs is referred to as the risk-neutral probability and E
is referred to as the risk-neutral expectations operator. In comparison, the true

probabilities, s , are frequently called the physical or statistical probabilities.
If the stochastic discount factor is interpreted as the marginal rate of substitution, then we see that
bs is higher than s in states where the marginal utility
of consumption is high (or the level of consumption is low). Thus, relative to
the physical probabilities, the risk-neutral probabilities place extra probability

weight on bad states and less probability weight on good states.
4.3.4
State Pricing Extensions
The complete markets pricing framework that we have just outlined is also
known as State Preference Theory and can be generalized to an infinite number
of states and elementary securities. Basically, this is done by defining probability
densities of states and replacing the summations in expressions like (4.43) and
(4.44) with integrals. For example, let states be indexed by all possible points
on the real line between 0 and 1, that is, the state s (0, 1). Also let p(s)
132
be the price (density) of a primitive security that pays 1 unit in state s, 0

otherwise. Further, define Xa (s) as the cashflow paid by security a in state s.
Then, analogous to (4.43) we can write
Z
p(s) ds =
1
Rf
(4.49)
and instead of (4.44) we can write the price of security a as
Pa =
p(s) Xa (s) ds.
(4.50)
In some cases, namely where markets are intertemporally complete, State

Preference Theory can be extended to allow assets cashflows to occur at different dates in the future. This generalization is sometimes referred to as Time
State Preference Theory.15 To illustrate, suppose that assets can pay cashflows
at both date 1 and date 2 in the future. Let s1 be a state at date 1 and let s2
be a state at date 2. States at date 2 can depend on which states were reached
at date 1.
For example, suppose there are two events at each date, economic recession
(r) or economic expansion (boom) (b). Then we could define s1 {r1 , b1 } and
s2 {r1 r2 , r1 b2 , b1 r2 , b1 b2 }. By assigning suitable probabilities and primitive
security state prices for assets that pay cashflows of 1 unit in each of these six
states, we can sum (or integrate) over both time and states at a given date to
obtain prices of complex securities. Thus, when primitive security prices exist at
all states for all future dates, essentially we are back to a single period complete
markets framework, and the analysis is the same as that derived previously.
15 See
Steward C. Myers (Myers 1968).
4.4. SUMMARY
4.4
133
Summary
This chapter began by extending an individuals portfolio choice problem to

include an initial consumption - savings decision. With this modification, we
showed that an optimal portfolio is one where assets expected marginal utilityweighted returns are equalized. Also, the individuals optimal level of savings
involves an intertemporal trade-o where the marginal utility of current consumption is equated to the expected marginal utility of future consumption.
The individuals optimal decision rules can be reinterpreted as an asset pricing formula.
This formula values assets returns using a stochastic discount
factor equal to the marginal rate of substitution between present and future
consumption. Importantly, the stochastic discount factor is independent of the
asset being priced and determines the assets risk premium based on the covariance of the assets return with the marginal utility of consumption. Moreover,
this consumption-based stochastic discount factor approach places restrictions
on assets risk premia relative to the volatility of consumption. However, these
restrictions appear to be violated when empirical evidence is interpreted using
standard utility specifications.
This contrary empirical evidence does not automatically invalidate the stochastic discount factor approach to pricing assets.
Rather than deriving dis-
count factors as the marginal rate of substituting present for future consumption, we showed that they can be derived based on the alternative assumptions
of market completeness and an absence of arbitrage.
When assets returns
spanned the economys states of nature, state prices for valuing any derivative
asset could be derived. Finally, we showed how an alternative risk-neutral pricing formula could be derived by transforming the states physical probabilities
to reflect an adjustment for risk. Risk-neutral pricing is an important valuation
tool in many areas of asset pricing, and it will be applied frequently in future
134
chapters.
4.5
Exercises
1. Consider the one-period model of consumption and portfolio choice. Suppose that individuals can invest in a one-period bond that pays a riskless
real return of Rrf and in a one-period bond that pays a riskless nominal
return of Rnf . Derive an expression for Rrf in terms of Rnf , E [I01 ], and
Cov (M01 , I01 ).
2. Assume there is an economy with k states of nature and where the following asset pricing formula holds:
Pa
k
X
s ms Xsa
s=1
= E [mXa ]
Let an individual in this economy have the utility function ln (C0 ) +

E [ ln (C1 )], and let C0 be her equilibrium consumption at date 0 and Cs
be her equilibrium consumption at date 1 in state s, s = 1, ..., k. Denote
the date 0 price of elementary security s as ps , and derive an expression
for it in terms of the individuals equilibrium consumption.
3. Consider the one-period consumption - portfolio choice problem. The
individuals first order conditions leads to the general relationship
1 = E [m01 Rs ]
where m01 is the stochastic discount factor between dates 0 and 1 and Rs
is the one-period stochastic return on any security in which the individual
can invest. Let there be a finite number of date 1 states where s is the
4.5. EXERCISES
135
probability of state s. Also assume markets are complete and consider the
above relationship for primitive security s, that is, let Rs be the rate of
return on primitive (or elementary) security s. The individuals elasticity
of intertemporal substitution is defined as
Rs d (Cs /C0 )
Cs /C0
dRs
where C0 is the individuals consumption at date 0 and Cs is the individuals consumption at date 1 in state s. If the individuals expected utility
is given by
h i
e1
U (C0 ) + E U C
where utility displays constant relative risk aversion, U (C) = C /, solve

for the elasticity of intertemporal substitution, I .
4.
16
Consider an economy with k = 2 states of nature, a "good" state and a
"bad" state. There are two assets, a risk free asset with Rf = 1.05, and a
second risky asset that pays cashflows
10
X2 =
5
The current price of the risky asset is 6.
4.a Solve for the prices of the elementary securities, p1 and p2 and the riskneutral probabilities of the two states.
4.b Suppose that the physical probabilities of the two states are 1 = 2 = 0.5.
What is the stochastic discount factor for the two states?
16 I
thank Michael Cli of Virginia Tech for suggesting this example.
136
5. Consider a one-period economy with two end-of-period states. An option

contract pays 3 in state 1 and 0 in state 2 and has a current price of
1. A forward contract pays 3 in state 1 and -2 in state 2. What is the
one-period risk-free return and the risk-neutral probabilities of the two
states?
6. This question asks you to relate the stochastic discount factor pricing
relationship to the CAPM. The CAPM can be expressed as
E [Ri ] = Rf + i
where E [] is the expectation operator, Ri is the realized return on asset

i, Rf is the risk-free return, i is asset is beta, and is a positive market
risk premium. Now, consider a stochastic discount factor of the form
m = a + bRm
where a and b are constants and Rm is the realized return on the market
portfolio. Also, denote the variance of the return on the market portfolio
as 2m .
6.a Derive an expression for as a function of a, b, E [Rm ], and 2m . (Hint:
you may want to start from the equilibrium expression 0 = E [m (Ri Rf )].)
6.b Note that the equation 1 = E [mRi ] holds for all assets.
Consider the
case of the risk-free asset and the case of the market portfolio, and solve
for a and b as a function of Rf , E [Rm ], and 2m .
6.c Using the formula for a and b in part b, show that = E [Rm ] Rf .
7. Consider a two factor economy with multiple risky assets and a risk-free
asset whose return is denoted Rf . The economys first factor is the return
4.5. EXERCISES
137
on the market portfolio, Rm , and the second factor is the return on a zero
net-investment portfolio, Rz . In other words, one can interpret the second
factor as the return on a portfolio that is long one asset and short another
asset, where the long and short positions are equal in magnitude (e.g.,
Rz = Ra Rb ) where Ra and Rb are the returns on the assets that are
long and short, respectively. It is assumed that Cov (Rm , Rz ) = 0. The
expected returns on all assets in the economy satisfy the APT relationship
E [Ri ] = 0 + im m + iz z
(*)
where Ri is the return on an arbitrary asset i, im = Cov (Ri , Rm ) / 2m ,

iz = Cov (Ri , Rz ) / 2z , and m and z are the risk premiums for factors
1 and 2, respectively.
Now suppose you are given the stochastic discount factor for this economy, m, measured over the same time period as the above asset returns.
It is given by
m = a + bRm + cRz
(**)
where a, b, and c are known constants. Given knowledge of this stochastic

discount factor in (**), show how you can solve for 0 , m , and z in
equation (*) in terms of a, b, c, m , and z . Just write down the conditions
that would allow you to solve for the 0 , m , and z . You need not derive
explicit solutions for the s since the conditions are non-linear and may
be tedious to manipulate.
Part II
Multiperiod Consumption, Portfolio Choice, and
Asset Pricing
Chapter 5
A Multi-Period Discrete
Time Model of
Consumption and Portfolio
Choice
This chapter considers an expected utility maximizing individuals consumption and portfolio choices over many periods. In contrast to our previous single
period or static models, here the intertemporal or dynamic nature of the problem is explicitly analyzed. Solving an individuals multi-period consumption
and portfolio choice problem is of interest in that it provides a theory for an
individuals optimal life-time savings and investment strategies. Hence, it has
normative value as a guide for individual financial planning. In addition, just as
our single-period mean-variance portfolio selection model provided the theory
of asset demands for the Capital Asset Pricing Model, a multi-period portfolio
141
142
CHAPTER 5. A MULTI-PERIOD DISCRETE TIME MODEL
choice model provides a theory of asset demands for a general equilibrium theory of intertemporal capital asset pricing. Combining this model of individuals
preferences over consumption and securities with a model of firm production
technologies can lead to an equilibrium model of the economy that determines
asset price processes.1
In the 1920s, Frank Ramsey (Ramsey 1928) derived optimal multi-period
consumption - savings decisions, but assumed that the individual could invest in
only a single asset paying a certain return. It was not until the late 1960s that
Paul A. Samuelson (Samuelson 1969) and Robert C. Merton (Merton 1969) were
able to solve for an individuals multi-period consumption and portfolio choice
decisions under uncertainty, that is, where both a consumption - savings choice
and a portfolio allocation decision involving risky assets was assumed to occur
each period.2
ming.
Their solution technique involves stochastic dynamic program-
While this dynamic programming technique is not the only approach
to solving problems of this type, it can sometimes be the most convenient and
intuitive way of deriving solutions.3
The model we present allows an individual to make multiple consumption
and portfolio decisions over a single planning horizon. This planning horizon,
which can be interpreted as the individuals remaining lifetime, is composed of
many decision periods, with consumption and portfolio decisions occurring once
each period.
The richness of this problem cannot be captured in the single-
period models that we presented earlier. This is because with only one period,
an investors decision period and planning horizon coincide. Still, the results
1 Important examples of such models were developed by John Cox, Jonathan Ingersoll, and
Stephen Ross (Cox, Ingersoll, and Ross 1985a) and Robert Lucas (Lucas 1978).
2 Jan Mossin (Mossin 1968) solved for an individuals optimal multi-period portfolio decisions but assumed the individual had no interim consumption decisions, only a utility of
terminal consumption.
3 An alternative martingale approach to solving consumption and portfolio choice problems
is given by John C. Cox and Chi-Fu Huang (Cox and Huang 1989). This approach will be
presented in Chapter 12 in the context of a continuous time consumption and portfolio choice
problem.
143
from our single-period analysis will be useful because often we can transform
multi-period models into a series of single-period ones, as will be illustrated
below.
The consumption - portfolio choice model presented in this chapter assumes
that the individuals decision interval is a discrete time period. Later in this
book, we change the assumption to make the interval instantaneous, that is,
the individual may make consumption and portfolio choices continuously. This
latter assumption often simplifies problems and can lead to sharper results.
When we move from discrete time to continuous time, continuous time stochastic
processes are used to model security prices.
The next section outlines the assumptions of the individuals multi-period

consumption - portfolio problem.
Perhaps the strongest assumption that we
make is that utility of consumption is time separable.4
The following section
shows how this problem can be solved. It introduces an important technique

for solving multi-period decision problems under uncertainty, namely, stochastic
dynamic programming. The beauty of this technique is that decisions over a
multi-period horizon can be broken-up into a series of decisions over a single
period horizon. This allows us to derive the individuals optimal consumption
and portfolio choices by starting at the end of the individuals planning horizon
and working backwards toward the present. In the last section, we complete
our analysis by deriving explicit solutions for the individuals consumption and
portfolio holdings when utility is assumed to be logarithmic.
4 Time-inseparable utility, where current utility can depend on past or expected future
consumption, is discussed in Chapter 14.
144
5.1
Assumptions and Notation of the Model
Consider an environment in which an individual chooses his level of consumption

and the proportions of his wealth invested in n risky assets plus a risk-free
asset.
As was the case in our single-period models, it is assumed that the
individual takes the stochastic processes followed by the prices of the dierent
assets as given. The implicit assumption is that security markets are perfectly
competitive in the sense that the (small) individual is a price-taker in security
markets. An individuals trades do not impact the price (or the return) of
the security. For most investors trading in liquid security markets, this is a
reasonably realistic assumption. In addition, it is assumed that there are no
transactions costs or taxes when buying or selling assets, so that security markets
can be described as friction-less.
An individual is assumed to make consumption and portfolio choice decisions
at the start of each period during a T period planning horizon. Each period is
of unit length, with the initial date being 0 and the terminal date being T .5
A.1 Preferences:
The individual is assumed to maximize an expected utility function defined
over consumption levels and a terminal bequest. Denote consumption at date
t as Ct , t = 0, ..., T 1, and the terminal bequest as WT , where Wt indicates
the individuals level of wealth at date t.
A general form for a multi-period
expected utility function would be E0 [ (C0 , C1 , ..., CT 1 , WT )] where we could

simply assume that is increasing and concave in its arguments.
However,
as a starting point, we will assume that has the following time-separable or

additively-separable form:
5 The following presentation borrows liberally from Samuelson (Samuelson 1969) and
Robert C. Mertons unpublished MIT course 15.433 class notes "Portfolio Theory and Capital
Markets."
5.1.
ASSUMPTIONS AND NOTATION OF THE MODEL
E0 [ (C0 , C1 , ..., CT 1 , WT )] = E0
"T 1
X
U (Ct , t) + B (WT , T )
t=0
145
(5.1)
where U and B are assumed to be increasing, concave functions of consumption

and wealth, respectively. Equation (5.1) restricts utility at date t, U (Ct , t), to
depend only on consumption at that date and not previous levels of consumption or expected future levels of consumption.
While this is the traditional
assumption in multi-period models, in later chapters we loosen this restriction

and investigate utility formulations that are not time-separable.6
A.2 The Dynamics of Wealth:
At date t, the value of the individuals tangible wealth held in the form of
assets equals Wt . In addition, the individual is assumed to receive wage income
of yt .7 This beginning-of-period wealth and wage income are divided between
consumption and savings, and then savings is allocated between n risky assets
as well as a risk-free asset. Let Rit be the random return on risky asset i over
the period starting at date t and ending at date t + 1. Also let Rf t be the return
on an asset that pays a risk-free return over the period starting at date t and
ending at date t + 1. Then if the proportion of date t saving allocated to risky
asset i is denoted it , we can write the evolution of the individuals tangible
wealth as
Wt+1
= (Wt + yt Ct ) Rf t +
= St Rt
n
X
i=1
it (Rit Rft )
(5.2)
6 Dynamic programming, the solution technique presented in this chapter, can also be
applied to consumption and portfolio choice problems where an individuals utility is time
inseparable.
7 Wage income can be random. The present value of wage income, referrred to as human
capital, is assumed to be a non-tradeable asset. The individual can re-balance how his financial
wealth is allocated among risky assets, but cannot trade his human capital.
146
S e q u e n c e o f In d iv id u a ls C o n su m p tio n a n d P o rtfo lio C h o ic e s

W0
y0
R
f 0
FR f | I 0
F y | I 0
F R i | I 0
In fo rm a tio n V a ria b le s
Wt
W T 1
y1
y T 1
yt
R f1
R ft
R f ,T 1
FR | I1
| I T 1
FR | I t FR
F y | I 1
F y | I t
FR | I1
FR | I t
W1
i ,T 1
D e c isio n s
C0
C1
{ i 0 }
{ i 1 }
Ct
{ it }
C T 1
{ i , T 1 }
T -1
T
D a te
Figure 5.1: Multi-Period Decisions
where St Wt + yt Ct is the individuals savings at date t and Rt Rf t +

Pn
i=1 it (Rit Rf t ) is the total return on the individuals invested wealth over
the period from date t to t + 1.
Note that we have not restricted the distribution of asset returns in any
way.
In particular, the return distribution of risky asset i could change over
time, so that the distribution of Rit could dier from the distribution of Ri
for t 6= . Moreover, the one-period risk-free return could be changing, so that
Rf t 6= Rf . Asset distributions that vary from one period to the next means
that the individual faces changing investment opportunities. Hence, in a multiperiod model, the individuals current consumption and portfolio decisions may
be influenced not only by the asset return distribution for the current period,
but also by the possibility that asset return distributions could change in the
future.
5.2. SOLVING THE MULTIPERIOD MODEL
147
The information and decision variables available to the individual at each

date are illustrated in 5.1. At date t, the individual knows her wealth at the
start of the period, Wt , her wage income received at date t, yt , and the risk-free
interest rate for investing or borrowing over the period from date t to date t + 1,
Rf t . Conditional on information at date t, denoted by It , she also knows the
distributions of future one-period risk-free rates and wage income, FRf |It and
Fy |It , respectively, for dates = t + 1, ..., T 1. Lastly, the individual also
knows the date t conditional distributions of the risky asset returns for dates
= t, ..., T 1, given by FRi |It . Date t information, It , includes all realizations
of wage income and risk-free rates for all dates up until and including date t.
It also includes all realizations of risky asset returns for all dates up until and
including date t 1.
Moreover, It could include any other state variables
known at date t that aect the distributions of future wages, risk-free rates, and
risky asset returns. Based on this information, the individuals date t decision
variables are consumption, Ct , and the portfolio weights for the n risky assets,
{ it } for i = 1, ..., n.
5.2
Solving the Multiperiod Model
We begin by defining an important concept that will help us simplify the solution
to this multi-period optimization problem. Let J (Wt , t) denote the derived
utility of wealth function. It is defined as follows:
J (Wt , It , t)
max
Et
Cs ,{is },s,i
T 1
P
s=t
U (Cs , s) + B (WT , T )
(5.3)
where max means to choose the decision variables Cs and { is } for s =

t, t + 1, ..., T 1 and i = 1, ..., n so as to maximize the expected value of the
term in brackets. Note that J is a function of current wealth and all information
148
up until and including date t.
This information could reflect state variables
describing a changing distribution of risky asset returns and/or a changing riskfree interest rate, where these state variables are assumed to be exogenous to
the individuals consumption and portfolio choices. However, by definition J is
not a function of the individuals current or future decision variables since they
are assumed to be set to those values that maximize lifetime expected utility.
Hence, J can be described as a derived utility of wealth function.
We will solve the individuals consumption and portfolio choice problem using backward dynamic programming. This entails considering the individuals
multi-period planning problem starting from her final set of decisions, because,
with one period remaining in the individuals planning horizon, the multi-period
problem has become a single period one. We know from Chapter 4 how to solve
for consumption and portfolio choices in a single period context. Once we characterize the last periods solution for some given wealth and distribution of asset
returns faced by the individual at date T 1, we can solve for the individuals
optimal decisions for the preceding period, those decisions made at date T 2.
This procedure is continued until we can solve for the individuals optimal decisions at the current date 0.
As will be clarified below, by following this
recursive solution technique, the individuals current decisions properly account

for her future optimal decisions that she will make in response to the evolution
of uncertainty in asset returns and labor income.
5.2.1
The Final Period Solution
From the definition of J, note that8

8 To keep notation manageable, we suppress making information, I , an explicit argument of
t
the indirect utility function. We use the short-hand notation J (Wt , t) to refer to J (Wt , It , t).
J (WT , T ) = ET [B (WT , T )] = B (WT , T )
149
(5.4)
Now working backwards, consider the individuals optimization problem when,

at date T 1, she has a single period left in her planning horizon.
J (WT 1 , T 1) =
max
ET 1 [U (CT 1 , T 1) + B (WT , T )](5.5)
max
U (CT 1 , T 1) + ET 1 [B (WT , T )]
CT 1 ,{i,T 1 }
CT 1 ,{i,T 1 }
To clarify how WT depends explicitly on CT 1 and { i,T 1 }, substitute equation

(5.2) for t = T 1 into equation (5.5)
J (WT 1 , T 1) =
max
U (CT 1 , T 1) + ET 1 [B (ST 1 RT 1 , T )]
CT 1 ,{i,T 1 }
(5.6)
where it should be recalled that ST 1 WT 1 + yT 1 CT 1 and RT 1

Pn
Rf,T 1 + i=1 i,T 1 (Ri,T 1 Rf,T 1 ). Equation (5.6) is a standard single-
period consumption - portfolio choice problem. To solve it, we dierentiate

with respect to each decision variable, CT 1 and { i,T 1 }, and set the resulting
expressions equal to zero:
UC (CT 1 , T 1) ET 1 [BW (WT , T ) RT 1 ] = 0
(5.7)
ET 1 [BW (WT , T ) (Ri,T 1 Rf,T 1 )] = 0, i = 1, ..., n
(5.8)
150
where the subscripts on U and B denote partial dierentiation.9
Using the
results in (5.8), we see that (5.7) can be re-written as
n
P
UC (CT 1 , T 1) = ET 1 BW (WT , T ) Rf,T 1 +
i,T 1 (Ri,T 1 Rf,T 1 )
i=1
= Rf,T 1 ET 1 [BW (WT , T )]
Conditions (5.8) and (5.9) represent n + 1 equations that determine the opti
mal choices of CT 1 and i,T 1 . They are identical to the single-period
model conditions (4.6) and (4.10) derived in the previous chapter but with the
utility of bequest function, B, replacing the end-of-period utility function, U .

If we substitute these optimal decision variables back into equation (5.6) and
dierentiate totally with respect to WT 1 , we have:
JW
CT 1
dWT
+ ET 1 BWT
WT 1
dWT 1
"
n
X
CT 1
WT
WT i,T 1
+ ET 1 BWT
+
= UC
WT 1
WT 1 i=1 i,T 1 WT 1
WT CT 1
+
CT 1 WT 1
n
"
X
i,T 1
CT 1
= UC
+ ET 1 BWT
[Ri,T 1 Rf,T 1 ] ST 1
WT 1
WT 1
i=1
CT 1
+RT 1 1
(5.10)
WT 1
= UC
Using the first order condition (5.8), ET 1 [BWT (Ri,T 1 Rf,T 1 )] = 0,

as well as (5.9), UC = Rf,T 1 ET 1 [BWT ], we see that (5.10) simplifies to JW =
Rf,T 1 ET 1 [BWT ]. Using (5.9) once again, this can be re-written as
9 Note that we apply the chain rule when dierentiating B (W , T ) with respect to
T
CT 1 since WT =ST 1 RT 1 depends on CT 1 through ST 1 .
(5.9)
JW (WT 1 , T 1) = UC CT 1 , T 1
151
(5.11)
which is known as the envelope condition. It says that the individuals optimal
policy equates her marginal utility of current consumption, UC , to her marginal
utility of wealth (future consumption).
5.2.2
Deriving the Bellman Equation
Having solved the individuals problem with one period to go in her planning
horizon, we next consider her optimal consumption and portfolio choices with
two periods remaining, at date T 2. The individuals objective at this date is
J (WT 2 , T 2) = max U (CT 2 , T 2) + ET 2 [U (CT 1 , T 1)

+B (WT , T )]
(5.12)
The individual must maximize the above expression by choosing CT 2 as well as

{ i,T 2 }. However, note that she wishes to maximize an expression that is an
expectation over utilities U (CT 1 , T 1) + B (WT , T ) that depend on future
decisions, namely, CT 1 and { i,T 1 }. What should the individual assume
these future values of CT 1 and {i,T 1 } to be? The answer comes from the
Principle of Optimality. It states:
An optimal set of decisions has the property that given an initial
decision, the remaining decisions must be optimal with respect to
the outcome that results from the initial decision.
The max in (5.12) is over all remaining decisions, but the principle of optimality says that whatever decision is made in period T 2, given the outcome,
152
the remaining decisions (for period T 1) must be optimal (maximal). In other

words
max
{(T 2),(T 1)}
(Y ) = max
max
{T 2} {T 1,| outcome from (T 2)}
(Y )
(5.13)
This principle allows us to rewrite (5.12) as
J (WT 2 , T 2) =
max
CT 2 ,{i,T 2 }
ET 2
{U (CT 2 , T 2) +
max
(5.14)
ET 1 [U (CT 1 , T 1) + B (WT , T )]
CT 1 ,{i,T 1 }
Then, using the definition of J (WT 1 , T 1) from (5.5), equation (5.14) can
be re-written as
J (WT 2 , T 2) =
max
U (CT 2 , T 2) + ET 2 [J (WT 1 , T 1)]
CT 2 ,{i,T 2 }
(5.15)
The recursive condition (5.15) is known as the (Richard) Bellman equation

(Bellman 1957). It characterizes the individuals objective at date T 2. What
is important about this characterization is that if we compare it to equation
(5.5), the individuals objective at date T 1, the two problems are quite similar.
The only dierence is that in (5.15) we replace the known function of wealth next
period, B, with another (known in principle) function of wealth next period, J.
But the solution to (5.15) will be of the same form as that for (5.5).10
5.2.3
The General Solution
10 Using the envelope condition, it can be shown that the concavity of U and B ensures
that J (W, t) is a concave and continuously dierentiable function of W . Hence, an interior
solution to the second-to-last period problem exists.
153
Thus, the optimality conditions for (5.15) are:
UC CT 2 , T 2 = ET 2 [JW (WT 1 , T 1) RT 2 ]
= Rf,T 2 ET 2 [JW (WT 1 , T 1)]

= JW (WT 2 , T 2)
(5.16)
ET 2 [Ri,T 2 JW (WT 1 , T 1)] = Rf,T 2 ET 2 [JW (WT 1 , T 1)] ,

i = 1, ..., n
(5.17)
Based on the above pattern, inductive reasoning implies that for any t =
0, 1, ..., T 1, we have the Bellman equation
J (Wt , t) = max U (Ct , t) + Et [J (Wt+1 , t + 1)]

Ct ,{i,t }
(5.18)
and, therefore, the date t optimality conditions are
UC (Ct , t) = Et [JW (Wt+1 , t + 1) Rt ]

= Rf,t Et [JW (Wt+1 , t + 1)]
= JW (Wt , t)
Et [Ri,t JW (Wt+1 , t + 1)] = Rf,t Et [JW (Wt+1 , t + 1)] , i = 1, ..., n
(5.19)
(5.20)
154
The insights of the multi-period model conditions (5.19) and (5.20) are similar
to those of a single-period model of Chapter 4.
The individual chooses to-
days consumption such that the marginal utility of current consumption equals
the derived marginal utility of wealth (the marginal utility of future consumption).
Furthermore, the portfolio weights should be adjusted to equate all
assets expected marginal utility weighted asset returns. However, solving for
the individuals actual consumption and portfolio weights at each date, Ct and
{ i,t }, t = 0, ..., T 1, is more complex than for a single-period model. The
conditions dependence on the derived utility of wealth function implies that
they depend on future contingent investment opportunities (the distributions
of future asset returns (Ri,t+j , Rf,t+j , j 1), future income flows, yt+j , and,
possibly, states of the world that might aect future utilities (U (, t + j)).
Solving the above system involves starting from the end of the planning
horizon and dynamically programing backwards toward the present. Thus, for
the last period, T , we know that J (WT , T ) = B (WT , T ). As we did previously,
we substitute B (WT , T ) for J (WT , T ) in conditions (5.18) to (5.20) for date
T 1 and solve for J (WT 1 , T 1). This is then substituted into conditions
(5.18) to (5.20) for date T 2 and one then solves for J (WT 2 , T 2). If
we proceed in this recursive manner, we eventually obtain J (W0 , 0) and the
solution is complete. These steps are summarized in the following table.
Step
Action
Construct J (WT , T )
Solve for CT 1 and { i,T 1 }, i = 1, ..., n
Substitute the decisions in step 2 to construct J (WT 1 , T 1)
Solve for CT 2 and { i,T 2 }, i = 1, ..., n
Substitute the decisions in step 4 to construct J (WT 2 , T 2)
Repeat steps 4 and 5 for date T 3.
Repeat step 6 for all prior dates until date 0 is reached.
5.3. EXAMPLE USING LOG UTILITY
155
By following this recursive procedure, we find that the optimal policy will
be of the form:11
Ct = g [Wt , yt , It , t]
(5.21)
it = h [Wt , yt , It , t]
(5.22)
Deriving analytical expressions for the functions g and h are not always possible, in which case numerical solutions satisfying the first order conditions at
each date can be computed. However, for particular assumptions regarding the
form of utility, wage income, and the distribution of asset returns, such explicit
solutions may be possible. The next section considers an example where this
is the case.
5.3
Example Using Log Utility
To illustrate how solutions of the form (5.21) and (5.22) can be obtained, consider the following example where the individual has log utility and no wage
income. Assume that U (Ct , t) t ln [Ct ], B (WT , T ) T ln [WT ], and yt 0
t, where =
1
1+
and is the individuals subjective rate of time preference.
Now at date T 1, using condition (5.7), we have

11 When asset returns are serially correlated, that is, the date t distribution of asset returns
depends on realized asset returns from periods prior to date t, the decision rules in 5.21 and
5.22 may depend on this prior, conditioning information. They will also depend any other
state variables know at time t and included in the date t information set It . This, however,
does not aect the general solution technique. These prior asset returns are exogenous state
variables that influence only the conditional expectations in the optimality conditions (5.19)
and (5.20).
156
(5.23)
UC (CT 1 , T 1) = ET 1 [BW (WT , T ) RT 1 ]
1
RT 1
RT 1
T 1
= ET 1 T
= ET 1 T
CT 1
WT
ST 1 RT 1
=
T
T
=
ST 1
WT 1 CT 1
or
CT 1 =
1
WT 1
1+
(5.24)
It is noteworthy that consumption for this log utility investor is a fixed

proportion of wealth and is independent of investment opportunities, that is,
independent of the distribution of asset returns. This is reminiscent of the result
derived in Chapters 1 and 4: the income and substitution eects from a change
in investment returns exactly oset each other for the log utility individual.
Turning to the first order conditions with respect to the portfolio weights,
conditions (5.8) imply
ET 1 [BWT Ri,T 1 ] = Rf,T 1 ET 1 [BWT ] , i = 1, ..., n
Ri,T 1
1
T ET 1
= T Rf,T 1 ET 1
ST 1 RT 1
ST 1 RT 1
Ri,T 1
1
ET 1
= Rf,T 1 ET 1
RT 1
RT 1
(5.25)
Furthermore, for the case of log utility we see that equation (5.25) equals unity,
since from (5.9) we have
UC (CT 1 , T 1) = Rf,T 1 ET 1 [BW (WT , T )]
T 1
1
T
= Rf,T 1 ET 1
CT 1
ST 1 RT 1
CT 1 Rf,T 1
1
1 =
E
T
1
WT 1 CT 1
RT 1
1
1 = Rf,T 1 ET 1
RT 1
157
(5.26)
where we have substituted equation (5.24) in going from the third to the fourth
line of (5.26). While we would need to make specific assumptions regarding the
distribution of asset returns in order to derive the portfolio weights { i,T 1 }
satisfying (5.25), note that the conditions in (5.25) are rather special in that
they do not depend on WT 1 , CT 1 , or , but only on the particular distribution
of asset returns that one assumes. The implication is that a log utility investor
chooses assets in the same relative proportions, independent of his initial wealth.
This, of course, is a consequence of log utility being a special case of constant
relative risk-aversion utility.12
The next step is to solve for J (WT 1 , T 1) by substituting in the date

T 1 optimal consumption and portfolio rules into the individuals objective
Pn
function. Denoting Rt Rf,t + i=1 it (Rit Rf t ) as the individuals total
portfolio return when assets are held in the optimal proportions, we have
12 Recall from section 1.3 that a one-period investor with constant relative risk aversion
places constant proportions of wealth in a risk-free and a single risky asset.
158
J (WT 1 , T 1) = T 1 ln CT 1 + T ET 1 ln RT 1 WT 1 CT 1
= T 1 ( ln [1 + ] + ln [WT 1 ]) +
ET 1 ln RT 1 + ln
+ ln [WT 1 ]
1+
= T 1 [(1 + ) ln [WT 1 ] + HT 1 ]
where HT 1 ln [1 + ] + ln
1+
+ ET 1 ln RT 1 .
(5.27)
Notably, from
equation (5.25) we saw that i,T 1 did not depend on WT 1 , and, therefore,
RT 1 and HT 1 do not depend on WT 1 .
Next, lets move back one more period and consider the individuals optimal
consumption and portfolio decisions at time T 2. From equation (5.15) we
have
J (WT 2 , T 2) =
=
max
U (CT 2 , T 2) + ET 2 [J (WT 1 , T 1)]
max
T 2 ln [CT 2 ]
CT 2 ,{i,T 2 }
CT 2 ,{i,T 2 }
+ T 1 ET 2 [(1 + ) ln [WT 1 ] + HT 1 ]
(5.28)
Thus, using (5.16), the optimality condition for consumption is
UC CT 2 , T 2 = ET 2 [JW (WT 1 , T 1) RT 2 ]
T 2
RT 2
T 1
= (1 + )
ET 2
CT 2
ST 2 RT 2
=
or
(1 + ) T 1
WT 2 CT 2
(5.29)
CT 2 =
1
WT 2
1 + + 2
159
(5.30)
Using (5.17), we then see that the optimality conditions for { i,T 2 } turn out
to be of the same form as at T 1:
ET 2
Ri,T 2
1
E
=
R
, i = 1, ..., n
f,T 2 T 2
RT 2
RT 2
(5.31)
and, as in the case of T 1, equation (5.31) equals unity, since
UC (CT 2 , T 2) = Rf,T 2 ET 2 [JW (WT 1 , T 1)]
T 2
1+
T 1
= Rf,T 2
ET 2
CT 2
ST 2 RT 2
(1 + ) CT 2 Rf,T 2
1
1 =
E
T
2
WT 2 CT 2
RT 2
1
1 = Rf,T 2 ET 2
RT 2
(5.32)
Recognizing the above pattern, we see that the optimal consumption and
portfolio rules for any prior date, t, are
1
1
Wt =
Wt
1 + + ... + T t
1 T t+1

Ri,t
1
E
Et
=
R
= 1, i = 1, ..., n
f
t
t
Rt
Rt
Ct =
(5.33)
(5.34)
Hence, we find that the consumption and portfolio rules are separable for a
log utility individual. Equation (5.33) shows that the consumption-savings decision does not depend on the distribution of asset returns. Moreover, equation
(5.34) indicates that the optimal portfolio proportions depend only on the distribution of one-period returns and not the distribution of asset returns beyond
the current period. This is described as myopic behavior because investment
160
allocation decisions made by the multi-period log investor are identical to those
of a one-period log investor. Hence, the log utility individuals current period
decisions are independent of the possibility of changing investment opportunities
in future periods. It should be emphasized that these independence results are
highly specific to the log utility assumption and do not occur with other utility
functions.
In general, it will be optimal for the individual to choose todays
portfolio in a way that hedges against possible changes in tomorrows investment

opportunities. Such hedging demands for assets will become transparent when
in Chapter 12 we consider the individuals consumption and portfolio choice
problem in a continuous-time setting.
The consumption rule (5.33) shows that consumption is positive whenever
wealth is.
Since utility of consumption is undefined for logarithmic (or any
other constant relative risk aversion) utility when consumption is non-positive,

what ensures that wealth is always positive? The individuals optimal portfolio
choices will reflect this concern. While this example has not specified a specific
distribution for asset returns, portfolio decisions in a discrete-time model can
be quite sensitive to the requirement that wealth exceed zero.
For example,
suppose that the distribution of a risky assets return had no lower bound, as
would be the case if the distribution was normal.
With logarithmic utility,
the optimality conditions (5.34) imply that the individual avoids holding any
normally distributed risky asset, since there is positive probability that a large
negative return would make wealth negative as well.13
In Chapter 12, we
revisit the individuals intertemporal consumption and portfolio choices in a

continuous time environment.
There we will see that the individuals ability
to continuously reallocate her portfolio can lead to fundamental dierences in

asset demands.
Individuals can maintain positive wealth even though they
13 Note that this would not be the case for a risky asset having a return distribution that is
bounded at zero, such as the lognormal distribution.
5.4. SUMMARY
161
hold assets having returns that are instantaneously normally distributed. The
intuition behind this dierence in the discrete- versus continuous-time results
is that the probability of wealth becoming negative decreases when the time
interval between portfolio revisions decreases.
5.4
Summary
An individuals optimal strategy for making lifetime consumption-savings and

portfolio allocation decisions is a topic having practical importance to financial
planners.
This chapters analysis represents a first step in formulating and
deriving a lifetime financial plan. We showed that an individual could approach

this problem by a backward dynamic programing technique that first considered
how decisions would be made when he reached the end of his planning horizon.
For prior periods, consumption and portfolio decisions were derived using the
recursive Bellman equation which is based on the concept of a derived utility
of wealth function. The multi-period planning problem was transformed into a
series of easier-to-solve one period problems. While the consumption-portfolio
choice problem in this chapter assumed that lifetime utility was time-separable,
in future chapters we show that the Bellman equation solution technique often
can apply to cases of time-inseparable lifetime utility.
Our general solution technique was illustrated for the special case of an
individual having logarithmic utility and no wage income. It turned out that
this individuals optimal consumption decision was to consume a proportion
of wealth each period where the proportion was a function of the remaining
periods in the individuals planning horizon but not of the current or future
distributions of asset returns. In other words, future investment opportunities
did not aect the individuals current consumption - savings decision. Optimal
portfolio allocations were also relatively simple because they depended only on
162
the current periods distribution of asset returns.

Deriving an individuals intertemporal consumption and portfolio decisions
has value beyond the application to financial planning. By summing all individuals demands for consumption and assets, a measure of aggregate consumption
and asset demands can be derived. When coupled with a theory of production
technologies and asset supplies, these aggregate demands can provide the foundation for a general equilibrium theory of asset pricing. We turn to this topic
in the next chapter.
5.5
Exercises
1. Consider the following consumption and portfolio choice problem. As
sume that U (Ct , t) = t aCt bCt 2 , B (WT , T ) = 0, and yt 6= 0, where

=
1
1+
and 0 is the individuals subjective rate of time preference.
Further, assume that n = 0 so that there are no risky assets but there is a
single-period riskless asset yielding a return of Rf t = 1/ that is constant
each period. (Equivalently, the risk-free interest rate rf = .) Note that
in this problem labor income is stochastic and there is only one (riskless)
asset for the individual consumer-investor to hold. Hence, the individual
has no portfolio choice decision but must decide only what to consume
each period. In solving this problem, assume that the individuals optimal level of consumption remains below the bliss point of the quadratic
utility function, that is, Ct < 12 a/b, t.
1.a Write down the individuals wealth accumulation equation from period t
to period t + 1.
1.b Solve for the individuals optimal level of consumption at date T 1 and
evaluate J (WT 1 , T 1). (This is trivial.)
5.5. EXERCISES
163
1.c Continue to solve the individuals problem at date T 2, T 3, etc.

and notice the pattern that emerges. From these results, solve for the
individuals optimal level of consumption for any arbitrary date, T t, in
terms of the individuals expected future levels of income.
2. Consider the consumption and portfolio choice problem with power utility
U (Ct , t) t Ct / and a power bequest function B (WT , T ) T WT /.
/ and a constant risk-free
Assume there is no wage income ( yt 0 t)
return equal to Rf t = Rf . Also, assume that n = 1 and the return of the
single risky asset, Rrt , is independently and identically distributed over
time. Denote the proportion of wealth invested in the risky asset at date
t as t .
2.a Derive the first order conditions for the optimal consumption level and
portfolio weight at date T 1, CT 1 and T 1 , and give an explicit expression for CT 1 .
2.b Solve for the form of J (WT 1 , T 1) .
2.c Derive the first order conditions for the optimal consumption level and
portfolio weight at date T 2, CT 2 and T 2 , and give an explicit expression for CT 2 .
2.d Solve for the form of J (WT 2 , T 2).
Based on the pattern for T 1
and T 2, provide expressions for the optimal consumption and portfolio

weight at any date T t, t = 1, 2, 3, ....
3. Consider the multi-period consumption and portfolio choice problem
max Et
Cs ,s s
T 1
P
s=t
U (Cs , s) + B (WT , T )
164

Assume negative exponential utility U (Cs , s) s ebCs and a bequest
function B (WT , T ) T ebWT where = e and > 0 is the (continuously compounded) rate of time preference.
Assume there is no wage
income ( ys 0 s) and a constant risk-free return equal to Rf s = Rf .

Also, assume that n = 1 and the return of the single risky asset, Rrs ,
has an identical and independent normally distribution of N R, 2 each
period.
Denote the proportion of wealth invested in the risky asset at
date s as s .
3.a Derive the optimal portfolio weight at date T 1, T 1 . Hint: It might be

easiest to evaluate expectations in the objective function prior to taking
the first order condition.
3.b Solve for the optimal level of consumption at date T 1, CT 1 . CT 1

will be a function of WT 1 , b, , Rf , R, and 2 .
3.c Solve for the indirect utility function of wealth at date T 1, J (WT 1 , T 1).
3.d Derive the optimal portfolio weight at date T 2, T 2 .

3.e Solve for the optimal level of consumption at date T 2, CT 2 .
4. An individual faces the following consumption and portfolio choice problem
max E0
Ct ,t t
T 1
P
t=0
ln [Ct ] + ln [WT ]
t
(4.1)
where each period the individual can choose between a risk-free asset
paying a time-varying return of Rf t over the period from t to t + 1 and
5.5. EXERCISES
165
a single risky asset. The individual receives no wage income. The risky
assets return over the period from t to t + 1 is given by
Rrt
(1 + ut ) Rf t
=
(1 + d ) R
where ut > 0 and 1 < dt < 0.
ft
with probability
1
2
with probability
1
2
(4.2)
Let t be individuals proportion of
wealth invested in the risky asset at date t.

optimal portfolio weight t for t = 0, ..., T 1.
Solve for the individuals
Chapter 6
Multi-Period Market
Equilibrium
The previous chapter showed how stochastic dynamic programming can be used
to solve for an individuals optimal multi-period consumption and portfolio decisions. In general, deriving an individuals decision rules for particular forms
of utility and distributions of asset returns can be complex.
However, even
though simple solutions for individuals decision rules may not exist, a number
of insights regarding equilibrium asset pricing relationships often can be derived
for an economy populated by such optimizing individuals. This is the topic of
the first section of this chapter. Similar to what was shown in the context of
Chapter 4s single period consumption - portfolio choice model, here we find
that an individuals first order conditions from the multi-period problem can be
re-interpreted as equilibrium conditions for asset prices. This leads to empirically testable implications even when analytical expressions for the individuals
lifetime consumption and portfolio decisions cannot be derived.
As we shall
see, these equilibrium implications generalize those that we derived earlier for a
167
168
CHAPTER 6. MULTI-PERIOD MARKET EQUILIBRIUM
single period environment.

In the second section, we consider an important and popular equilibrium
asset pricing model derived by Nobel laureate Robert E. Lucas (Lucas 1978).
It is an endowment economy model of infinitely-lived, representative individuals. The assumptions of the model, which determine individuals consumption
process, are particularly convenient for deriving the equilibrium price of the
market portfolio of all assets.
As will be shown, the models infinite horizon
gives rise to the possibility of speculative bubbles in asset prices. The last section of the chapter examines the nature of rational bubbles and considers what
conditions could give rise to these non-fundamental price dynamics.
6.1
Asset Pricing in the Multi-Period Model
Recall that the previous chapters Samuelson-Merton model of multi-period consumption and portfolio choices assumed that an individuals objective was
max
Et
Cs ,{is },s,i
T 1
P
s=t
U (Cs , s) + B (WT , T )
(6.1)
and that this problem of maximizing time separable multi-period utility could
be transformed into a series of one-period problems where the individual solved
the Bellman equation
J (Wt , t) = max U (Ct , t) + Et [J (Wt+1 , t + 1)]

Ct ,{i,t }
(6.2)
This led to the first order conditions
UC (Ct , t) = Rf,t Et [JW (Wt+1 , t + 1)]

= JW (Wt , t)
(6.3)
6.1. ASSET PRICING IN THE MULTI-PERIOD MODEL
Et [Rit JW (Wt+1 , t + 1)] = Rf,t Et [JW (Wt+1 , t + 1)] , i = 1, ..., n
169
(6.4)
By making specific assumptions regarding the form of the utility function,

the nature of wage income, and the distributions of asset returns at each date,
explicit formulas for Ct and it may be derived using backward dynamic programing. The previous chapter provided an example of such a derivation for
the case of an individual with log utility and no wage income. However, under
more general assumptions, the multi-period model may have equilibrium implications even when analytical expressions for consumption and portfolio choices
are not possible. This is the topic that we now consider.
6.1.1
The Multi-Period Pricing Kernel
Let us illustrate how equilibrium asset pricing implications can be derived from
the individuals envelope condition (6.3), UC (Ct , t) = JW (Wt , t). This condition conveys that under an optimal policy, the marginal value of financial wealth
equals the marginal utility of consumption. Substituting the envelope condition
evaluated at date t + 1 into the right-hand-side of the first line of (6.3), we have
UC (Ct , t) = Rf,t Et [JW (Wt+1 , t + 1)]

= Rf,t Et UC Ct+1
,t +1
(6.5)
Furthermore, substituting (6.4) into (6.3) and, again, using the envelope condition at date t + 1 allows us to write
170
UC (Ct , t) = Et [Rit JW (Wt+1 , t + 1)]
= Et Rit UC Ct+1
,t +1
(6.6)
or
1 = Et [mt,t+1 Rit ]
= Rf,t Et [mt,t+1 ]
(6.7)
where mt,t+1 UC Ct+1

, t + 1 /UC (Ct , t) is the stochastic discount factor
or pricing kernel between dates t and t + 1.
Equation (6.7) indicates that
our previous asset pricing results derived from a single period consumption portfolio choice problem, such as equation (4.18), hold on a period-by-period
basis even when we allow the consumption - portfolio choice problem to be a
more complex multi-period one.
As before, we can interpret (6.6) and (6.7)
as showing that the marginal rate of substitution between consumption at any

two dates, such as t and t + 1, equals the marginal rate of transformation.
Consumption at date t can be transformed into consumption at date t + 1 by
investing in the riskless asset having return Rf,t or by investing in a risky asset
having the random return Rit .
A similar relationship can be derived for asset returns for any holding period,
not just one of unit length. Note that if equation (6.6) for risky asset j is

, t + 1 = Et+1 Rj,t+1 UC Ct+2

, t + 2 , and then
updated one period, UC Ct+1
this is substituted into the right-hand-side of the original (6.6), one obtains
,t + 2
UC (Ct , t) = Et Rit Et+1 Rj,t+1 UC Ct+2
= Et Rit Rj,t+1 UC Ct+2

,t + 2
(6.8)
6.1. ASSET PRICING IN THE MULTI-PERIOD MODEL
171
or
1 = Et [Rit Rj,t+1 mt,t+2 ]
(6.9)
, t + 2 /UC (Ct , t) is the marginal rate of substitution,

where mt,t+2 UC Ct+2
or the stochastic discount factor, between dates t and t + 2.
In the above
expressions, Rit Rj,t+1 is the return from a trading strategy that first invests in
asset i over the period from t to t + 1 then invests in asset j over the period
t + 1 to t + 2. Of course, i could equal j but need not, in general. By repeated
substitution, (6.9) can be generalized to
1 = Et [Rt,t+k mt,t+k ]
(6.10)
, t + k /UC (Ct , t) and Rt,t+k is the return from any

where mt,t+k UC Ct+k
trading strategy involving multiple assets over the period from dates t to t + k.
Equation (6.10) says that optimizing consumers equate their expected marginal
utilities across all time periods and all states. Its equilibrium implication is that
the stochastic discount relationship holds for multi-period returns generated
from any particular trading strategy. This result implies that empirical tests
of multi-period, time - separable utility models using consumption data and
asset returns can be constructed using a wide variety of investment returns
and holding periods. Expressions such as (6.10) represent moment conditions
that are often tested using generalized method of moments techniques.1
As
mentioned in Chapter 4, such consumption-based tests typically reject models

that assume standard forms of time-separable utility.
This has motivated a
search for alternative utility specifications, a topic we will re-visit in future

chapters.
Lets us now consider a general equilibrium structure for this multi-period
1 See Lars Hansen and Kenneth Singleton (Hansen and Singleton 1983) and Lars Hansen
and Ravi Jagannathan (Hansen and Jagannathan 1991).
172
consumption portfolio choice model.
6.2
The Lucas Model of Asset Pricing
The Lucas model (Lucas 1978) derives the equilibrium prices of risky assets for
an endowment economy.
An endowment economy is one where the random
process generating the economys real output (e.g., Gross Domestic Product or
GDP) is taken to be exogenous. Moreover, it is assumed that output obtained
at a particular date cannot be reinvested to produce more output in the future.
Rather, all output on a given date can only be consumed immediately, implying
that equilibrium aggregate consumption equals the exogenous level of output
at each date. Assets in this economy represent ownership claims on output, so
that output (and consumption) on a given date can also be interpreted as the
cash dividends paid to asset holders. Because re-investment of output is not
permitted, so that the scale of the production process is fixed, assets can be
viewed as being perfectly inelastically supplied.2
As we will make explicit below, these endowment economy assumptions essentially fix the process for aggregate consumption. Along with the assumption
that all individuals are identical, that is, that there is a representative individual, the endowment economy assumptions fix the processes for individuals
consumptions. Thus, individuals marginal rates of substitution between current and future consumptions are pinned down, and the economys stochastic
discount factor becomes exogenous. Furthermore, since the exogenous outputconsumption process also represents the process for the market portfolios ag2 An endowment economy is sometimes described as a "fruit tree" economy. The analogy
refers to an economy whose production is represented by a fixed number of fruit trees. Each
season (date), the trees produce a random amount of output in the form of perishable fruit.
The only value to this fruit is to consume it immediately, as it cannot be reinvested to produce
more fruit in the future. (Planting seeds from the fruit to increase the number of fruit trees is
ruled out.) Assets represent ownership claims on the fixed number of fruit trees (orchards),
so that the fruit produced on each date also equals the dividend paid to asset holders.
6.2. THE LUCAS MODEL OF ASSET PRICING
173
gregate dividends, that, too, is exogenous. This makes it easy to solve for the
equilibrium price of the market portfolio.
In contrast, a production economy is, in a sense, the polar opposite of an endowment economy. A production economy allows for an aggregate consumption
- savings (investment) decision. Not all of current output need be consumed,
but some can be physically invested to produce more output using constant
returns to scale (linear) production technologies. The random distribution of
rates of return on these productive technologies is assumed to be exogenous.
Assets can be interpreted as ownership claims on these technological processes
and, therefore, their supplies are perfectly elastic, varying in accordance to the
individuals reinvestment decision. Hence, the main dierence between production and endowment economies is that production economies pin down assets
rates of return distribution and make consumption (and output) endogenous
while endowment economies pin down consumption and make assets rates of
return distribution endogenous. Probably the best known asset pricing model
based on a production economy was derived John C. Cox, Jonathan E. Ingersoll, and Stephen A. Ross (Cox, Ingersoll, and Ross 1985a). We will study this
continuous time, general equilibrium model in Chapter 13.
6.2.1
Including Dividends in Asset Returns
The Lucas model builds on the multi-period, time separable utility model of
consumption and portfolio choice.
We continue with the stochastic discount
factor pricing relationship of (6.7) but put more structure on the returns of each
asset. Let the return on the ith risky asset, Rit , include a dividend payment
made at date t + 1, di,t+1 , along with a capital gain, Pi,t+1 Pit . Hence Pit
174
denotes the ex-dividend price of the risky asset at date t.
Rit =
di,t+1 + Pi,t+1
.
Pit
(6.11)
Substituting (6.11) into (6.7) and rearranging gives
Pit = Et
"
,t +1
UC Ct+1
(di,t+1 + Pi,t+1 )
UC (Ct , t)
(6.12)
Similar to what was done in equation (6.8), if we substitute for Pi,t+1 using equation (6.12) updated one period, and use the properties of conditional
expectation, we have
Pit
!#

UC Ct+2
,t +1
,t +2
UC Ct+1

(di,t+2 + Pi,t+2 )
= Et
di,t+1 +
UC (Ct , t)
UC Ct+1
,t +1
#
"

UC Ct+2
,t + 1
,t +2
UC Ct+1
= Et
di,t+1 +
(di,t+2 + Pi,t+2 ) (6.13)
UC (Ct , t)
UC (Ct , t)
"
Repeating this type of substitution, that is, solving forward the dierence equation (6.13) gives us
Pit
T
X
UC Ct+j
,t +j
UC Ct+T
,t +T
di,t+j +
Pi,t+T
= Et
UC (Ct , t)
UC (Ct , t)
j=1
where the integer T reflects a large number of future periods.
(6.14)
Now suppose
utility reflects a rate of time preference, so that U (Ct , t) = t u (Ct ), where

=
1
1+
< 1, so that the rate of time preference > 0. Then (6.14) becomes
Pit

C
u
C
u
C
C
t+j
t+T
T
= Et
j
) di,t+j +
) Pi,t+T
u
(C
u
(C
C
C
t
t
j=1
T
X
(6.15)
175
If we have an infinitely lived individual or, equivalently, an individual whose

utility includes a bequest that depends on the utility of his or her ospring,
then we can consider the solution to (6.15) as the planning horizon, T , goes to
T uC (Ct+T )
infinity. If limT Et uC (C ) Pi,t+T = 0, which, as discussed in the next
t
section, is equivalent to assuming the absence of a speculative price bubble,

then
Pit
X
C
u
C
t+j
= Et
j
) di,t+j
u
(C
C
t
j=1
X
mt, t+j di,t+j
= Et
(6.16)
j=1
Equation (6.16) is a present value formula, where the stochastic discount factors
are the marginal rates of substitution between the present and the dates when
the dividends are paid. This "discounted dividend" asset pricing formula holds
for any individual following an optimal consumption - portfolio choice policy.
Thus far, we have not made any strong assumptions about consumer homogeneity or the structure of the economy. For example, (6.16) would hold for a
production economy with heterogeneous individuals.
6.2.2
Equating Dividends to Consumption
The Lucas model makes (6.16) into a general equilibrium model of asset pricing
by assuming there is an infinitely-lived representative individual, meaning that
all individuals are identical with respect to utility and initial wealth. It also
assumes that each asset is a claim on a real output process, where risky asset i
pays a real dividend of dit at date t. Moreover, the dividend from each asset is
assumed to come in the form of a non-storable consumption good that cannot
176
be reinvested. In other words, this dividend output cannot be transformed into

new investment in order to expand the scale of production. The only use for
each assets output is consumption. A share of risky asset i can be interpreted
as an ownership claim on an exogenous dividend-output process that is fixed in
supply. Assuming no wage income, it then follows that aggregate consumption
at each date must equal the total dividends paid by all of the n assets at that
date.
Ct =
n
X
dit
(6.17)
i=1
Given the assumption of a representative individual, this individuals consumption can be equated to aggregate consumption.3
6.2.3
Asset Pricing Examples
With these endowment economy assumptions, the specific form of utility for
the representative agent and the assumed distribution of the assets dividend
processes fully determine equilibrium asset prices. For example, if the representative individual is risk-neutral, so that uC is a constant, then (6.16) becomes
Pit
X
= Et
j di,t+j
(6.18)
j=1
In words, the price of risky asset i is the expected value of dividends discounted
by a constant constant factor, reflecting the constant rate of time preference.
Consider another example where utility is logarithmic, u (Ct ) = ln Ct . Also
Pn
denote dt = i=1 dit be the economys aggregate dividends, which we know by
(6.17) equals aggregate consumption. Then the price of risky asset i is given
3 If one assumes that there are many representative individuals, each will have indentical
per capita consumption and receive identical per capita dividends. Hence, in (6.17) Ct and
dit can be interpreted as per capita quantities.
177
by
Pit
X
C
= Et
j t di,t+j
Ct+j
j=1
X
d
t
j
di,t+j
= Et
d
t+j
j=1
(6.19)
Given assumptions regarding the distribution of the individual assets, the expectation in (6.19) can be computed. However, under this logarithmic utility
assumption, we can obtain the price of the market portfolio of all assets even
without any distributional assumptions. To see this, let Pt represent a claim
on aggregate dividends. Then (6.19) becomes
Pt
X
dt
= Et
j
dt+j
d
t+j
j=1
= dt
(6.20)
implying that the value of the market portfolio moves in step with the current
level of dividends. It does not depend on the distribution of future dividends.
Why?
Higher expected future dividends, dt+j , are exactly oset by a lower
expected marginal utility of consumption, mt, t+j = j dt /dt+j , leaving the value
of a claim on this output process unchanged. This is consistent with our earlier
results showing that a log utility individuals savings (and consumption) are
independent of the distribution of asset returns. Since aggregate savings equals
the aggregate demand for the market portfolio, no change in savings implies no
change in asset demand. Note that this will not be the case for the more general
specification of power (constant relative risk aversion) utility. If u (Ct ) = Ct /,
then
178
Pt
= Et
X
j=1
dt+j
dt
X
Et
j dt+j
= d1
t
dt+j
(6.21)
j=1
which does depend on the distribution of future aggregate dividends (output).
Note from (6.21) that for the case of certainty (Et dt+j = dt+j ), when < 0
higher future aggregate dividends reduce the value of the market portfolio, that
1
< 0. While this seems counter-intuitive, recall
is, Pt /dt+j = j (dt+j /dt )
that for < 0, individuals desire less savings (and more current consumption)
when investment opportunities improve. Since current consumption is fixed at
dt in this endowment economy, the only way to bring higher desired consumption
back down to dt is for total wealth to decrease.
In equilibrium, this occurs
when the price of the market portfolio falls as individuals attempt to sell some
of their portfolio in an (unsuccessful) attempt to raise consumption. Of course
the reverse story occurs when 0 < < 1, as a desired rise in savings is oset by
an increase in wealth via an appreciation of the market portfolio.
If we continue to assume power utility, we can also derive the value of a
hypothetical riskless asset that pays a one-period dividend of $1:
Pf t
1
=
= Et
Rf t
"
dt+1
dt
1 #
(6.22)
Using aggregate U.S. consumption data, Rajnish Mehra and Edward C. Prescott
(Mehra and Prescott 1985) used equations such as (6.21) and (6.22) with dt =
Ct to see if a reasonable value of would produce a risk-premium (excess
average return over a risk-free return) for a market portfolio of U.S. common
179
stocks that matched these stocks historical average excess returns. They found
that for reasonable values of , they could not come close to the historical riskpremium, which at that time they estimated to be around 6 percent.
described this finding as the equity premium puzzle.
They
As mentioned earlier in
Chapter 4, the problem is that for reasonable levels of risk aversion, aggregate
consumption appears to vary too little to justify the high Sharpe ratio for the
market portfolio of stocks. The moment conditions in (6.21) and (6.22) require
a highly negative value of to fit the data.
6.2.4
A Lucas Model with Labor Income
The Lucas endowment economy model has been modified to study a wide array
of issues. For example, Gurdip Bakshi and Zhiwu Chen (Bakshi and Chen 1996)
study a monetary endowment economy by assuming that a representative individual obtains utility from both real consumption and real money balances. In
future chapters, we will present other examples of Lucas-type economies where
utility is non-time separable and where utility reflects psychological biases. In
this section, we present a simplified version of a model by
Stephen Cecchetti, Pok-sang Lam, and Nelson Mark (Cecchetti, Lam, and
Mark 1993) that modifies the Lucas model to consider non-traded labor income.4
As before, suppose that there is a representative agent whose financial wealth
consists of a market portfolio of traded assets that pays an aggregate real dividend of dt at date t. We continue to assume that these assets are in fixed supply
and their dividend consists of a non-storable consumption good. However, now
4 They use a regime-switching version of this model to analyze the equity premium and
risk-free rate puzzles. Based on Generalized Method of Moments (GMM) tests, they find
that their model fits the first moments of the risk-free rate and the return to equity, but not
the second moments.
180
we also permit each individual to be endowed with non-tradeable human capital

that is fixed in supply. The agents return to human capital consists of a wage
payment of yt at date t that also takes the form of the non-storable consumption
good. Hence, equilibrium per capita consumption will equal
Ct = dt + yt
(6.23)
so that it is no longer the case that equilibrium consumption equals dividends.

However, assuming constant relative risk-aversion utility, the value of the market
portfolio can still be written in terms of future consumption and dividends
Pt
X
C
u
C
t+j
Et
j
dt+j
u
(C
C
t)
j=1
X
C
t+j
j
dt+j
Et
C
t
j=1
(6.24)
Because wage income creates a dierence between aggregate dividends and

equilibrium consumption, its presence allows us to assume separate random
processes for dividends and consumption.
For example, one might assume
dividends and equilibrium consumption follow the lognormal processes

/Ct = c + c t+1
ln Ct+1
(6.25)
ln (dt+1 /dt ) = d + d t+1
where the error terms are serially uncorrelated and distributed as
t
0 1
N ,
t
0
1
(6.26)
181
It is left as an end-of-chapter exercise to show that with these assumptions
and dt+j , one can compute the expectation

regarding the distributions of Ct+j
in (6.24) to be
Pt = dt
e
1 e
(6.27)
where
d (1 ) c +
i
1h
(1 )2 2c + 2d (1 ) c d
2
(6.28)
We can confirm that (6.27) equals (6.20) when = 0, d = c , c = d , and

= 1, which is the special case of log utility and no labor income. With no
labor income (d = c , c = d , = 1) but 6= 0, we have = c + 12 2 2c ,
which is increasing in the growth rate of dividends (and consumption) when
> 0. As discussed in Chapter 4, this occurs because greater dividend growth
leads individuals to desire increased savings since they have high intertemporal
elasticity ( = 1/ (1 ) > 1).
An increase in desire savings reflects the
substitution eect exceeding the income or wealth eect. Market clearing then
requires the value of the market portfolio to rise, raising income or wealth to
make desired consumption rise to equal the fixed supply. The reverse occurs
when < 0, as the income or wealth eect will exceed the substitution eect.
For the general case of labor income where is given by (6.28), note that a
lower correlation between consumption and dividends (decline in ) increases .
Since Pt / > 0, this lower correlation raises the value of the market portfolio. Intuitively, this greater demand for the market portfolio results because it
provides better diversification with uncertain labor income.
182
6.3
Rational Asset Price Bubbles
In this section, we examine whether there are solutions other than (6.16) that
can satisfy the asset price dierence equation (6.15).
Indeed, we will show
that there are, and that these alternative solutions can be interpreted as bubble
solutions where the asset price deviates from its fundamental value. Potentially,
these bubble solutions may be of interest because there appear to be numerous
historical episodes during which movements in asset prices appear inconsistent
with reasonable dynamics for dividends or outputs. In other words, assets do
not appear to be valued according to their fundamentals.
Examples include
the Dutch tulip bulb bubble during the 1620s, the Japanese stock price bubble
during the late 1980s, and the U.S. stock price bubble (particularly internetrelated stocks) during the late 1990s.5
While some may conclude that these
bubbles represent direct evidence of irrational behavior on the part of individual

investors, might an argument be made that bubbles could be consistent with
rational actions and beliefs? It is this possibility that we now consider.6
Let us start by defining pt Pit uC (Ct ) as the product of the asset price
and the marginal utility of consumption, excluding the time preference discount
factor, . Then equation (6.12) can be written as the dierence equation
Et [pt+1 ] = 1 pt Et uC Ct+1
di,t+1
(6.29)
where 1 = 1 + > 1 with being the individuals subjective rate of time

preference. The solution (6.16) to this equation is referred to as the fundamental
solution. Let us denote it as ft :
5 Charles P. Kindleberger (Kindleberger 2001) gives an entertaining account of numerous
asset price bubbles.
6 Of course, another possibility is that asset prices always equal their fundamental values,
and sudden rises and falls in these prices reflect sudden changes in perceived fundamentals.
6.3. RATIONAL ASSET PRICE BUBBLES
di,t+j
j uC Ct+j
pt = ft Et
183
(6.30)
j=1
The sum in (6.30) converges as long as the marginal utility-weighted dividends

are expected to grow slower than the time preference discount factor. For the
Lucas endowment economy, assumptions regarding the form of utility and the
distribution of the assets dividends can ensure that this solution has a finite
value.
While ft satisfies (6.29), it is not the only solution. Solutions that satisfy
(6.29) take the general form pt = ft + bt where the bubble component of the
solution is any process that satisfies
Et [bt+1 ] = 1 bt
(6.31)
This is easily verified by substitution into (6.29):

di,t+1
Et [ft+1 + bt+1 ] = 1 (ft + bt ) Et uC Ct+1

Et [ft+1 ] + Et [bt+1 ] = 1 ft + 1 bt Et uC Ct+1

di,t+1
Et [bt+1 ] = 1 bt
(6.32)
where in the last line of (6.32) we use the fact that ft satisfies the dierence
equation. Note that since 1 > 1, bt explodes in expected value:
+ if bt > 0
lim Et [bt+i ] = lim i bt =
i
i
if bt < 0
(6.33)
The exploding nature of bt provides a rationale for interpreting the general

solution pt = ft + bt , bt 6= 0, as a bubble solution. Only when bt = 0, do we get
the fundamental solution.
184
6.3.1
Examples of Bubble Solutions
Suppose that bt follows a deterministic time trend, that is,
bt = b0 t
(6.34)
pt = ft + b0 t
(6.35)
Then the solution
implies that the marginal utility-weighted asset price grows exponentially forever. In other words, we have an ever-expanding speculative bubble.
Next, consider a possibly more realistic modeling of a "bursting" bubble
proposed by Olivier Blanchard (Blanchard 1979).
bt+1
(q)1 bt + et+1
=
z
t+1
with probability q
(6.36)
with probability 1 q
with Et [et+1 ] = Et [zt+1 ] = 0. Note that this process satisfies the condition in
(6.31), so that pt = ft + bt is again a valid bubble solution. In this case, the
bubble continues with probability q each period but bursts with probability
1 q. If it bursts, it returns in expected value to zero, but then a new bubble
would start.
To compensate for the probability of a crash, the expected
return, conditional on not crashing, is higher than in the previous example of

a never ending bubble. The disturbance et allows bubbles to have additional
noise and allows new bubbles to begin after the previous bubble has crashed.
This bursting bubble model can be generalized to allow q to be stochastic.7
7 The
reader is asked to show this in an exercise at the end of the chapter.
6.3. RATIONAL ASSET PRICE BUBBLES
6.3.2
185
The Likelihood of Rational Bubbles
While these examples of bubble solutions indeed satisfy the asset pricing dierence equation in (6.29), there may be additional economic considerations that
rule them out. One issue involves negative bubbles, that is, cases where bt < 0.
From (6.33) we see that individuals must expect that at some future date > t
that the marginal utility-weighted price p = f + b will become negative. Of
course since marginal utility is always positive, this implies that the asset price,
Pit = pt /uC (Ct ) will also be negative. A negative price would be inconsistent
with limited-liability securities, such as typical shareholders equity (stocks).
Moreover, if an individual can freely dispose of an asset, its price cannot be
negative. Hence, negative bubbles can be ruled out.
Based on similar reasoning, Behzad Diba and Herschel Grossman (Diba and
Grossman 1988) argue that many types of bubble processes, including bubbles
that burst and start again, can also be ruled out. Their argument is as follows.
Note that the general process for a bubble can be written as
bt = t b0 +
t
P
st s
(6.37)
s=1
where s , s = 1, ..., t, are mean zero innovations. To avoid negative values of bt

(and negative expected future prices), then realizations of t must satisfy
t 1 bt1 , t 0
(6.38)
For example, suppose that bt = 0, implying that at the current date t a bubble
does not exist.
Then from (6.38) and the requirement that t+1 have mean
zero, it must be the case that t+1 = 0 with probability one.
This implies
that if a bubble currently does not exist, it cannot get started next period or
any future period. The only possibility would be if a positive bubble exists on
186
the first day of trading of the asset, that is, b0 > 0.8
Moreover, the bursting
and then restarting bubble in (6.36) could only avoid a negative value of bt+1 if
zt+1 = 0 with probability 1 and et+1 = 0 whenever bt = 0.
Hence, this type of
bubble would need to be positive on the first trading day, and once it bursts it
could never restart.
Note, however, that arbitrage trading is unlikely to be a strong argument
against a bursting bubble. While short selling an asset with bt > 0 would result
in a profit when the bubble bursts, the short seller could incur substantial losses
beforehand. Over the near term, if the bubble continues, the market value of
the short sellers position could become suciently negative so as to wipe out
his personal wealth.
Other arguments have been used to rule out positive bubbles.
Similar to
the assumptions underlying the Lucas model of the previous section, Jean Tirole
(Tirole 1982) considers a situation with a finite number of rational individuals
and where the dividend processes for risky assets are exogenously given.
In
such an economy, individuals who trade assets at other than their fundamental
prices are playing a zero-sum game, since the aggregate amounts of consumption
and wealth are exogenous. Trading assets at prices having a bubble component
only transfers claims on this fixed supply of wealth between individuals. Hence,
a rational individual will not purchase an asset whose price already reflects a
positive bubble component. This is because at a positive price, previous traders
in the asset have already realized their gains and left a negative-sum game to
the subsequent traders. The notion that an individual would believe that he
can buy an asset at a positive bubble price and later sell it to another at a price
reflecting an even greater bubble component might be considered a "greater
8 An implication is that an initial public oering (IPO) of stock should have a first-day
market price that is above its fundamental value. Interestingly, Jay Ritter (Ritter 1991)
documents that many IPOs initially appear to be overpriced since their subsequent returns
tend to be lower than comparable stocks.
6.4. SUMMARY
187
fool" theory of speculative bubbles.
However, this theory is not consistent
with a finite number of fully rational individuals in most economic settings.

Manual Santos and Michael Woodford (Santos and Woodford 1997) consider the
possibility of speculative bubbles in a wide variety of economies, including those
with overlapping generations of individuals. They conclude that the conditions
necessary for rational speculative bubbles to exist are relatively fragile. Under
fairly general assumptions, equilibria displaying rational price bubbles can be
excluded.9
6.4
Summary
When individuals choose lifetime consumption and portfolio holdings in an optimal fashion, a multi-period stochastic discount factor can be used to price
assets. This is an important generalization of our earlier single-period pricing
result. We also demonstrated that if an assets dividends (cashflows) are modeled explicitly, the assets price satisfies a discounted dividend formula.
The
Lucas endowment economy model took this discounted dividend formula a step
further by equating aggregate dividends to aggregate consumption. This simplified valuing a claim on aggregate dividends, since now the value of this market
portfolio could be expressed as an expectation of a function of only the future
dividend (output) process.
In an infinite horizon model, the possibility of rational asset price bubbles
needs to be considered.
In general, there are multiple solutions for the price
of a risky asset. Bubble solutions represent non-stationary alternatives to the

assets fundamental value. However, when additional aspects of the economic
9 Of course, other considerations that are not fully consistent with rationality may give
rise to bubbles. Jos Scheinkman and Wei Xiong (Scheinkman and Xiong 2003) present a
model where individuals with heterogeneous beliefs think that particular information is more
informative of asset fundamentals than it truely is. Bubbles arise due to a premium reflecting
the option to sell assets to the more optimistic individuals.
188
environment are considered, the conditions that would give rise to rational bubbles appear to be rare.
6.5
Exercises
1. Two individuals agree at date 0 to a forward contract that matures at date

2. The contract is written on an underlying asset that pays a dividend
at date 1 equal to D1 . Let f2 be the date 2 random payo (profit) to the
individual who is the long party in the forward contract.
Also let m0i
be the stochastic discount factor over the period from dates 0 to i where
i = 1, 2, and let E0 [] be the expectations operator at date 0. What is
the value of E0 [m02 f2 ]? Explain your answer.
2. Assume that there is an economy populated by infinitely-lived representative individuals who maximize the lifetime utility function
E0
"
X
t=0
t act
where ct is consumption at date t and a > 0, 0 < < 1.
The economy is
a Lucas (1978) endowment economy having multiple risky assets paying date t
dividends that total dt per capita. Write down an expression for the equilibrium
per capita price of the market portfolio in terms of the assets future dividends.
3. For the Lucas model with labor income, show that assumptions (6.25) and
(6.26) lead to the pricing relationship (6.27) and (6.28).
4. Consider a special case of the model of rational speculative bubbles discussed in this chapter.
Assume that infinitely-lived investors are risk-
neutral and that there is an asset paying a constant, one-period risk-free
6.5.
EXERCISES
189
return of Rf = 1 > 1. There is also an infinitely-lived risky asset with

price pt at date t.
The risky asset is assumed to pay a dividend of dt
which is declared at date t and paid at the end of the period, date t + 1.
Consider the price pt = ft + bt where
ft =
X
Et [dt+i ]
i=0
Rf i+1
(1)
and
bt+1 =
Rf
qt bt
zt+1
+ et+1
with probability qt
(2)
with probability 1 qt
where Et [et+1 ] = Et [zt+1 ] = 0 and where qt is a random variable as of date

t 1 but realized at date t and is uniformly distributed between 0 and 1.
4.a Show whether or not pt = ft + bt subject to the specifications in (1) and
(2) is a valid solution for the price of the risky asset.
4.b Suppose that pt is the price of a barrel of oil. If pt psolar , then solar
energy, which is in perfectly elastic supply, becomes an economically ecient perfect substitute for oil. Can a rational speculative bubble exist for
the price of oil? Explain why or why not.
4.c Suppose pt is the price of a bond that matures at date T < . In this
context, the dt for t T denotes the bonds coupon and principal payments. Can a rational speculative bubble exist for the price of this bond?
Explain why or why not.
5. Consider an endowment economy with representative agents who maxi-
190

mize the following objective function:
max
Cs ,{is },s,i
where T < .
Et
T
P
s=t
s u (Cs )
Explain why a rational speculative asset price bubble
could not exist in such an economy.
Part III
Contingent Claims Pricing
Chapter 7
Basics of Derivative Pricing

Chapter 4 showed how general pricing relationships for contingent claims could
be derived in terms of an equilibrium stochastic discount factor or in terms of
elementary securities. This chapter takes a more detailed look at this important
area of asset pricing.1 The field of contingent claims pricing experienced explosive growth following the seminal work on option pricing by Fischer Black and
Myron Scholes (Black and Scholes 1973), and by Robert Merton (Merton 1973b).
Research on contingent claims valuation and hedging continues to expand, with
significant contributions coming from both academics and finance practitioners.
This research is driving, and is being driven by, innovations in financial markets.
Because research has given new insights into how potential contingent
securities might be priced and hedged, financial service providers are more willing to introduce such securities to the market. In addition, existing contingent
securities motivate further research by academics and practitioners whose goal
is to improve the pricing and hedging of these securities.
1 The topics in this chapter are covered in greater detail in undergraduate- and masterslevel financial derivatives texts such as (McDonald 2002) and (Hull 2000). Readers with a
background in derivatives at this level may wish to skip this chapter. For others without this
knowledge, this chapter is meant to present some fundamentals of derivatives that provide a
foundation for more advanced topics covered in later chapters.
193
194
CHAPTER 7. BASICS OF DERIVATIVE PRICING
We begin by considering two major categories of contingent claims, namely

forward contracts and option contracts. These securities are called derivatives
because their cashflows derive from another underlying variable, such as an
asset price, interest rate, or exchange rate.2 For the case of a derivative whose
underlying is an asset price, we will show that the absence of arbitrage opportunities places restrictions on the derivatives value relative to that of its
underlying asset.3
In the case of forward contracts, arbitrage considerations
alone may lead to an exact pricing formula. However, in the case of options,
these no-arbitrage restrictions cannot determine an exact price for the derivative, but only bounds on the options price. An exact option pricing formula
requires additional assumptions regarding the probability distribution of the
underlying assets returns. The second section of this chapter illustrates how
options can be priced using the well-known binomial option pricing technique.
This is followed by a section covering dierent binomial model applications.
The next section begins with a re-examination of forward contracts and how
they are priced. We then compare them to option contracts and analyze how
the absence of arbitrage opportunities restricts option values.
7.1
Forward and Option Contracts
Chapter 3s discussion of arbitrage derived the link between spot and forward
contracts for foreign exchange. Now we show how that result can be generalized
to valuing forward contracts on any dividend-paying asset. Following this, we
compare option contracts to forward contracts and see how arbitrage places
limits on option prices.
2 Derivatives have been written on a wide assortment of other variables, including commodity prices, weather conditions, catastrophic insurance losses, and credit (default) losses.
3 Thus, our approach is in the spirit of considering the underlying asset as an elementary
security and using no-arbitrage restrictions to derive implications for the derivatives price.
7.1. FORWARD AND OPTION CONTRACTS
7.1.1
195
Forward Contracts on Assets Paying Dividends
Similar to the notation introduced previously, let F0 be the current date 0

forward price for exchanging one share of an underlying asset periods in the
future. Recall that this forward price represents the price agreed to at date 0
but to be paid at future date > 0 for delivery at date of one share of the
asset. The long (short) party in a forward contract agrees to purchase (deliver)
the underlying asset in return for paying (receiving) the forward price. Hence,
the date > 0 payo to the long party in this forward contract is S F0 where
S is the spot price of one share of the underlying asset at the maturity date
of the contract.4
The short partys payo is simply the negative of the long
partys payo. When the forward contract is initiated at date 0, the parties set
the forward price, F0 , to make the value of the contract equal zero. That is,
by setting F0 at date 0, the parties agree to the contract without one of them
needing to make an initial payment to the other.
Let Rf > 1 be one plus the per period risk-free rate for borrowing or lending
over the time interval from date 0 to date . Also, let us allow for the possibility
that the underlying asset might pay dividends during the life of the forward
contract, and use the notation D to denote the date 0 present value of dividends
paid by the underlying asset over the period from date 0 to date .5
The
assets dividends over the life of the forward contract are assumed to be known
at the initial date 0, so that D can be computed by discounting each dividend
payment at the appropriate date 0 risk-free rate corresponding to the time until
the dividend payment is made.
In the analysis that follows, we also assume
that risk-free interest rates are non-random, though most of our results in this
section and the next continue to hold when interest rates are assumed to change
4 Obviously
S is, in general, random as of date 0 while F0 is known as of date 0.

our context, "dividends" refers to any cashflows paid by the asset. For the case of a
coupon-paying bond, the cashflows would be its coupon payments.
5 In
196
randomly through time.6

Now we can derive the equilibrium forward price, F0 , to which the long
and short parties must agree in order for there to be no arbitrage opportunities.
This is done by showing that the long forward contracts date payos can
be exactly replicated by trading in the underlying asset and the risk-free asset.
Then we argue that that in the absence of arbitrage, the date 0 values of the
forward contract and the replicating trades must be the same.
The table below outlines the cashflows of a long forward contract as well as
the trades that would exactly replicate its date payos.
Date 0 Trade
Date 0 Cashflow
Date Cashflow
Long Forward Contract
S F0
1) Buy Asset and Sell Dividends
S0 + D
2) Borrow
R
f F0
F0
Net Cashflow
S0 + D + Rf F0
S F0
Replicating Trades
Note that the payo of the long forward party involves two cashflows: a
positive cashflow of S , which is random as of date 0, and a negative cashflow
equal to F0 , which is certain as of date 0.
The former cashflow can be
replicated by purchasing one share of the underlying asset but selling ownership
of the dividends paid by the asset between dates 0 and .7
This would cost
S0 D, where S0 is the date 0 spot price of one share of the underlying asset.
6 This is especially true for cases in which the underlying asset pays no dividends over the
life of the contract, that is, D = 0. Also, some results can generalize to cases where the
underlying asset pays dividends that are random, such as the case when dividend payments
are proportional to the assets value.
7 In the absence of an explict market for selling the assets dividends, the individual could
borrow the present value of dividends, D, and repay this loan at the future dates when the
dividends are received. This will generate a date 0 cashflow of D, and net future cashflows
of zero since the dividend payments exactly cover the loan repayments.
197
The latter cashflow can be replicated by borrowing the discounted value of F0 .

This would generate current revenue of Rf F0 .
Therefore, the net cost of
replicating the long partys cashflow is S0 D Rf F0 .
In the absence of
arbitrage, this cost must be the same as the cost of initiating the long position
in the forward contract, which is zero.8
Hence, we obtain the no-arbitrage
condition
S0 D Rf F0 = 0
(7.1)
F0 = (S0 D) Rf
(7.2)
or
Equation (7.2) determines the equilibrium forward price of the contract. Note
that if this contract had been initiated at a previous date, say date 1, at the
forward price F1 = X, then the date 0 value (replacement cost) of the long
partys payo, which we denote as f0 , would still be the cost of replicating the
two cashflows:
f0 = S0 D Rf X
8 If
(7.3)
S0 D Rf F0 < 0, the arbitrage would be to perform the following trades at date
0: 1) purchase one share of the stock and sell ownership of the dividends; 2) borrow R
f F0 ;
3) take a short position in the forward contract. The date 0 net cashflow of these three
transactions is (S0 D) + Rf F0 + 0 > 0, by assumption. At date the individual
would:1) deliver the one share of the stock to satisfy the short forward position; 2) receive F0
as payment for delivering this one share of stock; 3) repay borrowing equal to F0 . The date
net cashflow of these three transactions is 0+ F0 F0 = 0. Hence, this arbitrage generates a
positive cashflow at date 0 and a zero cashflow at date . Conversely, if S0 DR
f F0 > 0 ,
an arbitrage would be to perform the following trades at date 0: 1) short sell one share of
the stock and purchase rights to the dividends to be paid to the lender of the stock. (In
the absence of an explict market for buying the assets dividends, the individual could lend
out the present value of dividends, D, and receive payment on this loan at the future dates
when the dividends to be paid.); 2) lend R
f F0 ; 3) take a long position in the forward
contract. The date 0 net cashflow of these three transactions is (S0 D) Rf F0 + 0 > 0,
by assumption. At date the individual would:1) obtain one share of the stock from the long
forward position and deliver it to satisfy the short sale obligation; 2) pay F0 to short party
in forward contract; 3) receive F0 from lending agreement. The date net cashflow of these
three transactions is 0 F0 + F0 = 0. Hence, this arbitrage generates a positive cashflow
at date 0 and a zero cashflow at date .
198
However, as long as date 0 is following the initiation of the contract, the value
of the payo would not, in general, equal zero. Of course, the replacement cost
of the short partys payo would be simply f0 = Rf X +D S0 .
It should be pointed out that our derivation of the forward price in (7.2) did
not require any assumption regarding the random distribution of the underlying
asset price, S . The reason for this is due to our ability to replicate the forward
contracts payo using a static replication strategy: all trades needed to replicate
the forward contracts date payo were done at the initial date 0. As we shall
see, such a static replication strategy is not possible, in general, when pricing
other contingent claims such as options. Replicating option payos will entail,
in general, a dynamic replication strategy: trades to replicate a options payo
at date will involve trades at multiple dates during the interval between dates
0 and .
As will be shown, such a dynamic trading strategy requires some
assumptions regarding the stochastic properties of the underlying assets price.

Typically, assumptions are made that result in the markets for the contingent
claim and the underlying asset being dynamically complete.
As a prerequisite to these issues of option valuation, let us first discuss the
basic features of option contracts and compare their payos to those of forward
contracts.9
7.1.2
Basic Characteristics of Option Prices
The owner of a call option has the right, but not the obligation, to buy a given
asset in the future at a pre-agreed price, known as the exercise price or strike
price. Similarly, the owner of a put option has the right, but not the obligation,
to sell a given asset in the future at a pre-agreed price. For each owner (buyer)
9 Much of the next sections results are due to Robert C. Merton (Merton 1973b).
greater details, see this article.
For
199
of an option, there is an option seller, also referred to as the option writer.

If the owner of a call (put) option chooses to exercise, the seller must deliver
(receive) the underlying asset or commodity in return for receiving (paying) the
pre-agreed exercise price. Since an option always has a non-negative payo to
the owner, this buyer of the option must make an initial payment, called the
options premium, to the seller of the option.10
Options can have dierent features regarding which future date(s) that exercise can occur.
A European option can be exercised only at the maturity
of the option contract, while an American option can be exercised at any time
prior to the maturity of the contract.
Let us define the following notation, similar to that used to describe a forward
contract. Let S0 denote the current date 0 price per share of the underlying
asset, and let this assets price at the maturity date of the option contract, ,
be denoted as S . We let X be the exercise price of the option and denote the
date t price of European call and put options be ct and pt , respectively. Then
based on our description of the payos of call and put options, we can write the
maturity values of European call and put options as:
max [S X, 0]
(7.4)
max [X S , 0]
(7.5)
Now we recall that the payos to the long and short parties of a forward contract
are S F0 and F0 S , respectively. If we interpret the pre-agreed forward
price, F0 , as analogous to an options pre-agreed exercise price, X, then we
see that a call options payo equals that of long forward payo whenever the
long forward payo is positive, and it equals 0 when the long forward payo is
10 The owner of an option will choose to exercise it only if it is profitable to do so. The
owner can always let the option expire unexercised, in which case its resulting payo would
be zero.
200
negative. Similarly, the payo of the put option equals the short forward payo
when this payo is positive, and it equals 0 when the short forward payo is
negative.
Hence, assuming X = F0 , we see that the payo of a call option
weakly dominates that of a long forward position while the payo of a put
option weakly dominates that of a short forward position.11 This is due to the
consequence of option payos always being non-negative while forward contract
payos can be of either sign.
Lower Bounds on European Option Values

Since a European call options payo is at least as great as that of a comparable
long forward position, this implies that the current value of a European call must
be at least as great as the current value of a long forward position. Hence, since
(7.3) is the current value of a long forward position contract, the European calls
value must satisfy
c0 S0 D Rf X
(7.6)
Furthermore, since the call options payo is always non-negative, its current
value must also be non-negative, that is, c0 0.
with (7.6) implies
Combining this restriction
h
i
c0 max S0 D Rf X, 0
(7.7)
By comparing a European put options payo to that of a short forward position,

a similar argument can be made to prove that
h
i
p0 max R
f X + D S0 , 0
(7.8)
11 A payo is said to dominate another when its value is strictly greater in all states of
nature. A payo weakly dominates another when its value is greater in some states of nature
and the same in other states of nature.
201
An alternative proof is as follows. Consider constructing two portfolios at date

0:
Date 0:
Portfolio A = a put option having value p0 and a share of the underlying
asset having value S0
Portfolio B = a bond having initial value of R
f X +D
Then at date , these two portfolios are worth:
Date :
Portfolio A = max [ X S , 0] + S + DRf = max [X, S ] + DRf
Portfolio B = X + DRf
Since portfolio As value at date is always at least as great as that of
portfolio B, the absence of arbitrage implies that its value at date 0 must always
be at least as great as that of portfolio B at date 0. Hence, p0 +S0 R
f X +D,
proving result (7.8).
Put-Call Parity
Similar logic can be used to derive an important relationship that links the value
of European call and put options that are written on the same underlying asset
and that have the same maturity date and exercise price. This relationship is
referred to as put-call parity:
c0 + Rf X + D = p0 + S0
To show this, consider forming the following two portfolios at date 0:
Date 0:
(7.9)
202
Portfolio A = a put option having value p0 and a share of the underlying

asset having value S0
Portfolio B = a call option having value c0 and a bond with initial value
of Rf X + D
Then at date , these two portfolios are worth
Date :
Portfolio A = max [X S , 0] + S + DRf = max [X, S ] + DRf
Portfolio B = max [0, S X] + X + DRf = max [X, S ] + DRf
Since portfolios A and B have exactly the same payo, in the absence of arbitrage their initial values must be the same, proving the put-call parity relation
(7.9). Note that if we rearrange (7.9) as c0 p0 = S0 R
f X D = f0 , we
see that the value of a long forward contract can be replicated by purchasing a
European call option and writing (selling) a European put option.
American Options
Relative to European options, American options have the additional right that
allows the holder (owner) to exercise the option prior to the maturity date.
Hence, all other things equal, an American option must be at least as valuable
as a European option.
Thus if we let the upper case letters C0 and P0 be
the current values of American call and put options, respectively, then comparing them to European call and put options having equivalent underlying asset,
maturity, and exercise price features, it must be the case that C0 c0 and
P0 p0 .
There are, however, cases where an American options early exercise feature
has no value because it would not be optimal to exercise the option early. This
situation occurs for the case of an American call option written on an asset
7.2. BINOMIAL OPTION PRICING
203
that pays no dividends over the life of the option. To see this, note that (7.7)
says that prior to maturity, the value of a European call option must satisfy
c0 S0 R
f X.
However, if an American call option is exercised prior to
maturity, its value equals C0 = S0 X < S0 Rf X < c0 . This contradicts

the condition C0 c0 . Hence, if a holder of an American call option wished
to liquidate his position, it would always be better to sell the option, receiving
C0 , rather than exercising it for the lower amount S0 X. By exercising early,
the call option owner loses the time value of money due to paying X now rather
than later.
Note, however, that if the underlying asset pays dividends, early
exercise of an American call option just prior to a dividend payment may be

optimal.
In this instance, early exercise would entitle the option holder to
receive the assets dividend payment, a payment that would be lost if exercise
was delayed.
For an American put option that is suciently in the money, that is, S0 is
significantly less than X, it may be optimal to exercise the option early, selling
the asset immediately and receiving $X now, rather than waiting and receiving
$X at date (which would have a present value of Rf X).
Note that this
does not necessarily violate (7.8), since at exercise P0 = X S0 , which could be

greater than Rf X + D S0 if the remaining dividends are suciently small.
7.2
Binomial Option Pricing
The previous section demonstrated that the absence of arbitrage restricts the
price of an option in terms of its underlying asset. However, the no-arbitrage
assumption, alone, cannot determine an exact option price as a function of the
underlying asset price. To do so, one needs to make an additional assumption
regarding the distribution of returns earned by the underlying asset. As we shall
see, particular distributional assumptions for the underlying asset can lead to a
204
situation where the options payo can be replicated by trading in the underlying
asset and a risk-free asset and, in general, this trading occurs at multiple dates.
When such a dynamic replication strategy is feasible, the option market is said
to be dynamically complete.
Assuming the absence of arbitrage then allows
us to equate the value of the options payo to the prices of more primitive
securities, namely, the prices of the underlying asset and the risk-free asset.
We now turn to a popular discrete-time, discrete-state model that produces this
result.
The model presented in this section was developed John Cox, Stephen Ross,
and Mark Rubinstein (Cox, Ross, and Rubinstein 1979). It makes the assumption that the underlying asset, hereafter referred to as a stock, takes on one
of only two possible values each period. While this may seem unrealistic, the
assumption leads to a formula that often can accurately price options. This binomial option pricing technique is frequently applied by finance practitioners to
numerically compute the prices of complex options. Here, we start by considering the pricing of a simple European option written on a non-dividend-paying
stock.
In addition to assuming the absence of arbitrage opportunities, the binomial
model assumes that the current underlying stock price, S, either moves up, by
a proportion u, or down, by a proportion d, each period. The probability of an
up move is , so that the probability of a down move is 1 . This two-state
stock price process can be illustrated as
uS
S
with probability
(7.10)
&
dS
with probability 1
205
Denote Rf as one plus the risk-free interest rate for the period of unit length.
This risk-free return is assumed to be constant over time. To avoid arbitrage
between the stock and the risk-free investment, we must have d < Rf < u.12
7.2.1
Valuing a One Period Option
Our valuation of an option whose maturity can span multiple periods will use
a backward dynamic programming approach.
First, we will value the option
when it has only one period left until maturity, then we will value it when it has
two periods left until maturity, and so on until we establish an option formula
for an arbitrary number of periods until maturity.
Let c equal the value of a European call option written on the stock and
having a strike price of X. At maturity, c = max[0, S X]. Thus:
One period prior to maturity:
cu max [0, uS X] with probability

c
(7.11)
&
cd max [0, dS X]
with probability 1
What is c one period before maturity? Consider a portfolio containing shares

of stock and $B of bonds. It has current value equal to S + B. Then the value
of this portfolio evolves over the period as
12 If R < d, implying that the return on the stock is always higher than the risk-free return,
f
an arbitrage would be to borrow at the risk free rate and use the proceeds to purchase the
stock. A profit is assured because the return on the stock would always exceed the loan
repayment. Conversely, if u < Rf , implying that the return on the stock is always lower
than the risk free return, an arbitrage would be to short sell the stock and use the proceeds to
invest at the risk free rate. A profit is assured because the risk free return will always exceed
the value of the stock to be repaid to the stock lender.
206
uS + Rf B with probability
S + B
(7.12)
&
dS + Rf B
with probability 1
With two securities (the bond and stock) and two states of nature (up or down),
and B can be chosen to replicate the payo of the call option:
uS + Rf B = cu
(7.13)
dS + Rf B = cd
(7.14)
Solving for and B that satisfy these two equations, we have
cu cd
(u d) S
(7.15)
B =
ucd dcu
(u d) Rf
(7.16)
Hence, a portfolio of shares of stock and $B of bonds produces the

same cashflow as the call option.13 This is possible because the option market
is complete.
As was shown in Chapter 4, here there are equal numbers of
states and assets having independent returns so that trading in the stock and
bond produce payos that span the two states. Now since the portfolios return
replicates that of the option, the absence of arbitrage implies
13 , the number of shares of stock per option contract needed to replicate (or hedge) the
options payo, is referred to as the options hedge ratio. It can be verified from the formulas
that for standard call options, this ratio is always between 0 and 1. For put options, it is
always between -1 and 0. B , the investment in bonds, is negative for call options but positive
for put options. In other words, the replicating trades for a call option involve buying shares
in the underlying asset partially financed by borrowing at the risk-free rate. The replicating
trades for a put option involve investing at the risk-free rate partially financed by short-selling
the underlying asset.
c = S + B
207
(7.17)
This analysis provides practical insights for option traders. Suppose an option
writer wishes to hedge her position from selling an option, that is, insure that
she will be able to cover her liability to the option buyer in all states of nature.
Then her appropriate hedging strategy is to purchase shares of stock and
$B of bonds, since, from equations (7.13) and (7.14) the proceeds from this
hedge portfolio will cover her liability in both states of nature. Her cost for this
hedge portfolio is S + B , and in a perfectly competitive options market, the
premium received for selling the option, c, will equal this hedging cost.
Example: If S = $50, u = 2, d = .5, Rf = 1.25, X = $50, then
uS = $100, dS = $25, cu = $50, cd = $0.
Therefore:
B =
2
50 0
=
(2 .5) 50
3
40
0 25
=
(2 .5) 1.25
3
so that
c = S + B =
40
60
2
(50)
=
= $20
3
3
3
If c < S + B , then an arbitrage is to short sell shares of stock, invest

$ B in bonds, and buy the call option. Conversely, if c > S + B , then an
arbitrage is to write the call option, buy shares of stock, and borrow $ B .
The resulting option pricing formula has an interesting implication. It can
be re-written as
208
ucd dcu
cu cd
+
c = S + B =
(u d) (u d) Rf
h
i
Rf d
uRf
ud max [0, uS X] + ud max [0, dS X]
=
Rf
(7.18)
which does not depend on the probability of an up or down move of the stock, .
Thus, given S, investors will agree on the no-arbitrage value of the call option
even if they do not agree on . The call option formula does not directly depend
on investors attitudes toward risk. It is a relative (to the stock) pricing formula.
This is reminiscent of Chapter 4s result (4.44) in which contingent claims could
be priced based on state prices but without knowledge of the probability of
dierent states occurring.
Since determines the stocks expected rate of
return, u + d(1 ) 1, this does not need to be known or estimated in order

to solve for the no-arbitrage value of the option, c.
However, we do need to
know u and d, the size of movements per period, which determine the stocks
volatility.
Note also that we can re-write c as
c=
where
b
Rf d
ud .
1
[b
cu + (1
b) cd ]
Rf
(7.19)
Since 0 <
b < 1,
b has the properties of a probability. In fact, this is the
risk-neutral probability, as defined in Chapter 4, of an up move in the stocks

price. To see that
b equals the true probability if individuals are risk-neutral,
note that if the expected return on the stock equals the risk free return, Rf ,
then
209
[u + d (1 )] S = Rf S
(7.20)
which implies that

=
Rf d
=
b.
ud
(7.21)
so that
b does equal under risk-neutrality. Thus, (7.19) can be expressed as
ct =
1 b
E [ct+1 ]
Rf
b [] denotes the expectation operator

where, as in Chapter 4s equation (4.46), E
evaluated using the risk-neutral probabilities

b rather than the true or physical
probabilities .
7.2.2
Valuing a Multi-Period Option
Next, consider the options value with:

Two periods prior to maturity:
The stock price process is
u2 S
uS
%
&
duS
&
dS
%
&
d2 S
so that the option price process is
(7.22)
210
cu
cuu max 0, u2 S X
%
&
cdu max [0, duS X]
&
cd
(7.23)
%
&
cdd max 0, d2 S X
Using the results from our analysis when there was only one period to maturity, we know that
cu =
cd =
bcuu + (1
b) cdu
Rf
bcdu + (1
b) cdd
Rf
(7.24)
(7.25)
With two periods to maturity, the one period to go cashflows of cu and

cd can be replicated once again by the stock and bond portfolio composed of
=
cu cd
(ud)S
shares of stock and B =
c = S + B =
which, as before says that ct =
1
Rf
ucd dcu
(ud)Rf
of bonds. No arbitrage implies
1
[b
cu + (1
b) cd ]
Rf
(7.26)
b [ct+1 ]. The market is not only complete

E
over the last period but over the second-to-last period as well. Substituting in
for cu and cd , we have
211
i
1 h 2
2
b
(7.27)
c
+
2b
(1
b
)
c
+
(1
b
)
c
uu
ud
dd
Rf2
i
2
1 h 2
=
b
max
0,
u
S
X
+
2b
(1
b
)
max
[0,
duS
X]
Rf2
i
1 h
b)2 max 0, d2 S X
+ 2 (1
Rf
c =
which can also be interpreted as ct =
1
R2f
b [ct+2 ]. This illustrates that when a

E
market is complete each period, it becomes complete over the sequence of these
individual periods. In other words, the option market is said to be dynamically
complete. Even though the tree diagrams in (7.22) and (7.23) indicate that there
are four states of nature two periods in the future (and three dierent payos
for the option), these states can be spanned by a dynamic trading strategy
involving just two assets. That is, we have shown that by appropriate trading
in just two assets, payos in greater than two states can be replicated.
Note that c depends only on S, X, u, d, Rf , and the time until maturity,
two periods. Repeating this analysis for three, four, five, ..., n periods prior to
maturity, we always obtain
c = S + B =
1
[b
cu + (1
b) cd ]
Rf
By repeated substitution for cu , cd , cuu , cud , cdd , cuuu , etc., we obtain the
formula:
n periods prior to maturity:
n
X
1
n!
nj
b)
max 0, uj dnj S X
c= n
bj (1
Rf j=0 j! (n j)!
Similar to before, equation (7.28) can be interpreted as ct =
1 b
E [ct+n ],
Rn
f
(7.28)
imply-
212
ing that the market is dynamically complete over any number of periods prior
to the options expiration. The formula in (7.28) can be further simplified by
defining a as the minimum number of upward jumps of S for it to exceed X.
Thus a is the smallest non-negative integer such that ua dna S > X. Taking the
natural logarithm of both sides, a is the minimum integer > ln(X/Sdn )/ln(u/d).
Therefore for all j < a (the option matures out-of-the money),
max 0, uj dnj S X = 0,
(7.29)
while for all j > a (the option matures in-the-money),
max 0, uj dnj S X = uj dnj S X
(7.30)
Thus, the formula for c can be re-written:
n
X
1
n!
c= n
b)nj uj dnj S X
bj (1
Rf j=a j! (n j)!
(7.31)
Breaking (7.31) up into two terms, we have:
n
X
"
j nj
u d
n!
b)nj
bj (1
j!
(n
j)!
Rnf
j=a
n
X
n!
b)nj
XRfn
bj (1
j!
(n
j)!
j=a
c = S
(7.32)
The terms in brackets in (7.32) are complementary binomial distribution func-
tions, so that we can write it as
7.3. BINOMIAL MODEL APPLICATIONS
where
b0
u
Rf
b]
c = S[a; n,
b0 ] XRn
f [a; n,
213
(7.33)
b and [a; n,
b] = the probability that the sum of n random
variables which equal 1 with probability

b and 0 with probability 1
b will be
a. These formulas imply that c is the discounted expected value of the calls
terminal payo under the risk-neutral probability distribution.

If we define as the time until maturity of the call option and 2 as the
variance per unit time of the stocks rate of return (which depends on u and d),
then by taking the limit as the number of periods n , but the length of
each period
0, the Cox-Ross-Rubinstein binomial option pricing formula
converges to the well-known Black-Scholes-Merton option pricing formula14
% #
ln
where z
S
XR
f

c = SN (z) XR
f N z
+ 12 2
( )
distribution function.
7.3
&
(7.34)
and N () is that cumulative standard normal
Binomial Model Applications
The Cox, Ross, and Rubinsteins binomial technique is useful for valuing relatively complicated options, such as those having American (early exercise) features. In this section we show how the model can be used to value an American
put option and an option written on an asset that pays dividends.
Similar to our earlier presentation, assume that over each period of length
14 The intuition for why (7.34) is a limit of (7.33) is due to the Central Limit Theorem. As
the number of periods becomes large, the sum of binomially distributed random stock rates
of return becomes normally distributed. Note that in the Black-Scholes-Merton formula,
Rf is now the risk-free return per unit time rather than the risk-free return for each period.
The relationship between and u and d will be discussed shortly. The binomial model (??)
also can have a dierent continuous-time limit, namely, the jump diusion model that will be
presented in Chapter 11.
214
t, stock prices follow the process
uS
S
with probability
(7.35)
&
dS
with probability 1
The results of our earlier analysis showed that the assumption of an absence
of arbitrage allowed us to apply risk-neutral valuation techniques to derive the
price of an option. Recall that, in general, this method of valuing a derivative
security can be implemented by:
i) setting the expected rate of return on all securities equal to the risk-free
rate
ii) discounting the expected value of future cashflows generated from i) by
this risk-free rate
For example, suppose we examine the value of the stock, S, in terms of the
risk-neutral valuation method. Similar to the previous analysis, define Rf as the
risk-free return per unit time, so that the risk-free return over a time interval
t is Rft . Then we have
[St+t ]
= Rt
E
f
(7.36)
= Rt
[b
uS + (1
b)dS]
f
[] is the expectations operated under the condition that the expected

where E
rate of return on all assets equals the risk-free interest rate, which is not necessarily the assets true expected rates of return. Re-arranging (7.36) we obtain
Rt
bu + (1
b)d
f =
(7.37)
215
which implies
b=
Rft d
ud
This is the same formula for

b as was derived earlier.
(7.38)
Hence, risk-neutral
valuation is consistent with this simple example.
7.3.1
Calibrating the Model
To use the binomial model to value actual options, the parameters u and d
must be calibrated to fit the variance of the underlying stock. When estimating a stocks volatility, it is often assumed that stock prices are lognormally
distributed. This implies that the continuously-compounded rate of return on
the stock over a period of length t, given by ln (St+t ) ln (St ), is normally
distributed with a constant per-period variance of t2 .
As we shall see in
Chapter 9, this constant variance assumption is also used in the Black-Scholes

option pricing model. Thus, the sample standard deviation of a time-series of
historical log stock price changes provides us with an estimate of . Based on
this value of , approximate values of u and d that result in the same variance
for a binomial stock price distribution are15
15 That the values of u and d in (7.39) result in a variance of stock returns given by 2 t
for suciently small t can be verified by
k noting
l that, in the binomial model, the vari2
ance of the end-of-period stock price is E St+t
E [St+t ]2 = u2 S 2 + (1 ) d2 S 2
q
r
2
[S + (1 ) dS] = S 2 u2 + (1 ) d2 [u + (1 ) d]2
k

l
= S 2 et e t + e t 1 e2t , where = et and is the (continuously-
compounded) expected rate of return kon the stock per unit time.
This implies
that the

l
variance of the return on the stock is et e t + e t 1 e2t . Expanding
this expression in a series using ex = 1 + x + 12 x2 + 16 x3 + ... and then ignoring all terms of
order (t)2 and higher, it equals t 2 .
216
u = e
(7.39)
1
=e
u
d =
Hence, condition (7.39) provides a simple way of calibrating u and d to the

stocks volatility, .
Now consider the path of the stock price. Because we assumed u = d1 , the
binomial process for the stock price has the simplified form:
u4 S
u3 S
%
&
u2 S
u2 S
&
uS
uS
&
S
&
S
&
(7.40)
&
dS
dS
&
%
&
d2 S
d2 S
&
d3 S
%
&
d4 S
Given the stock price, S, and its volatility, , the above tree or lattice can be

calculated for any number of periods using u = e
7.3.2
217
and d = e
Valuing an American Option
We can numerically value an option on this stock by starting at the last period
and working back toward the first period. Recall that an American put option
that is not exercised early will have a final period (date ) value
P = max [0, X S ]
(7.41)
The value of the put at date t is then the risk-neutral expected value
discounted by Rft .
P t
[P ]
= Rft E
bP ,u + (1
= Rft
b) P ,d
(7.42)
where P ,u is the date value of the option if stock price changes by proportion
u, while P ,d is the date value of the option if stock price changes by proportion
d. However, with an American put option, we need to check whether this value
exceeds the value of the put if it were exercised early. Hence, the put options
value can be expressed as
i
bP ,u + (1
b) P ,d
P t = max X S t , Rft
(7.43)
Let us illustrate this binomial valuation technique with the following example:
A stock has a current price of S = $80.50 and a volatility = 0.33.
t =
1
9
year, then u = e
.33
= e.11 = 1.1163 and d =
1
u
= .8958.
If
218
Thus the 3-period tree for the stock price is
Date : 0
3
111.98
100.32
%
&
89.86
89.86
&
S = 80.50
80.50
&
%
&
72.12
72.12
&
64.60
%
&
57.86
Next, consider valuing an American put option on this stock that matures in
=
1
3
years (4 months) and has an exercise price of X = $75. Assume that the
risk-free return is Rf = e0.09 , that is, the continuously-compounded risk-free

interest rate is 9%. This implies
b=
0.09
Rt
e 9 .8958
f d
=
= .5181
ud
1.1163 .8958
We can now start at date 3 and begin filling in the tree for the put option.
Date : 0
219
3
Puuu
Puu
%
&
Pu
Puud
&
P
Pud
%
&
&
Pd
Pudd
&
Pdd
%
&
Pddd
Using P3 = max [0, X S3 ], we have
220
Date : 0
3
0.00
Puu
%
&
Pu
0.00
&
P
Pud
%
&
&
Pd
2.88
&
Pdd
%
&
17.14
Next, using
i
bP3,u + (1
b) P3,d
P2 = max X S2 , Rft
Date : 0
221
3
0.00
0.00
%
&
Pu
0.00
&
P
1.37
&
%
&
Pd
2.88
&
10.40
%
&
17.14
Note that at Pdd the option is exercised early since
bP3,u + (1
b) P3,d
Pdd = max X S2 , Rt
f
= max [75 64.60, 9.65] = $10.40

Next, using
b
P
+
(1
b
)
P
P1 = max X S1 , Rt
2,u
2,d
f
222
Date : 0
3
0.00
0.00
%
&
0.65
0.00
&
P
1.37
&
%
&
5.66
2.88
&
10.40
%
&
17.14
Note that the option is not exercised early at Pd since

h
bP2,u + (1
b) P2,d
Pd = max X S1 , Rt
f
= max [75 72.12, 5.66] = $5.66
Finally, we calculate the value of the put at date 0 using
h
i
bP1,u + (1
b) P1,d
P0 = max X S0 , Rft
= max [5.5, 3.03] = $3.03
and the final tree for the put is
Date : 0
223
3
0.00
0.00
%
&
0.65
0.00
&
3.03
1.37
&
%
&
5.66
2.88
&
10.40
%
&
17.14
7.3.3
Options on Dividend-Paying Assets
One can generalize the above procedure to allow for the stock (or portfolio of
stocks such as a stock index) to continuously pay dividends that have a per
unit time yield equal to , that is, for t suciently small, the owner of the
stock receives a dividend of St. For this case of a dividend-yielding asset, we
simply redefine
t
d
Rf e
b=
ud
(7.44)
This is because when the asset pays a dividend yield of , its expected riskt
neutral appreciation is Rf e
rather than Rft .
For the case in which a stock is assumed to pay a known dividend yield, , at
a single point in time, then if date it is prior to the stock going ex-dividend,
the nodes of the stock price tree equal
224
uj dij S
j = 0, 1, ..., i.
(7.45)
If the date it is after the stock goes ex-dividend, the nodes of the stock price
tree equal
uj dij S (1 )
j = 0, 1, ..., i.
(7.46)
The value of an option is calculated as before. We work backwards and again

check for the optimality of early exercise.
7.4
Summary
In an environment where there is an absence of arbitrage opportunities, the

price of a contingent claim is restricted by the price of its underlying asset.
For some derivative securities, such as forward contracts, the contracts payo
can be replicated by the underlying asset and a riskless asset using a static
trading strategy. In such a situation, the absence of arbitrage leads to a unique
link between the derivatives price and that of its underlying asset without
the need for additional assumptions regarding the assets return distribution.
For other types of derivatives, including options, static replication may not be
possible.
An additional assumption regarding the underlying assets return
distribution is necessary for valuing such derivative contracts. An example is

the assumption that the underlying assets returns are binomially distributed.
In this case, an options payo can be dynamically replicated by repeated trading
in a portfolio consisting of its underlying asset and a risk-free asset. Consistent
with our earlier analysis, this situation of a dynamically complete market allows
us to value derivatives using the risk-neutral approach. We also illustrated the
flexibility of this binomial model by applying it to value options having an early
7.5. EXERCISES
225
exercise feature as well as options written on a dividend-paying asset.

As will be shown in Chapter 9, the binomial assumption is not the only way
to obtain market completeness and a unique option pricing formula.
If one
assumes that investors can trade continuously in the underlying asset, and the
underlyings returns follow a continuous time diusion process, then these alternative assumptions can also lead to market completeness. The next chapter
prepares us for this important topic by introducing the mathematics of continuous time stochastic processes.
7.5
Exercises
1. In light of this chapters discussion of forward contracts on dividend paying

assets, re-interpret Chapter 3s example of a forward contract on a foreign currency. In particular, what are the "dividends" paid by a foreign
currency?
2. What is the lower bound for the price of a three-month European put
option on a dividend-paying stock when the stock price is $58, the strike
price is $65, the annualized risk-free return is Rf = e0.05 , and the stock is
to pay a $3 dividend two months from now?
3. Suppose that c1 , c2 , and c3 are the prices of European call options with
strike prices X1 , X2 , and X3 , respectively, where X3 > X2 > X1 and
X3 X2 = X2 X1 . All options are written on the same asset and have
the same maturity. Show that
c2
1
(c1 + c3 )
2
226
(Hint: Consider a portfolio that is long the option having a strike price of X1 ,
long the option having the strike price of X3 , and short two options having the
strike price of X2 .)
4. Consider the binomial (Cox-Ross-Rubinstein) option pricing model. The

underlying stock pays no dividends and has the characteristic that u = 2
and d = 1/2. In other words, if the stock increases (decreases) over a
period its value doubles (halves). Also, assume that one plus the risk-free
interest rate satisfies Rf = 5/4. Let there be two periods and three dates:
0, 1, and 2. At the initial date 0, the stock price is S0 = 4. The following
option is a type of Asian option referred to as an average price call. The
option matures at date 2 and has a terminal value equal to
c2 = max
S1 + S2
5, 0
2
where S1 and S2 are the prices of the stock at dates 1 and 2, respectively. Solve
for the no-arbitrage value of this call option at date 0, c0 .
5. Calculate the price of a three-month American put option on a nondividend-paying stock when the stock price is $60, the strike price is $60,
the annualized, risk-free return is Rf = e0.10 , and the annual standard deviation of the stocks rate of return is = .45, so that u = 1/d = e
.45
. Use a binomial tree with a time interval of 1 month.
6. Let the current date be t and let T > t be a future date, where T t
is the number of periods in the interval. Let A (t) and B (t) be the date
t prices of single shares of assets A and B, respectively. Asset A pays no
dividends but asset B does pay dividends, and the present (date t) value
of asset Bs known dividends per share paid over the interval from t to T
7.5. EXERCISES
227
equals D. The per period risk-free return is assumed to be constant and

equal to Rf .
6.a Consider a type of forward contract that has the following features. At
date t an agreement is made to exchange at date T one share of asset A
for F shares of asset B. No payments between the parties are exchanged
at date t. Note that F is negotiated at date t and can be considered a
forward price. Give an expression for the equilibrium value of this forward
price and explain your reasoning.
6.b Consider a type of European call option that gives the holder the right to
buy one share of asset A in exchange for paying X shares of asset B at
date T .
Give the no-arbitrage lower bound for the date t value of this
call option, c (t).

6.c Derive a put-call parity relation for European options of the type described
in part b.
Chapter 8
Essentials of Diusion
Processes and Its Lemma
This chapter covers the basic properties of continuous time stochastic processes
having continuous sample paths, commonly referred to as diusion processes.
It describes the characteristics of these processes that are helpful for modeling
many financial and economic time series. Modeling a variable as a continuous time, rather than a discrete time, random process can allow for dierent
behavioral assumptions and sharper model results. A variable that follows a
continuous time stochastic process can display constant change yet be observable at each moment in time. In contrast, a discrete time stochastic process
implies that there is no change in the value of the variable over a fixed interval,
or that the change cannot be observed between the discrete dates. If an asset
price is modeled as a discrete time process, it is natural to presume that no trading in the asset occurs over the discrete interval. Often this makes problems
that involve hedging the assets risk dicult, since portfolio allocations cannot
be re-balanced over the non-trading period. Thus, hedging risky asset returns
229
230
CHAPTER 8. DIFFUSION PROCESSES AND ITS LEMMA
may be less than perfect when a discrete time process is assumed.1

Instead, if one assumes that asset prices follow continuous time processes,
prices can be observed and trade can take place continuously. When asset prices
follow continuous sample paths, dynamic trading strategies that can fully hedge
an assets risk are possible. Making this continuous hedging assumption often
simplifies optimal portfolio choice problems and problems of valuing contingent
claims (derivative securities). It permits asset returns to have a continuous
distribution (an infinite number of states), yet market completeness is possible because payos may be dynamically replicated through continuous trading.
Such markets are characterized as dynamically complete.
The mathematics of continuous time stochastic processes can be traced to
Louis Bacheliers 1900 Sorbonne doctoral thesis Theory of Speculation. He developed the mathematics of diusion processes as a by-product of his modeling
of option values. While his work pre-dated Albert Einsteins work on Brownian motion by five years, it fell into obscurity until being uncovered by Leonard
J. Savage and Paul A. Samuelson in the 1950s. Samuelson (Samuelson 1965)
used Bacheliers techniques to develop a precursor of the Black Scholes option
pricing model, but it was Robert C. Merton who pioneered the application of
continuous time mathematics to solve a wide variety of problems in financial
economics.2
The popularity of modeling financial time series by continuous
time processes continues to this day.

This chapters analysis of continuous time processes is done at an intuitive
1 Imperfect hedging may, indeed, be a realistic phenomenon. However, in many situations
it may not be caused by the inability to trade during a period of time but due to discrete
movements (jumps) in asset prices. We examine how to model an asset price process that is
a mixture of a continuous process and a jump process in Chapter 11. Imperfect hedging can
also arise because transactions costs lead an individual to choose not to trade to hedge small
price movements. For models of portfolio choice in the presence of transactions costs, see
work by George Constantinides (Constantinides 1986) and Bernard Dumas and Elisa Luciano
(Dumas and Luciano 1991).
2 See a collection of Mertons work in (Merton 1992).
8.1. PURE BROWNIAN MOTION
231
level rather than a mathematically rigorous one.3 The first section examines
Brownian motion, which is the fundamental building block of diusion processes.
We show how Brownian motion is a continuous time limit of a discrete time
random walk. How diusion processes can be developed by generalizing a pure
Brownian motion process is the topic of the second section. The last section
introduces Its Lemma, which tells us how to derive the stochastic process for
a function of a variable that follows a diusion process. Its Lemma is applied
extensively in continuous time financial modeling.
It will be used frequently
during the remainder of this book.
8.1
Pure Brownian Motion
Here we show how a Brownian motion process can be defined as the limit of
a discrete time process.4 Consider the following stochastic process observed at
date t, z(t). Let t be a discrete change in time, that is, some time interval.
The change in z(t) over the time interval t is given by
z(t + t) z(t) z =
(8.1)
where is a random variable with E[ ] = 0, V ar[ ] = 1, and Cov[ z(t + t)

z(t), z(s + t) z(s) ] = 0 if (t, t + t) and (s, s + t) are non-overlapping
intervals. z(t) is an example of a random walk process. Its standard deviation
equals the square root of the time between observations.
Given the moments of , we have E[z] = 0, V ar[z] = t, and z(t) has
3 The following books (in order of increasing rigor and diculty) provide more in-depth
coverage of the chapters topics: (Neftci 1996),(Karlin and Taylor 1975),(Karlin and Taylor
1981), and (Karatzas and Shreve 1991).
4 Brownian motion is named after botanist Robert Brown, who in 1827 observed that pollen
suspended in a liquid moved in a continuous, random fashion. In 1905, Albert Einstein
explained this motion as the result of random collisions of water molecules with the pollen
particle.
232
serially uncorrelated (independent) increments. Now consider the change in z(t)

over a fixed interval, from 0 to T . Assume T is made up of n intervals of length
t. Then
z(T ) z(0) =
where zi z(i t) z( [i 1] t)
n
X
zi
(8.2)
i=1
t i , and i is the value of over
the ith interval. Hence (8.2) can also be written as
z(T ) z(0) =
n
n
X
X
t i = t
i .
i=1
(8.3)
i=1
Now note that the first two moments of z(T ) z(0) are
E0 [ z(T ) z(0) ] =
V ar0 [ z(T ) z(0) ]
n
X
E0 [ i ] = 0
(8.4)
i=1
n
2 X
t
V ar0 [i ] = t n 1 = T (8.5)
i=1
where Et [] and V art [] are the mean and variance operators, respectively, conditional on information at date t. We see that holding T , the length of the time
interval, fixed, the mean and variance of z(T ) z(0) are independent of n.
8.1.1
The Continuous Time Limit
Now let us perform the following experiment. Suppose we keep T fixed but let
n, the number of intervening increments of length t, go to infinity. Can we
say something else about the distribution of z(T ) z(0) besides what its first
two moments are? The answer is yes. Assuming that the i are independent
8.1. PURE BROWNIAN MOTION
233
and identically distributed, we can state
p lim (z(T ) z(0)) = p lim (z(T ) z(0)) N (0, T ).
(8.6)
t0
In other words, z(T ) z(0) has a normal distribution with mean zero and
variance T .
This follows from the Central Limit Theorem which states that
the sum of n independent, identically distributed random variables has a distribution that converges to the normal distribution as n .
Thus, the
distribution of z(t) over any finite interval, [ 0, T ], can be thought of as the
sum of infinitely many small independent increments, zi = t i , which are

realizations from an arbitrary distribution.
However, when added together,
these increments result in a normal distribution.
Therefore, without loss of
generality, we can assume that each of the i have a standard (mean 0 variance
1) normal distribution.5
The limit of one of these minute independent increments can be defined as
dz(t) lim z = lim

t0
t0
(8.7)
where N (0, 1). Hence, E[ dz(t) ] = 0 and V ar[ dz(t) ] = dt.6 dz is referred
to as a pure Brownian motion process or a Wiener process, named after the
mathematician Norbert Wiener who in 1923 first proved its existence. We can
now write the change in z(t) over any finite interval [ 0, T ] as
z(T ) z(0) =
dz(t) N (0, T ).
(8.8)
The integral in (8.8) is a stochastic or It integral, not the usual Riemann or

5 Note that sums of normally distributed random variables are also normally distributed.
Thus, the Central Limit Theorem also applies to sums of normals.
6 That the V ar [dz (t)] = dt can be confirmed by noting that the sum of the variance over
U
the interval from 0 to T is 0T dt = T .
234
Figure 8.1: Random Walk and Brownian Motion

Lebesgue integrals that measure the area under deterministic functions.7 Note
that z(t) is a continuous process but constantly changing (by over each infinitesimal interval t), such that over any finite interval it has unbounded variation.8 Hence, it is nowhere dierentiable (very jagged), that is, its derivative
dz(t)/dt does not exist.
The step function in Figure 8.1 illustrates a sample path for z (t) as a discrete
time random walk process with T = 2 and n = 20, so that t = 0.1.
As n
, so that t 0, this random walk process becomes the continuous time

Brownian motion process also shown in the figure.
Brownian motion provides the basis for more general continuous time stochastic processes. We next analyze such processes known as diusion processes.
Diusion processes are widely used in financial economics and are characterized
as continuous-time Markov processes having continuous sample paths.9
7 Kiyoshi It was a Japanese mathematician who developed the calculus of stochastic
processes (It 1944), (It 1951).
8 This means that if you measured the length of the continuous processs path over a finite
interval, it would be infinitely long.
9 A stochastic process is said to be Markov if the date t probability distribution of its future
date T > t value depends only on the processs date t value, and not values at prior dates
8.2. DIFFUSION PROCESSES
8.2
235
Diusion Processes
To illustrate how we can build on the basic Wiener process, consider the process
for dz multiplied by a constant, . Define a new process x(t) by
dx(t) = dz(t)
(8.9)
Then over a discrete interval, [0, T ], x(t) is distributed
x(T ) x(0) =
dx =
dz(t) =
dz(t) N(0, 2 T ).
(8.10)
Next, consider adding a deterministic (non-stochastic) change of (t) per unit

of time to the x(t) process.
dx = (t)dt + dz
(8.11)
Now over any discrete interval, [0, T ], we have
x(T ) x(0)
dx =
(t)dt +
0
(t)dt +
dz(t)
(8.12)
Z
dz(t) N (
(t)dt, 2 T ).
For example, if (t) = , a constant, then x(T ) x(0) = T +
RT
0
dz(t)
N(T, 2 T ). Thus, we have been able to generalize the standard trend-less

Wiener process to have a non-zero mean as well as any desired variance. The
process dx = dt + dz is referred to as arithmetic Brownian motion.
In general, both and can be time varying.
We permit them to be
functions of calendar time, t, and/or functions of the contemporaneous value

of the random variable, x(t). In this case, the stochastic dierential equation
s < t.
In other words, the processs future states are conditionally independent of past
states, given information on its current state.
236
describing x(t) is
dx(t) = [x(t), t] dt + [x(t), t] dz
(8.13)
and is a continuous time Markov process, by which we mean that the instantaneous change in the process at date t has a distribution that depends only on
t and the current level of the state variable x (t), and not prior values of the
x (s), for s < t. The function [x(t), t] which denotes the processs instantaneous expected change per unit time is referred to as processs drift while the
instantaneous standard deviation per unit time, [x(t), t], is also described as
the processs volatility.
The process in (8.13) can also be written in terms of its corresponding integral equation
x(T ) x(0) =
dx =
[x(t), t] dt +
[x(t), t] dz
(8.14)
In this general case, dx(t) could be described as being instantaneously normally

distributed with mean [x(t), t] dt and variance 2 [x(t), t] dt, but over any finite
interval, x(t) will not, in general, be normally distributed. One needs to know
the functional form of [x(t), t] and [x(t), t] to determine the discrete time
distribution of x(t) implied by its continuous time process. Shortly, we will
show how this discrete time probability can be derived.
8.2.1
Definition of an It Integral
An It integral is formally defined as a mean square limit of a sum involving the

discrete zi processes. For example, when [x(t), t] is a function of x (t) and t,
RT
the It integral in (8.14), 0 [x(t), t] dz, is defined from the relationship
8.2. DIFFUSION PROCESSES
237
Z
n
X
lim E0
[x ([i 1] t) , [i 1] t] zi
i=1
!2
[x(t), t] dz = 0
(8.15)
where we see that within the parentheses of (8.15) is the dierence between the
It integral and its discrete-time approximation. An important It integral
RT
that will be used below is 0 [dz (t)]2 . In this case, (8.15) gives its definition as
Z
n
X
lim E0
[zi ]2
i=1
To better understand the properties of
V ar0 [z (T ) z (0)] = V ar0

= E0
"
RT
0
" n
X
i=1
n
X
i=1
!2
[dz (t)]2 = 0
(8.16)
[dz (t)]2 , recall from (8.5) that
zi
!2
n
X
zi
= E0
i=1
[zi ]2 = T
because increments of z are serially uncorrelated.
(8.17)
Further, straightforward
algebra shows that10
!2
n
X
2
E0
[zi ] T = 2T t
(8.18)
i=1
Hence, taking the limit as t 0, or n , of the expression in (8.18), one

obtains
k
l
calculation uses the result that E0 (zi )2 (zj )2 = (t) (t) = (t)2 for i 6= j
k
l
and E0 (zi )4 = 3 (t)2 because the fourth moment of a normally distributed random
variable equals 3 times its squared variance.
10 This
238
!2
n
X
[zi ]2 T = lim 2T t = 0
lim E0
(8.19)
t0
i=1
Comparing (8.16) with (8.19) implies that in the sense of mean-square convergence, we have the equality
[dz (t)]2
= T
(8.20)
T
dt
Since
RT
0
[dz (t)]2 converges to
RT
0
dt for any T , we can see that over an infini-
tesimally short time period [dz (t)]2 converges to dt.

To further generalize continuous-time processes, suppose that we have some
variable, F , that is a function of the current value of a diusion process, x(t),
and (possibly) also is a direct function of time. Can we then characterize the
stochastic process followed by F (x(t), t), which now depends on the diusion
process, x(t)? The answer is yes, and Its lemma shows us how to do it.
8.3
Functions of Continuous Time Processes and

Its Lemma
Its Lemma also is known as the fundamental theorem of stochastic calculus.

It gives the rule for finding the dierential of a function of variables that follow
stochastic dierential equations containing Wiener processes.
Here, we state
Its lemma for the case of a function of a single variable that follows a diusion
process.
Its Lemma (univariate case): Let the variable x(t) follow the stochastic
8.3. ITS LEMMA
239
dierential equation dx(t) = (x, t) dt + (x, t) dz.
Also let F (x(t), t) be at
least a twice dierentiable function. Then the dierential of F (x, t) is given by:
dF =
F
1 2F
F
dx +
dt +
(dx)2
x
t
2 x2
(8.21)
where the product (dx)2 = (x, t)2 dt. Hence, substituting in for dx and (dx)2 ,
the above can be re-written:
dF =
F
1 2F 2
F
F
(x, t) +
+
(x, t) dz.
(x,
t)
dt +
x
t
2 x2
x
(8.22)
Proof: A formal proof is rather lengthy and only a brief, intuitive outline of
a proof is given here.11
Let us first expand F (x(t + t), t + t) in a Taylor
series around date t and the value of x at date t:

F
F
1 2F
x +
t +
(x)2
x
t
2 x2
1 2F
2F
xt +
+
(t)2 + H
(8.23)
xt
2 t2
F (x(t + t), t + t) = F (x (t) , t) +
where x x(t+t)x (t) and H refers to terms that are multiplied by higher
orders of x and t. Now a discrete time approximation of x can be written
as
x = (x, t) t + (x, t) t
(8.24)
Defining F F (x(t + t), t + t) F (x (t) , t) and substituting (8.24) in for
11 For more details, see Chapter 3 of Merton (Merton 1992), Chapter 16 of Ingersoll (Ingersoll
1987) or Chapter 10 in Neftci (Neftci 1996). A rigorous proof is given in Karatzas and Shreve
(Karatzas and Shreve 1991).
240
x, equation (8.23) can be re-written as
F
(x, t) t + (x, t) t +
t
x
t
2
1 2F
+
(x,
t)
t
+
(x,
t)
t
(8.25)
2 x2
2
2
F
1 F
2
(x, t) t + (x, t) t t +
+
(t) + H
xt
2 t2
The final step is to consider the limit of equation (8.25) as t becomes infini
tesimal, that is, t dt and F dF . Recall from (8.7) that t becomes
i h
i
h
t
t becomes [dz (t)]2 and converges to dt.
dz and from (8.20) that
Furthermore, it can be shown that all terms of the form (t)n where n > 1 go to
3
zero as t dt. Hence, terms that are multiplied by (t) 2 , (t)2 , (t) 2 , ...,
including all of the terms in H, vanish.
The result is equation (8.22). Similar
arguments show that12
(dx)2
= ((x, t) dt + (x, t) dz)2

=
(8.26)
(x, t)2 ( dz)2 = (x, t)2 dt
Note from (8.22) that the dF process is similar to the dx process in that
both depend on the same Brownian motion dz.
Thus, while dF will have a
mean (drift) and variance (volatility) that diers from dx, they both depend on
the same source of uncertainty.
it may be helpful to remember that in the continuous-time limit (dz)2 = dt, but
dzdt = 0, and dtn = 0 for n >1. This follows from thinking of the discrete approximation
of dz as being proportional to t, and any product that results in (t)n will go to zero as
t dt when n is strictly greater than 1.
12 Thus,
8.3. ITS LEMMA
8.3.1
241
Geometric Brownian Motion
A process that is used in many applications is the geometric Brownian motion

process. It is given by
dx = x dt + x dz
(8.27)
where and are constants. It is an attractive process because if x starts

at a positive value, it always remains positive. This is because its mean and
variance are both proportional to its current value, x. Hence a process like dx
is often used to model the price of a limited-liability security, such as a common
stock. Now consider the following function F (x, t) = ln(x). For example, if x
is a securitys price, then dF = d (ln x) represents this securitys continuouslycompounded return. What type of process does dF = d (ln x) follow? Applying
Its lemma, we have
dF
= d (ln x) =
(ln x)
1 2 (ln x)
(ln x)
2
x +
+
(x) dt
x
t
2 x2
(ln x)
x dz
+
x
1 2
=
+ 0 dt + dz.
2
(8.28)
Thus we see that if x follows geometric Brownian motion, then F = ln x

follows arithmetic Brownian motion. Since we know that
1
F (T ) F (0) N ( 2 ) T, 2 T
2
(8.29)
then x(t) = eF (t) has a lognormal distribution over any discrete interval (by the
definition of a lognormal random variable). Hence, geometric Brownian motion
is lognormally distributed over any time interval.
Figure 8.2 illustrates 300 simulated sample paths of geometric Brownian
242
Figure 8.2: Geometric Brownian Motion Sample Paths

motion for = 0.10 and = 0.30 over the period from t = 0 to 2, with x (0) = 1 .
These drift and volatility values are typical for a U.S. common stock. As can
be seen, the sample paths determine a frequency distribution at T = 2 which is
skewed upward, as it should be since the discrete time distribution is lognormal
and bounded at zero.
8.3.2
Kolmogorov Equation
There are many instances where knowledge of a diusion processs discrete time
probability distribution is very useful.
As we shall see in future chapters,
valuing a contingent claim often entails computing an expected value of its

discounted terminal payo at a specific future date. This discounted terminal
payo frequently depends on the value of a diusion process, so that computing
its expected value requires knowledge of the processs discrete time probability
distribution. Another situation where it is helpful to know a diusion processs
discrete-time distribution occurs when one wishes to estimate the processs drift
8.3. ITS LEMMA

and volatility parameters using time series data.
243
Because time series data is
typically sampled discretely, rather than continuously, empirical techniques such

as maximum likelihood estimation often require use of the processs discrete time
distribution.
A general method for finding the implied discrete time probability distribution for a continuous time process is to use the backward Kolmogorov equation.
A heuristic derivation of this condition is as follows. Let x (t) follow the general diusion process given by (8.13). Also let p (x, T ; xt , t) be the probability
density function for x at date T given that it equals xt at date t, where T t.
Applying Its lemma to this density function, one obtains13
dp =
1 2p 2
p
p
p
+
(xt , t) +
(xt , t) dt +
(xt , t) dz
xt
t
2 x2t
xt
(8.30)
Intuitively, one can see that only new information that was unexpected at date
t should change the probability density of x at date T. In other words, for small
< T t, E [p (x, T ; xt+ , t + ) |x (t) = xt ] = p (x, T ; xt , t).14 This implies
that the expected change in p should be zero, that is, the drift term in (8.30)
should be zero:
2p
p
p
1 2
(xt , t) 2 + [xt , t]
=0
+
2
xt
xt
t
(8.31)
Condition (8.31) is referred to as the backward Kolmogorov equation. This partial dierential equation for p (x, T ; xt , t) can be solved subject to the boundary
condition that when t becomes equal to T , then x must equal xt with probability
1. Formally, this boundary condition can be written as p (x, t; xt , t) = (x xt ),
13 In order to invoke Its lemma, we assume that the density function p (x, T, x , t) is diert
entiable in t and twice dierentiable in xt . Under particular conditions, the dierentiability
of p can be proved, but this issue will not be dealt with here.
14 Essentially, this result derives from the Law of Iterated Expectations.
244
where () is the Dirac delta function, which is defined as (0) = , (y) = 0

R
for all y 6= 0, and (y) dy = 1.
For example, recall that if xt follows geometric Brownian motion, then
[xt , t] = xt and 2 (xt , t) = 2 x2t where and are constants. In this

case the Kolmogorov equation becomes
p
p
1 2 2 2p
xt 2 + xt
=0
+
2
xt
xt
t
(8.32)
By substitution into (8.32), it can be verified that the solution to this partial dierential equation subject to the boundary condition that p (x, t; xt , t) =
(x xt ) is15
"
2 #
ln x ln xt 12 2 (T t)
1
p (x, T, xt , t) = p
exp
22 (T t)
x 22 (T t)
(8.33)
which is the lognormal probability density function for the random variable
x (0, ).
Hence, the backward Kolmogorov equation verifies that a vari-
able following geometric Brownian motion is lognormally distributed.
For a
diusion process with general drift and volatility functions, (x, t) and (x, t),
it may not be easy or possible to find a closed-form expression solution for
p (x, T, xt , t) such as in (8.33). Still, there are a number of instances where the
Kolmogorov equation is valuable in deriving or verifying a diusions discrete
time distribution.16
15 Methods for solving partial dierential equations are beyond the scope of this course.
However, if one makes the change in variable yt = ln (xt ), then equation (8.32) can be transformed to a more simple partial dierential equation with constant coecients. Its solution
is the probability density function of a normally distributed random variable. Reversing the
change in variables to xt = eyt results in the lognormal density function.
16 Andrew Lo (Lo 1988) provides additional examples where the backward Kolmogorov equation is used to derive discrete time distributions. These examples include not only diusion
processes, but the type of mixed jump-diusion processes that we will examine in Chapter 11.
8.3. ITS LEMMA
8.3.3
245
Multi-Variate Diusions and Its Lemma
In a number of portfolio choice and asset pricing applications that we will encounter in future chapters, one needs to derive the stochastic process for a function of several variables, each of which follows a diusion process. So suppose
we have m dierent diusion processes of the form:17
dxi = i dt + i dzi
i = 1, . . . , m,
(8.34)
and dzi dzj = ij dt, where ij has the interpretation of a correlation coecient
of the two Wiener processes. What is meant by this correlation? Recall that
dzi dzi = (dzi )2 = dt. Now the Wiener process dzj can be written as a linear
combination of two other Wiener processes, one being dzi , and another process
that is uncorrelated with dzi , call it dziu :
dzj = ij dzi +
q
1 2ij dziu
(8.35)
Then from this interpretation of dzj , we have
dzj dzj
= 2ij (dzi )2 + 1 2ij (dziu )2 + 2ij 1 2ij dzi dziu (8.36)
= 2ij dt + 1 2ij dt + 0
= dt
and
17 Note and may be functions of calendar time, t, and the current values of x , j =
i
j
i
1, ..., m.
246
dzi dzj
= dzi ij dzi + 1 2ij dziu

q
= ij (dzi )2 + 1 2ij dzi dziu
(8.37)
= ij dt + 0
Thus, ij can be interpreted as the proportion of dzj that is perfectly correlated

with dzi .
We can now state, without proof, a multivariate version of Its Lemma.
Its Lemma (multivariate version): Let F (x1 , . . . , xm , t) be at least a twice

dierentiable function. Then the dierential of F (x1 , . . . , xm , t) is given by:
dF =
m
m m
X
1 X X 2F
F
F
dt +
dxi +
dxi dxj
xi
t
2 i=1 j=1 xi xj
i=1
(8.38)
where dxi dxj = i j ij dt. Hence, the above can be re-written
dF
m
m X
m
2
2
X
X
1
F
F
F
F
+
i +
i j ij dt
+
2 i
x
2
x
t
xi xj
i
i
i=1
i=1 j>i
+
m
X
F
i dzi .
xi
i=1
(8.39)
Equation (8.39) generalizes our earlier statement of Its lemma for a univariate
diusion, equation (8.22).
Notably, we see that the process followed by a
function of several diusion processes inherits each of the processes Brownian

motions.
8.4
Summary
8.4. SUMMARY
247
Diusion processes and Its Lemma are important tools for modeling financial
time series, especially when individuals are assumed to be able to trade continuously.
Brownian motion is the foundation of diusion processes and is a
continuous time limit of a particular discrete time random walk process. By

modifying Brownian motions instantaneous mean and variance, a wide variety
of diusion processes can be created. Its Lemma tells us how to find the differential of a function of a diusion process. As we shall see in the next chapter,
Its Lemma is essential for valuing a contingent claim when its payo depends
on the price of an underlying asset that follows a diusion. This is because the
contingent claims value becomes a function of the underlyings assets value.
This chapter also showed that Its Lemma could be used to derive the
Kolmogorov equation, an important relation for finding the discrete time distribution of a random variable that follows a diusion process.
Finally, we
saw that multivariate diusions are natural extensions of univariate ones, and
the process followed by a function of several diusions can be derived from a
multi-variate version of Its Lemma.
8.5
Exercises
1. A variable, x (t), follows the process
dx = dt + dz
where and are constants.
Find the process followed by y (t) =
ex(t)t .
2. Let P be a price index, such as the Consumer Price Index (CPI). Let M
equal the nominal supply (stock) of money in the economy. For example,
M might be designated as the amount of bank deposits and currency in
248

circulation. Assume P and M each follow geometric Brownian motion
processes
dP
P
dM
M
= p dt + p dzp
= m dt + m dzm
with dzp dzm = dt. Monetary economists define real money balances, m, to be
m=
M
P .
Derive the stochastic process for m.
3. The value (price) of a portfolio of stocks, S(t), follows a geometric Brownian motion process
dS/S = s dt + s dzs
while the dividend yield for this portfolio, y(t), follows the process
dy = (S y) dt + y y 2 dzy
where dzs dzy = dt and , , and y are positive constants. Solve for the
process followed by the portfolios dividends paid per unit time, D(t) = yS.
4. The Ornstein-Uhlenbeck process can be useful for modeling a time series
whose value changes stochastically but which tends to revert to a longrun value (its unconditional or steady state mean). This continuous time
process is given by
dy(t) = [ y(t)] dt + dz(t)
8.5. EXERCISES
249
The process is sometimes referred to as an elastic random walk. y(t) varies

stochastically around its unconditional mean of /, and is a measure of the
strength of the variables reversion to this mean.
Find the distribution of y(t) given y(t0 ), where t > t0 . In particular, find
E [y(t) | y(t0 )] and V ar [y(t) | y(t0 )]. Hint: make the change in variables
(tt0 )
x(t) = y(t)
e
and apply Its lemma to find the stochastic process for x(t). The distribution and first two moments of x(t) should be obvious. From this, derive the
distribution and moments of y(t).
Chapter 9
Dynamic Hedging and PDE

Valuation
Having introduced diusion processes and Its lemma in the previous chapter,
we now apply these tools to derive the equilibrium prices of contingent claims.
In this chapter, asset prices are modeled as following diusion processes. Because prices are permitted to vary continuously, it is feasible to also assume
that individuals can choose to trade assets continuously. With the additional
assumption that markets are "frictionless," this environment can allow the markets for a contingent claim, its underlying asset, and the risk-free asset to be
dynamically complete.1
Though the returns of the underlying asset and its
contingent claim have a continuous distribution over any finite time interval,
implying an infinite number of states for their future values, the future values of
the contingent claim can be replicated by a dynamic trading strategy involving
its underlying asset and the risk-free asset.
1 Frictionless markets are characterized as having no direct trading costs or restrictions,
that is, markets for which there are no transactions costs, taxes, short sales restrictions, or
indivisibilities when trading assets.
251
252
CHAPTER 9. DYNAMIC HEDGING AND PDE VALUATION
By way of three examples, we illustrate the Black-Scholes-Merton portfolio

hedging argument that results in a partial dierential equation (PDE) for a contingent claims price. Solving this PDE subject to the appropriate boundary
condition determines a unique price for the contingent security. Our first example is the well-known Fischer Black - Myron Scholes (Black and Scholes 1973)
option pricing model. The second is the equilibrium term structure model of
Oldrich Vasicek (Vasicek 1977).
first two.
The final example combines aspects of the
It is Robert Mertons (Merton 1973b) option pricing model with
stochastic interest rates.

As the next chapter will show, contingent claims prices also can be derived
using alternative solution techniques: the martingale pricing approach which
involves computing expectations of a risk-neutral probability distribution; and
the stochastic discount factor (pricing kernel) approach where expectations are
computed for the physical probability distribution. In some situations, it may
be easier to derive contingent claims prices by solving the equilibrium PDE. In
others, the martingale technique or stochastic discount factor approach may be
simplest. All of these methods should be in a financial economists toolbox.
9.1
Black-Scholes Option Pricing
The major insight of Black and Scholes (Black and Scholes 1973) is that when assets follow diusion processes, an options payo can be replicated by continuous
trading in its underlying asset and a risk-free asset. In the absence of arbitrage,
the ability to replicate or "hedge" the option with the underlying stock and a
risk-free asset restricts the options value to bear a particular relationship to its
underlying asset and the risk-free return. The Black-Scholes hedging argument
is similar to that presented earlier in the context of the binomial option pricing model.
The main dierence is that the appropriate replicating portfolio
9.1. BLACK-SCHOLES OPTION PRICING
253
changed only once per period in the binomial model, while in the Black-Scholes
environment the replicating portfolio changes continuously.
In the binomial
model, market completion resulted from the assumption that at the end of each
period there was only two states for the underlying assets value. Under BlackScholes assumptions, markets become dynamically complete due to the ability
to trade continuously in the underlying asset whose price follows a continuous
sample path.
9.1.1
Portfolio Dynamics in Continuous Time
A prerequisite for analyzing the Black-Scholes hedging of contingent claims is

to consider the dynamics of a security portfolio in continuous time. The BlackScholes hedge portfolio consists of a position in the contingent claim and its
underlying asset, but we will begin by examining the general problem of an
investor who can trade in any n dierent assets whose prices follow diusion
processes. Let us define Si (t) as the price per share of asset i at date t, where
i = 1, ..., n.
The instantaneous rate of return on the ith asset is assumed to
satisfy the process

dSi (t) / Si (t) = i dt + i dzi
(9.1)
where its instantaneous expected return and variance, i and 2i , may be functions of time and possibly other asset prices or state variables that follow diffusion processes. For simplicity, assets are assumed to pay no cashflows (dividends or coupon payments), so that their total returns are given by their price
changes.2
An investor is assumed to form a portfolio of these assets and, in
general, the portfolio may experience cash inflows and outflows. Thus, let F (t)
2 This is not a critical assumption. What matters is the assets expected rates of return and
covariances, rather than their price changes, per se. If an asset, such as a common stock or
mutual fund, paid a dividend that was re-invested into new shares of the asset, then equation
(9.1) would represent the percentage change in the value of the asset holding and, thus, the
total rate of return.
254
be the net cash outflow per unit time from the portfolio at date t. For example, F (t) may be positive because the individual chooses to liquidate some of
the portfolio to pay for consumption expenditures.
Alternatively, F (t) may
be negative because the individual receives wage income that is invested in the
securities.
To derive the proper continuous time dynamics for this investors portfolio,
we will first consider the analogous discrete time dynamics where each discrete
period is of length h.
We will then take the limit as h 0.
Therefore, let
wi (t) be the number of shares held by the investor in asset i from date t to t + h.
The value of the portfolio at the beginning of date t is denoted as H (t) and
equals the prior periods holdings at date t prices:
H (t) =
n
X
i=1
wi (t h)Si (t)
(9.2)
Given these date t prices, the individual may chose to liquidate some of the
portfolio or augment it with new funds. The net cash outflow over the period
is F (t) h, which must equal the net sales of assets. Note that F (t) should be
interpreted as the average liquidation rate over the interval from t to t + h.
F (t) h =
n
X
[wi (t) wi (t h)] Si (t)
(9.3)
i=1
To properly derive the limits of (9.2) and (9.3) as of date t and as h 0,

we need to convert backward dierences, such as wi (t) wi (t h), to forward
dierences. We do this by updating one period, so that at the start of the next
255
period, t + h, we have:
F (t + h) h =
n
X
[wi (t + h) wi (t)] Si (t + h)
i=1
n
X
=
[wi (t + h) wi (t)] [Si (t + h) Si (t)]
i=1
n
X
[wi (t + h) wi (t)] Si (t)
(9.4)
i=1
and
H (t + h) =
n
X
wi (t) Si (t + h)
(9.5)
i=1
Taking the limits of (9.4) and (9.5) as h 0 gives the results

F (t) dt =
n
X
dwi (t) dSi (t) +
i=1
n
X
dwi (t) Si (t)
(9.6)
i=1
and
H (t) =
n
X
wi (t) Si (t)
(9.7)
i=1
Applying Its Lemma to (9.7), we can derive the dynamics of the portfolios
value to be
dH (t) =
n
X
wi (t) dSi (t) +
i=1
n
X
dwi (t)Si (t) +
i=1
n
X
dwi (t) dSi (t)
(9.8)
i=1
Substituting (9.6) into (9.8), we obtain:
dH (t) =
n
X
i=1
wi (t) dSi (t) F (t) dt
(9.9)
Equation (9.9) says that the portfolios value changes due to capital gains income
256
less net cash outflows. Substituting (9.1) into (9.9) we arrive at
dH (t)
n
X
=
=
i=1
n
X
i=1
wi (t) dSi (t) F (t) dt
(9.10)
wi (t) [i Si dt + i Si dzi ] F (t) dt
Now, in some cases, rather than write a portfolios dynamics in terms of the
number of shares of each asset, wi (t), i = 1, . . . , n, we may wish to write it in
terms of each assets proportion of the total portfolio value. If we define the
proportion of H (t) invested in asset i as i (t) = wi (t)Si (t)/H (t), then (9.10)
becomes
dH (t) =
n
X
i=1
i (t) H (t) [i dt + i dzi ] F (t) dt
(9.11)
or
dH (t) =
"
n
X
i=1
i (t) H (t) i F (t) dt +
Note from (9.7) that
Pn
i=1 i
n
X
i (t) H (t) i dzi
(9.12)
i=1
(t) = 1, that is, the portfolio proportions
invested in the n risky assets must sum to 1. However, consider the introduction
of a new risk-free asset. If, in addition to n risky assets, there is an asset that
pays an instantaneously risk-free rate of return, then this would correspond to
an asset having an instantaneous standard deviation, i , of zero and an expected
rate of return, i , equal to the instantaneous risk-free rate, which we denote as
r (t). In this case, the portfolio proportion invested in the risk-free asset equals
Pn
1 i=1 i (t). With this extension, equation (9.12) becomes
dH (t) =
"
n
X
i=1
i (t) (i r) H (t) i + rH (t) F (t) dt +
n
X
i (t) H (t) i dzi
i=1
(9.13)
Having derived the continuous time dynamics of an investment portfolio, we
257
now turn to the Black-Scholes approach to valuing contingent claims.
9.1.2
Black-Scholes Model Assumptions
The Black-Scholes model assumes that there is a contingent claim whose underlying asset pays no dividends. We will refer to this underlying asset as a stock,
and its date t price per share, S(t), is assumed to follow the diusion process
dS = S dt + S dz.
(9.14)
where the instantaneous expected rate of return on the stock, , may be a

function of S and t, that is, (S, t).
However, the standard deviation of the
stocks rate of return, , is assumed to be constant.
It is also assumed that
there is a risk-free asset that earns a constant rate of return equal to r per unit
time.
Hence, if an amount B (t) is invested in the risk-free asset, this value
follows the process

dB = rBdt
(9.15)
Now consider a European call option on this stock that matures at date T and
has an exercise price of X. Denote the options date t value as c(S, t).
We
assume it is a function of both calendar time, t, and the current stock price,
S (t), since at the maturity date t = T , the options payo depends on S (T ):
c(S(T ), T ) = max[ 0, S(T ) X].
(9.16)
Given that the options value depends on the stock price and calendar time,
what process does it follow prior to maturity?3
Let us assume that c (S, t)
is a twice dierentiable function of S and is dierentiable in t. Later, we will

3 The options value also depends on the risk-free rate, r, but since r is assumed to be
constant, it need not be an explicit argument of the options value.
258
verify that the no-arbitrage value of c (S, t) does, indeed, satisfy these conditions.
Then we can apply Its lemma to state that the options value must follow a
process of the form
dc =
c
1 2c 2 2
c
c
S +
+
S dz.
S
dt +
2
S
t
2 S
S
(9.17)
Hence, the call option inherits the same source of risk as the underlying stock,
reflected in the Wiener process dz.
9.1.3
The Hedge Portfolio
Now consider forming a portfolio that includes 1 unit of the option and a
position in the underlying stock and the risk-free asset. Such a portfolio would
reflect the wealth position of an option dealer who has just sold one call option
to a customer and now attempts to hedge this liability by purchasing some of the
underlying stock and investing or borrowing at the risk-free rate. We restrict
this portfolio to require zero net investment, that is, after selling one unit of the
call option and taking a hedge position in the underlying stock, the remaining
surplus or deficit of funds is made up by borrowing or lending at the risk-free
rate. Moreover, we require that the portfolio be self-financing, that is, F (t) = 0
t, by which we mean that any surplus or deficit of funds from the option and
stock positions are made up by investing or acquiring funds at the risk-free rate.
Hence, if we let w(t) be the number of shares invested in the stock, then this
zero net investment, self-financing restriction implies that the amount invested
in the risk-free asset for all dates t must be B (t) = c (t)w (t) S (t). Therefore,
denoting the value of this hedge portfolio as H (t) implies that its instantaneous
return satisfies
dH (t) = dc(t) + w (t) dS (t) + [c (t) w (t) S (t)] rdt
(9.18)
259
Substituting (9.14) and (9.17) into (9.18), we obtain
dH (t) =
c
c
1 2c 2 2
c
S +
+
S dz
S
dt
S
t
2 S 2
S
+w (t) (S dt + S dz) + [c (t) w (t) S (t)] rdt
(9.19)
Now consider selecting the number of shares invested in the stock in such a way
as to oset the risk of the return on the option. Specifically, suppose that the
option dealer chooses w (t) = c/S units (shares) of the stock, which is the
local sensitivity of the options value to the value of the underlying stock, also
known as the "hedge ratio."4 Hence, the hedging portfolio involves maintaining
a unit short position in the option and a position of c/S shares of stock, with
any surplus or deficit of funds required to maintain this hedge being invested
or acquired at the risk-free rate.
As will be verified, since c (S, t) is a non-
linear function of S and t, w (t) = c/S varies continuously over time as S

and t change: the hedge portfolios number of shares invested in the stock is
not constant, but is continuously rebalanced.5 However, as long as a position
of c/S shares of stock are held, we can substitute w (t) = c/S into (9.19)
to obtain
c
1 2c 2 2
c
c
S +
+
S dz
S
dt
S
t
2 S 2
S
c
c
+
(S dt + S dz) + c (t)
S (t) rdt
S
S
2
c
1
c
c
=
2 S 2 2 + rc (t) rS (t)
dt
t
2
S
S
dH (t) =
(9.20)
4 c/S is analogous to the hedge ratio in the binomial option pricing model. Recall that
the optimal choice of this hedge ratio was = (cu cd ) / (uS dS), which is essentially the
same partial derivative.
5 Since c (S, t) is yet to be determined, the question arises as to how w (t) = c/S would
be known to create the hedge portfolio. We will verify that if such a position in the stock
is maintained, then a no-arbitrage value for the option, c (S, t), is determined which, in turn,
makes known the hedge ratio w (t) = c/S.
260
Note that, by design, the return on this portfolio is instantaneously riskless.

Not only do the dz terms in the first line of (9.20) drop out, but so do the terms
that depend on the stocks drift, . By continually re-adjusting the number
of shares held in the stock so that it always equals c/S, the risk of the
option is perfectly hedged. Dynamic trading in the stock is able to replicate
the risk of the option because both the option and stock depend on the same
(continuous-time) Brownian motion process, dz.
In this sense, when assets
follow continuous-time stochastic processes, dynamic (continuous) trading can

lead to a complete market and price contingent claims.
9.1.4
No-Arbitrage Implies a PDE
Since the rate of return on this hedge portfolio is riskless, to avoid arbitrage
it must equal the competitive risk-free rate of return, r. But since we restricted
the hedge portfolio to require zero-net investment at the initial date, say t = 0,
then H (0) = 0 and
dH (0) = rH (0) dt = r0dt = 0
(9.21)
This implies H (t) = 0 t so that dH (t) = 0 t. This no-arbitrage condition

along with (9.20) allows us to write
1
c
2c
c
+ 2S 2 2 + r S
r c = 0.
t
2
S
S
(9.22)
which is the Black-Scholes partial dierential equation. The call options value
must satisfy this partial dierential equation subject to the boundary condition
c(S(T ), T ) = max[ 0, S(T ) X]
(9.23)
261
The solution to (9.22) and (9.23) is6
c(S(t), t) = S(t) N (d1 ) X er (T t) N(d2 )
(9.24)
where
d1
d2
ln (S(t)/X) + r + 12 2 (T t)
T t
d1 T t.
(9.25)
and N () is the standard normal distribution function. Similar to the binomial

option pricing formula, the value of the call option does not depend on the
stocks expected rate of return, , only its current price, S(t), and volatility, .
The value of a European put option follows immediately from put - call parity:7
p(S(t), t) = c(S(t), t) + X er (T t) S(t)
(9.26)
= S(t) N (d1 ) X er (T t) N (d2 ) + X er (T t) S(t)

= X er (T t) N (d2 ) S(t)N (d1 )
By taking the partial derivatives of (9.24) and (9.26) with respect to S (t), the
call and put options hedge ratios are shown to be8
6 The solution can be derived using a separation of variables method (Churchill and Brown
1978) or a LaPlace transform method (Shimko 1992). Also, in Chapter 10, we will show how
(9.24) can be derived using risk-neutral valuation.
7 The last line uses the symmetry property of the normal distribution 1 N (x) = N (x).
8 Deriving these partial derivatives is more tedious than it might first appear since d and d
1
2
1 Xer(T t) n (d ) d2
are both functions of S (t). Note that c/S = N (d1 ) +Sn (d1 ) d
2 S
S

where n (d) = 1 exp 12 d2 is the standard normal probability density function. This re2
1 = Xer(T t) n (d )
duces to (9.27) because it can be shown that Sn (d1 ) d
2
S
refer to the hedge ratios in (9.27) and (9.28) as the options deltas.
d2
.
S
Practioners
262
c
= N (d1 )
S
(9.27)
p
= N (d1 )
S
(9.28)
which implies 0 <c/S < 1 and 1 < p/S < 0.
Hence, hedging a call
option requires a long position in less than one share of the underlying stock,
while hedging a put option requires a short position in less than one share of
the underlying stock.
Since d1 is an increasing function of S (t), the hedge
portfolio for a call option increases the share amount in the stock as its price
rises.
A similar argument shows that the hedge portfolio for a put option
increases the share amount sold short as the price of the stock falls.
Thus,
because S (t) moves in a continuous fashion, so will the hedge portfolios position
in the stock. Finally, based on the solution in (9.24), we can verify that both
2 c/S 2 and c/t exist, which justifies our use of Its lemma in deriving the
process followed by the options price.9
We now turn to another application of the Black-Scholes-Merton hedging
argument for deriving security prices.
However, rather than derive the price
of a contingent security in terms of an underlying asset price, we next consider

pricing securities that pay known (fixed) cashflows at dierent future dates.
That is, we derive the relationship between the prices of dierent maturity
bonds, also known as fixed-income securities.
This provides an introduction
into the literature on the term structure of interest rates (or bond yields).
9 Using
(9.27) and (9.28), it is easy to see that 2 c/S 2 = 2 p/S 2 =

2
n (d1 ) / S T t > 0 where n (x) = N (x) /x = ex /2 / 2 is the standard normal
probability density function. Hence, both call and put options are convex functions of the
underlying asset price. Practitioners refer to this second derivative as the options gamma.
The larger is an options gamma, the larger is the required change in the hedge ratio for a
given change in the underlying assets price. The options theta or time decay is given by
c/ (T t) = Sn (d1 ) / [2 (T t)] rXer(T t) N (d2 ).
9.2. AN EQUILIBRIUM TERM STRUCTURE MODEL
9.2
263
An Equilibrium Term Structure Model
The previous section showed that in a continuous time environment, the absence
of arbitrage restricts a derivatives price in terms of its underlying assets price.
We now consider a second example of how the absence of arbitrage links security
prices.
When the prices of default-free bonds are assumed to be driven by
continuous time stochastic processes, continuous trading and the no-arbitrage

condition can lead to equilibrium relationships between the prices of dierent
maturity bonds.
The simplest equilibrium bond pricing models assume that
a single source of uncertainty aects bonds of all maturities. For these onefactor bond pricing models, it is often convenient to think of this uncertainty as
being summarized by the yield on the shortest (instantaneous) maturity bond,
r (t).10
This is the assumption we make in presenting the Oldrich Vasicek
(Vasicek 1977) model of the term structure of interest rates.

Define P (t, ) as the date t price of a bond that makes a single payment of
$1 in periods, at date T = t + .
Hence, denotes this "zero-coupon" or
"pure discount" bonds time until maturity. The instantaneous rate of return
on the bond is given by
dP (t, )
P (t, ) .
Also note that, by definition, P (t, 0) = $1.
The instantaneous yield, r (t), is defined as
lim
dP (t, )
r (t) dt
P (t, )
(9.29)
The Vasicek model assumes r (t) follows an Ornstein-Uhlenbeck process
dr(t) = [r r (t)] dt + r dzr

where , r, and r are positive constants.
(9.30)
The parameter r measures the
10 Other approaches to modeling the term structure of interest rates are considered in Chapter 17. For example, we will discuss research by David Heath, Robert Jarrow, and Andrew
Morton (Heath, Jarrow, and Morton 1992) that assumes that forward interest rates of all
maturities are aected by one or more sources of risk.
264
Figure 9.1: Ornstein-Uhlenbeck Interest Rate Process
instantaneous volatility of r (t) while measures the strength of the processs

mean reversion to r, the unconditional mean value of the process.
In dis-
crete time, (9.30) is equivalent to a normally distributed, auto-regressive (1)

process.11 Figure (9.1) illustrates a typical sample path for r (t) that assumes
the annualized parameter values of r (0) = r = 0.05, = 0.3, and r = 0.02.
Now assume that bond prices of all maturities depend on only a single source
of uncertainty, and that this single "factor" is summarized by the current level
of r (t).12
Then, we can write a -maturity bonds price as P (r (t) , ), and
Its lemma implies that it follows the process
11 The
discrete time expected value and variance implied by the continuous time process

2
in (9.30) are Et [r (t + )] = r + e (r (t) r) and V art [r (t + )] = 2r 1 e2 . See
problem 4 at the end of Chapter 8.
12 For example, a central bank may implement monetary policy by changing the level of the
short-term interest rate. Other macro-economic eects on bond prices might be summarized
in the level of the short rate.
P
2P
P
dr +
dt + 12 2 (dr)2
r
t
r
= Pr (r r) + Pt + 12 Prr 2r dt + Pr r dzr
dP (r, ) =
265
(9.31)
= p (r, ) P (r, ) dt p ( ) P (r, ) dzr
where
the subscriptsl on P denote partial derivatives and where p (r, )
k
1
Pr (rr)+Pt + 2 Prr 2r
P (r, )
r r
and p () PP(r,
) are the mean and standard devi-
ation, respectively, of the bonds instantaneous rate of return.13

( )P (r, )
Consider forming a portfolio containing one bond of maturity 1 and pp ( 12 )P (r, 12 )

units of a bond with maturity 2 . In other words, we have a unit long position
in a bond of maturity 1 and a short position in a bond with maturity 2 in
an amount that reflects the ratio of bond 1s return standard deviation to that
of bond 2s.
Since both bonds are driven by the same Wiener process, dzr ,
this portfolio is a hedged position. If we continually re-adjust the amount of

( )P (r, )
the 2 -maturity bonds to equal pp ( 12 )P (r, 12 ) as r (t) changes, the value of this
hedge portfolio, H (t), is
p ( 1 ) P (r, 1 )
P (r, 2 )
H (t) = P (r, 1 )
p ( 2 ) P (r, 2 )
p ( 1 )
= P (r, 1 ) 1
p ( 2 )
(9.32)
Furthermore, the hedge portfolios instantaneous return is
13 We define ( ) P /P (r, ) rather than ( ) P /P (r, ) because it will

p
r r
p
r r
turn out that Pr < 0. Hence, if we want both r and p to denote standard deviations, we
need them to be positive. This choice of definition makes no material dierence since the
instantaneous variance of the change in the interest rate and the bonds rate of return will
always be 2r and 2p , respectively.
266
dH (t) = dP (r, 1 )
p ( 1 ) P (r, 1 )
dP (r, 2 )
p ( 2 ) P (r, 2 )
(9.33)
= p (r, 1 ) P (r, 1 ) dt p ( 1 ) P (r, 1 ) dzr
p ( 1 )
P (r, 1 ) p (r, 2 ) dt + p ( 1 ) P (r, 1 ) dzr
p ( 2 )
p ( 1 )
P (r, 1 ) p (r, 2 ) dt
= p (r, 1 ) P (r, 1 ) dt
p ( 2 )
where the second equality in (9.33) reflects substitution of (9.31).
Since the
portfolio return is riskless at each instant of time, the absence of arbitrage

implies that its rate of return must equal the instantaneous riskless interest
rate, r (t):
p ( 1 )
p (r, 2 ) P (r, 1 ) dt
p (r, 1 )
p ( 2 )
p ( 1 )
= r (t) H (t) dt = r (t) 1
P (r, 1 ) dt
p ( 2 )
dH (t) =
(9.34)
Equating the terms that precede P (r, 1 ) on the first and second lines of (9.34),
we see that an implication of this equation is
p (r, 1 ) r (t)
p (r, 2 ) r (t)
=
p ( 1 )
p ( 2 )
(9.35)
which relates the risk-premiums or Sharpe ratios on the dierent maturity

bonds.
9.2.1
A Bond Risk Premium
Equation (9.35) says that bonds expected rates of return in excess of the instantaneous maturity rate, divided by their standard deviations, must be equal at
267
all points in time. This equality of Sharpe ratios must hold for any set of bonds
1 , 2 , 3 , etc. Each of the dierent bonds reward-to-risk ratios (Sharpe ratios) derives from the single source of risk represented by the dzr process driving
the short-term interest rate, r (t). Hence, condition (9.35) can be interpreted
as a law of one price that requires all bonds to have a uniform market price of
interest rate risk.
To derive the equilibrium prices for bonds, we must specify the form of
this market price of bond risk.
Chapter 13 outlines a general equilibrium
model by John Cox, Jonathan Ingersoll, and Stephen Ross, (Cox, Ingersoll, and
Ross 1985a) and (Cox, Ingersoll, and Ross 1985b), that shows how this bond
risk premium can be derived from individuals preferences (utilities) and the
economys technologies. For now, however, we simply assume that the market
price of bond risk is constant over time and equal to q. Thus, we have for any
bond maturity, ,
p (r, ) r (t)
=q
p ()
(9.36)
p (r, ) = r (t) + qp ()
(9.37)
or
which says that the expected rate of return on a bond with maturity equals
the instantaneous risk-free rate plus a risk premium proportional to the bonds
standard deviation.
Substituting p (r, ) and p ( ) from Its lemma into
(9.37) and simplifying, we obtain
Pr (r r) + Pt + 12 Prr 2r = rP q r Pr
This can be re-written as
(9.38)
268
2r
2 Prr
+ (r + q r r) Pr rP + Pt = 0
(9.39)
Equation (9.39) is the equilibrium partial dierential equation that all bonds
must satisfy. Since T t, so that Pt
P
t
= P
P , equation (9.39)
can be re-written as
2r
2 Prr
+ [ (r r) + q r ] Pr rP P = 0
(9.40)
and solved subject to the boundary condition that at = 0, the bond price
equals $1, that is, P (r, 0) = 1. Doing so, gives the following solution14
P (r (t) , ) = A ( ) eB( )r(t)
(9.41)
where
B ()
1 e
"
(9.42)
2
2
r
2r B ( )
1 r
A () exp (B ( ) ) r + q
2 2
9.2.2
(9.43)
Characteristics of Bond Prices
Using equation (9.41) we see that
p () = r
r
Pr
= r B ( ) =
1 e
P
(9.44)
14 The solution can be derived by "guessing" a solution of the form in (9.41) and substituting
it into (9.40). Noting that the terms multiplied by r (t) and those terms not multiplied by
r (t) must each be zero for all r (t) leads to simple ordinary dierential equations for A ( ) and
B ( ). These equations are solved sub ject to the boundary condition P (r, = 0) = 1, which
implies A ( = 0) = 1 and B ( = 0) = 0. See Chapter 17 for details and a generalization to
bond prices that are influenced by multiple factors.
269
which implies that a bonds rate of return standard deviation (volatility) is an

increasing but concave function of its maturity, . Moreover, (9.44) confirms
as the bond approaches its maturity date, its price volatility shrinks to zero,
p ( = 0) = 0, since the instantaneous maturity bonds return is riskless. The
tendency for price volatility to decrease over time is a fundamental property of
finitely-lived, fixed-income securities that distinguishes them from potentially
infinitely-lived securities such as common or preferred stocks. While it may be
reasonable to assume as in (9.14) that the volatility of a stocks price need not
be a function of calendar time, this cannot be the case for a zero-coupon bond.
Given that p ( ) is an increasing function maturity, equation (9.37) says
that a bonds expected rate of return increases (decreases) with its time until
maturity if the market price of risk, q, is positive (negative). Since historical
returns on longer-maturity bonds have exceeded those of shorter-maturity ones
in most, though not all countries, this suggests that q is likely to be positive.15
Additional evidence on the value of q can be gleaned by observing the yieldsto-maturity on dierent maturity bonds.
A -maturity bonds continuously
compounded yield to maturity, denoted Y (r (t) , ), can be derived from its

price in (9.41):
1
Y (r (t) , ) ln [P (r (t) , )]
B ()
1
r (t)
= ln [A ( )] +
B () 2r B ( )2
+
= Y + [r (t) Y ]
(9.45)
where Y r+q r 12 r2 . Note that lim Y (r (t) , ) = Y , so that the yield to

maturity on a very long maturity bond approaches Y . Hence, the yield curve,
15 See
(Dimson, Marsh, and Staunton 2002) for an account of the historical evidence.
270
which is the graph of Y (r (t) , ) as a function of , equals r (t) at = 0 and

2
r
3r
r
asymptotes to Y for large. When r (t) Y 4
2 = r +q 42 , the yield
2
curve is monotonically increasing. When Y 4r2 < r (t) < Y + 2r2 = r+q r ,
the yield curve has a humped shape. An "inverted" or monotonically downward
sloping yield curve occurs when r+q r r (t). Since the unconditional mean of
the short rate is r, and, empirically, the yield curve is normally upward sloping,
32
then this suggests that r < r + q r 4r2 , or q >
3r
4 .
Therefore, a yield curve
that typically is upward sloping is also evidence of a positive market price of

bond risk.
9.3
Option Pricing with Random Interest Rates
This last example of the Black-Scholes hedging argument combines aspects of

the first two in that we now consider option pricing in an environment where
interest rates can be random.
We follow Robert Merton (Merton 1973b) in
valuing a European call option when the risk-free interest rate is stochastic and
bond prices satisfy the Vasicek model. The main alteration to the Black-Scholes
derivation is to realize that the call options payo, max [S (T ) X, 0], depends
not only on the maturity date, T , and the stock price at that date, S (T ), but
on the present value of the exercise price, X, which can be interpreted as the
value of a default-free bond that pays X at its maturity date of T . Given the
randomness of interest rates, even the value of this exercise price is stochastic
prior to the options maturity. This motivates us to consider the process of a
bond maturing in T t periods to be another underlying asset, in addition
to the stock, aecting the options value. Writing this bond price as P (t, ),
the options value can now be expressed as c (S (t) , P (t, ) , t). Consistent with
9.3. OPTION PRICING WITH RANDOM INTEREST RATES
271
the Vasicek model, we write this bonds process as
dP (t, ) = p (t, ) P (t, ) dt + p ( ) P (t, ) dzp
(9.46)
where from (9.31) we define dzp dzr . In general, the bonds return will be
correlated with that of the stock, and we allow for this possibility by assuming
dzp dz = dt. Given the options dependence on both the stock and the bond,
Its lemma says that the option price satisfies
dc
c
1 2c 2 2 1 2c 2 2
c
c
S +
p P +
+
S +
P
S
P
t
2 S 2
2 P 2 p
c
2c
c
p SP dt +
S dz +
p P dzp
+
(9.47)
SP
S
P
c
c
S dz +
p P dzp
c cdt +
S
P
where c c is defined as the terms in brackets in the first two lines of (9.47).
Similar to our first example in which a dealer wishes to hedge the sale of an
option, let us form a hedge portfolio consisting of a unit short position in the
option, and a purchase of ws (t) units of the underlying stock, and a purchase
of wp (t) units of the -maturity bond, where we also restrict the portfolio to
require a zero net investment. The zero net investment restriction implies
c (t) ws (t) S (t) wp (t) P (t, ) = 0
(9.48)
272
The hedge portfolios return can then be written as
dH (t) = dc(t) + ws (t) dS (t) + wp (t) dP (t, )
= c c + ws (t) S + wp (t) p P dt
c
S + ws (t) S dz
+
S
c
p P + wp (t) p P dzp
+
P

= ws (t) ( c ) S + wp (t) p c P dt
c
+ ws (t)
Sdz
S
c
+ wp (t)
p P dzp
P
(9.49)
where in the last equality of (9.49) we have substituted in for c using the zero
net investment condition (9.48). If ws (t) and wp (t) can be chosen to make the
hedge portfolios return to be riskless, then it must be the case that the terms
in brackets in the last line of (9.49) can be made to equal zero. In other words,
the following two conditions must hold:
ws (t) =
c
S
(9.50)
wp (t) =
c
P
(9.51)
but from the zero net investment condition (9.48), this can only be possible if
it happens to be the case that
c = ws (t) S + wp (t) P
= S
c
c
+P
S
P
(9.52)
9.3. OPTION PRICING WITH RANDOM INTEREST RATES
273
By Eulers Theorem, condition (9.52) holds if the option price is a homogenous

of degree 1 function of S and P .16 What this means is that if the stocks price
and the bonds price happened to increase by the same proportion, then the
options price would increase by that same proportion.
c (kS (t) , kP (t, ) , t) = kc (S (t) , P (t, ) , t).17
That is, for k > 0,
We assume this to be so, and
later verify that the solution indeed satisfies this homogeneity condition.
Given that condition (9.52) does hold, so that we can choose ws (t) = c/S
and wp (t) = c/P to make the hedge portfolios return to be riskless, then
as in the first example the zero net investment portfolios riskless return must
equal zero in the absence of arbitrage:
ws (t) ( c ) S + wp (t) p c P = 0
(9.53)
or
c
c
( c ) S +
c P = 0
S
P p
(9.54)
which, using (9.52), can be re-written as

c
c
S +
P c c = 0
S
P p
(9.55)
Substituting for c c from (9.47), we obtain
c
1 2c 2 2 1 2c 2 2
2c
p SP = 0
p
t
2 S 2
2 P 2
SP
(9.56)
16 A function f (x ..., x ) is defined to be homogeneous of degree r (where r is an integer)

n
1,
if for every k > 0 then f (kx1, ..., kxn ) = kr f (x1, ..., xn ). Eulers Theorem states that if
Sn
f
f (x1, ..., xn ) is homogeneous of degree r and dierentiable, then
i=1 xi xi = rf .
17 For example, suppose there was a general rise in inflation that increased the stocks and
bonds prices but did not change their relative price, S/P . Then the homogeneity condition
implies that the options price would rise by the same increase in inflation.
274
which since T t, can also be written as
2c 2 2
2c
c
1 2c 2 2
=0
S
+
P
+
2
SP
p
p
2
2
2 S
P
SP
(9.57)
Equation (9.57) is the equilibrium partial dierential equation that the options
value must satisfy.
Importantly, it does not depend on either the expected
rate of return on the stock, , or the expected rate of return on the bond, p .
The appropriate boundary condition for a European call option is similar to
before, with c (S (T ) , P (T, 0) , T ) = c (S (T ) , 1, T ) = max [S (T ) X, 0], where
we impose the condition P (t = T, = 0) = 1. Robert Merton (Merton 1973b)
shows that the solution to this equation is
c (S (t) , P (t, ) , ) = S(t) N (h1 ) P (t, ) XN (h2 )
(9.58)
where
ln
where
v2 =
h1
h2
S(t)
P (t, )X
1 2
2v
(9.59)
h1 v .
2 + p (y)2 2p (y) dy
(9.60)
The solution is essentially the same as the Black-Scholes constant interest rate
formula (9.24) but where the parameter v2 replaces 2 . v2 is the total variance
of the ratio of the stock price to the discounted exercise price over the life of
the option.18 In other words, it is the variance of the ratio
S(t)
P (t, )X
from date t
18 As one would expect, when interest rates are non-stochastic so that the volatility of bond
prices is zero, that is, p (y) = 0, then v 2 = 2 , and we obtain the standard Black-Scholes
formula.
9.4. SUMMARY
275
to date T , an interval of periods. Because the instantaneous variance of the

bond, and hence the variance of the discounted exercise price, shrinks as the
option approaches maturity, this changing variance is accounted for by making
p (y) a function of the time until maturity in (9.60). If we assume that the
bonds volatility is that of the Vasicek model, p (y) =
(1 ey ), then
(9.60) becomes
r
2r
2
y
2y
y
1e
2
=
+e
+ 2 1 2e
dt
(9.61)
2
1 e2
r
2 1 e
= 2 + r3 +
2 2 1 e
Finally, note that solution is homogeneous of degree 1 in S (t) and P (t, ),

which verifies condition (9.52).
9.4
Summary
Fischer Black, Myron Scholes, and Robert Merton made a fundamental discovery that profoundly changed the pricing of contingent securities. They showed
that when an underlying asset follows a diusion, and trade is allowed to occur continuously, a portfolio can be created that fully hedges the risk of the
contingent claim. Therefore, in the absence of arbitrage the hedge portfolios
return must be riskless, and this implies that the contingent claims price must
satisfy a particular partial dierential equation subject to a boundary condition
that its value must equal its terminal payo.
Solving this equation led to a
surprising result: the contingent claims value did not depend directly on the
underlying securitys expected rate of return, but only its volatility. This was
an attractive feature because estimating a risky assets expected rate of return
276
is much more dicult than estimating its volatility.19

As our second example illustrated, the Black-Scholes-Merton hedging argument can be used to derive models of the default-free term structure of interest
rates. The pricing of dierent maturity bonds and of fixed-income derivatives
is a large and ever growing field of asset pricing. Chapter 17 is devoted solely
to this subject. A related topic is the pricing of default-risky bonds. As the
title of Black and Scholes seminal paper suggests, it was readily recognized that
a satisfactory model of option pricing could be applied to valuing the liabilities
of corporations that were subject to default. This link between option pricing
and credit risk also will be explored in Chapter 18.
9.5
Exercises
1. Suppose that the price of a non-dividend-paying stock follows the process
dS = Sdt + S dz
where , , and are constants. The risk-free interest rate equals a

constant, r. Denote p(S(t), t) as the current price of a European put
option on this stock having an exercise price of X and a maturity date
of T . Derive the equilibrium partial dierential equation and boundary
condition for the price of this put option using the Black-Scholes hedging
argument.
2. Define P (r (t) , ) as the date t price of a pure discount bond that pays
$1 in periods. The bond price depends on the instantaneous maturity
19 The accuracy of estimates for a risky assets expected rate of return is proportional to
the time interval over which its average return is computed. In contrast, the accuracy of a
risky assets standard deviation of return is proportional to the number of times the return
is sampled over any fixed time interval. See Merton (Merton 1980) and Chapter 9.3.2 of
Campbell, Lo, and MacKinlay (Campbell, Lo, and MacKinlay 1997).
9.5. EXERCISES
277
yield, r (t), which follows the process
dr (t) = [r r (t)] dt + rdz

where , , and are positive constants. If the process followed by the
price of a bond having periods until maturity is
dP (r, ) /P (r, ) = (r, ) dt p (r, ) dz

and the market price of bond risk is
(r, ) r (t)
= r
p (r, )
then write down the equilibrium partial dierential equation and boundary
condition that this bond price satisfies.
3. The date t price of stock A, A (t), follows the process
dA/A = A dt + A dz
and the date t price of stock B, B (t), follows the process
dB/B = B dt + B dq
where A and B are constants and dz and dq are Brownian motion

processes for which dzdq = dt.
Let c (t) be the date t price of a Eu-
ropean option written on the dierence between these two stocks prices.
Specifically, at this options maturity date, T , the value of the option
equals
c (T ) = max [0, A (T ) B (T )]
278
3.a Using Its lemma, derive the process followed by this option.
3.b Suppose that you are an option dealer who has just sold (wrote) one of
these options for a customer. You now wish to form a hedge portfolio
composed of your unit short position in the option and positions in the
two stocks. Let H (t) denote the date t value of this hedge portfolio. Write
down an equation for H (t) that indicates the amount of shares of stocks
A and B that should be held.
3.c Write down the dynamics for dH (t), showing that its return is riskless.
3.d Assuming the absence of arbitrage, derive the equilibrium partial dierential equation that this option must satisfy.
4. Let S (t) be the date t price of an asset that continuously pays a dividend
that is a fixed proportion of its price.
Specifically, the asset pays a
dividend of S (t) dt over the time interval dt. The process followed by
this assets price can be written as
dS = ( ) Sdt + Sdz
where is the standard deviation of the assets rate of return and is the
assets total expected rate of return, which includes its dividend payment
and price appreciation. Note that the total rate of return earned by the
owner of one share of this asset is dS/S+ dt = dt+ dz. Consider a
European call option written on this asset that has an exercise price of X
and a maturity date of T > t. Assuming a constant interest rate equal to
r, use a Black-Scholes hedging argument to derive the equilibrium partial
dierential equation that this options price, c (t), must satisfy.
Chapter 10
Arbitrage, Martingales, and

Pricing Kernels
In chapters 4 and 7, we examined the asset pricing implications of market completeness in a discrete time model. It was shown that when the number of nonredundant assets equaled the number of states of nature, markets were complete
and the absence of arbitrage ensured that state prices and a state price deflator
would exist.
Pricing could be performed using risk-neutral valuation.
The
current chapter extends these results in a continuous time environment.
We
formally show that when asset prices follow diusion processes and trading is
continuous, then the absence of arbitrage may allow us to value assets using a
martingale pricing technique, a generalization of risk-neutral pricing.
Under
these conditions, a continuous-time stochastic discount factor or pricing kernel

also exists.
These results were developed by John Cox and Stephen Ross (Cox and Ross
1976), John Harrison and David Kreps (Harrison and Kreps 1979), and John
Harrison and Stanley Pliska (Harrison and Pliska 1981), and have proven to be
279
280
CHAPTER 10. ARBITRAGE, MARTINGALES, PRICING KERNELS
very popular approaches to valuing a wide variety of contingent claims. Valuing

contingent claims using risk-neutral pricing or a pricing kernel method can be
an alternative to the previous chapters partial dierential equation approach.
The first section reviews the derivation of the Black Scholes partial dierential equation and points out that this equation also implies that the market
price of risk must be uniform for a contingent claim and its underlying asset. It
also shows how the contingent claims price process can be transformed into a
driftless process by adjusting its Brownian motion process by the market price
of risk and then deflating the contingent claims price by that of a riskless asset.
This driftless (zero expected change) process is known as a martingale.
The
contingent claims value then can be computed as the expectation of its terminal
value under this transformed process.
The second section derives the form of a continuous time state price deflator that can also be used to price contingent claims.
It also demonstrates
how the continuous time state price deflator transforms actual probabilities into
risk-neutral probabilities. The third section shows how problems of valuing a
contingent claim sometimes can be simplified by deflating the contingent claims
price by that of another risky asset. It gives an example of this by valuing an
option written on the dierence between the prices of two risky assets.
The
final section of the chapter examines applications of the martingale approach.

It is used to value an option written on an asset that pays a continuous dividend, examples of which include an option written on a foreign currency and
an option written on a futures price. The martingale pricing technique is also
applied to re-deriving a model of the term structure of interest rates.
10.1. ARBITRAGE AND MARTINGALES
10.1
281
Arbitrage and Martingales
We begin by reviewing the Black-Scholes derivation of contingent claims prices.

Let S be the value of a risky asset which follows a general scalar diusion process
dS = Sdt + Sdz
(10.1)
where both = (S, t) and = (S, t) may be functions of S and t and dz is

a standard pure Brownian motion or Wiener process. For ease of presentation,
we assume that S (t) is a scalar process.
Later, we discuss how multivariate
processes can be handled by the theory, such that and can depend on
other variables that follow diusion processes (driven by additional Brownian
motions) in addition to S (t). In this way, asset values can depend on multiple
sources of uncertainty.
Next let c (S, t) denote the value of a contingent claim whose payo depends
solely on S and t. From Its lemma, we know that this value satisfies
dc = c cdt + c cdz
(10.2)
where c c = ct + ScS + 12 2 S 2 cSS and c c = ScS , and the subscripts on c

denote partial derivatives.
Similar to our earlier analysis, we employ a form of the Black-Scholes hedging
argument by considering a portfolio of 1 units of the contingent claim and cS
units of the risky asset. The value of this portfolio, H, satisfies1
H = c + cS S
(10.3)
and the change in value of this portfolio over the next instant is
1 Unlike last chapters derivation, we do not restrict this portfolio to be a zero-net investment portfolio. As will be clear, the lack of this restriction does not change the nature of our
results.
282
dH
= dc + cS dS
(10.4)
= c cdt c cdz + cS Sdt + cS Sdz

= [cS S c c] dt
Since the portfolio is riskless, the absence of arbitrage implies that it must
earn the risk-free rate. Denoting the (possibly stochastic) instantaneous risk-free
rate as r (t), we have2
dH = [cS S c c] dt = rHdt = r[c + cS S]dt
(10.5)
cS S c c = r[c + cS S]
(10.6)
which implies
If we substitute c c = ct + ScS + 12 2 S 2 cSS into (10.6), we obtain the BlackScholes equilibrium partial dierential equation (PDE).
1 2 2
S cSS + rScS rc + ct = 0
2
(10.7)
However, consider a dierent interpretation of equation (10.6). From Its

lemma, we can substitute cS =
c c
S
into (10.6) and re-arrange to obtain
2 For simplicity, we have assumed that the contingent claims value depends only on a single
risky asset price, S (t). However, when the interest rate is stochastic, the contingent claims
value might be also a function of r (t), that is, c (S, r, t). If, for example, the interest rate
followed the process dr = r (r) dt + r (r) dzr where dzr is an additional Wiener process
aecting interest rate movements, then the contingent claims process would be given by a
bi-variate version of Its lemma. Also, to create a portfolio that earns an instantaneous riskfree rate, the portfolio would need to include a bond whose price is driven by dzr . Later,
we discuss how our results generalize to multiple sources of uncertainty. However, the current
univariate setting can be fully consistent with stochastic interest rates if the risky asset is,
itself, a bond so that S (r, t) and dz = dzr . The contingent claim could then be interpreted
as a fixed-income (bond) derivative security.
r
r
= c
(t)
283
(10.8)
Condition (10.8) is the familiar no-arbitrage condition that requires a unique

market price of risk, which we denote as (t). Then the stochastic process for
the contingent claim can be written as
dc = c cdt + c cdz = [rc + c c] dt + c cdz
(10.9)
Note that the drift of this process depends on the market price of risk, (t),
which may not be directly observable or easily estimated. We now consider an
approach to valuing contingent claims which is an alternative to solving the PDE
in (10.7) but which shares with it the benefit of not having to know (t). The
next topic discusses how a contingent claims risk-premium can be eliminated
by reinterpreting the probability distribution generating asset returns.
10.1.1
A Change in Probability: Girsanovs Theorem
Girsanovs Theorem says that by shifting the Brownian motion process, one can
change the drift of a diusion process when this process is interpreted under a
new probability distribution. Moreover, this shift in Brownian motion changes
the future probability distribution for asset prices in a particular way. To
Rt
zt =
see how this works, consider a new process zbt = zt + 0 (s) ds, so that db
zt (t) dt in (10.9), it can be rewritten:
dzt + (t) dt. Then substituting dzt = db
dc = [rc + c c] dt + c c [db
z dt]
= rcdt + c cdb
z
(10.10)
284
Hence, converting from the Brownian motion process dz to db

z , which removes
the risk-premium c c from the drift term on the right-hand side of (10.9), results in the expected rate of return of c being equal to the risk-free rate if we
now viewed db
z , rather than dz, as a Brownian motion process. The probability distribution of future values of c that are generated by db
z , a probability
distribution that we define as the Q probability measure, is referred to as the
risk-neutral probability measure.3 This is in contrast to the actual probability
distribution for c generated by the dz Brownian motion in (10.9), the original physical or statistical probability distribution that is denoted as the P
measure.
Girsonovs Theorem states that as long as (t) is well-behaved in the sense
that it follows a process that does not vary too much over time, then the probability density function for a random variable at some future date T , such as
c (T ), under the risk-neutral Q distribution bears a particular relationship to
that of the physical P distribution.4 Specifically, denote dPT as the instantaneous change in the physical distribution function at date T generated by dzt ,
which makes it the physical probability density function at date T .5 Similarly,
zt . Then
let dQT be the risk-neutral probability density function generated by db
3 The idea of a probability measure (or distribution), P , is as follows. Define a set function,
f , which assigns a real number to a set E, where E could be a set of real numbers, such as
anV
intervalon the
Sn real line. Formally, f (E) R. This function is countably additive if
n
f
i=1 Ei =
i=1 f (Ei ) where hEi i is a finite or countably infinite sequence of disjoint
sets. A measure is defined as a non-negative set function that is countably additive. Note
that probabilities are measures since they assign a non-negative probability to a particular set.
For example, let the domain ofU a continuous probability distribution for a random variable, x,
dP (x) = 1 where P is the probability measure (probability

be the entire real line, that is,
distribution function). Now let a set E1 = [a, b] be an interval on this line. The probability
U
of x E1 is f (E1 ) = ab dP (x) 0. Similarly, if E2 = [c, d] which is assumed to be an
U
U
V
interval that does not overlap with E1 , then f (E1 E2 ) = ab dP (x) + cd dP (x) = f (E1 )
+f (E2 ). Hence, probabilities are non-negative
l additive.
k
U and countably
T
2
4 The restriction on (t) is that E exp
(u)
du
< , which is known as the
t
t
Novikov condition. Ioannis Karatzas and Steven Shreve (Karatzas and Shreve 1991) give a
formal statement and proof of Girsonovs Theorem.
5 Recall that since a probability distribution function, P , is an integral over the probabilU
ity density function, dP , the density function can be interpreted as the derivative of the
probability distribution function.
285
Girsonovs theorem says that at some date t where 0 < t < T , the relationship
between the two probability densities at date T is
dQT
" Z
= exp
T
t
1
(u) dz
2
= ( T / t ) dPT
(u) du dPT
(10.11)
where t is positive random process that depends on (t) and zt and is given
by
Z
= exp
1
(u) dz
2
(u) ds
2
(10.12)
In other words, by multiplying the physical probability density at date T by

the factor T / t , we can determine the risk-neutral probability density at date
T.
Since from (10.12) we see that T / t > 0, equation (10.11) implies that
whenever dPT has positive probability, so does dQT . Because they share this
characteristic, the physical P measure and the risk-neutral Q measure are called
equivalent probability measures in that any future value of c that has positive
probability (density) under the physical measure also has positive probability
(density) under the risk-neutral measure.6 We can re-arrange (10.11) to obtain
dQT
= T / t
dPT
(10.13)
which clarifies that T / t can be interpreted as the derivative of the risk-neutral

measure QT with respect to the physical measure PT . Indeed, T / t is known
as the Radon-Nikodym derivative of Q with respect to P . Later in this chapter
we will return to an interpretation of this derivative T / t following a discussion
6 An example illustrates this equivalency.
Suppose in (10.1) that and are constant
and the risk-free interest rate, r, is constant. Then the process dS/S = dt + dz has a
discrete time lognormal distribution under the P measure. Under the Q measure the process
is, dS/S = rdt + de
z, which is also lognormally distributed but with r replacing . Since
these lognormal distributions both have positive probability density over the domain from 0
to , they are referred to as equivalent.
286
of the continuous time pricing kernel approach to valuing contingent securities.

In summary, we have seen that a transformation of a Brownian motion by
the market price of risk transforms a securitys expected rate of return to equal
the risk-free rate. This transformation from the physical Brownian motion to
a risk-neutral one also transforms the probability density functions for random
variables at future dates.
10.1.2
Money Market Deflator
As a final step in deriving a new valuation formula for contingent claims, we

now show that the contingent claims appropriately deflated price process can
be made driftless (a martingale) under the probability measure Q. Let B (t)
be the value of an investment in a money market fund, that is, an investment
in the instantaneous maturity risk-free asset.7 Then
dB/B = r(t)dt
UT
Note that B (T ) = B (t) e
r(u)du
for any date T t.
(10.14)
Now define C(t)
c(t)/B(t) as the deflated price process for the contingent claim.
Essentially,
C (t) is the value of the contingent claim measured in terms of the value of the
riskless safe investment that grows at rate r (t). A trivial application of Its
lemma gives
dC
=
=
1
dc
B
rc
dt +
B
c
dB
B2
cc
c
db
z r dt
B
B
(10.15)
= c Cdb
z
7 An investment that earns the instantaneous maturity risk-free rate is sometimes referred
to as a money market fund because money market mutual funds invest in short maturity, high
credit quality (nearly risk free) debt instruments.
287
Thus, the deflated price process under the equivalent probability measure generated by db
z is a driftless process: its expected change is zero. An implication
of (10.15) is that the expectation under the risk-neutral or Q measure of any
future value of C is the current value of C. This can be stated as
bt [C (T )]
C (t) = E
T t
(10.16)
bt [] denotes the expectation operator under the probability measure

where E
generated by db
z .8 The mathematical name for a process such as (10.16) is a
martingale, which is essentially a random walk in discrete time.9
To summarize, we showed that the absence of arbitrage implies the existence

of an equivalent probability measure such that the deflated price process is a
martingale. Note that (10.16) holds for any deflated contingent claim, including
the deflated underlying risky asset, S/B, since we could define the contingent
claim as c = S.
10.1.3
Feynman-Kac Solution
Now if we re-write (10.16) in terms of the undeflated contingent claims price,

we obtain
bt c (T ) 1
c(t) = B(t)E
B (T )
h UT
i
bt e t r(u)du c (T )
= E
(10.17)
common notation for this risk-neutral or Q measure expectation is EtQ [].

formally, define a family of information sets, It , that start at date t = 0 and continue
for all future dates, {It , t [0, ]}. Also, assume that information at date t includes all
information from previous dates, so that for t0 < t1 < t2 , It0 It1 It2 . Such a family of
information sets is referred to as a filtration. A process is a martingale with respect to It if it
satisfies E [C (T ) |It ] = C (t) t < T where It includes the value of C (t), and E [|C (T )|] < ,
that is, the unconditional expectation of the process is finite.
8 Another
9 More
288
Equation (10.17) can be interpreted as a solution to the Black-Scholes partial

dierential equation (10.7) and, indeed, is referred to as the Feynman-Kac solution.10 From a computational point of view, equation (10.17) says that we
can price (value) a contingent security by taking the expected value of its discounted payo, where we discount at the risk-free rate but also assume that
when taking the expectation of c (T ) that the rate of return on c (and all other
asset prices, such as S) equals the risk-free rate, a rate that may be changing
over time. As when the contingent securitys value is found directly from the
partial dierential equation (10.7), no assumption regarding the market price of
risk, (t), is required because it was eliminated from all assets return processes
when converting to the Q measure. Equivalently, one can use equation (10.16)
to value c (t) /B (t) by taking expectations of the deflated price process, where
this deflated process has zero drift. Both of these procedures are continuoustime extensions of the discrete-time risk-neutral valuation technique that we
examined in Chapters 4 and 7.
10.2
Arbitrage and Pricing Kernels
This is not the first time that we have computed an expectation to value a
security. Recall from the single- or multi-period consumption - portfolio choice
problem with time separable utility that we obtained an Euler condition of the
form11
10 To solve (10.7), a boundary condition for the derivative is needed. For example, in the
case of a European call option it would be c (T ) = max [0, S (T ) X]. The solution given by
(10.17) incorporates this boundary condition, c (T ).
11 In equation (10.18) we are assuming that the contingent claim pays no dividends between
dates t and T .
10.2. ARBITRAGE AND PRICING KERNELS
289
c (t) = Et [mt,T c (T )]
MT
= Et
c (T )
Mt
(10.18)
where date T t, mt,T MT /Mt and Mt = Uc (Ct , t) was the marginal utility
of consumption at date t.
In Chapter 4, we also showed in a discrete time
- discrete state model that the absence of arbitrage implies that a stochastic
discount factor, mt,T , exists whenever markets are complete.
We now show
that this same result applies in a continuous time environment whenever markets
are dynamically complete. The absence of arbitrage opportunities, which earlier
guaranteed the existence of an equivalent martingale measure, also determines a
pricing kernel or state price deflator, Mt . In fact, the concepts of an equivalent
martingale measure and state pricing kernel are one and the same.
Note that we can rewrite (10.18) as
c (t) Mt = Et [c (T ) MT ]
(10.19)
which says that the deflated price process, c (t) Mt , is a martingale. But note the
dierence here versus our earlier analysis: the expectation in (10.19) is taken
under the physical probability measure, P , while in (10.16) and (10.17) the
expectation is taken under the risk-neutral measure, Q.
Since in the standard time-separable utility portfolio choice model, Mt is
the marginal utility of consumption, this suggests that Mt should be a positive
process even when we consider more general environments where a stochastic
discount factor pricing relationship would hold.
Hence, we assume that the
state price deflator, Mt , follows a strictly positive diusion process of the general
form:
290
dM = m dt + m dz
(10.20)
Now consider the restrictions that the Black-Scholes no-arbitrage conditions

place on m and m if (10.19) and (10.20) hold. For any arbitrary security or
contingent claim, c, define cm = cM and apply Its lemma:
dcm
= cdM + Mdc + (dc) (dM )
(10.21)
= [cm + M c c + c c m ] dt + [c m + M c c] dz
If cm = cM satisfies (10.19), that is, cm is a martingale, then its drift in (10.21)
must be zero, implying
c =
m c m
M
M
(10.22)
Now consider the case in which c is the instantaneously riskless asset, that is,
c (t) = B (t) is the money market investment following the process in equation
(10.14). This implies that c = 0 and c = r (t). Using (10.22), requires
r (t) =
m
M
(10.23)
In other words, the expected rate of change of the pricing kernel must equal
minus the instantaneous risk-free interest rate. Next, consider the general case
where the asset c is risky, so that c 6= 0. Using (10.22) and (10.23) together,
we obtain
c = r (t)
or
cm
M
(10.24)
m
c r
=
c
M
291
(10.25)
Comparing (10.25) to (10.8), we see that
m
= (t)
M
(10.26)
Thus, the no-arbitrage condition implies that the form of the pricing kernel
must be
dM/M = r (t) dt (t) dz
(10.27)
Note that if we define mt ln Mt , then dm= r + 12 2 dt dz. Hence, in
using the pricing kernel to value any contingent claim, we can re-write (10.18)
as
c (t) = Et [c (T ) MT /Mt ] = Et c (T ) em T m t
h
i
UT
UT
1 2
= Et c (T ) e t [r(u)+ 2 (u)]du t (u)dz
(10.28)
Given processes for r (t), (t) and the contingent claims payo, c (T ), in some
instances it may be easier to compute (10.28) rather than, say, (10.16) or (10.17).
Of course, in computing (10.28), we need to use the actual drift for c, that is,
we compute expectations under the P measure, not the Q measure.
10.2.1
Linking the Valuation Methods
To better understand the connection between the pricing kernel (stochastic discount factor) approach and the martingale (risk-neutral) valuation approach,
we now show how Mt is related to the change in probability distribution accom-
292
plished using Girsonovs theorem. Equating (10.17) to (10.28), we have

h U
i
bt e tT r(u)du c (T ) = Et [c (T ) MT /Mt ]
E
h UT
UT
= Et e t r(u)du c (T ) e t
(10.29)
i
UT
1 2
2 (u)du t (u)dz
and then if we substitute using the definition of from (10.12) we have

h U
h U
i
i
bt e tT r(u)du c (T ) = Et e tT r(u)du c (T ) ( T / t )
E
Z
bt [C (T )] = Et [C (T ) ( T / t )]
E
Z
C (T ) dQT =
C (T ) ( T / t ) dPT
(10.30)
where recall that C (t) = c (t) /B (t). From the first two lines of (10.30), we see
that on both sides of the equation, the terms in brackets are exactly the same
except that the expectation under P includes the Radon-Nikodym derivative
T / t . As predicted by Girsonovs theorem, this factor transforms the physical
probability density at date T to the risk-neutral probability density at date T .
Furthermore, relating (10.29) to (10.30) implies
MT /Mt = e
UT
t
r(u)du
( T / t )
(10.31)
so that the continuous time pricing kernel (stochastic discount factor) is the
product of a risk-free rate discount factor and the Radon-Nikodym derivative.
Hence, MT /Mt can be interpreted as providing both discounting at the risk-free
rate and transforming the probability distribution to the risk-neutral one. Indeed, if contingent security prices are deflated by the money market investment,
thereby removing the risk-free discount factor, the second line of (10.30) shows
that the pricing kernel, MT /Mt , and the Radon-Nikodym derivative, T / t , are
exactly the same.
293
Similar to the discrete time case discussed in Chapter 4, the role of this
derivative ( T / t or MT /Mt ) is to adjust the risk-neutral probability, Q, to give
it greater probability density for "bad" outcomes and less probability density for
"good" outcomes relative to the physical probability, P . In continuous time,
the extent to which an outcome, as reflected by a realization of dz, is "bad"
or "good" depends on the sign and magnitude of its market price of risk, (t).
This explains why in equation (10.27) the stochastic component of the pricing
kernel is of the form (t) dz.
10.2.2
The Multivariate Case
The previous analysis has assumed that contingent claims prices depend on only
a single source of uncertainty, dz. In a straightforward manner, the results can
be generalized to permit multiple independent sources of risk. Suppose we had
asset returns depending on an n 1 vector of independent Brownian motion
processes, dZ = (dz1 ...dzn )0 where dzi dzj = 0 for i 6= j.12 A contingent claim
whose payo depended on these asset returns then would have a price that
followed the process
dc/c = c dt + c dZ
where c is a 1 n vector c = ( c1 ...cn ).13
(10.32)
Let the corresponding n 1
vector of market prices of risks associated with each of the Brownian motions
0
be = (1 ... n ) . Then, it is straightforward to show that we would have the
no-arbitrage condition
c r = c
(10.33)
12 The independence assumption is not important. If there are correlated sources of risk
(Brownian motions), they can be re-defined by a linear transformation to be represented by
n orthogonal risk sources.
13 Both and the elements of may be functions of state variables driven by the Brownian
c
c
motion components of dZ.
294
Equations (10.16) and (10.17) would still hold, and now the pricing kernels
process would be given by
dM/M = r (t) dt (t) dZ
10.3
(10.34)
Alternative Price Deflators
In previous sections, we found it convenient to deflate a contingent claim price by

the money market funds price, B (t). Sometimes, however, it may be convenient
to deflate or normalize a contingent claims price by the price of a dierent
type of security. Such a situation can occur when a contingent claims payo
depends on multiple risky assets. Lets now consider an example of this, in
particular, where the contingent claim is an option written on the dierence
between two securities (stocks) prices. The date t price of stock 1, S1 (t),
follows the process
dS1 /S1 = 1 dt + 1 dz1
(10.35)
and the date t price of stock 2, S2 (t), follows the process
dS2 /S2 = 2 dt + 2 dz2
(10.36)
where 1 and 2 are assumed to be constants and dz1 and dz2 are Brownian
motion processes for which dz1 dz2 = dt.
Let C (t) be the date t price of
a European option written on the dierence between these two stocks prices.
Specifically, at this options maturity date, T , the value of the option equals
C (T ) = max [0, S1 (T ) S2 (T )]
(10.37)
10.3. ALTERNATIVE PRICE DEFLATORS
295
Now define c (t) = C (t) /S2 (t) , s (t) S1 (t) /S2 (t), and B (t) = S2 (t) /S2 (t)
= 1 as the deflated price processes, where the prices of the option, stock 1, and
stock 2 all are normalized by the price of stock 2. With this normalized price
system, the terminal payo corresponding to (10.37) is now
c (T ) = max [0, s (T ) 1]
(10.38)
Applying Its lemma, the process for s (t) is given by
ds/s = s dt + s dz3
(10.39)
where s 1 2 + 22 1 2 , s dz3 1 dz1 2 dz2 , and 2s = 21 + 22

21 2 .
Further, when prices are measured in terms of stock 2, the deflated
price of stock 2 becomes the riskless asset, with the riskless rate of return given
by dB/B = 0dt. That is, because the deflated price of stock 2 never changes,
it returns a riskless rate of zero.
Using Its lemma once again, the deflated
option price, c (s (t) , t), follows the process
1
2 2
dc = cs s s + ct + css s s dt + cs s s dz3 .
2
(10.40)
With this normalized price system, the usual Black-Scholes hedge portfolio can
be created from the option and stock 1. The hedge portfolios value is given by
H = c + cs s
and the instantaneous change in value of the portfolio is
(10.41)
296
dH
=
=
=
(10.42)
dc + cs ds
1
cs s s + ct + css 2s s2 dt cs s s dz3 + cs s s dt + cs s s dz3
2
1
2 2
ct + css s s dt
2
When measured in terms of stock 2s price, the return on this portfolio is

instantaneously riskless. In the absence of arbitrage, it must earn the riskless
return, which as noted above, equals zero under this deflated price system. Thus
we can write
1
2 2
dH = ct + css s s dt = 0
2
(10.43)
which implies
ct +
1
css 2s s2 = 0
2
(10.44)
which is the Black-Scholes partial dierential equation but with the risk-free
rate, r, set to zero. Solving it subject to the boundary condition (10.38), which
implies a unit exercise price, gives the usual Black-Scholes formula
c(s, t) = s N (d1 ) N (d2 )
(10.45)
where
d1
d2
ln (s(t)) + 12 2s (T t)
s T t
d1 s T t.
(10.46)
To convert back to the undeflated price system, we simply multiply (10.45) by

S2 (t) and obtain
10.4. APPLICATIONS
297
C( t) = S1 N (d1 ) S2 N (d2 )
(10.47)
Note that the option price does not depend on the non-deflated price systems
risk-free rate, r (t). Hence, the formula holds even for stochastic interest rates.
10.4
Applications
This section illustrates the usefulness of the martingale pricing technique. The
first set of applications deal with options written on assets that continuously
pay dividends. Examples include an option written on a foreign currency and
an option written on a futures price. The second application is to value bonds
of dierent maturities, which determines the term structure of interest rates.
10.4.1
Continuous Dividends
Many types of contingent claims depend on an underlying asset that can be

interpreted as paying a continuous dividend that is proportional to the assets
price. Let us apply the risk-neutral pricing method to value an option on such
an asset. Denote as S (t) the date t price of an asset that continuously pays
a dividend that is a fixed proportion of its price. Specifically, the asset pays
a dividend of S (t) dt over the time interval dt. The process followed by this
assets price can be written as
dS = ( ) Sdt + Sdz
(10.48)
where is the standard deviation of the assets rate of return and is the
assets total expected rate of return, which includes its dividend payment and
price appreciation. Similar to the assumptions of Black and Scholes, and
298
are assumed to be constant, but may be a function of S and t.
Now note
that the total rate of return earned by the owner of one share of this asset is
dS/S+ dt = dt+ dz. Consider a European call option written on this asset
that has an exercise price of X and a maturity date of T > t, where we define
T t. Assuming a constant interest rate equal to r, we use equation (10.17)
to write the date t price of this option as
bt er c (T )
c (t) = E
(10.49)
bt [max [S (T ) X, 0]]
= er E
To calculate the expectation in (10.49), we need to consider the distribution of

S (T ). Note that because could be a function of S and t, the distribution of
S (T ) under the physical P measure cannot be determined until this functional
relationship (S, t) is specified. However, (10.49) requires the distribution of
S (T ) under the risk-neutral Q measure, and given the assumption of a constant
risk-free rate, this distribution already is determined. As in (10.10), converting
from the physical measure generated by dz to the risk-neutral measure generated
by db
z removes the risk-premium from the assets expected rate of return. Hence,
the risk-neutral process for the stock price becomes
dS = (r ) Sdt + Sdb
z
(10.50)
Since r and are constants, we know that S follows geometric Brownian

motion, and hence is lognormally distributed, under Q.
From our previous
results, we also know that the risk-neutral distribution of ln[S (T )] is normal:
ln [S (T )] N
1
ln [S (t)] + (r 2 ) , 2
2
(10.51)
10.4. APPLICATIONS
299
Equation (10.49) can now be computed as
c (t)
bt [max [S (T ) X, 0]]
= er E
Z
= er
(S (T ) X) g(S (T )) dS (T )
(10.52)
where g(ST ) is the lognormal probability density function. This integral can
be evaluated by making the change in variable
ln [S (T ) /S (t)] r 12 2
Y =

(10.53)
which from (10.51) transforms the lognormally distributed S (T ) into the variable Y distributed N (0, 1). The result is the modified Black-Scholes formula
c = Se N (d1 ) Xer N (d2 )
(10.54)
where
d1
d2
ln (S/X) + r + 12 2
= d1
=
(10.55)
Comparing this formula to (9.24) and (9.25), the value of an option written
on an asset that pays no dividends, the only dierence is that the non-dividend
paying assets price, S (t), is replaced with the dividend-discounted price of the
dividend paying asset, S (t) e .
The intuition behind this can be seen by
t [S (T )] = S (t) er . However,
realizing that if no dividends are paid, then E
with dividends, the risk-neutral expected asset price appreciates at rate r ,
rather than r. This is because with dividends paid out at rate , expected price
300
appreciation must be at rate r to keep the total expected rate of return equal
to + r = r. Thus, the risk-neutral expectation of S (T ) is
t [S (T )] = S (t) e(r)
E
(10.56)
= S (t) e er = S (t) er
where we define S (t) S (t) e . This shows that the value of an option on
a dividend-paying asset with current price S equals the value of an option on a
non-dividend paying asset having current price S = Se .
Formula (10.54) can be applied to an option on a foreign currency. If S (t)
is defined as the domestic currency value of a unit of foreign currency, that is,
the spot exchange rate, then assuming this rate has a constant volatility gives
it a process satisfying (10.48).
Since purchase of a foreign currency allows
the owner to invest it in an interest-earning asset yielding the foreign currency

interest rate, rf , the dividend yield will equal this foreign currency rate, = rf .
t [S (T )] = S (t) e(rrf ) , where the domestic and foreign currency
Hence, E
interest rates are those for a risk-free investment having a maturity equal to
that of the option.
Note that this expression is the no-arbitrage value of
the date t forward exchange rate having a time until maturity of , that is,
Ft, = Se(rrf ) .14 Therefore, (10.54) can be written as
c (t) = er [Ft, N (d1 ) XN (d2 )]

2
where d1 =
ln[Ft, /X]+ 2
(10.57)
and d2 = d1 .
A final example is an option written on a futures price. Options are written

on the futures prices of commodities, equities, bonds, and currencies. Futures
14 This
is the same formula as (3.19) or (7.2) but with continuously-compouned yields.
10.4. APPLICATIONS
301
prices are similar to forward prices.15 Like a forward contract, futures contracts
involve long and short parties, and if both parties maintain their positions until
the maturity of the contract, their total profits equal the dierence between the
underlying assets maturity value and the initial future price. The main dierence between futures contracts and forward contracts is that a futures contract
is marked-to-market daily: that is, the futures price for a particular maturity
contract is re-computed daily and profits equal to the dierence between todays and yesterdays future price are transferred (settled) from the short party
to the long party on a daily basis.
Thus, if Ft,t is the date t futures price
for a contract maturing at date t , then the undiscounted profit (loss) earned
by the long (short) party over the period from date t to date T t is simply
FT,t Ft,t . Like forward contracts, there is no initial cost for the parties who
enter into a futures contract.
Hence, in a risk-neutral world, their expected
profits must be zero. This implies that
t [FT,t Ft,t ] = 0
E
(10.58)
or that under the Q measure, the futures price is a martingale
t [FT,t ] = Ft,t
E
(10.59)
Thus, while under the Q measure a non-dividend paying asset price would be
expected to grow at rate r, a futures price would be expected to grow at rate
0. Hence, futures are like assets with a dividend yield = r.
From this one
can derive the value of a futures call option that matures in periods where
15 See (Cox, Ingersoll, and Ross 1981) and (Jarrow and Oldfield 1981) for a comparison of
forward and futures contracts. If markets are frictionless, there are no arbitrage opportunities,
and default-free interest rates are non-stochastic, then it can be shown that forward and
futures prices are equivalent for contracts written on the same underlying asset and having
the same maturity date. When interest rates are stochastic, then futures price will be greater
(less) than equivalent contract forward prices if the underlying asset is positively (negatively)
correlated with short-term interest rates.
302
(t t) as:
c (t) = er [Ft,t N (d1 ) XN (d2 )]
2
where d1 =
ln[Ft,t /X ]+ 2
(10.60)
and d2 = d1 . Note that this is similar in
form to an option on a foreign currency written in terms of the forward exchange

rate.
10.4.2
The Term Structure Revisited
The martingale pricing equation (10.17) can be applied to deriving the date
t price of a default-free bond that matures in periods and pays $1 at the
maturity date T = t + .
This allows us to value default-free bonds in a
manner that is an alternative to the partial dierential equation approach of

the previous chapter.
Using the same notation as in Chapter 9, let P (t, )
denote this bonds current price.

(10.17) becomes
Then since c (T ) = P (T, 0) = 1, equation
h U
i
bt e tT r(u)du 1
P (t, ) = E
We now re-derive the Vasicek model using this equation.
(10.61)
To apply equation
(10.61), we need to find the risk-neutral (Q measure) process for the instantaneous maturity interest rate, r (t). Recall that the physical (P measure) process
for the interest rate was assumed to be the Ornstein-Uhlenbeck process
dr(t) = [r r (t)] dt + r dzr
(10.62)
and that the market price of bond risk, q, was assumed to be a constant. This
implied that the expected rate of return on all bonds satisfied
p (r, ) = r (t) + qp ( )
(10.63)
10.4. APPLICATIONS
303
where p () = Pr r /P . Thus, the physical process for a bonds price, given

by equation (9.31), can be re-written as
dP (r, ) /P (r, ) = p (r, ) dt p ( ) dzr
(10.64)
= [r (t) + qp ()] dt p ( ) dzr

Now note that if we define the transformed Brownian motion process db
zr =
dzr qdt, then equation (10.64) becomes
zr + qdt]
dP (t, ) /P (t, ) = [r (t) + q p ( )] dt p () [db
(10.65)
= r (t) dt p ( ) db
zr
which is the risk-neutral, Q measure process for the bond price.
This is so
because under this transformation all bond prices now have an expected rate
of return equal to the instantaneously risk-free rate, r (t). Therefore, applying
this same Brownian motion transformation to equation (10.62), we find that the
instantaneous maturity interest rate process under the Q measure is
zr + qdt]
dr(t) = [r r (t)] dt + r [db
i
h
qr
r (t) dt + r db
= r+
zr
(10.66)
Hence, we see that the risk-neutral process for r (t) continues to be an OrnsteinUhlenbeck process but with a dierent unconditional mean, r + q r /. Thus,
we can use the valuation equation (10.61) to compute the discounted value of
h
R
i
bt exp T r (u) du assuming r (t) follows
the bonds $1 payo, P (t, ) = E
t
the process in (10.66). Doing so leads to the same solution given in the previous
304
chapter, equation (9.41).16

The intuition for why (10.66) is the appropriate risk-neutral process for r (t)
is as follows.
Note that if the market price of risk, q, is positive, then the
risk-neutral mean, r + q r /, exceeds the physical processs mean, r. In this

h
R
i
bt exp T r (u) du , the
case, when we use valuation equation P (t, ) = E
t
expected risk-neutral discount rate is greater than the physical expectation of

r (t). Therefore, ceteris paribus, the greater is q, the lower will be the bonds
price, P (t, ), and the greater will be its yield to maturity, Y (t, ). Thus, the
greater the market price of interest rate risk, the lower are bond prices and the
greater are bond yields.
10.5
Summary
This chapter has covered much ground.
Yet, many of its results are similar
to discrete time counterparts derived in Chapter 4.
The martingale pricing
method essentially is a generalization of risk-neutral pricing, and is applicable

in complete market economies when arbitrage opportunities are not present. A
continuous time state price deflator can also be derived when asset markets are
dynamically complete. We demonstrated that this pricing kernel is expected
to grow at minus the short term interest rate and that the standard deviation
of its growth is equal to the market price of risk. We also saw that contingent
claims valuation often can be simplified by an appropriate normalization of asset
prices. In some cases this is done by deflating by the price of a riskless asset,
and in others by deflating by a risky asset price. A final set of results included
showing how the martingale approach can be applied to valuing a contingent
claim written on an asset that pays a continuous, proportional dividend. Im16 Since
the Ornstein-Uhlenbeck process in (10.66) is normally distributed, the integral

r
(u)
du
is also normally distributed based on the idea that sums (an integral) of normals
t
l
k U
are normal. Hence, exp tT r (u) du is lognormally distributed.
UT
10.6. EXERCISES
305
portant examples of this included options on foreign exchange and on futures

prices. Also included was an illustration of how the martingale method can be
applied to deriving the term structure of interest rates.
10.6
Exercises
1. In this problem, you are asked to derive the equivalent martingale measure
and the pricing kernel for the case to two sources of risk. Let S1 and S2
be the values of two risky assets which follow the processes
dSi /Si = i dt + i dzi , i = 1, 2
where both i and i may be functions of S1 , S2 , and t, and dz1 and dz2
are two independent Brownian motion processes, implying dz1 dz2 = 0. Let
f (S1 , S2 , t) denote the value of a contingent claim whose payo depends
solely on S1 , S2 , and t. Also let r (t) be the instantaneous risk-free interest
rate. From Its lemma, we know that the derivatives value satisfies
df = f fdt + f 1 fdz1 + f 2 f dz2

where f f = f3 + 1 S1 f1 + 2 S2 f2 + 12 21 S12 f11 + 12 22 S22 f22 , f 1 f =
1 S1 f1 , f 2 f = 2 S2 f2 and where the subscripts on f denote the partial
derivatives with respect to its three arguments, S1 , S2 , and t.
1.a By forming a riskless portfolio composed of the contingent claim and the
two risky assets, show that in the absence of arbitrage an expression for
f can be derived in terms of r, 1
1 r
1 ,
and 2
2 r
2 .
1.b Define the risk-neutral processes db

z1 and db
z2 in terms of the original
306

Brownian motion processes, and then give the risk-neutral process for
z2 .
df in terms of db
z1 and db
1.c Let B (t) be the value of a money market fund which invests in the
instantaneous maturity risk-free asset. Show that F (t) f (t) /B (t) is a
martingale under the risk-neutral probability measure.
1.d Let M (t) be the state price deflator such that f (t) M (t) is a martingale
under the physical probability measure. If
dM = m dt + m1 dz1 + m2 dz2 ,
what must be the values of m , m1 , and m2 that preclude arbitrage?

Show how you solve for these values.
2. The Cox, Ingersoll, and Ross (Cox, Ingersoll, and Ross 1985b) model of
the term structure of interest rates assumes that the process followed by
the instantaneous maturity risk-free interest rate is
dr = ( r) dt + rdz
where , , and are constants.
Let P (t, ) be the date t price of a
zero-coupon bond paying $1 at date t + . It is assumed that r(t) is the

only source of uncertainty aecting P (t, ). Also, let p (t, ) and p (t, )
be the instantaneous mean and standard deviation of the rate of return
on this bond and assume
p (t, ) r (t)
= r
p (t, )
where is a constant.
10.6. EXERCISES
307
2.a Write down the stochastic process followed by the pricing kernel (state
price deflator), M (t), for this problem, that is, the process dM/M. Also,
apply Its lemma to derive the process for m(t) ln (M ), that is, the
process dm.
2.b Let the current date be 0 and write down the formula for the bond price,
P (0, ), in terms of an expectation of m m0 . Show how this can be
written in terms of an expectation of functions of integrals of r (t) and .
3. If the price of a non-dividend paying stock follows the process dS/S =
dt + dz where is constant, and there is a constant risk-free interest rate equal to r, then the Black-Scholes showed that the no-arbitrage
value of a standard call option having periods to maturity and an exercise price of X is given by c = SN (d1 ) Xer N (d2 ) where d1 =

ln (S/X) + r + 12 2 / ( ) and d2 = d1 .
A forward start call option is similar to this standard option but with the
dierence that the option exercise price, X, is initially a random variable.

The exercise price is set equal to the contemporaneous stock price at a
future date prior to the maturity of the option. Specifically, let the current
date be 0 and the option maturity date be . Then at date t where
0 < t < , the options exercise price, X, is set equal to the date t value
of the stock, denoted as S (t). Hence, X = S (t) is a random variable as
of the current date 0.
For a given date t, derive the date 0 value of this forward start call option.
(Hint: Note the value of a standard call option when S = X, and then
use a simple application of risk-neutral pricing to derive the value of the
forward start option.)
308
4. If the price of a non-dividend paying stock follows the process dS/S =

dt + dz where is constant, and there is a constant risk-free interest rate equal to r, then the Black-Scholes showed that the no-arbitrage
value of a standard call option having periods to maturity and an exercise price of X is given by c = SN (d1 ) Xer N (d2 ) where d1 =

ln (S/X) + r + 12 2 / ( ) and d2 = d1 . Based on this re-
sult and a simple application of risk-neutral pricing, derive the value of

the following binary options.
Continue to assume that the underlying
stock price follows the process dS/S = dt + dz, the risk-free interest
rate equals r, and the options time until maturity equals .
4.a Consider the value of a cash-or-nothing call, cnc. If S (T ) is the stocks

price at the options maturity date of T , the payo of this option is
F if S (T ) > X
cncT =
0 if S (T ) X
where F is a fixed amount. Derive the value of this option when its time
until maturity is and the current stock price is S. Explain your reasoning.
4.b Consider the value of an asset-or-nothing call, anc. If S (T ) is the stocks

price at the options maturity date of T , the payo of this option is
S (T ) if S (T ) > X
ancT =
0 if S (T ) X
Derive the value of this option when its time until maturity is and the
current stock price is S. Explain your reasoning.
5. Outline a derivation of the form of the multi-variate state price deflator
10.6. EXERCISES
309
given in equations (10.33) and (10.34).

6. Consider a continuous-time version of a Lucas (1978) endowment economy.
It is assumed that there is a single risky asset (e.g., fruit tree) that produces
a perishable consumption good that is paid out as a continuous dividend,
gt . This dividend satisfies the process
dgt /gt = dt + dz
where and are constants. There is a representative agent who at date

0 maximizes lifetime consumption given by
E0
U (Ct , t) dt
where U (Ct , t) = et Ct /, < 1. Under the Lucas endowment economy

assumption, we know that in equilibrium Ct = gt .
6.a Let Pt () denote the date t price of a riskless discount (zero-coupon) bond
that pays one unit of the consumption good in periods. Derive an (Euler
equation) expression for Pt ( ) in terms of an expectation of a function of
future dividends.
6.b Let mt,t+ Mt+ /Mt be the stochastic discount factor (pricing kernel)
for this economy. Based on your answer in part a, write down the stochastic process for Mt . (Hint: find an expression for Mt and then use Its
lemma.)
6.c Based on your previous answers, write down the instantaneous risk-free
real interest rate. Is it constant or time-varying?
Chapter 11
Mixing Diusion and Jump

Processes
We have studied the nature and application of diusion processes, which are
continuous time stochastic processes whose uncertainty derives from Brownian
motions. While these processes have proven useful in modeling many dierent
types of economic and financial time series, they may be unrealistic for modeling
random variables whose values can change very significantly over a short period
of time. This is because diusion processes have continuous sample paths and
cannot model discontinuities or jumps in their values.
In some situations,
it may be more accurate to allow for large, sudden changes in value.
For
example, when the release of significant new information results in an immediate,

substantial change in the market value of an asset, then we need to augment the
diusion process with another type of uncertainty to capture this discontinuity
in the assets price. This is where Poisson jump processes can be useful. In
particular, we can model an economic or financial time series as the sum of
diusion (Brownian motion-based) processes and jump processes.
311
312
CHAPTER 11. MIXING DIFFUSION AND JUMP PROCESSES
The first section of this chapter introduces the mathematics of a process

that is mixture of a jump process and a diusion process.
Section 2 shows
how Its lemma can be extended to derive the process of a variable that is a
function of a mixed jump-diusion process. It comes as no surprise that this
function inherits the risk of both the Brownian motion component as well as the
jump component of the underlying process. Section 3 revisits the problem of
valuing a contingent claim, but now assumes the underlying assets price follows
a mixed jump diusion process.
Our analysis follows that of Robert Merton
(Merton 1976) who first analyzed this subject. In general, the inclusion of a
jump process means that a contingent claims risk cannot be perfectly hedged
by trading in the underlying asset. In this situation of market incompleteness,
additional assumptions regarding the price of jump risk need to be made in
order to value derivative securities.
We show how an option can be valued
when the underlying assets jump risk is perfectly diversifiable. The problem of
option valuation when the underlying asset is the market portfolio of all assets
is also discussed.
11.1
Modeling Jumps in Continuous Time
Consider the following continuous time process
dS/S = ( k) dt + dz + (Y ) dq
(11.1)
where dz is a standard Wiener (Brownian motion) process and q (t) is a Poisson counting process that increases by 1 whenever a Poisson-distributed event
11.1. MODELING JUMPS IN CONTINUOUS TIME
313
occurs. Specifically, dq (t) satisfies:
1 if a jump occurs
dq =
0 otherwise
(11.2)
During each time interval, dt, the probability that q (t) will augment by 1
is (t) dt, where (t) is referred to as the Poisson intensity. When a Poisson
event does occur, say at date b
t, then there is a discontinuous change in S equal
to dS = (Y ) S where is a function of Y b
t , which may be a random variable
t, then
realized at date b
t.1 In other words, if a Poisson event occurs at date b

+
t = (Y ) S b
t or
dS b
t =S b
t S b

+
t
S b
t = [1 + (Y )] S b
(11.3)
Thus, if (Y ) > 0, there is an upward jump in S while if (Y ) < 0, there is a

downward jump in S. Now we can define k E[ (Y )] as the expected proportional jump given that a Poisson event occurs, so that the expected change in S
from the jump component (Y ) dq over the time interval dt is k dt. Therefore,
if we wish to let the parameter denote the instantaneous total expected rate
of return (rate of change) on S, we need to subtract o k dt from the drift
term of S:
E[dS/S]
E[( k) dt] + E[ dz] + E[ (Y ) dq]
( k) dt + 0 + k dt = dt
(11.4)
The sample path of S(t) for a process described by equation (11.1) will be

1 The date or "point" of a jump, e
t . Hence,
t, is associated with the attribute or "mark" Y e

e
t, Y e
t is referred to as a marked point process or space-time point process.
314
continuous most of the time, but can have finite jumps of diering signs and
amplitudes at discrete points in time, where the timing of the jumps depends
on the Poisson random variable q (t) and the jump sizes depend on the random
variable Y (t). If S(t) is an asset price, these jump events can be thought of as
times when important information aecting the value of the asset is released.
Jump-diusion processes can be generalized to a multivariate setting where

the process for S (t) can depend on multiple Brownian motion and Poisson jump
components. Moreover, the functions , , , and may be time varying and
depend on other variables that follow diusion or jump-diusion processes. In
particular, if (t) depends on a random state variable x (t), where, for example,
dx (t) follows a diusion process, then (t, x (t)) is called a doubly stochastic
Poisson process or Cox process.
Wolfgang Runggaldier (Runggaldier 2003)
gives an excellent review of univariate and multivariate specifications for jumpdiusion models.
univariate models.2
For simplicity, in this chapter we restrict our attention to

Let us next consider an extension of Its lemma that
covers univariate jump-diusion processes.
11.2
Its Lemma for Jump-Diusion Processes
Let c(S, t) be the value of a variable that is a twice-dierentiable function of

S(t), where S (t) follows the jump-diusion process in (11.1). For example,
c(S, t) might be the value of a derivative security whose payo depends on an
underlying asset having the current price S(t). Its lemma can be extended to
the case of mixed jump-diusion processes, and this generalization implies that
2 In Chapter 18, we consider examples of credit risk models where and are permitted
to be functions of other state variables that follow diusion processes.
11.2. ITS LEMMA FOR JUMP-DIFFUSION PROCESSES
315
the value c (S, t) follows the process
dc
cs [ ( k)S dt + S dz ] +
1
css 2 S 2 dt + ct dt
2
+ {c ([1 + (Y )] S, t) c(S, t)} dq

where subscripts on c denote its partial derivatives.
(11.5)
Note that the first line
on the right-hand-side of equation (11.5) is the standard form for Its lemma
when S (t) is restricted to following a diusion process.
The second line is
what is new. It states that when S jumps, the contingent claims value has a
corresponding jump and moves from c(S, t) to c ([1 + (Y )] S, t). Now define
c dt as the instantaneous expected rate of return on c per unit time, that is,
E[dc/c].
Also, define c as the standard deviation of the instantaneous rate
of return on c, conditional on a jump not occurring.
Then we can re-write
equation (11.5) as
dc/c = [c kc (t)] dt + c dz + c (Y ) dq
(11.6)
where
1
1
2 2
cs ( k) S + css S + ct + kc (t)
c
2
cs
S
c
= [c ([1 + (Y )] S, t) c (S, t)] /c (S, t)
kc (t) Et [c ([1 + (Y )] S, t) c (S, t)] /c (S, t)
(11.7)
(11.8)
(11.9)
(11.10)
Here, kc (t) is the expected proportional jump of the variable c (S, t) given that
a Poisson event occurs. In general, kc (t) is time varying. Let us now apply
these results to valuing a contingent claim that depends on an asset whose price
316
follows a jump-diusion process.
11.3
Valuing Contingent Claims
This section follows work by Robert Merton (Merton 1976).
For simplicity,
the analysis that follows assumes that is constant over time and that (Y ) =
(Y 1) . Thus, if a jump occurs, the discontinuous change in S is dS = (Y 1)S.
+

t =YS b
t , where b
t is the date of the jump.
In other words, S b
t goes to S b
It is also assumed that successive random jump sizes, (Ye 1), are independently
and identically distributed.
Note that if and are constants, so that the continuous component of

S(t) is lognormally distributed, then conditional upon there being n jumps in
the interval (0, t)
1 2
S(t)
= S(0) e( 2 k) t + (hzt z0 ) y(n)
(11.11)
where zt z0 N (0, t) is the change in the Brownian motion process from date
0 to date t. Jump uncertainty is reflected in the random variable ye (n), where
n
Y
Yi for n 1 where {Yi }ni=1 is a set of independent
y(0) = 1 and y(n) =
i=1
identically distributed jumps. A verification of (11.11) is left as an exercise.
Similar to a Black-Scholes hedge portfolio, let us now consider an investment

that includes a contingent claim (for example, a call option), its underlying
asset, and the riskless asset.3 Let the contingent claims price be c and assume
3 Our analysis regarding the return on a portfolio containing the underlying asset, the
contingent claim, and the risk-free asset diers somewhat from our orginal Black-Scholes
presentation because here we write the portfolios return in terms of the assets portfolio
proportions instead of units of their shares. To do this, we do not impose the requirement
that the portfolio require zero net investment (H (t) = 0), since then portfolio proportions
would be undefined. However, as before, we do require that the portfolio to be self-financing.
11.3. VALUING CONTINGENT CLAIMS
317
the underlying assets price follows the jump-diusion process given in equation
(11.1) with (Y ) = (Y 1). Furthermore, assume the risk-free interest rate
is a constant equal to r per unit time. Denote the proportions of the portfolio
invested in the underlying asset, contingent claim, and risk-free asset as 1 , 2 ,
and 3 = 1 1 2 , respectively. The instantaneous rate of return on this
portfolio, denoted dH/H, is given by:
dH/H
1 dS/S + 2 dc/c + (1 1 2 )r dt
[ 1 ( r) + 2 (c r) + r (1 k + 2 kc ) ] dt
(11.12)
+ ( 1 + 2 c ) dz + [ 1 (Y ) + 2 c (Y )] dq
11.3.1
An Imperfect Hedge
Consider the possibility of choosing 1 and 2 in order to eliminate the risk

from jumps.
Note that while jumps occur simultaneously in the asset and
the contingent claim, that is, jump risk is perfectly dependent for these two
securities, these risks are not necessarily linearly dependent. This is because
the contingent claim price, c(S, t), is, in general, a nonlinear function of the
asset price.
Unlike Brownian motion generated movements, jumps result in
non-local changes in S and c(S, t).
When the underlying assets jump size
(Y 1) is random, the ratio between the size of the jump in S and the size

of the jump in c, which is Ye / c Ye is unpredictable. Hence, a pre-
determined hedge ratio, 1 /2 , that would eliminate all portfolio risk does not
exist.4
The implication is that one cannot perfectly replicate the contingent
claims payo by a portfolio composed of the underlying asset and the risk-free
4 If the size of the jump is deterministic, a hedge that eliminates jump risk is possible.
Alternatively, Phillip Jones (Jones 1984) shows that if the underlying assets jump size has a
discrete (finite state) distribution, and a sucient number of dierent contingent claims are
written on this asset, a hedge portfolio that combines the underlying asset and these multiple
contingent claims could also eliminate jump risk.
318
asset. In this sense, the market for the contingent claim is incomplete.
Instead, suppose we pick 1 and 2 to eliminate only the risk from the continuous Brownian motion movements. This Black-Scholes hedge implies setting
1 / 2 = c / = cs S/c from our definition of c . This leads to the process
for the value of the portfolio:
dH/H
[ 1 ( r) + 2 (c r) + r ( 1 k + 2 kc )] dt
+ [ 1 (Y ) + 2 c (Y )] dq
(11.13)
The return on this portfolio is a pure jump process. The return is deterministic,
except when jumps occur. Using the definitions of , c , and 1 = 2 cs S/c,
we see that the portfolio jump term, [ 1 (Y ) + 2 c (Y )] dq, equals
h
i
c(S Y , t)c(S, t) cs (S, t) S Y S

if a jump occurs
2
c(S, t)
c(S, t)
(11.14)
otherwise
Now consider the case when the contingent claim is a European option on a
stock with a time until expiration of and a strike price X.
What would
be the pattern of profits and losses on the (quasi-) hedge portfolio?
We can
answer this question by noting that if the rate of return on the underlying asset
is independent of its price level, as is the case in (11.1), then the absence of
arbitrage restricts the option price to be a convex function of the asset price.5
The options convexity implies that c(SY, t) c(S, t) cs (S, t)[SY S] 0 for
all Y and t. This is illustrated in Figure 11.1 where the convex solid line gives
the value of a call option as a function of its underlying assets price.
From this fact and (11.14), we see that the unanticipated return on the
hedge portfolio has the same sign as 2 . This means that 1 k + 2 kc , the
5 For a proof, see Theorem 8.10 in Chapter 8 of Robert Merton (Merton 1992) which
reproduces (Merton 1973b).
319
Option
Price
c(SY)
c S[SY-S]
c(S)
S
Xe -r
SY
Asset Price
Figure 11.1: Hedge Portfolio Return with Jump

expected portfolio value jump size, also has the same sign as 2 .
Therefore,
an option writer who follows this Black-Scholes hedge by being short the option
( 2 < 0) and long the underlying asset earns, most of the time, more than the
portfolios expected rate of return. However, on those rare occasions when the
underlying asset price jumps, a relatively large loss is incurred. Thus in quiet
times, option writers appear to make positive excess returns. However, during
infrequent active times, option writers suer large losses.
11.3.2
Diversifiable Jump Risk
Since the hedge portfolio is not riskless, but is exposed to jump risk, we cannot
use the previous no-arbitrage argument to equate the hedge portfolios rate of
return to the risk-free rate. The hedge portfolio is exposed to jump risk, and, in
general, there may be a market price to such risk. One assumption might be
that this jump risk is the result of purely firm specific information and, hence,
the jump risk is perfectly diversifiable.
This would imply that the market
320
price of jump risk is zero. In this case, all of the risk of the hedge portfolio is
diversifiable, so that its expected rate of return must equal the risk-free rate, r.
Making this assumption implies
1 ( r) + 2 (c r) + r = r
(11.15)
1 / 2 = c / = (c r)/( r)
(11.16)
or
Now denote T as the maturity date of the contingent claim, and let us use the
time until maturity T t as the second argument for c (S, ) rather than
calendar time, t. Hence, c (S, ) is the price of the contingent claim when the
current asset price is S and the time until maturity of the contingent claim is .
With this redefinition, note that c = ct . Using (11.16) and substituting in
for c and c from the definitions (11.7) and (11.8), we obtain the equilibrium
partial dierential equation
h
i
1 2 2
S css + (r k)Scs c rc + Et c(S Y , ) c(S, ) = 0 (11.17)
2
For a call option, this is solved subject to the boundary conditions c(0, ) = 0
and c(S (T ) , 0) = max[S (T ) X, 0]. Note that when = 0, equation (11.17)
is the standard Black-Scholes equation which we know has the solution
b(S, , X, 2 , r) S N (d1 ) Xer N (d2 )
(11.18)
where d1 = [ ln(S/X) + (r + 12 2 ) ] / ( ) and d2 = d1 . Robert C.

Merton (Merton 1976) shows that the general solution to (11.17) is
c(S, ) =
e ( )n
Et b(S y(n) ek , , X, 2 , r)
n!
n=0
(11.19)

where, recall that y(0) = 1 and y(n) =
n
Y
i=1
321
Yi for n 1. The intuition behind the
formula in (11.19) is that the option is a probability weighted average of expected

Black-Scholes option prices. Note that if the underlying asset price followed
(11.1), then conditional on no jumps occurring over the life of the option, riskneutral valuation would imply that the Black-Scholes option price would be
b(Se k , , X, 2 , r).6 Similarly, conditional on 1 jump occurring, risk-neutral
valuation would imply that the option price would be b(Sy(1)e k , , X, 2 , r).
Conditional on two jumps, it would be b(Sy(2)e k , , X, 2 , r), and thus for
n jumps, it would be b(Sy(n)e k , , X, 2 , r).
Since
e ( )n
n!
is the probability of n jumps occurring, we see that (11.19)
is the jump-probability weighted average of expected option values conditioned

over all possible numbers of jumps.
11.3.3
Lognormal Jump Proportions
Under particular assumptions regarding the distribution of Y , solutions to

(11.19) can be calculated numerically or, in some cases, in closed-form. Here, we
consider a case which leads to a closed-form solution, namely the case in which
Y is lognormally distributed. Thus, if E[ln Y ] 12 2 where var[ln Y ] 2 ,
then E[Y ] = e = 1 + k. Hence, ln(1 + k). Given this assumption, if is
assumed to be constant, the probability density for ln[S(t + )] conditional on
the value of S(t) is
n=0
g(ln[S(t + )/S(t)] | n)h(n)
(11.20)
where g( | n) is the conditional density function given that n jumps occur during
the interval between t and t + , and h(n) is the probability that n jumps occur
between t and t + . The values of these expressions are
6 Recall that since the drift is k, and risk-neutral valuation sets = r, then k is like
a dividend yield. Hence b(Se k , , X, 2 , r) is the Black-Scholes formula for an asset with
a dividend yield of k.
322

S (t + )
g ln
|n
S(t)
h (n)
exp

2
2
)
n
n
ln[ S(t+
S(t) ] k+ 2
e ()n
n!
2 2n
p
2 2n
(11.21)
(11.22)
where 2n 2 + n 2 / is the average variance per unit time. From (11.21),

we see that conditional on n jumps occurring, ln[S(t + )/S(t)] is normally
distributed. Using the Cox-Ross risk-neutral (equivalent martingale) transformation which allows us to set = r, we can compute the date t risk-neutral
expectation of max[S (T ) X, 0], discounted by the risk-free rate, and conditional on n jumps occurring. This is given by
Et [ b(S y(n)e k , , X, 2 , r) ]
= e k (1 + k)n b(S, , X, 2n , rn )
= e k (1 + k)n bn (S, )
(11.23)
where bn (S, ) b(S, , X, 2n , rn ) and where rn r k + n/. The actual

value of the option is then the weighted average of these conditional values,
where each weight equals the probability that a Poisson random variable with
characteristic parameter will take on the value n. Defining 0 (1 + k),
this equals
c(S, )
=
=
11.3.4
X
e ( )n k
e
(1 + k)n bn (S, )
n!
n=0
0
X
e (0 )n
bn (S, )
n!
n=0
Non-Diversifiable Jump Risk
(11.24)
323
In some circumstances, it is unrealistic to assume that jump risk is non-priced

risk. For example, David Bates (Bates 1991) investigates the U.S. stock market
crash of 1987, an event that certainly was not firm specific but aected the entire
market for equities. Similar work by Vasanttilak Naik and Moon Lee (Naik and
Lee 1990) considers non-diversifiable jump risk. The models in these articles
assume that aggregate wealth in the economy follows a mixed jump-diusion
process. This could result from a representative agent Cox, Ingersoll, and Ross type production economy in which technologies follow a jump-diusion process
and individuals select investments in these technologies such that their optimally
invested aggregate wealth follows a mixed jump-diusion process (Bates 1991).
Or it can simply be assumed that the economy is a Lucas-type endowment
economy and there is an exogenous firm dividend process that follows a mixedjump diusion process, and these dividends cannot be invested but must be
consumed (Naik and Lee 1990).
In both articles, jumps in aggregate wealth or consumption (endowment)
are assumed to be of the lognormal type that we assumed earlier. Further,
representative individuals are assumed to have constant relative risk aversion
utility. These assumptions allow the authors to solve for the general equilibrium
price of jump risk. Given this setup, contingent claims, which are assumed to
be in zero net supply, can be priced. For example, the formula for a call option
derived by Bates has a series solution that is similar in form to (11.24).
11.3.5
Black-Scholes versus the Jump-Diusion Model
Having derived a model for pricing options written on an underlying asset whose
price follows a jump-diusion process, the natural question to ask is whether this
makes any dierence vis-a-vis the Black-Scholes option pricing model which does
not permit the underlyings price to jump. The answer is yes, and the jump-
324
diusion model appears to better fit the actual prices of many options written
on stocks, stock indices, and foreign exchange. In most types of options, the
Black-Scholes model underprices out-of-the money and in-the-money options
relative to at-the-money-options. What this means is that the prices of actual
options whose exercise price is substantially dierent from the current price of
the underlying are priced higher than the theoretical Black-Scholes price while
the prices of actual options whose exercise price is close to the current price of
the underlying are priced lower than the theoretical Black-Scholes price. This
phenomenon has been described as a volatility smile or volatility smirk.7
This empirical deficiency can be traced to the Black-Scholes models assumption that the underlyings terminal price has a risk-neutral distribution
that is lognormal.
Apparently, investors price actual options under the be-
lief that the risk-neutral distribution has much fatter "tails" than those of the
lognormal distribution. In other words, investors price securities as if they believe that extreme asset prices are more likely than what would be predicted by
a lognormal distribution because actual in- and out-of-the-money options are
priced relatively high versus the Black-Scholes theoretical prices. A model that
permits the underlying assets price to jump, with jumps possibly being both
positive and negative, can generate a distribution for the assets price that has
fatter tails than the lognormal. The possibility of jumps makes extreme price
changes more likely and, indeed, the jump-diusion option pricing model can
7 Note that if the Black-Scholes model correctly priced all options having the same maturity
date and the same underlying asset but dierent exercise prices, there would be one volatility
parameter, , consistent with all of these options. However, the implied volatilities, , needed
to fit in- and, especially, out-of the money call options are greater than the volatility parameter
needed to fit at-the-money options. Hence, when implied volatility is graphed versus call
options exercise prices, it forms an inverted hump or "smile," or, in the case of equity index
options, a downward sloping curve or "smirk." These characteristics of option prices are
equivalent to the Black-Scholes model giving relatively low prices for in- and out-of-the-money
options because options prices are increasing functions of the underlyings volatility, . The
Black-Scholes model needs relatively high estimated volatility for in- and out-of-the-money
options versus at-the-money options. If a (theoretically correct) single volatility parameter
was used for all options, in- and out-of-the-money options would be relatively underpriced by
the model. See Hull (Hull 2000) for a review of this issue.
11.4. SUMMARY
325
better match the market prices of many types of options.

However, there are other aspects of actual option prices for which even
the standard jump-diusion model cannot account.
The volatility parame-
ters implied by actual option prices change over time and appear to follow a
mean-reverting stochastic process. To account for this empirical time variation,
stochastic volatility option pricing models have been developed. These models
start by assuming that the underlying asset price follows a diusion process such
as dS/S = dt + dz but where the volatility, , is stochastic. The volatility
follows a mean reverting process of the form d = () dt + () dz where
dz is another Brownian motion process possibly correlated with dz. Similar
to the jump-diusion model, one must assign a market price of risk associated
with the volatility uncertainty reflected in the dz term.8
While stochastic volatility option pricing models also produce fatter-tailed
distributions relative to the lognormal, empirically these distributions do not
tend to be fat enough to explain volatility smiles and smirks. To capture both
time variation in volatilities and cross-sectional dierences in volatility due to
dierent degrees of "moneyness" (volatility smiles or smirks), it appears that
an option pricing model that allows for both stochastic volatility and jumps
is required.9 For recent reviews of the empirical option pricing literature, see
(Bates 2002) and (Bakshi, Cao, and Chen 1997).
11.4
Summary
Allowing for the possibility of discontinuous movements can add realism to the
modeling of asset prices.
For example, a firms stock price might experience
a sudden, large change upon the public announcement that it is involved in a

8 Steven
Heston (Heston 1993) developed a popular stochastic volatility model.

Bates (Bates 1996) derived an option pricing model that combines both jumps and
stochastic volatility and estimated its parameters using options on foreign exchange.
9 David
326
corporate merger. While the mixed jump-diusion process captures such asset
price dynamics, it complicates the valuation of contingent claims written on
such an asset. In general, we showed that the contingent claims payo cannot
be perfectly replicated by a dynamic trading strategy involving the underlying
asset and risk-free asset. In this situation of market incompleteness, additional
theory that assigns a market risk premium to jump risk is required to determine
the contingent claims value.
The additional complications in deriving jump-diusion models of option
pricing appear worthwhile. Because jumps increase the likelihood of extreme
price movements, they generate a risk-neutral distribution of asset prices whose
tails are fatter than the Black-Scholes models lognormal distribution. Since the
actual prices of many types of options appear to reflect significant probabilities
of extreme movements in the underlyings price, the jump-diusion model has
better empirical performance.
Having seen that pricing contingent claims sometimes requires specifying
market prices of risk, the following chapters turn to the subject of deriving
equilibrium risk premia for assets in continuous time economies.
As a pre-
liminary, we re-visit the individuals consumption and portfolio choice problem

when asset prices, and the individuals consumption and portfolio choices, can
change continuously.
Based on this structure of consumption and asset de-
mands, we then derive assets risk premia in a general equilibrium, continuous

time economy.
11.5
Exercises
1. Verify that (11.11) holds by using Its lemma to find the process followed
by ln (S (t)).
11.5. EXERCISES
327
2. Let S (t) be the U.S. dollar price of a stock. It is assumed to follow the
process

dS/S = [s k] dt + s dzs + Ye dq
(*)
where dzs is a standard Wiener process, q (t) is a Poisson counting process,

and Ye = Ye 1 . The probability that q will jump during the time

h
i
interval dt is dt. k E Y 1 is the expected jump size. Let F be the
foreign exchange rate between U.S. dollars and Japanese yen, denominated
as U.S. dollars per yen. F follows the process
dF/F = f dt + f dzf
where dzs dzf = dt. Define x (t) as the Japanese yen price of the stock
whose U.S. dollar price follows the process in (*). Derive the stochastic
process followed by x (t).
3. Suppose that the instantaneous-maturity, default-free interest rate follows
the jump-diusion process
dr(t) = [ r(t)] dt + dz + r (Y ) dq
where dz is a standard Wiener process and q (t) is a Poisson counting process
having the arrival rate of dt. The arrival of jumps is assumed to be independent
of the Wiener process, dz. (Y ) = (Y 1) where Y > 1 is a known positive
constant.
3.a Define P (r, ) as the price of a default-free discount bond that pays $1
in periods. Using Its lemma for the case of jump-diusion processes,
write down the process followed by dP (r, ).
328
3.b Assume that the market price of jump risk is zero, but that the market
price of Brownian motion (dz) risk is given by , so that = [p r(t)] /p ,
where p (r, ) is the expected rate of return on the bond and p ( ) is
the standard deviation of the bonds rate of return from Brownian motion
risk (not including the risk from jumps). Derive the equilibrium partial
dierential equation that the value P (r, ) must satisfy.
4. Suppose that a securitys price follows a jump-diusion process and yields
a continuous dividend at a constant rate of dt. For example, its price,
S (t), follows the process

dS/S = [ (S, t) k ] dt + (S, t) dz + Ye dq
where q (t) is a Poisson counting process and Ye = Ye 1 . Also let

h
i
k E Ye 1 , let the probability of a jump be dt, and denote (S, t) as the
assets total expected rate of return. Consider a forward contract written on this
security that is negotiated at date t and matures at date T where = T t > 0.

Let r(t, ) be the date t continuously-compounded risk-free interest rate for
borrowing or lending between dates t and T .
Assuming that one can trade
continuously in the security, derive the equilibrium date t forward price using
an argument that rules out arbitrage. (Hint: Some information in this problem
is extraneous. The solution is relatively simple.)
Part IV
Asset Pricing in Continuous Time
Chapter 12
Continuous Time
Consumption and Portfolio
Choice
Until now, our applications of continuous time stochastic processes have focused
on the valuation of contingent claims. We now re-consider the topic introduced
in Chapter 5, namely, an individuals intertemporal consumption and portfolio choice problem.
However, rather than assume a discrete time setting, we
now examine this problem where asset prices are subject to continuous, random
changes and an individual can adjust consumption and portfolio allocations at
any time.
Specifically, this chapter assumes that an individual maximizes a
time-separable expected utility function that depends on the rate of consumption at all future dates.
The savings of this individual are allocated among
assets whose prices follow the sorts of continuous time stochastic processes first
introduced in Chapter 8. Hence, in this environment, the values of the individ325
326
CHAPTER 12. CONTINUOUS TIME PORTFOLIO CHOICE
uals portfolio holdings and total wealth change constantly and, in general, it is
optimal for the individual to make continuous re-balancing decisions.
The continuous time consumption and portfolio choice problem described
above was formulated and solved in two papers by Robert Merton (Merton 1969)
and (Merton 1971). This work was the foundation of his model of intertemporal asset pricing (Merton 1973a) that we will study in the next chapter. As is
discussed below, allowing individuals to rebalance their portfolios continuously
can lead to qualitatively dierent portfolio choices compared to restricting rebalancing to only discrete dates. This, in turn, means that the asset pricing
implications of individuals decisions in continuous time can dier from those
of a discrete time model.
Continuous trading may enable markets to be dy-
namically complete and lead to sharper asset pricing results. For this reason,
continuous time consumption and portfolio choice models are often used in financial research on asset pricing. Much of our analysis in later chapters will
be based on such models.
By studying consumption and portfolio choices in continuous time, the eects
of time variation in assets return distributions, that is, changing investment
opportunities, become transparent.
As will be shown, individuals portfolio
choices include demands for assets that are the same as those derived from
the single period mean-variance analysis of Chapter 2.
However, portfolio
choices also include demands for assets that hedge against changes in investment
opportunities. This is a key insight which dierentiates single-period and multiperiod models and has implications for equilibrium asset pricing.
The next section outlines the assumptions of an individuals consumption
and portfolio choice problem for a continuous time environment.
Then, sim-
ilar to what was done in solving for an individuals decisions in discrete time,
we introduce and apply a continuous time version of stochastic dynamic pro-
12.1. MODEL ASSUMPTIONS
327
gramming to derive consumption and portfolio demands. Portfolio behaviors

both for constant investment opportunities and for changing investment opportunities are analyzed. We next present an alternative martingale approach to
finding an individuals optimal consumption and portfolio choices.
This al-
ternative martingale technique is most applicable to situations when markets

are dynamically complete. We illustrate this solution method for an example
where an individual faces risky asset returns that are negatively correlated with
investment opportunities.
12.1
Model Assumptions
Let us assume that an individual chooses levels of consumption and savings,

where his savings can be invested in n dierent risky assets plus a risk-free
asset.
We define Ct as the individuals level of consumption per unit time
at date t, and Si (t) as the price of risky asset i at date t, where i = 1, ..., n.
The instantaneous rate of return on the ith risky asset is assumed to satisfy the
process
dSi (t) / Si (t) = i (x, t) dt + i (x, t) dzi
where i = 1, ..., n, and (i dzi )(j dzj ) = ij dt.
(12.1)
In addition, we denote the
instantaneous risk-free return as r (x, t). For simplicity, risky assets are assumed
to pay no cashflows (dividends or coupon payments), so that their total returns
are given by their price changes.1
It is assumed that i , i , and r can be
functions of time and a k 1 vector of state variables, which we denote by

x(t). When the i , i , and/or r are time-varying, the investor is said to face
changing investment opportunities. The state variables aecting the moments
of the asset prices, x (t), can, themselves, follow diusion processes. Let the ith
1 As mentioned in Chapter 9, this is not a critical assumption since if an asset paid a
dividend that was re-invested into new shares of the asset, then equation (12.1) would represent
the percentage change in the value of the asset holding and, thus, the total rate of return.
328
state variable follow the process
dxi = ai (x, t) dt + bi (x, t) d i
(12.2)
i = 1, ..., k, where d i is a Brownian motion process and (bi d i )(bj d j ) =

bij dt and (i dzi )(bj d j ) = ij dt. Hence, equations (12.1) and (12.2) indicate
that up to n + k sources of uncertainty (Brownian motion processes) aect the
distribution of asset returns.
We denote the value of the individuals wealth portfolio at date t as Wt .
Similar to our analysis in Chapter 9 and treating consumption as a net cash
outflow from the individuals wealth portfolio, we can write the dynamics of
wealth as2
dW
=
=
"
n
X
i=1
n
X
i=1
i dSi /Si + 1
n
X
i=1
! #
i r W Cdt
i (i r)W dt + (rW C) dt +
n
X
(12.3)
i W i dzi
i=1
We can now state the individuals intertemporal consumption and portfolio

choice problem:
max
Cs ,{ i,s },s,i
Et
"Z
U (Cs , s) ds + B(WT , T )
(12.4)
subject to the constraint (12.3).

The utility function, U (Ct , t), is assumed to be strictly increasing and concave in Ct and the bequest function, B(WT , T ), is assumed to be strictly in2 Our presentation assumes that there are no other sources of wealth, such as wage income.
If the model is extended to include a flow of non-traded wage income received at date t, say
yt , it could be incorporated into the individuals intertemporal budget constraint in a manner
similar to that of consumption but with an opposite sign. In other words, the term (Ct yt )
would replace Ct in our derivation of the individuals dynamic budget constraint. Due,
Fleming, Soner, and Zariphopoulou (Due, Fleming, Soner, and Zariphopoulou 1997) solve
for optimal consumption and portfolio choices when the individual receives stochastic wage
income.
12.2. CONTINUOUS TIME DYNAMIC PROGRAMMING

creasing and concave in terminal wealth, WT .
329
This problem, in which the
individual has time-separable utility of consumption, is analogous to the discrete time problem studied in Chapter 5. The variables Ws and x (s) are the
date s state variables while the individual chooses the control variables Cs and
i (s), i = 1, ..., n, at each date s over the interval from dates t to T .
Note that some possible constraints have not been imposed. For example,
one might wish to impose the constraint Ct 0 (nonnegative consumption)
and/or i 0 (no short sales). However, for some utility functions, negative
consumption is never optimal, so that solutions satisfying Ct 0 would result
even without the constraint.3
Before we attempt to solve this problem, lets digress to consider how stochastic dynamic programing applies to a continuous time setting.
12.2
Continuous Time Dynamic Programming
To illustrate the principles of dynamic programming in continuous time, consider

a simplified version of the problem specified in (12.3) to (12.4) where there is
only one choice variable:
max Et
{c}
"Z
U (cs , xs ) ds
(12.5)
subject to
dx = a(x, c) dt + b(x, c) dz
(12.6)
where ct is a control variable (such as a consumption and/or vector of portfolio

proportions) and xt is a state variable (such as wealth and/or a variable that
3 For
example, if lim
Ct 0
U (Ct ,t)
Ct
= , as would be the case if the individuals utility
displayed constant relative risk aversion (power utility), then the individual would always
avoid non-positive consumption. However, other utility functions, such as constant absoute
risk aversion (negative exponential) utility, do not display this property.
330
changes investment opportunities, that is, a variable that aects the i s and/or
i s). Define the indirect utility function, J(xt , t), as
J(xt , t)
= max Et
{c}
"Z
U (cs , xs ) ds
= max Et
{c}
"Z
t+t
U(cs , xs ) ds +
(12.7)
Z
U(cs , xs ) ds .
t+t
Now let us apply Bellmans Principle of Optimality. Recall that this concept
says that an optimal policy must be such that for a given future realization of
the state variable, xt+t , (whose value may be aected by the optimal control
policy at date t and earlier), any remaining decisions at date t + t and later
must be optimal with respect to xt+t . In other words, an optimal policy must
be time consistent. This allows us to write
J(xt , t) = max Et
{c}
max Et
{c}
"Z
t+t
U (cs , xs ) ds + max Et+t

{c}
"Z
t+t
"Z
##
U (cs , xs ) ds
t+t
U (cs , xs ) ds + J(xt+t , t + t) .
(12.8)
Now think of t as a short interval of time and approximate the first integral
as U (ct , xt ) t. Also expand J(xt+t , t + t) around the points xt and t in a
Taylor series to get
J(xt , t)
= max Et [ U (ct , xt ) t + J(xt , t) + Jx x + Jt t
(12.9)
1
1
+ Jxx (x)2 + Jxt (x)(t) + Jtt (t)2 + o(t)
2
2
{c}
12.2. CONTINUOUS TIME DYNAMIC PROGRAMMING
331
y(t)
t0 t
where o (t) represents higher order terms, say y (t), where lim
= 0.
The state variable process (12.6) can be approximated as
x a(x, c)t + b(x, c)z + o(t)

where z =
(12.10)
te
and e
N (0, 1). Substituting (12.10) into (12.9), and sub-
tracting J(xt , t) from both sides, one obtains:
0 = max Et [U (ct , xt )t + J + o(t)]

{c}
where
J =
1
Jt + Jx a + Jxx b2 t + Jx bz.
2
Equation (12.12) is just a discrete time version of Its Lemma.
(12.11)
(12.12)
Next, note
that in equation (12.11) the term Et [Jx bz] = 0 and then divide both sides of
(12.11) by t. Finally, take the limit as t 0 to obtain
1
0 = max U (ct , xt ) + Jt + Jx a + Jxx b2
{c}
2
(12.13)
which is the stochastic, continuous-time Bellman equation analogous to the

discrete time Bellman equation (5.16). It is sometimes written as
0 = max [U(ct , xt ) + L[J] ]

{c}
(12.14)
where L[] is the Dynkin operator. This operator is the drift term (expected
change per unit time) in dJ(x, t) that one obtains by applying Its Lemma to
J(x, t).
In summary, equation (12.14) gives us a condition that the optimal
stochastic control policy, ct , must satisfy. Let us now return to the complete
consumption and portfolio choice problem and apply this solution technique.
332
12.3
Solving the Continuous Time Problem
Define the indirect utility of wealth function, J(W, x, t), as
J(W, x, t) =
max
Cs ,{ i,s },s,i
Et
"Z
U (Cs , s) ds + B(WT , T )
(12.15)
and define L as the Dynkin operator with respect to the state variables W and
xi , i = 1, . . . , k. In other words
L [J]
J
+
t
+
"
n
X
i=1
n
n
1 XX
i (i r)W + (rW C)
k
ij i j W 2
i=1 j=1
k X
n
X
+
W i ij
j=1 i=1
k
X
J
J
+
ai
W
xi
i=1
1 XX
2J
2J
+
b
ij
W 2
2 i=1 j=1 xi xj
2J
W xj
(12.16)
Thus, using (12.14), we have
0 =
max [U (Ct , t) + L[J]] .
Ct ,{ i,t }
(12.17)
Given the concavity of U and B, equation (12.17) implies that the optimal
choices of Ct and i,t satisfy the conditions we obtain from dierentiating
U (Ct , t) + L[J] and setting the result equal to zero. Hence, the first order
conditions are:
0 =
0 =
J (W, x, t)
U (C , t)
C
W
(12.18)
n
k
J
2J X
2J X
2
(i r)W +
W
+
W ij , i = 1, . . . , n
ij
j
W
W 2 j=1
xi W j=1
(12.19)
12.3. SOLVING THE CONTINUOUS TIME PROBLEM
333
Equation (12.18) is the envelope condition that we earlier derived in a discrete time framework as equation (5.20) while equation (12.19) has the discrete time analogue (5.21).
1
G = [U/C]
Defining the inverse marginal utility function as
, condition (12.18) can be re-written as
C = G (JW , t)
where we write JW as short hand for J/W .
(12.20)
Also, the n linear equations
in (12.19) can be solved in terms of the optimal portfolio weights.
Denote
[ij ] to be the n n covariance matrix whose i, j th element is ij , and

denote the i, j th element of the inverse of to be ij , that is, 1 [ ij ].
Then the solutions to (12.19) can be written as
i =
JW
JW W W
n
X
j=1
ij (j r)
n
k X
X
JW xm
ij , i = 1, . . . , n (12.21)
J
W jm
m=1 j=1 W W
Note that the optimal portfolio weights in (12.21) depend on JW / (JW W W )

which is the inverse of relative risk-aversion for lifetime utility of wealth.
Given particular functional forms for U and the i s, ij s, and ij s, equations (12.20) and (12.21) are functions of the state variables W , x, and derivatives of J, that is, JW , JW W , and JW xi . They can be substituted back into
equation (12.17) to obtain a non-linear partial dierential equation (PDE) for
J. For some specifications of utility and the processes for asset returns and the
state variables, this PDE can be solved to obtain an analytic expression for J
which, in turn, allows for explicit solutions for C and the i based on (12.20)
and (12.21). Examples of such analytical solutions are given in the next two
sections.
In general, however, one must resort to numerical solutions for J,
and, therefore, C and the i .4

4 Techniques for solving partial dierential equations numerically are covered in Carrier and
Peason (Carrier and Pearson 1976), Judd (Judd 1998), and Rogers and Talay (Rogers and
334
12.3.1
Constant Investment Opportunities
Let us consider the special case for which asset prices are lognormally distributed, that is, all of the i s (including r) and i s are constants.5 This means
that each assets expected rate of return and variance of its rate of return do
not change; there is a constant investment opportunity set. Hence, investment
and portfolio choice decisions are independent of the state variables, x, since
they do not aect U , B, the i s, or the i s. The only state variable aecting
consumption and portfolio choice decisions is wealth, W .
This simplifies the
above analysis, since now the indirect utility function J depends only on W and
t, but not x.
For this constant investment opportunity set case, the optimal portfolio
weights in (12.21) simplify to:
JW
JW W W
n
X
j=1
ij (j r), i = 1, . . . , n.
(12.22)
Plugging (12.20) and (12.22) back into the optimality equation (12.17), and
using the fact that [ ij ] 1 , we have
0 = U (G, t) + Jt + JW (rW G)
n
n X
2
X
JW
ij (i r)(j r). (12.23)
2JW W i=1 j=1
The non-linear partial dierential equation (12.23) may not have an analytic
solution for an arbitrary utility function, U. However, we can still draw some
conclusions about the individuals investment behavior by looking at equation
(12.22).
This expression for the individuals optimal portfolio weights has
Talay 1997).
5 Recall that if and are constants, then dS /S follows geometric Brownian motion and
i
i
i
i
1 2
Si (t) = Si (0) e(i 2 i )t+ i (zi (t)zi (0)) is lognormally distributed over any discrete period
since zi (t)zi (0) N (0, t). Therefore, the return on a unit initial investment over this period,
1 2
distributed.
The continuouslySi (t) /Si (0) = e(i 2 i )t+ i (zi (t)zi (0)) , is also lognormally

compounded rate of return, equal to ln [Si (t) /Si (0)] = i 12 2i t + i (zi (t) zi (0)) is
normally distributed.
335
an interesting implication, but one that might be intuitive given a constant

investment opportunity set. Since ij , j , and r are constants, the proportion
of each risky asset that is optimally held will be proportional to JW /(JW W W ),
which only depends on the total wealth state variable, W . Thus, the proportion
of wealth in risky asset i to risky asset k is a constant, that is,
n
X
i
j=1
= n
X
k
j=1
ij (j r)
(12.24)
kj (j r)
and the proportion of risky asset k to all risky assets is
k = Pn
i=1 i
n
X
kj (j
j=1
n X
n
X
i=1 j=1
r)
.
(12.25)
ij (j r)
This means that each individual, no matter what her utility function, allocates
her portfolio between the risk-free asset, paying return r, and a portfolio of
the risky assets that holds the n risky assets in constant proportions, given by
(12.25).
Hence, two mutual funds, one holding only the the risk-free asset
and the other holding a risky asset portfolio with the weights in (12.25) would
satisfy all investors. Only the investors preferences, current level of wealth, Wt ,
and the investors time horizon determine the amounts allocated to the risk-free
fund and the risky one.
The implication is that with a constant investment opportunity set, one can
think of the investment decision as being just a two-asset decision, where the
choice is between the risk-free asset paying rate of return r and a risky asset
336
having expected rate of return and variance 2 where
n
X
i i
i=1
n X
n
X
(12.26)
i j ij .
i=1 j=1
These results are reminiscent of those derived from the single period meanvariance analysis of Chapter 2. In fact, the relative asset proportions given in
(12.24) and (12.25) are exactly the same as those implied by the single period
mean-variance portfolio proportions given in (2.42).6 The instantaneous means
and covariances for the continuous time asset price processes simply replace the
previous means and covariances of the single period multi-variate normal asset
returns distribution.
Again, we can interpret all investors choosing along an
ecient frontier, where the tangency portfolio is given by the weights in (12.25).
But what is dierent in this continuous time analysis is the assumption regarding
the distribution in asset prices.
In the discrete time mean-variance analysis,
we needed to assume that asset returns were normally distributed, whereas in

the continuous time context we specified that asset returns were lognormally
distributed. This latter assumption is more attractive since most assets, like
bonds and common stocks, have limited liability so that their values cannot
become negative. The assumption of a lognormal return distribution embodies
this restriction, whereas the assumption of normality does not.
The intuition for why we obtain the single-period Markowitz results in a
continuous-time setting with lognormally distributed asset returns is as follows.
By allowing continuous-rebalancing, an individuals portfolio choice horizon is
that the ith element of (2.42) can be written as wi =
equals (12.22) when = JW / (JW W W ).
6 Note
Sn
j=1

ij Rj Rf , which
337
essentially a very short one, that is the "period" is instantaneous. Since diusion processes can be thought of as being instantaneously (locally) normally distributed, our continuous-time environment is as if the individual faces a infinite
sequence of similar short portfolio selection periods with normally distributed
asset returns.
Lets now look at a special case of the above general solution. Specifically,
we assume that utility is of the hyperbolic absolute risk aversion (HARA) class.
HARA Utility
Recall from Chapter 1 that HARA utility functions are defined by
U (C, t) = et
C
+
1
(12.27)
and that this class of utility nests power (constant relative risk-aversion) utility,
exponential (constant absolute risk aversion), and quadratic utility. Robert C.
Merton (Merton 1971) derived explicit solutions for this class of utility functions.
With HARA utility, optimal consumption given in equation (12.20) becomes
1
1 et JW 1
(1 )
=
(12.28)
and using (12.22) and (12.26), the proportion put in the risky asset portfolio is
JW r
.
JW W W 2
(12.29)
This solution is incomplete since C and are in terms of JW and JW W .

However, we solve for J in the following manner. Substitute (12.28) and (12.29)
into the optimality equation (12.17), or, alternatively, directly simplify equation
338
(12.23) to obtain

(1 )2 t et JW 1
e
+ Jt
2
(1 )
JW
( r)2
+ rW JW
+
.
JW W 22
(12.30)
This is the partial dierential equation for J that can be solved subject to
a boundary condition for J(W, T ).
Let us assume a zero bequest function,
B 0, so that the appropriate boundary condition is J(W, T ) = 0.
The
nonlinear partial dierential equation in (12.30) can be simplified by a change
in variable Y = J 1 . This puts it in the form of a Bernoulli-type equation and

an analytic solution exists. The expression for the general solution is lengthy
and can be found in (Merton 1971). Given this solution for J, one can then
calculate JW to solve for C and also calculate JW W to solve for .
It is
interesting to note that for this class of HARA utility, C is of the form
Ct = aWt + b
(12.31)
and
t = g +
h
Wt
(12.32)
where a, b, g, and h are, at most, functions of time. For the special case of
constant relative risk aversion where U (C, t) = et C /, the solution is
J (W, t) = e
Ct =
1 ea(T t)
a
a
ea(T t)
Wt
W /
(12.33)
(12.34)
and
=
r
(1 ) 2
(12.35)

where a
(r)2
2(1)2
339
i
. Note that the individuals optimal portfolio
weight for the risky asset is independent of the time horizon, T , and declines as
risk aversion increases ( becomes more negative). For the case of an infinite
horizon, a solution exists only if a > 0. In this case we can see that by taking
the limit at T becomes infinite J (W, t) = et a1 W / and consumption is a
constant proportion of wealth, Ct = aWt .
12.3.2
Changing Investment Opportunities
Next, let us generalize the individuals consumption and portfolio choice problem
by considering the eects of changing investment opportunities. To keep the
analysis fairly simple, assume that there is a single state variable, x. That is,
let k = 1 so that x is a scalar.
We also simplify the notation by writing its
process as
dx = a (x, t) dt + b (x, t) d
(12.36)
where b d i dzi = i dt. This allows us to write the optimal portfolio weights
in (12.21) as:
i =
n
n
JW X
JW x X
ij j r
ij j , i = 1, . . . , n
W JW W j=1
W JW W j=1
(12.37)
or, written in matrix form,
A 1
H 1
( re) +

W
W
(12.38)
where = ( 1 ... n )0 is the n1 vector of portfolio weights for the n risky assets,
= (1 ...n )0 is the n 1 vector of these assets expected rates of return, e is an
n-dimensional vector of ones, = (1 , ..., n )0 , A = JJWWW , and H = JJWWWx .
340
We will use bold type to denote vector or matrix variables, while normal type
is used for scalar variables.
Note that A and H will, in general, dier from one individual to another,
depending on the form of the particular individuals utility function and level
of wealth. Thus, unlike in the constant investment opportunity set case (where
JW x = H = 0), i /j is not the same for all investors, that is, a Two Mutual
Fund Theorem does not hold. However, with one state variable, x, a Three-Fund
Theorem does hold. Investors will be satisfied choosing between a fund holding
only the risk-free asset, a second fund of risky assets that provides optimal
instantaneous diversification, and a third fund composed of a portfolio of the
risky assets that has the maximum absolute correlation with the state variable,
x.
The portfolio weights of the second fund are 1 ( re) and are the
same ones representing the mean-variance ecient tangency portfolio that were
derived for the case of constant investment opportunities. The portfolio weights
for the third fund are 1 .
Note that these weights are of the same form
equation (2.64), which are the hedging demands derived in Chapter 2s crosshedging example.
They are the coecients from a regression (or projection)
of changes in the state variable on the returns of the risky assets. A/W and
H/W , which depend on the individuals preferences, then determine the relative
amounts that the individual invests in the second and third risky portfolios.
To gain more insight regarding the nature of the individuals portfolio holdings, recall the envelope condition JW = UC , which allows us to write JW W =
UCC C/W . Therefore, A can be re-written as
A=
UC
>0
UCC (C/W )
by the concavity of U. Also, since JW x = UCC C/x, we have
(12.39)
H=
C/x
R0
C/W
341
(12.40)
Now the first vector of terms on the right-hand side of (12.38) is the usual
demand functions for risky assets chosen by a single-period mean-variance utility
maximizer. Since A is proportional to the reciprocal of the individuals absolute
risk aversion, we see that the more risk averse is the individual, the smaller is
A and the smaller in magnitude is the individuals demand for any risky asset.
The second vector of terms on the right-hand side of (12.38) captures the
individuals desire to hedge against "unfavorable" shifts in investment opportunities that would reduce optimal consumption. An unfavorable shift is defined
as a change in x such that consumption falls for a given level of current wealth,
that is, an increase in x if C/x < 0 and a decrease in x if C/x > 0. For
example, suppose that is a diagonal matrix, so that ij = 0 for i 6= j, and
ii = 1/ii > 0, and also assume that i 6= 0.7 Then in this special case the
hedging demand term for risky asset i in (12.38) simplifies to
H ii i =
C/x
C
ii i > 0 i
<0
C/W
x i
(12.41)
Condition (12.41) says that if an increase in x leads to a decrease in optimal

consumption (C/x < 0) and if x and asset i are positively correlated (i > 0),
then there is a positive hedging demand for asset i, that is, H ii i > 0 and asset
i is held in greater amounts than what would be predicted based on a simple
single-period mean - variance analysis. The intuition for this result is that by
holding more of asset i, one hedges against a decline in future consumption
due to an unfavorable shift in x. If x increases, which would tend to decrease
consumption (C/x < 0), then asset i would tend to have a high return (i >
0), which by augmenting wealth, W , helps neutralize the fall in consumption
7 Alternatively,
assume is non-diagonal, but that j = 0 for j 6= i.
342
(C/W > 0). Hence, the individuals optimal portfolio holdings are designed
to reduce fluctuations in consumption over his planning horizon.
The Special Case of Logarithmic Utility
Let us continue to assume that there is a single state variable aecting investment opportunities, but now also specify that the individual have logarithmic utility and a logarithmic bequest function, so that in (12.15) U (Cs , s) =
es ln (Cs ) and B (WT , T ) = eT ln (WT ). Logarithmic utility is one of the
few cases in which analytical solutions for consumption and portfolio choices
can be obtained when investment opportunities are changing.
To derive the
solution to (12.17) for log utility, let us consider a "trial" solution for the
indirect utility function of the form J (W, x, t) = d (t) U (Wt , t) + F (x, t) =
d (t) et ln (Wt ) + F (x, t). Then optimal consumption in (12.20) would be
Ct =
Wt
d (t)
(12.42)
and the first order conditions for the portfolio weights (12.37) simplify to
i =
n
X
j=1
ij j r
(12.43)
since JW x = 0. Substituting these conditions into the Bellman equation (12.17),

it becomes
343
0 = U (Ct , t) + Jt + JW [rWt Ct ] + a (x, t) Jx

n X
n
2
X
1
JW
+ b (x, t)2 Jxx +
ij j r (i r)
2
2JW W i=1 j=1
Wt
t
t d (t)
d (t) ln [Wt ] + Ft + et d (t) r et
ln
= e
+e
d (t)
t
n
n
1
d (t) et X X
+a (x, t) Fx + b (x, t)2 Fxx
ij j r (i r)
2
2
i=1 j=1
(12.44)
or
d (t)
d (t) ln [Wt ] + et Ft + d (t) r 1
0 = ln [d (t)] + 1 +
t
n
n
1
d (t) X X
2
ij j r (i r)
+a (x, t) et Fx + b (x, t) et Fxx
2
2 i=1 j=1
(12.45)
Note that a solution to this equation must hold for all values of wealth. Hence,
it must be the case that
d (t)
d (t) + 1 = 0
t
(12.46)
subject to the boundary condition d (T ) = 1. The solution to this first order

ordinary dierential equation is
d (t) =
i
1h
1 (1 ) e(T t)
(12.47)
344
The complete solution to (12.45) is then to solve
0 = ln [d (t)] + et Ft + d (t) r 1 + a (x, t) et Fx

n
n
1
d (t) X X
+ b (x, t)2 et Fxx
ij j r (i r)
2
2 i=1 j=1
(12.48)
subject to the boundary condition F (x, T ) = 0 and where (12.47) is substituted

in for d (t).
The solution to (12.48) depends on how r, the i s, and are
assumed to depend on the state variable x.
However, whatever assumptions
are made regarding these variables relationships to the state variable x, they
will influence only the level of indirect utility via the value of F (x, t) and will
not change the form of the optimal consumption and portfolio rules. Thus, this
verifies that our trial solution is, indeed, a valid form for the solution to the
individuals problem. Substituting (12.47) into (12.42), consumption satisfies
Ct =
Wt
1 (1 ) e(T t)
(12.49)
which is the continuous time counterpart to the log utility investors optimal
consumption that we derived for the discrete time problem in Chapter 5, equation (5.34). Note, also, that the log utility investors optimal portfolio weights
given in (12.43) are of the same form as in the case of a constant investment
opportunity set, equation (12.35) with = 0. The log utility investor may be
described as behaving "myopically" in that, similar to the discrete time case,
she has no desire to hedge against changes in investment opportunities.8 However, note that even with log utility, a dierence from the constant investment
opportunity set case is that since r, the i s, and depend, in general, on the
constantly changing state variable xt , the portfolio weights in equation (12.43)
8 The portfolio weights for the discrete time case are given by (5.35). As discussed earlier,
the log utility investor acts myopically because income and substitution eects from changing
investment opportunities exactly cancel for this individual.
12.4. THE MARTINGALE APPROACH
345
vary over time.

Recall that log utility is a very special case, and, in general, other utility specifications lead to consumption and portfolio choices that reflect desires
to hedge against investment opportunities.
An example is given in the next
section. After introducing an alternative solution technique, we solve for the

consumption and portfolio choices of an individual with general power utility
who faces changing investment opportunities.
12.4
The Martingale Approach to Consumption
and Portfolio Choice

The preceding sections of this chapter showed how stochastic dynamic programming could be used to find an individuals optimal consumption and portfolio
choices. An alternative to this dynamic programming method was developed
by Cox and Huang (Cox and Huang 1989), Karatzas, Lehoczsky, and Shreve
(Karatzas, Lehoczky, and Shreve 1987) and Pliska (Pliska 1986). Their solution
technique uses a stochastic discount factor (state price deflator or pricing kernel) for valuation, and so it is most applicable to an environment characterized
by dynamically complete markets.9 Recall that Chapter 10 demonstrated that
when markets are complete, the absence of arbitrage ensures the existence of a
unique positive stochastic discount factor. Therefore, let us start by considering
the necessary assumptions for market completeness.
9 Hua He and Neil Pearson (He and Pearson 1991) have extended this martingale approach
to an incomplete markets environment. Although in this case there exists an infinity of
possible stochastic discount factors, their solution technique chooses what is referred to as a
"minimax" martingale measure. This leads to a pricing kernel such that agents do not wish
to hedge against the "unhedgeable" uncertainty.
346
12.4.1
Market Completeness Assumptions
As before, let there be n risky assets and a risk-free asset which has an instantaneous return r (t).
We modify the previous risky asset return specification
(12.1) to write the return on risky i as
dSi /Si = i dt + i dZ, i = 1, ..., n
(12.50)
where i = (i1 ... in ) is a 1 n vector of volatility components and dZ =

0
(dz1 ...dzn ) is an n 1 vector of independent Brownian motions.10 The scalar
i , the elements of i , and r (t) may be functions of state variables driven by the
Brownian motion elements of dZ. Further, we assume that the n risky assets
are non-redundant in the sense that their instantaneous covariance matrix is
non-singular. Specifically, if we let be the n n matrix whose ith row equals
i , then the instantaneous covariance matrix of the assets returns, 0 ,
has rank equal to n.
Importantly, we are assuming that any uncertain changes in the means and
covariances of the asset return processes in (12.50) are driven only by the vector
dZ.
This implies that changes in investment opportunities can be perfectly
hedged by the n assets, and such an assumption makes this market dynamically
complete. This diers from the assumptions of (12.1) and (12.2) because we
exclude state variables driven by other, arbitrary Brownian motion processes,
d i , that cannot be perfectly hedged by the n assets returns. Equivalently, if
we assume there is a state variable aecting asset returns, say xi as represented
10 Note
that in (12.50), the independent Brownian motion components of dZ, dzi , i = 1, ..., n
are dierent from the possibly correlated Brownian motion processes dzi defined in (12.1).
Accordingly, the return on asset i in (12.50) depends on all n of the independent Brownian
motion processes while the return on asset i in (12.1) depends on only one of the correlated
Brownian motion processes, namely, the ith one, dzi . These dierent ways of writting the
risky asset returns are not important because an orthogonal transformation of the n correlated
Brownian motion processes in (12.1) can allow us to write asset returns as (12.50) where each
asset return depends on all n independent processes. The reason for writing asset returns as
(12.50) is that individual market prices of risk can be identified with each of the independent
risk sources.
347
in (12.2), then its Brownian motion process, d i , must be a linear function of

the Brownian motion components of dZ. Hence, in this section, there can be
no more than n (not n + k) sources of uncertainty aecting the distribution of
asset returns.
Given this structure, we showed in Chapter 10 that when arbitrage is not
possible, a unique stochastic discount factor exists and follows the process
dM/M = rdt (t)0 dZ
(12.51)
0
where = (1 ... n ) is an n 1 vector of market prices of risks associated with
each Brownian motion and where satisfies
i r = i , i = 1, ..., n
(12.52)
Notice that if we take the form of the assets expected rate of returns and
volatilities as given, then equation (12.52) is a system of n linear equations that
determine the n market prices of risk, . Alternatively, if and the assets
volatilities are taken as given, (12.52) determines the assets expected rates of
return.
12.4.2
The Optimal Consumption Plan
Now consider the individuals original consumption and portfolio choice problem
in (12.4), (12.3), and (??).
A key to solving this problem is to view the
individuals optimally invested wealth as an asset (literally, a portfolio of assets)

that pays a continuous dividend equal to the individuals consumption. This
implies that the return on wealth, equal to its change in value plus its dividend,
can be priced using the stochastic discount factor. The current value of wealth
equals the expected discounted value of the dividends (consumption) that it
348
pays over the individuals planning horizon plus discounted terminal wealth.
Wt = Et
"Z
MT
Ms
Cs ds +
WT
Mt
Mt
(12.53)
Equation (12.53) can be interpreted as an intertemporal budget constraint. This

allows the individuals choice of consumption and terminal wealth to be transformed into a static, rather than dynamic, optimization problem. Specifically,
the individuals problem can be written as the following Lagrange multiplier
problem.
max
Cs s[t,T ],WT
Et
"Z
U (Cs , s) ds + B (WT , T )
t
+ Mt Wt Et
"Z
Ms Cs ds + MT WT
#!
(12.54)
Note that the problem in (12.54) does not explicitly address the portfolio choice
decision.
This will be determined later by deriving the individuals portfolio
allocation that is required to finance his optimal consumption plan.
By treating the integrals in (12.54) as summations over infinite points in

time, the first order conditions for optimal consumption at each date and for
terminal wealth are derived as
U (Cs , s)
= Ms ,
Cs
s [t, T ]
B (WT , T )
= MT
WT
(12.55)
(12.56)
Similar to what we did earlier, define the inverse marginal utility function as
G = [U/C]1 and the inverse marginal utility of bequest function GB =
349
[B/W ]1 . This allows us to re-write these first order conditions as
Cs = G (Ms , s) ,
s [t, T ]
WT = GB (MT , T )
(12.57)
(12.58)
Except for the yet-to-be-determined Lagrange multiplier , equations (12.57)

and (12.58) provide solutions to the optimal choices of consumption and terminal wealth. We can now solve for based on the condition that the discounted
optimal consumption path and terminal wealth must equal the individuals initial endowment of wealth, Wt . Specifically, we substitute (12.57) and (12.58)
into (12.53) to obtain
Wt = Et
"Z
#
MT
Ms
G (Ms , s) ds +
GB (MT , T )
Mt
Mt
(12.59)
Given the initial endowment of wealth, Wt , the distribution of the stochastic

discount factor based upon its process in (12.51), and the forms of the utility and
bequest functions (which determine G and GB ), the expectation in equation
(12.59) can be calculated to determine as a function of Wt , Mt , and any
date t state variables.
Moreover, there is an alternative way to solve for Wt
as a function of Mt , , and the date t state variables that may sometimes

be easier to compute than equation (12.59).
As demonstrated in Chapter
10, since wealth represents an asset or contingent claim that pays a dividend
equal to consumption, Wt must satisfy a particular Black-Scholes-Merton partial
dierential equation (PDE) similar to equation (10.7). The equivalence of the
stochastic discount factor relationship in (12.59) and this PDE solution was
shown to be a result of the assumptions of market completeness and an absence
of arbitrage.
To derive the PDE corresponding to (12.59), let us assume, for simplicity,
350
that there is a single state variable that aects the distribution of asset returns.
That is, i , the elements of i , and r (t) may be functions of a single state
variable, say xt . This state variable follows the process
dx = a (x, t) dt + B (x, t) 0dZ
(12.60)
where B (x, t) = (B1 ...Bn )0 is an n 1 vector of volatilities multiplying the

Brownian motion components of dZ. Based on (12.59) and the fact that the
processes for Mt in (12.51) and xt in (12.60) are Markov, we know that the
date t value of optimally invested wealth is a function of Mt and xt and the
individuals time horizon.11
Hence, by Its lemma, the process followed by
W (Mt , xt , t) satisfies
dW
1
W
dt + WMM (dM)2
t
2
1
+WMx (dM ) (dx) + Wxx (dx)2
2
= WM dM + Wx dx +
= W dt + 0W dZ
(12.61)
where
W rM WM + aWx +
1
W
1
+ 0 M 2 WMM 0 BM WMx + B0 BWxx
t
2
2
(12.62)
and
W WM M + Wx B
(12.63)
Following the arguments of Black and Scholes in Chapter 10, the expected
return on wealth must earn the instantaneous risk-free rate plus a risk premium,
11 This is because the expectation in (12.59) depends on the distribution of future values of
the pricing kernel. From (12.51) and (12.52), the distribution clearly depends on its initial
level, Mt , but also on r and which can vary with the state variable x.
351
where this risk premium equals the market prices of risk times the sensitivities
(volatilities) of wealth to these sources of risk. Specifically,
W + G (Mt , t) = rWt + 0W
(12.64)
Wealths expected return, given by the left-hand side of (12.64) equals the expected change in wealth plus its consumption dividend. Substituting in for W
and 0W leads to the PDE
0 =
1
1 0
M 2 WMM 0 BM WMx + B0 BWxx + (0 r) M WM
2
2
W
0
+ G (Mt , t) rW
+ (a B ) Wx +
(12.65)
t
which is solved subject to the boundary condition that terminal wealth is optimal given the bequest motive, that is, W (MT , xT , T ) = GB (MT , T ). Because
this PDE is linear, as opposed to the non-linear PDE for the indirect utility function, J (W, x, t), that results from the dynamic programming approach, it may
be relatively easy to solve, either analytically or numerically.
Thus, either equation (12.59) or (12.65) leads to the solution W (Mt , xt , t; )

= Wt that allows us to determine as a function of Wt , Mt , and xt , and this
solution for can then be substituted into (12.57) and (12.58). The result is
that consumption at any point in time and terminal wealth will depend only on
the contemporaneous value of the pricing kernel, that is, Cs (Ms ) and WT (MT ).
Note that when the individual follows this optimal policy, it is time-consistent in
the sense that should the individual re-solve the optimal consumption problem
at some future date, say s > t, the computed value of will be the same as that
derived at date t.
352
12.4.3
The Portfolio Allocation
The final step is to derive the portfolio allocation policy that finances these
consumption and terminal wealth rules.
We can do this by comparing the
process for wealth in (12.61) to the dynamics of wealth where the portfolio
weights in the n risky assets are explicitly represented. Based on the assumed
dynamics of asset returns in (12.50), equation (12.3) is
dW
n
X
i=1
0
i (i r)W dt + (rW Ct ) dt +
n
X
i i dZ
i=1
0
= ( re) W dt + (rW Ct ) dt + W dZ
(12.66)
where = (1 ...n )0 is the n 1 vector of portfolio weights for the n risky

0
assets and = (1 ...n ) is the n 1 vector of these assets expected rates of
return.
Equating the coecients of the Brownian motion components of the
0
wealth processes in (12.66) and (12.61), we obtain W 0 = W . Substituting
in (12.63) for W and rearranging results in
Wx 0 1
M WM 0 1
+
B
W
W
(12.67)
Next, recall equation (12.52), and note that it can be written in the following
matrix form
re =
(12.68)
Using (12.68) to substitute for , equation (12.67) becomes
M WM 1 0 1
Wx 0 1
( re) +
B
W
W
Wx 0 1
M WM 1
( re) +
B
=
W
W
=
(12.69)
353
These optimal portfolio weights are of the same form as what was derived earlier
in (12.38) for the case where the state variable is perfectly correlated with
asset returns.12
JW x /JW W .
A comparison shows that M WM = JW /JW W and Wx =

Thus, given the solution for W (M, x, t) in (12.59) or (12.65,
equation (12.69) represents a derivation of the individuals optimal portfolio

choices that is an alternative to the dynamic programming approach. Let us
now use this martingale technique to solve a specific consumption - portfolio
choice problem.
12.4.4
An Example
The following example was derived by Jessica Wachter (Wachter 2002).
It
considers an environment where a single state variable aects investment opportunities and, to ensure market completeness, this state variable is perfectly
correlated with asset returns. In particular, let there be a risk-free asset paying
a constant rate of return of r > 0, and also assume there is a single risky asset
so that equation (12.50) can be written simply as
dS/S = (t) dt + dz
(12.70)
The risky assets volatility, , is assumed to be a positive constant but the assets
drift is permitted to vary over time. Specifically, let the single market price of
risk be (t) = [ (t) r] /.
It is assumed to follow the Ornstein-Uhlenbeck
process
d = a dt bdz
12 In
this case 1 = 0 1 B.
(12.71)
354
where a, , and b are positive constants.
Thus, the market price of risk is
perfectly negatively correlated with the risky assets return.13 Wachter justifies
the assumption of perfect negative correlation as being reasonable based on
empirical studies of stock returns.
Since (t) = r + (t) and, therefore
d = d, this model implies that the expected rate of return on the risky asset
is mean-reverting, becoming lower (higher) after its realized return has been
high (low).14
The individual is assumed to have constant relative risk aversion utility and
a zero bequest function, so that (12.54) becomes
max Et
Cs s[t,T ]
"Z
s C
ds + Mt Wt Et
"Z
#!
Ms Cs ds
(12.72)
where we have used the fact that it is optimal to set terminal wealth to zero
in the absence of a bequest motive. The first order condition corresponding to
(12.57) is then
Cs = e 1 (Ms ) 1 ,
s [t, T ]
(12.73)
Therefore, the relationship between current wealth and this optimal consumption policy, equation (12.59), is
Wt
#
1
s
Ms 1
1
= Et
e
(Ms )
ds
Mt
t
Z T
h i
s
1
1
1
Mt
e 1 Et Ms 1 ds
=
"Z
(12.74)
13 Robert Merton (Merton 1971) considered a similar problem where the market price of
risk was perfectly positively correlated with a risky assets return.
14 Straightforward algebra shows that (t) follows the similar Ornstein-Uhlenbeck process

d = a + r dt bdz.
355
Since dM/M = rdt dz, the expectation in (12.74) depends only on Mt and
the distribution of which follows the Ornstein-Uhlenbeck process in (12.71).
A solution for Wt can be obtained by computing the expectation in (12.74)
directly. Alternatively, one can solve for Wt using the PDE (12.65). For this
example, the PDE is
0 =
1 2 2
M WMM + bMWM +
2

W
+ a + b W +
t
1 2
b W + 2 r M WM
2
t
+ e 1 (Mt ) 1 rW (12.75)
which is solved subject to the boundary condition W (MT , T , T ) = 0 since it is

assumed there is no utility from leaving a bequest. Wachter discusses how the
equations in (12.74) and (12.75) are similar to ones found in the literature on
the term structure of interest rates. When < 0, so that the individual has
risk-aversion greater than that a log utility maximizer, the solution is shown to
be15
1
Wt = (Mt ) 1 e 1
T t
H ( t , ) d
(12.76)
where H ( t , ) is the exponential of a quadratic function of t given by
H (t , ) e
1
1

2
A1 ( ) 2t +A2 ( )t +A3 ( )
(12.77)
and
2c1 (1 ec3 )
2c3 (c2 + c3 ) (1 ec3 )
2
4c1 a 1 ec3 /2
A2 ()
c3 [2c3 (c2 + c3 ) (1 ec3 )]
Z
b2
b2
A3 ()
A22 (s) + A1 (s) + aA2 (s) + r ds
2 (1 )
2
0
A1 ()
15 This
solution also requires c22 4c1 b2 / (1 ) > 0.
356
with c1 / (1 ), c2 2 (a + c1 b), and c3
p
c22 4c1 b2 / (1 ). Equa-
tion (12.76) can be inverted to solve for the Lagrange multiplier, , but since we
1
know from (12.73) that (Mt ) 1 e 1 = Ct , we can immediately re-write

(12.76) to derive the optimal consumption rule as
Ct = R T t
0
Wt
H (t , ) d
(12.78)
The positive function H (t , ) can be given an economic interpretation. Recall

that wealth equals the value of consumption from now until T t periods into
R T t
the future. Therefore, since 0 H ( t , ) d = Wt /Ct , the function H (t , )
equals the value of consumption periods in the future scaled by current consumption.
Wachter shows that when < 0 and t > 0, so that the excess return on the
risky asset, (t)r, is positive, then (Ct /Wt ) /t > 0, that is, the individual
consumes a greater proportion of wealth the larger is the excess rate of return
on the risky asset. This is what we would expect given our earlier analysis that
the "income" eect dominates the "substitution" eect when risk-aversion is
greater than that of log utility. The higher expected rate of return on the risky
asset allows the individual to aord more current consumption, which outweighs
the desire to save more in order to take advantage of the higher expected return
on wealth.
Let us next solve for this individuals optimal portfolio choice. The risky
assets portfolio weight that finances the optimal consumption plan is given by
(12.69) for the case of a single risky asset:
M WM (t) r W b
W
2
W
(12.79)
Using (12.76), we see that MWM /W = 1/ (1 ). Moreover, it is straight-
12.5. SUMMARY
357
forward to compute W from (12.76), and by substituting these two derivatives

into (12.79) we obtain
b
(t) r
(1 ) 2 (1 )
(t) r
b
(1 ) 2 (1 )
R T t
0
H ( t , ) [A1 ( ) t + A2 ()] d
R T t
H (t , ) d
0
T t
(12.80)
H ( t , )
[A1 ( ) t + A2 ( )] d
R T t
H ( t , ) d
0
The first term is the familiar risky asset demand whose form is the same as for
the case of constant investment opportunities, equation (12.35).
The second
term on the right-hand side of (12.80) is the demand for hedging against changing investment opportunities. It can be interpreted as a consumption weighted
average of separate demands for hedging against changes in investment opportunities at all horizons from 0 to T t periods in the future, where the weight
R T t
at horizon is H (t , ) / 0 H ( t , ) d.
It can be shown that A1 () and A2 ( ) are negative when < 0, so that if
t > 0, the term [A1 ( ) t + A2 ( )] is unambiguously negative and, therefore,
the hedging demand is positive. Hence, for individuals who are more risk averse
that log utility, they place more of their wealth in the risky asset than would
be the case if investment opportunities were constant. Because of the negative
correlation between risky asset returns and future investment opportunities,
overweighting ones portfolio in the risky asset means that unexpectedly good
returns today hedge against returns that are expected to be poorer tomorrow.
12.5
Summary
The eects of asset return dynamics on consumption and portfolio decisions

often become more transparent in a continuous time environment. Interestingly,
when asset returns are assumed to be lognormally distributed so that investment
358
opportunities are constant, the individuals optimal portfolio weights equal those
of Chapter 2s single period mean - variance model that assumed normally
distributed asset returns. The fact that the mean - variance optimal portfolio
weights could be derived in a multi-period model with lognormal returns is an
attractive result because lognormality is consistent with the limited liability
characteristics of most securities such as bonds and common stocks.
When assets means and variances are time varying, so that investment opportunities are randomly changing, we found that portfolio allocation rules no
longer satisfy the simple mean variance demands.
For cases other than log
utility, portfolio choices will include additional demand components that reflect
a desire to hedge against unfavorable shifts in investment opportunities.
We presented two techniques for finding an individuals optimal consumption
and portfolio decisions.
The first is a continuous time analog of the discrete
time dynamic programming approach studied in Chapter 5.
This approach
leads to a continuous time Bellman equation which, in turn, results in a partial

dierential equation for the derived utility of wealth. Solving for the derived
utility of wealth allows one to then derive the individuals optimal consumption
and portfolio choices at each point in time. The second, alternative martingale
solution technique is based on the insight that the individuals wealth represents
an asset portfolio that pays dividends in the form of a stream of consumption.
This permits valuation of the individuals optimal consumption stream using the
economys stochastic discount factor. After deriving the optimal consumption
rule, one can then find the portfolio decisions that finance the consumption plan.
This chapters analysis of an individuals optimal consumption and portfolio
decisions provides the foundation for considering the equilibrium prices of assets
in a continuous time economy. This is the topic that we address in the next
chapter.
12.6. EXERCISES
12.6
359
Exercises
1. Consider the following consumption and portfolio choice problem. An

individual must choose between two dierent assets, a stock and a short
(instantaneous) maturity default-free bond. In addition, the individual
faces a stochastic rate of inflation, that is, uncertain changes in the price
level (for example, the Consumer Price Index). The price level (currency
price of the consumption good) follows the process
dPt /Pt = dt + d.
The nominal (currency value) of the stock is given by St . This nominal stock
price satisfies
dSt /St = dt + dz.

The nominal (currency value) of the bond is given by Bt . It pays and instantaneous nominal rate of return equal to i. Hence, its nominal price satisfies
dBt /Bt = idt.

d and dz are standard Wiener processes with ddz = dt. Also assume , ,
, , and i are all constants.
1.a What processes do the real (consumption good value) rates of return on
the stock and the bond satisfy?
1.b Let Ct be the individuals date t real rate of consumption and be the
proportion of real wealth, Wt , that is invested in the stock. Give the
process followed by real wealth, Wt .
360
1.c Assume that the individual solves the following problem:
max E0
C,
U (Ct , t) dt
subject to the real wealth dynamic budget constraint given in part b. Assuming
U (Ct , t) is a concave utility function, solve for the individuals optimal choice
of in terms of the indirect utility of wealth function.
1.d How does vary with ? What is the economic intuition for this comparative static result?
2. Consider the individuals intertemporal consumption and portfolio choice

problem for the case of a single risky asset and an instantaneously risk-free
asset. The individual maximizes expected lifetime utility of the form
E0
"Z
T
t
u (Ct ) dt
The price of the risky asset, S, is assumed to follow the geometric Brownian
motion process
dS/S = dt + dz
where and are constants. The instantaneously risk-free asset pays an instantaneous rate of return of rt . Thus, an investment that takes the form of
continually re-investing at this risk-free rate has a value (price), Bt , that follows
the process
dB/B = rt dt
12.6. EXERCISES
361
where rt is assumed to change over time, following the Vasicek (1977) meanreverting process
drt = a [b rt ] dt + sd
where dzdq = dt.
2.a Write down the intertemporal budget constraint for this problem.
2.b What are the two state variables for this consumption-portfolio choice
problem? Write down the stochastic, continuous-time Bellman equation
for this problem.
2.c Take the first order conditions for the optimal choices of consumption and
the demand for the risky asset.
2.d Show how the demand for the risky asset can be written as two terms:
one term that would be present even if r where constant and another term
that exists due to changes in r (investment opportunities).
3. Consider the following resource allocation - portfolio choice problem faced

by a university. The university obtains utility (for example, an enhanced
reputation for its students, faculty, and alumni) from carrying out research
and teaching in two dierent areas: the arts and the sciences. Let Ca
be the number of units of arts activities consumed at the university and
let Cs be the number of science activities consumed at the university.
At date 0, the university is assumed to maximize an expected utility function
of the form
362
E0
u (Ca (t), Cs (t)) dt
where u (Ca , Cs ) is assumed to be increasing and strictly concave with respect

to the consumption levels. It is assumed that the cost (or price) of consuming a
unit of arts activity is fixed at one. In other words, in what follows we express
all values in terms of units of the arts activity, making units of the arts activity
the numeraire. Thus, consuming Ca units of the arts activity always costs Ca .
The cost (or price) of consuming one unit of science activity at date t is given by
S (t), implying that the universitys expenditure on Cs units of science activities
costs SCs . S (t) is assumed to follow the process
dS/S = s dt + s d
where s and s may be functions of S.
The university is assumed to fund its consumption of arts and sciences activities from its endowment. The value of its endowment is denoted Wt . It can
be invested in either a risk-free asset or a risky asset. The risk-free asset pays
a constant rate of return equal to r. The price of the risky asset is denoted P
and is assumed to follow the process
dP/P = dt + dz
where and are constants and dzd = dt. Let denote the proportion of
the universitys endowment invested in the risky asset, and thus (1 ) is the
proportion invested in the risk-free asset. The universitys problem is then to
maximize its expected utility by optimally selecting Ca , Cs , and .
3.a Write down the universitys intertemporal budget constraint, that is, the
dynamics for its endowment, Wt .
12.6. EXERCISES
363
3.b What are the two state variables for this problem? Define a derived
utility of endowment (wealth) function and write down the stochastic,
continuous-time Bellman equation for this problem.
3.c Write down the first order conditions for the optimal choices of Ca , Cs ,
and .
3.d Show how the demand for the risky asset can be written as two terms, a
standard (single-period) portfolio demand term and a hedging term.
3.e For the special case in which utility is given by u (Ca ,Cs ) = Ca Cs , solve
for the universitys optimal level of arts activity in terms of the level and
price of the science activity.
4. Consider an individuals intertemporal consumption, labor, and portfolio

choice problem for the case of a risk-free asset and a single risky asset.
The individual maximizes expected lifetime utility of the form
E0
(Z
et u (Ct , Lt ) dt + B (WT )
where Ct is the individuals consumption at date t and Lt is the amount of

labor eort that the individual exerts at date t. u (Ct , Lt ) is assumed to be an
increasing concave function of Ct but a decreasing concave function of Lt . The
risk-free asset pays a constant rate of return equal to r per unit time and the
price of the risky asset, S, satisfies the process
dS/S = dt + dz
where and are constants. For each unit of labor eort exerted at date t, the
364
individual earns an instantaneous flow of labor income of Lt yt dt. yt , the return

to eort or wage rate, is stochastic and follows the process
dy = y (y) dt + y (y) d
where dzd = dt.
4.a Letting be the proportion of wealth invested in the risky asset, write
down the intertemporal budget constraint for this problem.
4.b What are the state variables for this problem? Write down the stochastic,
continuous-time Bellman equation for this problem.
4.c Take the first order conditions with respect to each of the individuals
decision variables.
4.d Show how the demand for the risky asset can be written as two terms:
one term that would be present even if y were constant and another term
that exists due to changes in y.
4.e If u (Ct , Lt ) = ln [Ct ] + ln [Lt ], solve for the optimal amount of labor
eort in terms of the optimal level of consumption.
5. Consider an individuals intertemporal consumption and portfolio choice

problem for the case of two risky assets (with no risk-free asset). The
individual maximizes expected lifetime utility of the form
E0
et u (Ct ) dt
12.6. EXERCISES
365
where Ct is the individuals consumption at date t. The individuals portfolio

can be invested in a stock whose price, S, follows the process
dS/S = dt + dz
and a default-risky bond whose price, B, follows the process
dB = rBdt Bdq
where dq is a Poisson process defined as
dq =
1
0
if a default occurs
otherwise
The probability of a default occurring over time interval dt is dt. , , r,

and are assumed to be constants. Note that the bond earns a rate of return
equal to r when it does not default, but when default occurs the total amount
invested in the bond is lost, that is, the bond price goes to zero, dB = B.
We also assume that if default occurs, a new default-risky bond, following the
same original bond price process given above, becomes available, so that the
individual can always allocate her wealth between the stock and a default-risky
bond.
5.a Letting be the proportion of wealth invested in the stock, write down
the intertemporal budget constraint for this problem.
5.b Write down the stochastic, continuous-time Bellman equation for this
problem. (Hint: Recall that the Dynkin operator, L [J], reflects the driftterms from applying Its lemma to J. In this problem, these terms need
to include the expected change in J from jumps in wealth due to bond
default.)
366
5.c Take the first order conditions with respect to each of the individuals
decision variables.
5.d Since this problem reflects constant investment opportunities, it can be
shown that when u (Ct ) = c /, < 1, the derived utility of wealth function takes the form J (W, t) = aet W / where a is a positive constant.
For this constant relative risk-aversion case, derive the conditions for optimal C and in terms of current wealth and the parameters of the asset
price processes. (Note: An explicit formula for in terms of all of the
other parameters may not be possible because the condition is non-linear
in .)
5.e Maintaining the constant relative risk-aversion assumption, what is the
optimal if = 0? Assuming the parameters are such that 0 < < 1 for
this case, how would a small increase in aect , the proportion of the
portfolio held in the stock?
6. Show that a log utility investors optimal consumption for the continuous
time problem, (12.49) is comparable to that of the discrete time problem,
(5.34).
Chapter 13
Equilibrium Prices of
Assets
This chapter considers the equilibrium pricing of assets for a continuous time
economy when individuals have time separable utility. It derives the Intertemporal Capital Asset Pricing Model (ICAPM) that was developed by Robert
Merton (Merton 1973a). This model shows that if investment opportunities
are constant through time and asset prices are log-normally distributed, the
standard single-period CAPM results hold. This is an important modification
of the CAPM, not only because the results are extended to a multi-period environment but because the single-period models assumption of a normal asset
return distribution is replaced with a more attractive assumption of lognormally
distributed returns. Since assets such as stocks and bonds have limited liability, the assumption of lognormal returns, which bounds returns at zero, is more
realistic.
When investment opportunities are changing, the standard "single-beta"
CAPM no longer holds. Rather, a multi-beta ICAPM is necessary for pricing
367
368
CHAPTER 13. EQUILIBRIUM PRICES OF ASSETS
assets, where the additional betas reflect priced sources of risk from additional
state variables that aect investment opportunities. However, as was shown by
Douglas Breeden (Breeden 1979), the multi-beta ICAPM can be collapsed into
a single "consumption" beta model, the so-called Consumption Capital Asset
Pricing Model (CCAPM).
The Merton ICAPM is not a fully general equilibrium analysis, since it takes
the form of the asset prices processes faced by individuals as given. However,
as is also shown in this chapter, these asset price processes turn out to be of the
same form as those derived by John Cox, Jonathan Ingersoll, and Stephen Ross
(CIR) (Cox, Ingersoll, and Ross 1985a) in their general equilibrium production
economy model. The CIR model is an example of a production economy where
the form of the economys productive technologies are given. These technologies
are assumed to display constant returns to scale and provides us with a model of
asset supplies that is an alternative to the Lucas endowment economy discussed
in Chapter 6.
The final section of this chapter gives an example of how the
CIR model can be used to solve for the prices of various maturity bonds that
are assumed to be in zero net supply.
13.1
An Intertemporal Capital Asset Pricing Model
Mertons ICAPM is based on essentially the same assumptions made in the

previous chapter on individuals consumption and portfolio choice decisions.
Individuals can trade in a risk-free asset paying an instantaneous rate of return
of r (t), and n risky assets, where the instantaneous rates of return for the risky
assets satisfy
dSi (t) / Si (t) = i (x, t) dt + i (x, t) dzi
(13.1)
13.1. AN INTERTEMPORAL CAPITAL ASSET PRICING MODEL
369
where i = 1, ..., n, and (i dzi )(j dzj ) = ij dt. The risk-free return and the
means and standard deviations of the risky assets can be functions of time and
a k 1 vector of state variables that follow the processes
(13.2)
where i = 1, ..., k, and (bi d i )(bj d j ) = bij dt and ( i dzi )(bj d j ) = ij dt.
Now we wish to consider what must be the equilibrium relationships between
the parameters of the asset return processes characterized by (13.1) and (13.2).
Let us start by analyzing the simplest case first, namely, when investment opportunities are constant through time.
13.1.1
Constant Investment Opportunities
As shown in the previous chapter, when the risk-free rate and the parameters
of assets return processes are constants (r, the i s, i s, and ij s are all constants), the asset price processes in (13.1) are geometric Brownian motions and
asset returns are lognormally distributed.
In this case, the optimal portfolio
choices of all individuals lead them to choose the same portfolio of risky assets.
Individuals dier only in how they divide their total wealths between this common risky asset portfolio and the risk-free asset.
For this common risky asset
portfolio, it was shown in (12.25) that the proportion of risky asset k to all risky
assets is
k =
n
X
kj (j
j=1
n
n X
X
i=1 j=1
r)
.
ij (j r)
and in (12.26) that this portfolios mean and variance are given by
(13.3)
370
n
X
i i
i=1
n X
n
X
(13.4)
i j ij .
i=1 j=1
Similar to our derivation of the single period CAPM, in equilibrium, this

common risky asset portfolio must be the market portfolio, that is, = m and
2 = 2m . Moreover, the continuous time market portfolio is exactly the same
as that implied by the single - period CAPM, where the instantaneous means
and covariances of the continuous time asset return processes replace the means
and covariances of CAPMs multi-variate normal asset return distribution. This
implies that the equilibrium asset returns in this continuous time environment
satisfy the same relationship as the single period CAPM:
i r = i (m r) ,
i = 1, ...n
(13.5)
where i im /2m and im is the covariance between the ith assets rate of
return and the markets rate of return. Thus, the constant investment opportunity set assumption replicates the standard, single-period CAPM. Yet, rather
than asset returns being normally distributed as in the single-period CAPM,
the ICAPM has asset returns being lognormally distributed.
While the standard CAPM results continue to hold for this more realistic
intertemporal environment, the assumptions of a constant risk-free rate and
unchanging asset return means and variances are untenable. Clearly, interest
rates vary over time, as do the volatilities of assets such as common stocks.1
Moreover, there is substantial evidence that mean returns on assets display
1 Not only do nominal interest rates vary over time, but there is also evidence that real
interest rates do as well (Pennacchi 1991). Also, volatilities of stock returns have been found
to follow mean-reverting processes. See, for example, (Bollerslev, Chou, and Kroner 1992)
and (Andersen, Bollerslev, Diebold, and Ebens 2001).

predictable time variation.2
371
Let us next analyze equilibrium asset pricing for
a model that permits such changing investment opportunities.
13.1.2
Stochastic Investment Opportunities
To keep the analysis simple, let us start by assuming that there is a single
state variable, x. The system of n equations that a given individuals portfolio
weights satisfy is given by the previous chapters equation (12.19) with k = 1.
It can be re-written as
0 = A(i r) +
n
X
j=1
ij j W Hi , i = 1, . . . , n
(13.6)
where recall that A = JW /JW W = UC / [UCC (C/W )] and H = JW x /JW W =

(C/x) / (C/W ). Lets re-write (13.6) in matrix form, using bold type to
denote vectors and matrices while using normal type to indicate scalars. Also
let the superscript p denote the pth individuals (persons) value of wealth, vector
of optimal portfolio weights, and values of A and H. Then (13.6) becomes
Ap ( re) = p W p H p
(13.7)
where = (1 , ..., n )0 , e is an n-dimensional vector of ones, p = ( p1 , ..., pn )0

and = (1 , ..., n )0 . Now if we sum across all individuals and divide both sides
P
by p Ap , we obtain
re = a h
(13.8)
2 For example, empirical evidence by Narasimhan Jeegadeesh and Sheridan Titman

(Jegadeesh and Titman 1993) find that abnormal stock returns appear to display positive
serial correlation at short horizons up to about year, a phenomenon described as "momentum." In contrast, there is some evidence (e.g., (Poterba and Summers 1988) and (Fama and
French 1988)) that abnormal stock returns are negatively serially correlated over longer term
horizons.
372
where a

P
pW
p
pA ,
Hp/
p
pA ,
and
W p/
pW
is
the average investment in each asset across investors. These must be the market
weights, in equilibrium. Hence, the ith row (ith risky asset excess return) of
equation (13.8) is:
i r = aim hi
(13.9)
To find the excess return on the market portfolio, we can pre-multiply (13.8) by
0 and obtain:
m r = a2m h mx
(13.10)
where mx = 0 is the covariance between the market portfolio and the state
variable, x.
Next, define
1
e0 1 .
By construction, is a vector of
portfolio weights for the risky assets, where this portfolio has the maximum
absolute correlation with the state variable, x.
In this sense, it provides the
best possible hedge against changes in the state variable.3
To find the excess
return on this optimal hedge portfolio, we can pre-multiply (13.8) by 0 and

obtain
r = a m h x
(13.11)
where m is the covariance between the optimal hedge portfolio and the market
portfolio and x is the covariance between the optimal hedge portfolio and the
state variable, x. Equations (13.10) and (13.11) are two linear equations in the
two unknowns, a and h. Solving for a and h and substituting them back into
equation (13.9), we obtain:
3 Note that the numerator of , 1 , is the nx1 vector of coecients from a regression
of dx on the n risky asset returns, dSi /Si , i 1, ..., n. Dividing these individual coecients
by their sum, e0 1 , transforms them into portfolio weights.
i r =
im x i m
i 2m im mx
r
(
r)
+
m
2
2
m x mx m
m x mx m
373
(13.12)
While the derivation is somewhat lengthy, it can be shown that (13.12) is equivalent to
i r
im 2 i m
i 2m im m
r
(m r) +
2
2
2
2
2
2
m m
m m
m
i (m r) + i r
(13.13)
=
where i is the covariance between the return on asset i and that of the hedge
portfolio. Note that i = 0 if and only if i = 0. For the case in which the
state variable, x, is uncorrelated with the market, equation (13.13) simplifies to:
i r =
im
i
(m r) + 2 r
2
m
(13.14)
In this case, the first term on the right-hand side of (13.14) is that found in
the standard CAPM. The assumption that x is uncorrelated with the market is
not as restrictive as one might first believe, since one could re-define the state
variable x as a factor that cannot be explained by current market returns, that
is, a factor that is uncorrelated with the market.
An equation such as (13.13) can be derived when more than one state variable
exists. In this case, there will be an additional beta for each state variable.
The intertemporal capital asset pricing relations (ICAPM) given by (13.13)
and (13.14) have a form similar to the Arbitrage Pricing Model of Chapter 3.
Indeed, the multi-factor ICAPM has been used to justify empirical APT-type
factor models. The ICAPM predicts that APT risk factors should be related
to changes in investment opportunities. However, it should be noted that, in
374
general, the ICAPMs betas may be time varying and not easy to estimate in a
constant-coecients multi-factor regression model.
13.1.3
An Extension to State-Dependent Utility
It is possible that individuals utilities may be aected directly by the state of

the economy. Here, we briefly mention the consequences of allowing the state
of nature, x, to influence utility by making it an argument of the utility function, U (Ct , xt , t). It is straightforward to verify that the form of the individuals
continuous time Bellman equation (12.17) and the first order conditions for consumption, Ct , and the portfolio weights, the i s, remain unchanged from those
specified in Chapter 12. Hence our results on the equilibrium returns on assets,
equation (13.13), continue to hold. The only change is in the interpretation of
H, the individuals hedging demand coecient. With state-dependent utility,
by taking the total derivative of the envelope condition (12.18), one obtains
JW x = UCC
C
+ UCx
x
(13.15)
so that
H=
UCx
C/x
C
C/W
UCC W
(13.16)
It can be shown that in this case individuals do not hold consumption variance minimizing portfolios, but marginal utility variance-minimizing portfolios.
13.2
Breedens Consumption CAPM
Douglas T. Breeden (Breeden 1979) provided a way of simplifying the asset

return relationship given in Mertons ICAPM. Breedens model shows that
Chapter 4s single-period consumption-portfolio choice result that an assets
13.2. BREEDENS CONSUMPTION CAPM
375
expected rate of return depends upon its covariance with the marginal utility
of consumption can be generalized to a multi-period, continuous time context.
Breeden considers the same model as Merton, and hence in the case of multiple state variables, derives equation (12.38). Substituting in for A and H,
equation (12.38) can be written in matrix form, and for the case of k (multiple)
state variables the optimal portfolio weights for the pth investor are given by:
p W p =
UCp
p
1
( re) 1 Cp
p
p
x /CW
UCC
CW
p
where CW
= C p /W p , Cp
x =
C p
C p
x1 ... xk
(13.17)
, and is the n k matrix of
covariances of asset returns with changes in the state variables, that is, its i,j th
p
element is ij . Pre-multiplying (13.17) by CW
and rearranging terms, we have
UCp
p
p
p ( re) = Wp CW + Cx
UCC
(13.18)
where Wp is the n 1vector of covariances between asset returns with the

change in wealth of individual p. Now individual ps optimal consumption,
C p (W p , x, t) is a function of wealth, W p , the vector of state variables, x, and
time, t. Thus, from Its lemma, we know that the stochastic terms for dC p will
be
p
( p1 W p 1 dz1 + ... + pn W p n dzn ) + (b1 d 1 + ... + bk d k ) Cp
CW
x
(13.19)
Hence, the instantaneous covariances of asset returns with changes in individual

ps consumption are given by calculating the instantaneous covariance between
each asset (having stochastic term i dzi ) with the terms given in (13.19). The
result, in matrix form, is that the n 1 vector of covariances between asset
returns and changes in the individuals consumption, denoted Cp , is
376
p
+Cp
Cp = Wp CW
x
(13.20)
Note that the right-hand side of (13.20) equals the right-hand side of (13.18),
and therefore
Cp =
UCp
p ( re)
UCC
(13.21)
Equation (13.21) holds for each individual, p. Next, define C as the aggregate
rate of consumption and define T as an aggregate rate of risk tolerance, where
X
p
UCp
p
UCC
(13.22)
Then (13.21) can be aggregated over all individuals to obtain.
re = T 1 C
(13.23)
where C is the n1 vector of covariances between asset returns and changes in

aggregate consumption. If we multiply and divide the right hand side of (13.23)
by current aggregate consumption, one obtains
re = (T /C)
ln C
(13.24)
where ln C is the n 1 vector of covariances between asset returns and changes

in the logarithm of consumption (percentage rates of change of consumption).
Consider a portfolio, m, with vector of weights m . Pre-multiplying (13.24)
by m0 , we have:
m r = (T /C)1 m,ln C
(13.25)
13.2. BREEDENS CONSUMPTION CAPM
377
where m is the expected return on portfolio m and m,ln C is the (scalar) covariance between returns on portfolio m and changes in the log of consumption.
Using (13.25) to substitute for (T /C)1 in (13.24), we have
re = (ln C /m,ln C ) (m r)
= ( C / mC ) (m r)
(13.26)
where C and mC are the consumption betas of asset returns and of portfolio
ms return. The consumption beta for any asset is defined as
iC = cov (dSi /Si , d ln C) /var (d ln C)
(13.27)
Portfolio m may be any portfolio of assets, not necessarily the market portfolio. Equation (13.26) says that the ratio of expected excess returns on any
two assets or portfolios of assets is equal to the ratio of their betas measured
relative to aggregate consumption. Hence, the risk of a securitys return can
be summarized by a single consumption beta. Aggregate optimal consumption,
C (W, x, t), encompasses the eects of levels of wealth and the state variables,
and in this way is a sucient statistic for the value of asset returns in dierent
states of the world.
Breedens consumption CAPM (CCAPM) is a considerable simplification
relative to Mertons multi-beta ICAPM. Furthermore, while the multiple state
variables in Mertons model may not be directly identified or observed, and hence
the multiple state variable betas may not be computed, Breedens consumption beta can be computed given that we have data on aggregate consumption.
However, as discussed earlier, the results of empirical tests using aggregate consumption data are unimpressive. As in all of our earlier asset pricing models
378
based on individuals optimal consumption and portfolio choices, the CCAPM

and ICAPM rely on the assumption of time separable utility. When we depart
from this restriction on utility, as we do in the next chapter, consumption-based
models are able to better describe empirical distributions of asset prices.
The ICAPM and CCAPM would not be considered to be fully general equilibrium models of the economy.
While they model individuals "tastes" by
specifying the form of their utilities, they determine the equilibrium relations
between the parameters of the asset returns generation processes in (13.1) and
(13.2) by taking the form of these processes as given.
A fully general equi-
librium model would not start by specifying these assets return process but,
rather, specify the economys "technologies." In other words, it would specify
the economys productive opportunities that determine the supplies of assets
in the economy. By matching individuals asset demands with the asset supplies, the returns on assets would then be determined endogenously. The Lucas
endowment economy model in Chapter 6 was an example of this, and we now
turn to another general equilibrium model, namely, Cox, Ingersoll, and Rosss
production economy model.
13.3
A Cox, Ingersoll, and Ross Production Economy
In two companion articles (Cox, Ingersoll, and Ross 1985a)(Cox, Ingersoll, and
Ross 1985b), John Cox, Jonathan Ingersoll, and Stephen Ross (CIR) develop
a continuous time model of a production economy that is a general equilibrium framework for many of the asset pricing results of this chapter.
Their
model starts from basic assumptions regarding individuals preferences and the
economys production possibilities. Individuals are assumed to have identical
13.3. A COX, INGERSOLL, AND ROSS PRODUCTION ECONOMY
379
preferences and initial wealth, and to maximize standard, time separable utility similar to the lifetime utility previously specified in this and the previous
chapter, namely (12.4).4
The unique feature of their model is the economys
technologies. Recall that in the general equilibrium endowment economy model

of Robert Lucas (Lucas 1978), technologies were assumed to produce perishable
output (dividends) that could not be reinvested, only consumed. In this sense,
these Lucas-type technologies are inelastically supplied.
Individuals cannot
save output and physically re-invest it to increase the productive capacity of

the economy. In the Lucas economy, prices of the fixed technologies adjusted
to determine their equilibrium rates of return. In contrast, the CIR production
economy can be viewed as the opposite. Technologies are in perfectly elastic
supply.
Individuals can save some of the economys output and reinvest it,
thereby changing the productive capacity of the economy. Assets rates of return are pinned-down by the economys technologies, and the amounts invested
in these technologies become endogenous.
Specifically, CIR assume that there is a single good that can be either consumed or invested.
This "capital-consumption" good can be invested in any
of n dierent risky technologies that produce an instantaneous change in the

amount of the consumption good. If an amount i is invested in technology i,
then the change in the amount of this good that is produced is given by
d i (t) = i (x, t) i (t) dt + i (x, t) i (t) dzi , i = 1, ..., n
(13.28)
where (i dzi )(j dzj ) = ij dt. i i is the instantaneous expected change in the
amount of the invested good and i i is the instantaneous standard deviation
of the change. Note that because i and i are independent of i , the change
4 When individuals are assumed to have the same utility and initial wealth, we can think
of there being a "representative" individual.
380
in the quantity of the good is linear in the amount invested.
Hence, each
technology is characterized by "constant returns to scale." i and i can vary

with time and with a k 1 vector of state variables, x(t). Thus, the economys
technologies for transforming consumption into more consumption can reflect
changing (physical) investment opportunities. The ith state variable is assumed
to follow the process
(13.29)
where i = 1, ..., k, and (bi d i )(bj d j ) = bij dt and ( i dzi )(bj d j ) = ij dt.
Note that equations (13.28) and (13.29) are nearly identical to our earlier
modeling of financial asset prices, equations (13.1) and (13.2). The only difference is that Si (t) in (13.1) is the price of a financial asset while i (t) in
(13.28) is a physical quantity invested.
However, if each technology is inter-
preted as being owned by an individual firm, and each of these firms is financed
entirely by shareholders equity, then the value of shareholders equity of firm
i, Si (t), equals the value of the firms physical assets (capital), i (t). Hence,
i (t) can be interpreted as the value or price of firm i in terms of units of the
capital-consumption good.
In the CIR model, investors decide how much of their wealth (the capitalconsumption good) to consume versus save and, of the amount saved, how to
allocate it between the n dierent technologies (or firms). Because equations
(13.28) and (13.29) model an economys production possibilities as constant
returns to scale technologies, the distributions of assets rates of return available
to investors are exogenous.
In one sense, this situation is not dierent from
our earlier modeling of an investors optimal consumption and portfolio choices.

However, CIRs specification allows one to solve for the equilibrium prices of
securities other than those represented by the n risky technologies.
This is
381
done by imagining there to be other securities that have zero net supplies. For
example, there may be no technology that produces an instantaneously risk-free
return, that is, i 6= 0 i. However, one can solve for the equilibrium risk-less
borrowing or lending rate, call it r (t), for which the representative individuals
would be just indierent between borrowing or lending. In other words, r would
be the riskless rate such that individuals choose to invest zero amounts of the
consumption good at this rate. Since all individuals are identical, this amounts
to the riskless investment having a zero supply in the economy, so that r is really
a "shadow" riskless rate.
Yet, this rate would be consistent, in equilibrium,
with the specification of the economys other technologies.

Let us solve for this equilibrium riskless rate in the CIR economy. The individuals consumption and portfolio choice problem is similar to that in Chapter
12, (12.4), except that the individuals savings are now allocated, either directly or indirectly through firms, to the n technologies.
An equilibrium is
defined as a set of interest rate, consumption, and portfolio weight processes

{r, C , 1 , ..., n } such that the representative individuals first order condiPn
Note that because
tions hold and markets clear:
i=1 i = 1 and i 0.
Pn
i=1 i = 1, this definition of equilibrium implies that riskless borrowing and
lending at the equilibrium rate r has zero net supply. Further, since the capital-
consumption good is being physically invested in the technological processes, the

constraint against short selling, i 0, applies.
To solve for the representative individuals optimal consumption and portfolio weights, note that since in equilibrium the individual does not borrow or lend,
the individuals situation is exactly as if a riskless asset did not exist. Hence,
the individuals consumption and portfolio choice problem is the same one as in
the previous chapter but where the process for wealth excludes a risk-free asset.
Specifically, the individual solves
382
max
Cs ,{ i,s },s,i
Et
"Z
(13.30)
i W i dzi
(13.31)
U (Cs , s) ds + B(WT , T )
subject to
dW =
n
X
i=1
i W i dt Ct dt +
and also subject to the condition
Pn
i=1 i
n
X
i=1
= 1 and the constraint that i 0.
The individuals first order condition for consumption is the usual one
0 =
J (W, x, t)
U (C , t)
C
W
(13.32)
but the first order conditions with respect to the portfolio weights are modified
slightly. If we let be the Lagrange multiplier associated with the equality
Pn
i=1 i = 1, then the appropriate first order conditions for the portfolio weights
are
n
k
J
2J X
2J X
2
W+
ij j W +
W 0
W i
W 2 j=1
xi W j=1 ij
0 = i i
i = 1, . . . , n
(13.33)
The Kuhn-Tucker conditions in (13.33) imply that if i < 0, then i = 0, so

that in this case the ith technology would not be employed. Assuming that the
parameters in (13.28) and (13.29) are such that all technologies are employed,
that is, i = 0 i, then the solution to the system of equations in (13.33) is
i =
JW
JW W W
n
X
j=1
ij j
n
k X
n
X
X
JW xm
ij jm +
ij (13.34)
J
W
JW W W 2 j=1
m=1 j=1 W W
for i = 1, ..., n. Using our previously defined matrix notation, (13.34) can be
383
re-written as
k
X
A 1
A
Hj 1
1

j
e+
2
W
JW W
W
j=1
(13.35)
where A = JW /JW W , Hj = JW xj /JW W , and j = (1j , ..., nj )0 . These

portfolio weights can be interpreted as a linear combination of k + 2 portfolios.
The first two portfolios are mean variance ecient portfolios in a single-period
Markowitz portfolio selection model: 1 is the portfolio on the ecient frontier that is tangent to a line drawn from the origin (a zero interest rate) while
1 e is the global minimum variance portfolio.5 The last k portfolios, 1 j ,
j = 1, ..., k, are held to hedge against changes in the technological risks (investment opportunities). The proportions of these k + 2 portfolios chosen depends
on the individuals utility. A exact solution is found in the usual manner of substituting (13.35) and (12.18) into the Bellman equation. For specific functional
forms, a value for the indirect utility function, J (W, x, t) can be derived. This,
P
along with the restriction ni=1 i = 1, allows the specific optimal consumption
and portfolio weights to be determined.
Since in the CIR economy the riskless asset is in zero net supply, we know
that the portfolio weights in (13.35) must be those chosen by the representative
individual even if oered the opportunity to borrow or lend at rate r. Recall
from the previous chapters equations (12.21) that these conditions, re-written
in matrix notation, are
k
X
A 1
Hj 1
(re) +
j , i = 1, . . . , n
W
W
j=1
(13.36)
Hence, since the individual takes prices and rates as given, the portfolio choices
5 Recall that a linear combination of any two portfolios on the mean variance frontier can
create any other portfolio on the frontier.
384
given by the first order conditions in (13.36), the case when a riskless asset
exists, must be the same as (13.35). By inspection, the weights in (13.35) and
(13.36) are identical when r = / (JW W ). Hence, substituting for in terms
of the optimal portfolio weights, we can write the equilibrium interest as6
W JW
= 0
(13.37)
k
X
W 0
Hj 0

j
A
A
j=1
Note that equation (13.37) is the same as the previously derived relationship
(13.10) except that (13.37) is extended to k state variables. Hence, Mertons
ICAPM, as well as Breedens CCAPM, holds for the CIR economy.
The CIR model also can be used to find the equilibrium shadow prices of
other securities that are assumed to have zero net supplies. Such "contingent
claims" could include securities such as longer maturity bonds or options and
futures. For example, suppose a zero net supply contingent claim has a payo
whose value could depend on wealth, time, the state variables, P (W, t, {xi }).7
Its lemma implies that its price will follow a process of the form
dP = uP dt + PW W
n
X
i i dzi +
i=1
k
X
Pxi bi d i
(13.38)
i=1
where
uP
= PW (W 0 C) +
+
k
X
i=1
k
X
Pxi ai + Pt +
i=1
PW xi W 0 i +
PW W W 2 0

2
1 XX
Px x bij
2 i=1 j=1 i j
(13.39)
6 To derive the second line in (13.37), it is easiest to write in matrix form the first order
condtions in (13.33) and assume these conditions all hold as equalities. Then solve for by
pre-multiplying by 0 and noting that 0 e = 1.
7 A contingent claim whose payo depends on the returns or prices of the technologies can
be found by the Black-Scholes methodolgy described in Chapter 9.
385
Using the Merton ICAPM result (13.9) extended to k state variables, the expected rate of return on the contingent claim must also satisfy8
u=r+
k
X
W
Hi
Cov (dP/P, dW/W )
Cov (dP/P, dxi )
A
A
i=1
(13.40)
or
uP
k
X
1
Hi
Cov (dP, dW )
Cov (dP, dxi )
A
A
i=1
!
k
X
1
Pxi W 0 i
= rP +
PW W 2 0 +
A
i=1
k
k
X Hi
X
PW W 0 i +
Pxj bij
A
i=1
j=1
= rP +
(13.41)
where in (13.40) we make use of the fact that the market portfolio equals the
optimally invested wealth of the representative individual.
Equating (13.39)
and (13.41) and recalling the value of the equilibrium risk-free rate in (13.37),
we obtain a partial dierential equation for the contingent claims value.9
k
k
k
X
PW W W 2 0
1 XX
0
+
0 =
PW xi W i +
Px x bij + Pt +
2
2 i=1 j=1 i j
i=1
k
k
X
X
W 0
Hj bij
PW (rW C) +
i +
rP (13.42)
Pxi ai
A
A
i=1
j=1
The next section illustrates how (13.37) and (13.42) can be used to find
8 Condition (13.9) can be derived for the case of a contingent claim by using the fact that
the contingent claims weight in the market portfolio is zero.
9 It is straightforward to derive the valuation equation for a contingent claim that pays a
continuous dividend at rate (W, x, t) dt. In this case the additional term (W, x, t) appears
on the right hand side of equation (13.42).
386
the risk free rate and particular contingent claims for a specific case of a CIR
economy.
13.3.1
An Example Using Log Utility
The example in this section is based on (Cox, Ingersoll, and Ross 1985b). It assumes that the representative individuals utility and bequest functions are logarithmic and of the form U(Cs , s) = es ln (Cs ) and B (WT , T ) = eT ln (WT ).
For this specification, we showed in the previous chapter that the indirect utility function was separable and equalled J (W, x, t) = d (t) et ln (Wt ) + F (x, t)
where d (t) = 1 1 (1 ) e(T t) , so that optimal consumption satisfies

(12.49) and the optimal portfolio proportions equal (12.43). Since JW xi = 0,
Hi = 0, and A = W , the portfolio proportions in (13.35) simplify to
= 1 (re)
(13.43)
where we have used the result that r = / (JW W ). Using the market clearing
condition e0 = 1, we can solve for the equilibrium risk free rate.
r=
e0 1 1
e0 1 e
(13.44)
Substituting (13.44) into (13.43), we see that the optimal portfolio weights are
0 1

e 1
= 1
e
e0 1 e
(13.45)
Let us next assume that a single state variable, x (t) , aects all production
processes in the following manner.
bi x dt +
bi xdzi , i = 1, ..., n
di /i =
(13.46)
387
where
bi and
bi are assumed to be constants and the state variable follows the
square root process 10
dx = (a0 + a1 x) dt + b0 xd
(13.47)
where dzi d = i dt. Note that this specification implies that the means and
variances of the technologies rates of returns are proportional to the state variable. If a0 > 0 and a1 < 0, x is a non-negative, mean-reverting random variable.
A rise in x raises all technologies expected rates of return but also increases
their variances.
We can write the technologies nx1 vector of expected rates of return as =
b
b x and their nxn matrix of rate of return covariances as = x.
Using these
distributional assumptions in (13.44), we find that the equilibrium interest rate

is proportional to the state variable.
r=
b 1
b 1
e0
x = x
0
1
b
e e
b 1
b 1 e is a constant.
b 1 /e0
where e0
(13.48)
This implies that the risk
free rate follows a square root process of the form
dr = dx = (r r) dt + rd
(13.49)
where a1 > 0, r a0 /a1 > 0, and b0 / . CIR (Cox, Ingersoll,

and Ross 1985b) state that when the parameters satisfy 2r 2 , then if r (t)
is currently positive, it will remain positive at all future dates T t. This is an
attractive feature if the model is used to characterize a nominal interest rate.11
10 This process is a specific case of the more general constant elasticity of variance process
given by dx = (a0 + a1 x) dt + b0 xc dq where c [0, 1].
11 In contrast, recall from Chapter 9 that the Vasicek (Vasicek 1977) model assumes that the
risk free rate follows an Ornstein - Uhlenbeck process which implies that r has a discrete time
normal distribution. Hence, the Vasicek model may be preferred for modeling a real interest
388
Next, let us consider how to value contingent claims based on this examples
assumptions.
Specifically, let us consider the price of a default free discount
bond that pays one unit of the consumption good when it matures at date T t.
Since this bonds payo is independent of wealth, and since logarithmic utility
implies that the equilibrium interest rate and optimal portfolio proportions are
independent of wealth, the price of this bond will also be independent of wealth.
Hence, the derivatives PW , PW W , and PW x in the valuation equation (13.42)
will all be zero. Moreover, since r = x, it will be insightful to think of r as
the state variable rather than x, so that the date t bond price can be written
as P (r, t, T ). With these changes, the valuation equation (13.42) becomes12
2r
Prr + [ (r r) r] Pr rP + Pt = 0
2
(13.50)
b
b equals the right hand side of equation
b 0 .
where is a constant equal to
b while
b is an nx1
b and replaced by
(13.45) but with replaced by
vector of constants whose ith element is b
i i . r = 0 is the covariance of
interest rate changes with the proportional change in optimally invested wealth.
In other words, it is the interest rates "beta" (covariance with the market
portfolios return).
The partial dierential equation (13.50), when solved subject to the boundary condition P (r, T, T ) = 1, leads to the bond pricing formula
P (r, t, T ) = A ( ) eB( )r
(13.51)
rate since r can become negative. See Pennacchi (Pennacchi 1991) for such an application.
It can be shown that the discrete time distribution for the CIR interest rate process in (13.49)
is a non-central chi-square.
12 Recall that logarithmic utility implies A = W and H = 0.
389
where = T t,
2e(++) 2
A ( )
( + + ) (e 1) + 2
B ()
and
2r/2
2 e 1
( + + ) (e 1) + 2
(13.52)
(13.53)
q
( + )2 + 2 2 . This CIR bond price can be contrasted with that of
the Vasicek model derived in Chapter 9, equation (9.39). They are similar in
having the same structure given in equation (13.51) but with dierent values for
A ( ) and B ( ). Hence, the discount bond yield, Y (r, ) ln [P (r, t, T )] /
= ln [A ( )] / +B ( ) r/ is linear in the state variable for both models.13
But the two models dier in a number of ways. Recall that Vasicek directly
assumed that the short rate, r, followed an Ornstein Uhlenbeck process and
derived the result that, in the absence of arbitrage, the market price of interest
rate risk must be the same for bonds of all maturities.
Using the notation
of p (r, ) and p () to be the mean and standard deviation of the return on

a bond with periods to maturity, it was assumed that the market price of
interest rate risk, p (r, ) r /p (), was a constant.

In contrast, the CIR model derived an equilibrium square root process for
r based on assumptions of economic fundamentals (tastes and technologies).

Moreover, the derivation of bond prices did not focus on the absence of arbitrage
but rather the (zero net supply) market clearing conditions consistent with
individuals consumption and portfolio choices. Moreover, unlike the Vasicek
model, the CIR derivation required no explicit assumption regarding the form
of the market price of interest rate risk.
Rather, this market price of risk
was endogenous to the models other assumptions regarding preferences and

13 Models having bond yields that are linear in the state variables are referred to as ane
models of the term structure. Such models will be discussed further in Chapter 16.
390
technologies.
Lets solve for the market risk premium implicit in CIR bond
prices.
Note that Its lemma says that the bond price follows the process
dP
1
= Pr dr + Prr 2 rdt + Pt dt
2
1
Prr 2 r + Pr [ (r r)] + Pt dt + Pr rd
=
2
In addition, rearranging (13.50) implies that
1
2
2 Prr r
(13.54)
+ Pr [ (r r)] + Pt =
rP + rPr . Substituting this into (13.54), it can be rewritten as
dP/P
Pr
Pr
= r 1+
rd
dt +
P
P
= r (1 B ( )) dt B ( ) rd
(13.55)
where we have used equation (13.51)s result that Pr /P = B ( ) in the second

line of (13.55). Hence, we can write
p (r, ) r
rB ( )
r
=
=
p (r, )
rB ( )
(13.56)
so that the market price of interest rate risk is not constant as in the Vasicek
model, but is proportional to the square root of the interest rate. When < 0,
which occurs when the interest rate is negatively correlated with the return
on the market portfolio (and bond prices are positively correlated with the
market portfolio), bonds will carry a positive risk premium. CIR (Cox, Ingersoll,
and Ross 1985b) argue that their equilibrium approach to deriving a market
risk premium avoids problems that can occur when, following the no-arbitrage
approach, an arbitrary form for a market risk premium is assumed. They show
that some functional forms for market risk premia are inconsistent with the no
arbitrage assumption.
13.4. SUMMARY
13.4
391
Summary
In a multi-period, continuous time environment, the Merton ICAPM shows

that when investment opportunities are constant the expected returns on assets
satisfy the single-period CAPM relationship.
For the more interesting case
of changing investment opportunities, the CAPM relationship is generalized to

include risk premia reflecting an assets covariances with asset portfolios that
best hedge against changes in investment opportunities. However, this multibeta relationship can be simplified to express an assets expected return in terms
of a single consumption beta.
The Cox, Ingersoll, and Ross model of a production economy helps to justify
the ICAPM results by showing that they are consistent with a model that starts
from more primitive assumptions regarding the nature of an economys asset
supplies.
It also can be used to derive the economys equilibrium risk free
interest rate and the shadow prices of contingent claims that are assumed to be
in zero net supply. One important application of the model is a derivation of
the equilibrium term structure of interest rates.
The next chapter builds on our results to this point by generalizing individuals lifetime utility functions.
No longer will we assume that utility is time
separable. Allowing for time inseparable utility can lead to dierent equilibrium
relationships between asset returns that can better describe empirical findings.
13.5
Exercises
1. Consider a CIR economy similar to the log utility example given in this
chapter.
However, instead of the productive technologies following the
392

processes (13.46), assume that they satisfy
bi x dt + i dzi , i = 1, ..., n
di /i =
In addition, rather than assume that the state variable follows the process
(13.47), suppose that it is given by
dx = (a0 + a1 x) dt + b0 d
where dzi d = i dt. It is assumed that a0 > 0 and a1 < 0.

1.a Solve for the equilibrium risk-free interest rate, r, and the process it follows, dr. What parametric assumptions are needed for the unconditional
mean of r to be positive?
1.b Derive the optimal (market) portfolio weights for this economy, . How
does vary with r?
1.c Derive the partial dierential equation for P (r, t, T ), the date t price of
a default-free discount bond that matures at date T . Does this equation
look familiar?
2. Consider the intertemporal consumption - portfolio choice model and the

Intertemporal Capital Asset Pricing Model of Merton and its general equilibrium specification by Cox, Ingersoll, and Ross.
2.a What assumptions are needed for the single-period Sharpe-Treyner-LinterMossin CAPM results to hold in this multi-period, environment where
consumption and portfolio choices are made continuously?
13.5. EXERCISES
393
2.b Briefly discuss the portfolio choice implications of a situation in which

the instantaneous real interest rate, r (t), is stochastic, following a meanreverting process such as the square root process of Cox, Ingersoll, and
Ross, or the Ornstein-Ulhenbeck process of Vasicek. Specifically, suppose
that individuals can hold the instantaneous-maturity risk-free asset, a long
maturity default-free bond, and equities (stocks) and that a rise in r (t)
raises all assets expected rates of return. How would the results dier
from the single period Markowitz portfolio demands? In explaining your
answer, discuss how the results are sensitive to utility displaying greater
or lesser risk aversion compared to log utility.
Chapter 14
Time-Inseparable Utility
In previous chapters, individuals multi-period utility functions were assumed to
be time separable. For example, in a continuous time context, time separable
expected lifetime utility was specified as
Et
"Z
U (Cs , s) ds
(14.1)
where U (Cs , s) is commonly taken to be of the form
U (Cs , s) = e(st) u (Cs )
(14.2)
so that utility at date s depends only on consumption at date s and not consumption at previous or future dates. However, as was noted earlier, there is
substantial evidence that standard time separable utility appears inconsistent
with the empirical time series properties of U.S. consumption data and the average returns on risky assets (common stocks) and risk-free investments. These
empirical contradictions, referred to as the equity premium puzzle and the riskfree interest rate puzzle, have led researchers to explore lifetime utility functions
395
396
CHAPTER 14. TIME-INSEPARABLE UTILITY
that dier from (14.1) by permitting time-inseparable forms.

In this chapter we consider two types of lifetime utility functions that are
not time-separable.
The first type is a class of lifetime utility functions for
which past consumption plays a role in determining current utility.
These
utility functions can be interpreted as displaying habit persistence. We summarize two models of this type that are based the articles of George Constantinides
(Constantinides 1990) and of John Campbell and John Cochrane (Campbell and
Cochrane 1999). In addition to modeling habit persistence dierently, these
models provide an interesting contrast in terms of their assumptions regarding the economys aggregate supplies of assets and the techniques we use to
solve them. The model in the Constantinides paper is a simple example of a
Cox, Ingersoll, and Ross (Cox, Ingersoll, and Ross 1985a) production economy
where asset supplies are perfectly elastic. It is solved using a Bellman equation approach. In contrast, the Campbell - Cochrane paper assumes a Lucas
(Lucas 1978) endowment economy where asset supplies are perfectly inelastic.
Its solution is based on the economys stochastic discount factor.
The second type of time-inseparable utility that we discuss is known as
recursive utility. From one perspective, recursive utility is the opposite of habit
persistence because recursive utility functions make current utility depend on
expected values of future utility, which in turn depends on future consumption.
We illustrate this type of utility by considering the general equilibrium of an
economy where representative consumer-investors have recursive utility. The
specific model that we analyze is a continuous-time version of a discrete-time
model by Maurice Obstfeld (Obstfeld 1994).
By generalizing utility functions to permit habit persistence or to be recursive, we hope to provide better models of individuals actual preferences and
their resulting consumption and portfolio choice decisions. In this way, greater
14.1. CONSTANTINIDES INTERNAL HABIT MODEL
397
insights into the nature of equilibrium asset returns may be possible. Specifically, we can analyze these models in terms of their ability to resolve various
asset pricing "puzzles," such as the equity premium puzzle and the risk-free
rate puzzle that arise when utility is time-separable.
Let us first investigate
how utility can be extended from the standard time-separable, constant relative
risk aversion case to display habit persistence.
We then follow this with an
examination of recursive utility.
14.1
Constantinides Internal Habit Model
The notion of habit persistence can be traced to the writings of Alfred Marshall
(Marshall 1920), James Duesenberry (Duesenberry 1949), and, more recently,
Harl Ryder and Georey Heal (Ryder and Heal 1973).
It is based on the
idea that an individuals choice of consumption aects not only utility today,
but it directly aects utility in the near future because the individual becomes
accustomed to todays consumption standard.
Let us illustrate this idea by presenting Constantinides internal habit formation model, which derives a representative individuals consumption and portfolio choices in a simple production economy.
It is based on the following
assumptions.
14.1.1
Assumptions
A.1. Technology:
A single capital-consumption good can be invested in up to two dierent
technologies. The first is a risk-free technology whose output, Bt , follows the
process
dB/B = r dt.
(14.3)
398
The second is a risky technology whose output, t , follows the process
d/ = dt + dz.
(14.4)
Note that the specification of technologies fixes the expected rates of return
and variances of the safe and risky investments.1 In this setting, individuals
asset demands determine equilibrium quantities of the assets supplied rather
than asset prices. Since r, , and are assumed to be constants, there is a
constant investment opportunity set.
A.2 Preferences:
Representative agents maximize expected utility of consumption, Ct , of the
form
E0

bt dt
u C
(14.5)

bt /, < 1, C
bt = C
bt = Ct bxt , and
where u C
xt eat x0 +
ea(ts) Cs ds.
(14.6)
Note that if b = 0, utility is of the standard time separable form and displays
constant relative risk aversion with a coecient of relative risk aversion equal to
(1 ). The variable xt is an exponentially weighted sum of past consumption,
so that when b > 0, the quantity bxt can be interpreted as a subsistence
bt = Ct bxt can be interpreted as
or habit level of consumption and C
surplus consumption. In this case, the specification in (14.5) assumes that

the individuals utility depends on only the level of consumption in excess of
the habit level. This models the notion that an individual becomes accustomed
1 Note that in this model, the existence of a risk-free technology determines the risk-free
interest rate. This is unlike the model of Cox, Ingersoll, and Ross (Cox, Ingersoll, and
Ross 1985a) where risk-free borrowing and lending is assumed to be in zero net supply and
the interest rate is an equilibrium rate determined by risky investment opportunities and
individuals preferences.
399
to a standard of living (habit), and current utility derives from only the part
of consumption that is in excess of this standard.
Alternatively, if b < 0, so
that past consumption adds to, rather than subtracts from, current utility, then
the model can be interpreted as one displaying durability in consumption, rather
than habit persistence.2 Empirical evidence seems to support habit persistence,
rather than consumption durability, so our analysis focuses on the case of b > 0.3
With this modeling of habit persistence, current utility is a linear function
of not only current consumption but of past consumption through the variable
xt , so it is not time-separable. An increase in consumption at date t decreases
current marginal utility but it also increases the marginal utility of consumption
at future dates because it raises the level of subsistence consumption.
Of
course, there are more general ways of modeling habit persistence, for example,
u (Ct , wt ) where wt is any function of past consumption levels.4 However, we
use the linear habit persistence specification in (14.5) and (14.6) for its analytical
tractability.
A.3 Additional Parametric Assumptions:
Let W0 be the initial wealth of the representative individual. The following
parametric assumptions are made to have a well-specified consumption and
portfolio choice problem.
W0 >
bx0
> 0
r+ab
r+a > b > 0

r
2 Ayman
( r)2
> 0
2(1 )2
(14.7)
(14.8)
(14.9)
Hindy and Chi-Fu Huang (Hindy and Huang 1993) consider such a model.
asset pricing tests by Wayne Ferson and George Constantinides (Ferson and
Constantinides 1991) that used aggregate consumption data provided more support for habit
persistence relative to consumption durability.
4 Jerome Detemple and Fernando Zapatero (Detemple and Zapatero 1991) consider a model
that displays non-linear habit persistence.
3 Empirical
400

0 m
r
1
(1 ) 2
(14.10)
The reasons for making these parametric assumptions are the following. Note
that Ct needs to be greater than bxt for the individual to avoid infinite marginal utility.5 Conditions (14.7) and (14.8) ensure that an admissible (feasible)
consumption and portfolio choice strategy exists that enables Ct > bxt .6 To see
this, note that the dynamics for the individuals wealth is given by
dW = {[( r) t + r]W Ct } dt + t W dz
(14.11)
where t , 0 t 1 is the proportion of wealth that the individual invests

in the risky technology. Now if t = 0 for all t, that is, one invests only in
the riskless technology, and consumption equals a fixed proportion of wealth,
Ct = (r + a b)Wt , then
dW = {rW (r + a b)W } dt = (b a) W dt
(14.12)
which is a first order dierential equation in W having initial condition that it

equal W0 at t = 0. Its solution is
Wt = W0 e(ba)t > 0
(14.13)
so that wealth always stays positive. This implies Ct = (r +ab) W0 e(ba)t > 0
that limCt bxt (Ct bxt )(1) = .

ability to maintain Ct > bxt is possible when the underlying economy is assumed to be
a production economy because individuals have the freedom of determining the aggregate level
of consumption versus savings. This is not possible in an endowment economy where the path
of Ct and, therefore, its exponentially weighted average, xt , is assumed to be an exogenous
stochastic process. For many random processes, there will be a positive probability that
Ct < bxt . Based on this observation, David Chapman (Chapman 1998) argues that many
models that assume a linear habit persistence are incompatible with an endowment economy
equilibrium.
5 Note
6 The
401
and
Ct bxt
Z t
(r + a b)W0 e(ba)t b eat x0 +
ea(ts) (r + a b) W0 e(ba)s ds
0
Z t
(r + a b) W0 e(ba)t eat bx0 + b(r + a b)W0 eat
ebs ds
0
(r + a b) W0 e(ba)t eat bx0 + (r + a b)W0 eat (ebt 1)

eat [ (r + a b)W0 bx0 ]
(14.14)
which is greater than zero by assumption (14.7).

Condition (14.9) is a transversality condition.
It ensures that if the indi-
vidual follows an optimal policy (which will be derived below), the expected
utility of consumption over an infinite horizon is finite. As will be seen, condition (14.10) ensures that the individual wishes to put positive levels of wealth in
both the safe and risky technologies, that is, the individuals optimal portfolio
choice has an interior solution. Recall from (12.35) that m is the optimal choice
of the risky asset portfolio weight for the time-separable, constant relative risk
aversion case.
14.1.2
Consumption and Portfolio Choices
The solution technique presented here uses a dynamic programming approach

similar to that of (Sundaresan 1989) and our previous derivation of consumption
and portfolio choices under time-separable utility.7 The individuals maximiza7 Interestingly, Mark Schroder and Costis Skiadas (Schroder and Skiadas 2002) show that
consumption - portfolio choice models where an individual displays linear habit formation can
be transformed into a consumption - portfolio model where the individual does not exhibit
habit formation. This can often simplify solving such problems. Further, known solutions to
time-separable or recursive utility consumption-portfolio choice problems can be transformed
to obtain novel solutions that also display linear habit formation.
402
tion problem is
max Et
{Cs , s }
e s 1 [Cs bxs ] ds e t J(Wt , xt )
(14.15)
subject to the intertemporal budget constraint given by equation (14.11). Given

the assumption of an infinite horizon, we can simplify the analysis by separating
out the factor of the indirect utility function that depends on calendar time, t,
that is, Jb (Wt , xt , t) = e t J(Wt , xt ). The discounted indirect utility func-
tion depends on two state variables, wealth, Wt , and the state variable xt , the
current habit level of consumption. Since there are no changes in investment

opportunities (, , and r are all constant), there are no other relevant state
variables. Similar to wealth, xt is not exogenous but depends on past consumption. We can work out its dynamics by taking the derivative of equation
(14.6):
dx/dt = aeat x0 + Ct a
ea(ts) Cs ds,
or
(14.16)
dx = (Ct axt ) dt.
(14.17)
Thus, changes in xt are instantaneously deterministic. The Bellman equation is

then
0
max
{Ct ,t }
max
{Ct ,t }
U (Ct , xt , t) + L[et J]
t 1
e (Ct bxt ) + et JW [(( r) t + r)W Ct ]
1 t
e JW W 2 2t W 2 + et Jx (Ct axt ) et J }.
2
(14.18)
403
The first order conditions with respect to Ct and t are:

(C t bxt )1 = JW Jx ,
or
(14.19)
Ct = bxt +[JW Jx ] 1 ,
and
( r)W JW + t 2 W 2 JW W = 0,
or
(14.20)
( r)
JW
.
t =
2 (W JW W )
Note that the additional term Jx in (14.19) reflects the fact that an increase
in current consumption has the negative eect of raising the level of subsistence consumption, which decreases future utility. The form of (14.20), which
determines the portfolio weight of the risky asset, is more traditional.
Substituting (14.19) and (14.20) back into (14.18), we obtain the equilibrium
partial dierential equation:
J 2 ( r)2
1
[JW Jx ] 1 W
+ (rW bx)JW + (b a)xJx J = 0.
JW W 2 2
(14.21)
From our previous discussion of the time-separable, constant relative risk aversion case (a = b = x = 0), when the horizon is infinite, we saw from (12.33) that
a solution for J is of the form J(W ) = kW . For this previous case, u = C /,
uc = JW , and optimal consumption was a constant proportion of wealth:
1
( r)2
1
)
/ (1 )
C = (k) (1) W = W r (
2 1
2
(14.22)
and
= m
(14.23)
404
where m is defined above in condition (14.10).

These results for the time-separable case suggest that the derived utility of
wealth function for the time-inseparable case might have the form
J(W, x) = k0 [W + k1 x] .
(14.24)
Making this guess, substituting it into (14.21), and setting the coecients on x
and W equal to zero, we find
k0 =
where
(r + a b)h1
(r + a)
(14.25)
r+ab
( r)2
h
> 0
r
(r + a)(1 )
2(1 )2
(14.26)
and
k1 =
b
< 0.
r+ab
(14.27)
Using equations (14.19) and (14.20), this implies
Ct = bxt + h Wt
and
bxt
r+ab
bxt /Wt
t = m 1
.
r+ab
(14.28)
(14.29)
Interestingly, since r + a > b, by assumption, the individual always demands

less of the risky asset compared to the case of no habit persistence. Thus we
would expect lower volatility of wealth over time.
In order to study the dynamics of Ct , consider the change in the term
i
h
bxt
. Recall that the dynamics of Wt and xt are given in equations
Wt r+ab
405
(14.11) and (14.17), respectively. Using these, one finds
d Wt
bxt
r+ab
C axt
[ ( r) t + r]Wt Ct b t
dt + t Wt dz.
r+ab
(14.30)
Substituting in for t and Ct from (14.28) and (14.29), one obtains
d Wt
bxt
r+ab
Wt
bxt
[ n dt + m dz]
r+ab
(14.31)
where
n
r ( r)2 (2 )
+
.
1
2(1 )2 2
(14.32)
Using this and (14.28), one can show8

dCt
=
Ct
(n + a)bxt
Ct bxt
n+b
dt +
m dz.
Ct
Ct
For particular parametric conditions, the ratio
(14.33)
bxt
Ct bxt
has a stationary distrib
bxt
m dz,
ution.9 However, one sees from the stochastic term in (14.33), CtC
t
that consumption growth is smoother than in the case of no habit persistence.
For a given equity (risky asset) risk premium, this can imply relatively smooth
consumption paths, even though risk aversion, , may not be of a very high magnitude. To see this, recall from inequality (4.32) that the Hansen-Jagannathan
(H-J) bound for the time separable case can be written as
(1 ) c
(14.34)
In the current case of habit persistence, from (14.33) we see that the instanta-
8 See
9 See
Appendix A in (Constantinides 1990).

Theorem 2 in (Constantinides 1990).
406
neous standard deviation of consumption growth is
c,t
=
=
Ct bxt
m
Ct
!
bt
C
r
Ct
(1 ) 2
(14.35)
bt Ct bxt is defined as surplus consumption. If we define

where, recall, that C
bt /Ct as the surplus consumption ratio, we can rearrange equation (14.35)

St C
to obtain
(1 ) c,t
r
=
St
Since St
Ct bxt
Ct
(14.36)
is less than 1, we see by comparing (14.36) to (14.34) that
habit persistence can reconcile the empirical violation of the H-J bound. With
habit persistence, the lower demand for the risky asset, relative to the timeseparable case, can result in a higher equilibrium excess return on the risky
asset and, hence, help explain the puzzle of a large equity premium. However, empirical work by Wayne Ferson and George Constantinides (Ferson and
Constantinides 1991) that tests linear models of habit persistence suggests that
these models cannot produce an equity risk premium as large as that found in
historical equity returns.
Let us next turn to another approach to modeling habit persistence where

an individuals habit level depends on the behavior of other individuals and,
hence, is referred to as an external habit.
14.2. CAMPBELL AND COCHRANES EXTERNAL HABIT MODEL
14.2
407
Campbell and Cochranes External Habit

Model
The Campbell-Cochrane external habit persistence model is based on the following assumptions.
14.2.1
Assumptions
A.1 Technology:
Campbell and Cochrane consider a discrete-time endowment economy. Date
t aggregate consumption, which also equals aggregate output, is denoted Ct ,
and it is assumed to follow an independent and identically distributed lognormal
process
ln (Ct+1 ) ln (Ct ) = g + t+1
(14.37)
where vt+1 N 0, 2 .
A.2 Preferences:
It is assumed that there is a representative individual who maximizes expected utility of the form
E0
"
X
t (Ct
t=0
Xt ) 1
(14.38)
where < 1 and Xt denotes the habit level. Xt is related to past consumption
in the following non-linear manner. Define the surplus consumption ratio, St ,
as
St
Ct Xt
Ct
(14.39)
Then the log of surplus consumption is assumed to follow the auto-regressive
408
process10

ln (St+1 ) = (1 ) ln S + ln (St ) + (St ) t+1
(14.40)
where (St ), the sensitivity function, measures the proportional change in the
surplus consumption ratio resulting from a shock to output growth. It is assumed to take the form
(St ) =
1
S

1 2 ln (St ) ln S 1
(14.41)
and
S=
1
1
(14.42)
The lifetime utility function in (14.38) looks somewhat similar to (14.5) of the
Constantinides model. However, while Constantinides assumes that an individuals habit level depends on his or her own level of past consumption, Campbell
and Cochrane assume the an individuals habit level depends on everyone elses
current and past consumption. Thus, in the Constantinides model, the individuals choice of consumption, Ct , aects his future habit level, bxs , for all s > t,
and he takes this into account in terms of how it aects his expected utility when
he chooses Ct . This type of habit formation is referred to as internal habit. In
contrast, in the Campbell and Cochrane model, the individuals choice of consumption, Ct , does not aect her future habit level, Xs , for all s t, so that
she views Xt as exogenous when choosing Ct . This type of habit formation is
referred to as external habit or keeping up with the Joneses.11 The external
10
SThis iprocess is locally equivalent to ln (Xt ) = ln (Xt1 ) + ln (Ct ) or ln (Xt ) =

i=0 ln (Cti ). The reason for the more complicated form in (14.40) is that it ensures
that consumption is always above habit since St is always positive. This precludes infinite
marginal utility.
11 A similar modeling was developed by Andrew Abel (Abel 1990).
409
habit assumption simplifies the representative agents decision making because

habit becomes an exogenous state variable that depends on aggregate, not the
individuals, consumption.
14.2.2
Equilibrium Asset Prices
Because habit is exogenous to the individual, the individuals marginal utility

of consumption is
uc (Ct , Xt ) = (Ct Xt )1 = Ct1 St1
(14.43)
and the representative agents stochastic discount factor is
mt,t+1 =
uc (Ct+1 , Xt+1 )
=
uc (Ct , Xt )
Ct+1
Ct
St+1
St
(14.44)
If we define r as the continuously-compounded risk-free real interest rate between dates t and t + 1, then it equals
h
i
(14.45)
= ln (Et [mt,t+1 ]) = ln Et e(1) ln(Ct+1 /Ct )(1) ln(St+1 /St )
2
1
= ln e(1)Et [ln(Ct+1 /Ct )](1)Et [ln(St+1 /St )]+ 2 (1) V art [ln(Ct+1 /Ct )+ln(St+1 /St )]
(1 )2 2
[1 + (St )]2
= ln () + (1 ) g + (1 ) (1 ) ln S ln St
2
Substituting in for (St ) from (14.41), equation (14.45) becomes
r = ln () + (1 ) g
1
(1 ) (1 )
2
(14.46)
which, by construction, turns out to be constant over time. One can also derive
a relationship for the date t price of the market portfolio of all assets, denoted
Pt . Recall that since we have an endowment economy, aggregate consumption
410
equals the economys aggregate output, which equals the aggregate dividends
paid by the market portfolio. Therefore,
Pt = Et [mt,t+1 (Ct+1 + Pt+1 )]
(14.47)
or, equivalently, one can solve for the price - dividend ratio for the market
portfolio.
Pt
Ct
Ct+1
Pt+1
= Et mt,t+1
1+
Ct
Ct+1
"
1

#
St+1
Ct+1
Pt+1
= Et
1+
St
Ct
Ct+1
(14.48)
As in the Lucas model, this stochastic dierence equation can be solved

forward to obtain
Pt
Ct
"
1

1

!#
St+1
St+2
Ct+1
Ct+2
Pt+2
= Et
1+
1+
St
Ct
St+1
Ct+1
Ct+2
#
"
1
St+1
Ct+1
St+2
Ct+2
= Et
+ 2
+ ...
St
Ct
St
Ct
"
#
X i St+i 1 Ct+i
= Et
(14.49)
St
Ct
i=1
The solutions can then be computed numerically by simulating the lognormal

processes for Ct and St . The distribution of Ct+1 /Ct is lognormal and does not
depend on the level of consumption, Ct , whereas the distribution of St+1 /St
does depend on the current level of St .12
Hence, the value of the market
portfolio relative to current output, Pt /Ct , varies only with the current surplus
consumption ratio, St . By numerically calculating Pt /Ct as a function of St ,
12 Note that from (14.37) expected consumption growth, g, is a constant, but from (14.40) the

expected growth in the surplus consumption ratio, (1 ) ln S ln (St ) is mean-reverting.
411
Campbell and Cochrane can determine the market portfolios expected returns
and the standard deviation of returns as the level of St varies.
Note that in this model, the coecient of relative risk aversion is given by
Ct ucc
1
=
uc
St
(14.50)
and, as was shown in (4.32), the relationship between the Sharpe ratio for any
asset and the coecient of relative risk aversion when consumption is lognormally distributed is approximately:
E [ri ] r
Ct ucc c = (1 ) c
r
uc
St
i
(14.51)
which has a similar form to that of the Constantinides internal habit model
except, here, c is a constant and, for the case of the market portfolio E [ri ] and
ri will be time-varying functions of St . The coecient of relative risk aversion
will be relatively high when St is relatively low, that is, when consumption is
low (a recession). Moreover, the model predicts that the equity risk-premium
increases during a recession (when Ctuuccc is high), a phenomenon that seems to
be present in the post-war U.S. stock market. Campbell and Cochrane calibrate
the model to U.S. consumption and stock market data.13 Due to the dierent
(non-linear) specification for St vis-a-vis the model of Constantinides, they have
relatively more success in fitting this model to the data.
The next section introduces a class of time-inseparable utility that is much
dierent from habit persistence in that current utility depends on expected
future utility which, in turn, depends on future consumption.
Hence, unlike
habit persistence in which utility depends on past consumption and is backward13 They generalize the model to allow dividends on the (stock) market portfolio to dier from
consumption, so that dividend growth is not perfectly correlated with consumption growth.
Technically, this violates the assumption of an endowment economy but, empirically, there is
low correlation between growth rates of stock market dividends and consumption.
412
looking, recursive utility is forward looking.14
14.3
Recursive Utility
A class of time-inseparable utility known as recursive utility was developed by

David Kreps and Evan Porteus (Kreps and Porteus 1978) and Larry Epstein
and Stanley Zin (Epstein and Zin 1989).
They analyze this type of utility
in a discrete-time setting, while Darrell Due and Larry Epstein (Due and
Epstein 1992a) study the continuous-time limit. In continuous-time, recall that
standard, time-separable utility can be written as:
Vt = Et
"Z
U (Cs , s) ds
(14.52)
where U (Cs , s) is often taken to be of the form U (Cs , s) = e(st) u (Cs ).

Recursive utility, however, is specified as
Vt = Et
"Z
f (Cs , Vs ) ds
(14.53)
where f is known as an aggregator function. The specification is recursive in

nature because current utility, Vt , depends on expected values of future utility,
Vs , s > t. When f has appropriate properties, Darrell Due and Larry Epstein
(Due and Epstein 1992b) show that a Bellman-type equation can be derived
which characterizes the optimal consumption and portfolio choice policies for
utility of this type. For particular functional forms, they have been able to work
out a number of asset pricing models.
14 Utility can be both forward and backward looking in that it is possible to construct models
that are recursive and also display habit persistence (Schroder and Skiadas 2002).
14.3. RECURSIVE UTILITY
413
In the example to follow, we consider a form of recursive utility that is a generalization of standard power (constant relative risk aversion) utility in that it
separates an individuals risk-aversion from her elasticity of intertemporal substitution. This generalization is potentially important because, as was shown
in Chapter 4 equation (4.14), multi-period power utility restricts the elasticity
of intertemporal substitution, , to equal 1/ (1 ), the reciprocal of the coefficient of relative risk aversion. Conceptually, this may be a strong restriction.
Risk-aversion characterizes an individuals (portfolio) choices between assets of
dierent risks, and is well-defined concept even in an atemporal (single-period)
setting, as was illustrated in Chapter 1. In contrast, the elasticity of intertemporal substitution characterizes an individuals choice of consumption at dierent
points in time, and is inherently a temporal concept.
14.3.1
A Model by Obstfeld
Let us now consider the general equilibrium of an economy where representative

consumer-investors have recursive utility.
We analyze the simple production
economy model of Maurice Obstfeld (Obstfeld 1994).
This model makes the
following assumptions:
A.1 Technology:
A single capital-consumption good can be invested in up to two dierent
technologies. The first is a risk-free technology whose output, Bt , follows the
process
dB/B = rdt
(14.54)
The second is a risky technology whose output, t , follows the process
d/ = dt + dz
(14.55)
414
As in the Constantinides models production economy, the specification of

technologies fixes the expected rates of return and variances of the safe and risky
investments. Individuals asset demands will determine equilibrium quantities
of the assets supplied rather than asset prices. Since r, , and are assumed
to be constants, there is a constant investment opportunity set.
A.2 Preferences:
Representative, infinitely-lived households must choose between consuming
(at rate Cs at date s) and investing the single capital-consumption good in the
two technologies. The lifetime utility function at date t faced by each of these
households, denoted Vt , is
Vt = Et
f (Cs , Vs ) ds
(14.56)
where f, the aggregator function, is given by

1 1
Cs
[Vs ]
f(Cs , Vs ) =
1
1 1 [Vs ] 1
(14.57)
Clearly, this specification is recursive in that current lifetime utility, Vt , depends on expected values of future lifetime utility, Vs , s > t. The form of equation (14.57) is ordinally equivalent to the continuous-time limit of the discretetime utility function specified in (Obstfeld 1994). Recall that utility functions
are ordinally equivalent, that is, they result in the same consumer choices, if the
utility functions evaluated at equivalent sets of decisions produce values that
are linear transformations of each other. It can be shown (see (Epstein and
Zin 1989) and (Due and Epstein 1992a)) that > 0 is the continuously compounded subjective rate of time preference,
> 0 is the households elasticity
of intertemporal substitution, and 1 > 0 is the households coecient of

relative risk aversion. For the special case of
= 1/ (1 ) , the utility func-
415
tion given in (14.56) and (14.58) is (ordinally) equivalent to the time-separable,

constant relative risk-aversion case:
Vt = Et
es
Cs
ds
(14.58)
Let t be the proportion of each households wealth invested in the risky

asset (technology). Then the intertemporal budget constraint is given by
dW = [( r)W + rW C] dt + W dz
(14.59)
When the aggregator function, f , is put in a particular form by an ordinally

equivalent change in variables, what Due and Epstein (Due and Epstein
1992b) refer to as a normalization, then a Bellman equation can be used to
solve the problem. The aggregator in (14.57) is in normalized form.
As before, let us define J (Wt ) as the maximized lifetime utility at date t
J (Wt ) =
=
max Et
{Cs ,s }
max Et
{Cs ,s }
f (Cs , Vs ) ds
(14.60)
f (Cs , J (Ws )) ds
Since this is an infinite horizon problem with constant investment opportunities, and the aggregator function, f (C, V ), is not an explicit function of
calendar time, the only state variable is W .
The solution to the individuals consumption and portfolio choice problem
is given by the continuous-time stochastic Bellman equation
0 = max f [Ct , J (Wt )] + L [J (Wt )]

{Ct ,t }
or
(14.61)
416
0 =
1
max f [C, J (W )] + JW [ ( r) W + rW C] + JW W 2 2 W 2 (14.62)
{Ct ,t }
2
1
C 1 [J]
1
= max
+ JW [ ( r) W + rW C] + JW W 2 2 W 2
1
1
2
{Ct ,t }
1 1 [J]
Taking the first order condition with respect to C,
C
[J]
1
1
JW = 0
(14.63)
or
C=
JW
[J]
(14.64)
Taking the first order condition with respect to ,
JW ( r) W + JW W 2 W 2 = 0
(14.65)
or
JW r
JW W W 2
(14.66)
Substituting the optimal values for C and given by (14.64) and (14.66)
into the Bellman equation (14.62) we obtain the dierential equation:
+JW
"
JW
1
( 1)[1 1
]
[J]
[J]
1
1 1 [J] 1
JW ( r)2
+ rW
JW W
2
JW
[J]
(14.67)
2
1 JW
( r)2
=0
2 JW W
2
417
or
+JW
"
"
JW
[J]
JW
JW ( r)2
+ rW
JW W
2
[J]
(14.68)
2
1 JW
( r)2
=0
2 JW W
2
If one guesses that the solution is of the form J (W ) = (aW ) / and

substitutes this into (14.68), one finds that a = 1/(1
"
where
( r)2
+ (1 ) r +
2 (1 ) 2
#!
(14.69)
Thus, substituting this value for J into (14.64) we find that optimal consumption is a fixed proportion of wealth
= W
"
+ (1 ) r +
( r)
2 (1 ) 2
#!
(14.70)
W
and the optimal portfolio weight of the risky asset is
r
(1 ) 2
(14.71)
which is the same as for an individual with standard constant relative riskaversion and time-separable utility. The result that the optimal portfolio choice
depends only on risk-aversion turns out to be an artifact of the models assumption that investment opportunities are constant. Harjoat Bhamra and Raman
Uppal (Bhamra and Uppal 2003) demonstrate that when investment opportu-
418
nities are stochastic, the portfolio weight, , can depend on both and .
Note that if = 1/ (1 ), then equation (14.70) is the same as optimal consumption for the time separable, constant relative risk aversion, infinite horizon
case given in equation (12.34). Similar to the time separable case, for an infinite
horizon solution to exist we need consumption to be positive in (14.70) which
requires > 1 r + [ r]2 / 2 (1 ) 2 . This will be the case when the
elasticity of intertemporal substitution, , is suciently large.

assuming > 0, this inequality is always satisfied when
satisfied when
14.3.2
For example,
> 1 but will not be
is suciently close to zero.
Discussion of the Model
Let us examine how optimal consumption depends on the models parameters.
2
Note that the term r + [ r] / 2 (1 ) 2 in (14.70) can be rewritten using
(14.71) as
r+
( r)2
r
=r+
2
2 (1 )
2
(14.72)
and can be interpreted as relating to the risk-adjusted investment returns available to individuals. From (14.70) we see that an increase in (14.72) increases
consumption when
< 1 and reduces consumption when
> 1.
This result
provides intuition for the role of intertemporal substitution. When
< 1, the
income eect from an improvement in investment opportunities dominates the

substitution eect, so that consumption rises and savings fall. The reverse occurs when > 1: the substitution eect dominates the income eect and savings
rise.
We can also study how the growth rate of the economy depends on the
models parameters. Assuming 0 < < 1 and substituting (14.70) and (14.71)
into (14.59) we have that wealth follows the geometric Brownian motion process
dW/W
419
(14.73)
= [ ( r) + r ] dt + dz
"
!#
2
2
( r)
( r)
r
dz
=
+ r (1 ) r +
dt +
2
2
(1 )
2 (1 )
(1 )
"
!
#
( r)2
( r)2
r
=
r+
dz
+
dt +
2 (1 ) 2
2 (1 ) 2
(1 )
Since C = W , the drift and volatility of wealth in (14.73) are also the
drift and volatility of the consumption process, dC/C. Thus, consumption and
wealth are both lognormally distributed and their continuously-compounded
growth, d ln C, has a volatility, c , and mean, gc , equal to
c =
r
(1 )
(14.74)
and
gc
=
=
!
( r)2
( r)2
1
r+
+
2
2 (1 ) 2
2 (1 ) 2 2 c
!
( r)2
( r)2
r+
2 (1 ) 2
2 (1 )2 2
(14.75)
From (14.75), we see if r + [ r]2 / 2 (1 ) 2 > , an economys growth
rate is higher the higher is intertemporal substitution, , since individuals save

more.
Also, consider how an economys rate varies with the squared Sharpe
2
ratio, [ r] /2 , a measure of the relative attractiveness of the risky asset.
The sign of the derivative gc / [ r]2 /2 equals the sign of / (1 ).
For the time separable, constant relative risk aversion case of
= 1/ (1 ),
this derivative is unambiguously positive, indicating that a higher or a lower

would result in the economy growing faster. However, in the general case, the
420
economy could grow slower if
< / (1 ). The reason is that even though
from (14.71) individuals put a large proportion of their wealth into the faster
growing risky asset as the Sharpe ratio rises, a higher Sharpe ratio leads to
greater consumption (and less savings) when
< 1.
For
< / (1 ), the
eect of less savings dominates the portfolio eect and the economy is expected
to grow slower.
Obstfeld points out that the integration of global financial markets that
allows residents to hold risky foreign, as well as domestic, investments increases
diversification and eectively reduces individuals risky portfolio variance, 2 .
This reduction in would lead individuals to allocate a greater proportion of
their wealth to the higher yielding risky assets.
If
> / (1 ), financial
market integration would also predict that countries will tend to grow faster.
It is natural to ask whether this recursive utility specification, which distinguishes between risk-aversion and the intertemporal elasticity of substitution,
can provide a better fit to historical asset returns compared to time separable
power utility. In terms of explaining the equity premium puzzle, from (14.74)
we see that the risky asset Sharpe ratio, ( r) /, equals (1 ) c , the same
form as with time separable utility.
So, as discussed earlier, one would still
need to assume that the coecient of relative risk aversion (1 ) is quite high
in order to justify the equity risk premium. However, recursive utility has more
hope of explaining the risk-free rate puzzle because of the additional degree of
freedom added by the elasticity of substitution parameter, . If we substitute
(14.74) into (14.75) and solve for the risk-free rate, we find
r =+
gc
h
i 2c
1+
2
(14.76)
14.4. SUMMARY
421
Recall that for the time separable case of
= 1/ (1 ) we have
2
r = + (1 ) gc (1 )
2c
2
(14.77)
Because, empirically, gc 0.018 is large relative to 2c /2 0.032 /2 = 0.00045,

the net eect of higher risk-aversion, 1, needed to fit the equity risk premium
leads to too high a risk-free rate in (14.77). However, we see that the recursive
utility specification in (14.76) potentially circumvents this problem because gc
is divided by
rather than being multiplied by 1 .15
Empirical estimates of the elasticity of intertemporal substitution have been

obtained by regressing consumption growth, d ln C, on the real interest rate, r.
From equations (14.73) and (14.75), we see that if the risky asset Sharpe ratio,
( r) /, is assumed to be independent of the level of the real interest rate, r,
then the regression coecient on the real interest should provide an estimate of
. Tests using aggregate consumption data, such as (Hall 1988) and (Campbell
and Mankiw 1989) generally find that
zero.
is small, often indistinguishable from
However, other tests based on consumption data disaggregated at the
state level (Beaudry and van Wincoop 1996) or at the household level (Attanasio
and Weber 1993) find higher estimates for , often around 1. From (14.76) we
see that a value of
= 1 would make r independent of risk aversion, , and,
assuming is small, could produce a reasonable value for the real interest rate.
14.4
Summary
The models presented in this chapter generalize the standard model of timeseparable, power utility.
For particular function forms, an individuals con-
sumption and portfolio choice problem can be solved using the same techniques
15 Philippe Weil ((Weil 1989)) appears to be the first to examine the equity premium and
risk-free rate puzzles in the context of recursive utility.
422
that were previously applied to the time-separable case. For utility that displays
habit persistence, we saw that the standard coecient of relative risk aversion,
(1 ), is transformed to the expression (1 ) /St where St < 1 is the surplus
consumption ratio. Hence, habit persistence can make individuals behave in a
very risk averse fashion in order to avoid consuming below their habit or subsistence level. As a result, these models have the potential to produce sucient
aversion to holding risky assets that could justify a high equity risk premium.
An attraction of recursive utility is that it distinguishes between an individuals level of risk aversion and his elasticity of intertemporal substitution, a
distinction that is not possible with time-separable, power utility which makes
these characteristics reciprocals of one another.
As a result, recursive utility
can permit an individual to have high risk aversion while, at the same time,
have a high elasticity of intertemporal substitution. Such a utility specification
has the potential to produce both a high equity risk premium and a low risk-free
interest rate that is present in historical data.
While recursive utility and utility displaying habit persistence might be considered non-standard forms of utility, they are preference specifications that are
considered to be those of rational individuals. In the next chapter, we study
utility that is influenced by psychological biases which might be described as
irrational behavior. Such biases have been identified in experimental settings
but have also been shown to be present in the actual investment behavior of
some individuals. We examine how these biases might influence the equilibrium
prices of assets.
14.5. EXERCISES
14.5
423
Exercises
1. In the Constantinides habit persistence model, suppose that there are

three, rather than two, technologies. Assume that there is the risk-free
technology and two risky technologies:
dB/B
= rdt
dS1 /S1
= 1 dt + 1 dz1
dS2 /S2
= 2 dt + 2 dz2
where dz1 dz2 = dt. Also assume that the parameters are such that there
is an interior solution for the portfolio weights (all portfolio weights are
positive). What would be the optimal consumption and portfolio weights
for this case?
2. Consider an endowment economy where a representative agent maximizes

utility of the form
max
t=0
(Ct Xt )
where Xt is a level of external habit and equals Xt = C t1 , where C t1 is

aggregate consumption at date t 1.
2.a Write down an expression for the one-period risk-free interest rate at date
t, Rf,t .
2.b If consumption growth, Ct+1 /Ct follows an independent and identical distribution, is the one-period riskless interest rate, Rf,t constant over time?
424
3. The following problem is based on Menzly, Santos, and Veronesi (Menzly,

Santos, and Veronesi 2001). Consider a continuous time endowment economy where agents maximize utility that displays external habit persistence. Utility is of the form
Et
et ln (Ct Xt ) dt
and aggregate consumption (dividend output) follows the lognormal process
dCt /Ct = dt + dz
Define Yt as the inverse surplus consumption ratio, that is, Yt

=
1
1(Xt /Ct )
Ct
Ct Xt
> 1. It is assumed to satisfy the mean-reverting process
dYt = k Y Yt dt (Yt ) dz
where Y > 1 is the long run mean of the inverse surplus, k > 0 reflects
the speed of mean reversion, > 0. The parameter sets a lower bound
for Yt , and the positivity of (Yt ) implies that a shock to the aggregate output (dividend-consumption) process decreases the inverse surplus
consumption ratio (and increases the surplus consumption ratio). Let Pt
be the price of the market portfolio. Derive a closed-form expression for
the price - dividend ratio of the market portfolio, Pt /Ct . How does Pt /Ct
vary with an increase in the surplus consumption ratio?
4. Consider an individuals consumption and portfolio choice problem when
her preferences display habit persistence. The individuals lifetime utility
satisfies
Et
"Z
es u (Cs , xs ) ds
(1)
14.5. EXERCISES
425
where Cs is date s consumption and xs is the individuals date s level

of habit. The individual can choose among a risk-free asset that pays a
constant rate of return equal to r and n risky assets. The instantaneous
rate of return on risky asset i, satisfies
dPi /Pi = i dt + i dzi ,
i = 1, ..., n
(2)
where dzi dzj = ij dt and i , i , and ij are constants. Thus, the individuals level of wealth, W , follows the process
dW =
n
X
i=1
i (i r)W dt + (rW Ct ) dt +
n
X
i W i dzi .
(3)
i=1
where i is the proportion of wealth invested in risky asset i. The habit

level, xs , is assumed to follow the process
dx = f C t , xt dt
(4)
where C t is the date t consumption that determines the individuals habit.

4.a Let J (W, x, t) be the individuals derived utility of wealth function. Write
down the continuous-time Bellman equation that J (W, x, t) satisfies.
4.b Derive the first order conditions with respect to the portfolio weights, i .
Does the optimal portfolio proportion of risky asset i to risky asset j,
i / j , depend on the individuals preferences? Why or why not?
4.c Assume that the consumption, C t , in equation (4) is such that the individuals preferences display an internal habit, similar to the Constantinides
(1990) model. Derive the first order condition with respect to the individuals date t optimal consumption, Ct .
426
4.d Assume that the consumption, C t , in equation (4) is such that the individuals preferences display an external habit, similar to the Campbell
- Cochrane (1999) model.
("Keeping Up with the Jones" preferences).
Derive the first order condition with respect to the individuals date t
optimal consumption, Ct .
Part V
Additional Topics in Asset Pricing
Chapter 15
Behavioral Finance and

Asset Pricing
This chapter considers asset pricing when investors asset demands incorporate
some elements of irrationality. Irrationality can occur because investors preferences are subject to psychological biases or because investors make systematic
errors in judging the probability distribution of asset returns.
Incorporating
irrationality is a departure from von Neumann-Morgenstern expected utility

maximization and the standard or classical economic approach. A model that
incorporates some form of irrationality is unlikely to be useful for drawing normative conclusions regarding an individuals asset choice. Rather, such a model
attempts to provide a positive or descriptive theory of how individuals actually
behave. For this reason, the approach is referred to as "behavioral finance."
There is both experimental evidence as well as conventional empirical research documenting investor behavior that is inconsistent with von NeumannMorgenstern expected utility theory. Numerous forms of cognitive biases and
judgement errors appear to characterize the preferences of at least some indi429
430
CHAPTER 15. BEHAVIORAL FINANCE AND ASSET PRICING
viduals.
Surveys by (Daniel, Hirshleifer, and Teoh 2001) and (Barberis and
Thaler 2002) describe the evidence for these behavioral phenomena. However,
to date there have been relatively few models that analyze how irrationality
might aect equilibrium asset prices. This chapter examines two recent behavioral asset pricing models.
The first is an intertemporal consumption and portfolio choice model by
Nicholas Barberis, Ming Huang, and Jesus Santos (Barberis, Huang, and Santos
2001) that incorporates two types of biases that are prominent in the behavioral
finance literature. They are loss aversion and the house money eect. These
biases fall within the general category of Prospect Theory. Prospect Theory
deviates from von Neumann-Morgenstern expected utility maximization because
investor utility is a function of recent changes in, rather than simply the current
level of, financial wealth. In particular, investor utility characterized by Prospect
Theory may be more sensitive to recent losses than recent gains in financial
wealth, this phenomenon being referred to as loss aversion. Moreover, losses
following previous losses create more disutility than losses following previous
gains. After a run-up in asset prices, the investor is less risk-averse because
subsequent losses would be cushioned by the previous gains. This is the socalled house money eect.1
An implication of this intertemporal variation in risk-aversion is that after a
substantial rise in asset prices, lower investor risk aversion can drive prices even
higher. Hence, asset prices display volatility that is greater than that predicted
by observed changes in fundamentals, such as changes in dividends. This also
generates predictability in asset returns. A substantial recent fall (rise) in asset
prices increases (decreases) risk aversion and expected asset returns. It can also
1 This expression derives from the psychological misperception that a gamblers (unexpected) winnings are the casino houses money. The gambler views these winnings as dierent
from his initial wealth upon entering the casino. Hence, the gambler is willing to bet more
agressively in the future because if the houses money is lost, the disutility of this loss will be
small relative to the disutility of losing the same amount of his initial wealth.
431
imply a high equity risk premium because the excess volatility in stock prices
leads loss-averse investors to demand a relatively high average rate of return on
stocks.
Prospect theory assumes that investors are overly concerned with changes
in financial wealth, that is, they care about wealth changes more than would be
justified by how these changes aect consumption. This psychological notion
was advanced by Daniel Kahneman and Amos Tversky (Kahneman and Tversky
1979) and is based primarily on experimental evidence.2 For example, Richard
Thaler and Eric Johnson (Thaler and Johnson 1990) find that individuals faced
with a sequence of gambles are more willing to take risk if they have made
gains from previous gambles, evidence consistent with the house money eect.
However, in a recent study of the behavior of traders of the Chicago Board of
Trades Treasury bond futures, Joshua Coval and Tyler Shumway (Coval and
Shumway 2003) find evidence consistent with loss aversion but not the house
money eect.
The second model presented in this chapter examines how equilibrium asset prices are aected when some investors are rational but others suer from
systematic optimism or pessimism. Leonid Kogan, Stephen Ross, Jiang Wang,
and Mark Westerfied (Kogan, Ross, Wang, and Westerfield 2006) construct a
simple endowment economy where rational and irrational investors are identical except that the irrational investors systematically mis-perceive the expected
growth rate of the aggregate dividend process. Interestingly, it is shown that
this economy can be transformed into one where the irrational traders can be
viewed as acting rationally but that their utilities are state dependent.
This
transformation of the problem allows it to be solved using standard techniques.

Kogan, Ross, Wang, and Westerfields general equilibrium model shows that
investors having irrational beliefs regarding the economys fundamentals may
2 Daniel
Kahneman was awarded the Nobel Prize in Economics in 2002.
432
not necessarily lose wealth to rational investors and be driven out of the asset
market.3
Moreover, in those instances where irrational individuals do lose
wealth relative to the rational individuals, so that they do not survive in the long
run, their trading behavior can significantly aect asset prices for substantial
periods of time.
We now turn to the Barberis, Huang, and Santos model, which generalizes
a standard consumption and portfolio choice to incorporate aspects of Prospect
Theory.
15.1
The Eects of Psychological Biases on Asset Prices
The Barberis, Huang, and Santos model is based on the following assumptions.
15.1.1
Assumptions
The model is set in an endowment economy having the following characteristics.

A.1 Technology:
A discrete-time endowment economy is assumed. The risky asset (or portfolio of all risky assets) pays a dividend of perishable output of Dt at date t.
In the Economy I version of the Barberis, Huang, and Santos model, it is assumed that aggregate consumption equals dividends. This is the standard Lucas
(Lucas 1978) economy assumption. However, in the Economy II version of the
model, which is the main focus of our analysis, the risky assets dividends are
distinct from aggregate consumption because there is assumed to be additional
(non-traded) non-financial assets, such as labor income. In equilibrium, aggre3 This result was shown by Bradford De Long, Andrei Shleifer, Lawrence Summers, and
Robert Waldmann (DeLong, Shleifer, Summers, and Waldmann 1991) in a partial equilibrium
model.
15.1. THE EFFECTS OF PSYCHOLOGICAL BIASES ON ASSET PRICES433

gate consumption, C t , then equals dividends, Dt , plus nonfinancial income, Yt ,
because both dividends and nonfinancial income are assumed to be perishable.
Aggregate consumption and dividends are assumed to follow the joint lognormal
process
ln C t+1 /C t = gC + C t+1
(15.1)
ln (Dt+1 /Dt ) = gD + D t+1
where the error terms are serially uncorrelated and distributed as
t
0 1
N ,
t
0
1
(15.2)
The return on the risky asset from date t to date t + 1 is denoted Rt+1 . A oneperiod risk-free investment is assumed to be in zero-net supply, and its return
from date t to date t + 1 is denoted Rf,t .4 The equilibrium value for Rf,t is
derived below.
A.2 Preferences:
Representative, infinitely-lived individuals maximize lifetime utility of the
form
E0
"
X
t Ct
t+1
+ bt v (Xt+1 , wt , zt )
t=0
(15.3)
where Ct is the individuals consumption at date t, < 1, and is a time

discount factor. wt denotes the value of the individuals risky asset holdings at
date t. Xt+1 is defined as the total excess return or gain that the individual
4 Since the risk-free asset is in zero net supply, the representative individuals equilibrium
holding of this asset is zero. Similar to the case of the Cox, Ingersoll, and Ross model
presented in Chapter 13, Rf,t is interpreted as the shadow riskless return.
434
earned from holding the risky asset between date t and date t + 1. Specifically,
this risky asset gain is assumed to be measured relative to the alternative of
holding wealth in the risk-free asset and is given by:
Xt+1 wt (Rt+1 Rf,t )
(15.4)
zt is a measure of the individuals prior gains as a fraction of wt . zt < (>) 1

denotes a situation in which the investor has earned prior gains (losses) on the
risky asset. The prior gain factor, zt , is assumed to follow the process
zt = (1 ) + zt1
R
Rt
(15.5)
where 0 1 and R is a parameter, approximately equal to the average

risky asset return, that makes the steady state value of zt equal 1. If = 0, zt
= 1 for all t. However, if = 1, zt is smaller (larger) than zt1 when risky asset
returns were relatively high last period, Rt > R. In this case, the benchmark
rate, zt , adjusts slowly to prior asset returns. In general, the greater is , the
longer is the investors memory in measuring prior gains from the risky asset.
v () is a function characterizing the prospect theory eect of risky asset
gains on utility.5 For the case of zt = 1 (no prior gains or losses), this function
displays pure loss aversion:
Xt+1 if Xt+1 0
v (Xt+1 , wt , 1) =
X
t+1 if Xt+1 < 0
(15.6)
where > 1. Hence, ceteris paribus, losses have a disproportionately bigger

impact on utility. When zt 6= 1, the function v () reflects Prospect Theorys
house money eect. In the case of prior gains (zt 1), the function takes the
5 Since v () depends only on the risky assets returns, it is assumed that the individual is
not subject to loss aversion on nonfinancial assets.

form
Xt+1
if Rt+1 zt Rf,t
v (Xt+1 , wt , zt ) =
Xt+1 + ( 1) wt (Rt+1 zt Rf,t ) if Rt+1 < zt Rf,t
(15.7)
The interpretation of this function is that when a return exceeds the cushion
built by prior gains, that is, Rt+1 zt Rf,t , it aects utility one-for-one. However, when the gain is less than the amount of prior gains, Rt+1 < zt Rf,t , it
has a greater than one-for-one impact on disutility. In the case of prior losses
(zt > 1), the function becomes
Xt+1
v (Xt+1 , wt , zt ) =
(z ) X
t+1
if Xt+1 0
(15.8)
if Xt+1 < 0
where (zt ) = + k (zt 1), k > 0. Here we see that losses that follow previous
losses are penalized at the rate, (zt ), which exceeds and grows larger as prior
losses become larger (zt exceeds unity).
Finally, the prospect theory term in the utility function is scaled to make the
risky asset price-dividend ratio and the risky asset risk premium be stationary
variables as aggregate wealth increases over time.6 The form of this scaling
factor is chosen to be
bt = b0 C t
(15.9)
where b0 > 0 and C t is aggregate consumption at date t.7

6 Without
the scaling factor, as wealth (output) grows at rate gD , the prospect theory term
would dominate the conventional constant relative risk aversion term.
7 Because C is assumed to be aggregate consumption, the individual views b as an exot
t
geneous variable.
436
15.1.2
Solving the Model
The state variables for the individuals consumption - portfolio choice problem
are wealth, Wt , and zt . Intuitively, since the aggregate consumption - dividend
growth process in (15.1) is an independent, identical distribution, the dividend
level is not a state variable. We start by assuming that the ratio of the risky
asset price to its dividend is a function of only the state variable zt , that is
ft Pt /Dt = ft (zt ), and then show that an equilibrium exists in which this is
true.8 Given this assumption, the return on the risky asset can be written as
Rt+1
=
=
Pt+1 + Dt+1
1 + f (zt+1 ) Dt+1
=
Pt
f (zt )
Dt
1 + f (zt+1 ) gD +D t+1
e
f (zt )
(15.10)
It is also assumed that an equilibrium exists in which the risk-free return is

constant, that is, Rf,t = Rf . This will be verified by the solution to the agents
first order conditions. Making this assumption simplifies the form of the function
v. From (15.7) and (15.8) it can be verified that v is proportional to wt . Hence,
v (Xt+1 , wt , zt ) can be written as v (Xt+1 , wt , zt ) = wt vb (Rt+1 , zt ) where for
zt < 1
Rt+1 Rf
if Rt+1 zt Rf
vb (Rt+1 , zt ) =
Rt+1 Rf + ( 1) (Rt+1 zt Rf ) if Rt+1 < zt Rf
(15.11)
and for zt > 1
8 This is plausible because the standard part of the utility function displays constant relative
risk aversion. With this type of utility, optimal portfolio proportions would not be a function
of wealth.
Rt+1 Rf
if Rt+1 Rf
vb (Rt+1 , zt ) =
(z ) (R
t
t+1 Rf ) if Rt+1 < Rf
(15.12)
The individuals maximization problem is then
X
t Ct
t+1 1
+ b0 C t wt vb (Rt+1 , zt )
t=0
(15.13)
Wt+1 = (Wt + Yt Ct ) Rf + wt (Rt+1 Rf )
(15.14)
max E0
{Ct ,wt }
"
subject to the budget constraint
and the dynamics for zt given in (15.5). Define t J (Wt , zt ) as the derived utility
of wealth function. Then the Bellman equation for this problem is
h
i
Ct
1
+Et b0 C t wt vb (Rt+1 , zt ) + J (Wt+1 , zt+1 ) (15.15)
{Ct ,wt }
J (Wt , zt ) = max
Taking the first order conditions with respect to Ct and wt one obtains
0 = Ct1 Rf Et [JW (Wt+1 , zt+1 )]
(15.16)
h
i
1
0 = Et b0 C t vb (Rt+1 , zt ) + JW (Wt+1 , zt+1 ) (Rt+1 Rf )
1
= b0 C t
Et [b
v (Rt+1 , zt )] + Et [JW (Wt+1 , zt+1 ) Rt+1 ]
Rf Et [JW (Wt+1 , zt+1 )]
(15.17)
It is straightforward to show that (15.16) and (15.17) imply the standard envelope condition
438
Ct1 = JW (Wt , zt )
(15.18)
Substituting this into (15.16), one obtains the Euler equation
1 = Rf Et
"
Ct+1
Ct
1 #
(15.19)
Using (15.18) and (15.19) in (15.17) implies
h
i
h
i
1
1
Et [b
v (Rt+1 , zt )] + Et Ct+1
Rt+1 Rf Et Ct+1
h
i
1
1
= b0 C t Et [b
v (Rt+1 , zt )] + Et Ct+1
Rt+1 Ct1 /
(15.20)
1
0 = b0 C t
or
1 = b0
Ct
Ct
"
Et [b
v (Rt+1 , zt )] + Et Rt+1
Ct+1
Ct
1 #
(15.21)
In equilibrium, conditions (15.19) and (15.21) hold with the representative

agents consumption, Ct , replaced with aggregate consumption, C t . Using the
assumption in (15.1) that aggregate consumption is lognormally distributed, we
can compute the expectation in (15.19) to solve for the risk-free interest rate:
Rf = e(1)gC 2 (1)
2C
(15.22)
Using (15.1) and (15.10), condition (15.21) can also be simplified:
v (Rt+1 , zt )] + Et
1 = b0 Et [b
or
1 + f (zt+1 ) gD +D t+1 gC +C t+1 1

e
e
f (zt )
(15.23)
1 + f (zt+1 ) gD +D t+1
e
1 = b0 Et vb
, zt
(15.24)
f (zt )
2 2
2
1
1 + f (zt+1 ) (D (1)C )t+1
e
+egD (1)gC + 2 (1) C (1 ) Et
f (zt )
The price - dividend ratio, Pt /Dt = ft (zt ), can be computed numeri
1
and Rt+1 =
cally from (15.24). However, because zt+1 = 1 + zt RR
t+1
1+f(zt+1 ) gD + D t+1
,
f (zt ) e
zt+1 depends upon zt , f (zt ), f (zt+1 ), and t+1 , that is
Rf (zt ) egD D t+1

1
zt+1 = 1 + zt
1 + f (zt+1 )
(15.25)
Therefore, (15.24) and (15.25) need to be solved jointly. Barberis, Huang, and
Santos describe an iterative numerical technique for finding the function f ().
Given all other parameters, they guess an initial function, f (0) , and then use
it to solve for zt+1 in (15.25) for given zt and t+1 . Then, they find a new
candidate solution, f (1) , using the following recursion that is based on (15.24):
2 2
2
1
f (i+1) (zt ) = egD (1)gC + 2 (1) C (1 ) x
hh
i
i
Et 1 + f (i) (zt+1 ) e(D (1)C )t+1
(15.26)

1 + f (i) (zt+1 ) gD +D t+1

+f (i) (zt ) b0 Et vb
e
, zt , zt
f (i) (zt )
where the expectations are computed using a Monte Carlo simulation of the
t+1 . Given the new candidate function, f (1) , zt+1 is again found from (15.25).
The procedure is repeated until the function f (i) converges.
440
15.1.3
Model Results
For reasonable parameter values, Barberis, Huang, and Santos find that Pt /Dt =
ft (zt ) is a decreasing function of zt . The intuition was described earlier: if there
were prior gains from holding the risky asset (zt is low), then investors become
less risk averse and bid up the price of the risky asset.
Using their estimate of f (), the unconditional distribution of stock returns
is simulated from a randomly generated sequence of t s. Because dividends and
consumption follow separate processes and stock prices have volatility exceeding
that of dividend fundamentals, the volatility of stock prices can be made substantially higher than that of consumption. Moreover, because of loss aversion,
the model can generate a significant equity risk premium for reasonable values
of the consumption risk aversion parameter .
Thus, the model provides an
explanation for the "equity premium puzzle." Because the investor cares about
stock volatility, per se, a large premium can exist even though stocks may not
have a high correlation with consumption.9
The model also generates predictability in stock returns: returns tend to be
higher following crashes and smaller following expansions. An implication of
this is that stock returns are negatively correlated at long horizons, a feature
documented by empirical research such as (Fama and French 1988), (Poterba
and Summers 1988), and (Richards 1997).
The Barberis, Huang, and Santos model is one with a single type of representative individual who suers from psychological biases. The next model that
we consider assumes that their are two types of representative individuals, ones
with rational beliefs and others with irrational beliefs regarding the economys
fundamentals.
Important insights are obtained by analyzing the interactions
of these two groups of investors.

9 Recall that in standard consumption asset pricing models, an assets risk premium depends
only on its returns covariance with consumption.
15.2. THE IMPACT OF IRRATIONAL TRADERS ON ASSET PRICES 441
15.2
The Impact of Irrational Traders on Asset

Prices
The Kogan, Ross, Wang, and Westerfield model is based on the following assumptions.
15.2.1
Assumptions
The model is a simplified endowment economy with two dierent types of representative individuals, where one type of the individuals suers from irrational
optimism or pessimism regarding risky asset returns. Both types of individuals
maximize utility of consumption at a single, future date.10
A.1 Technology:
There is a risky asset that represents a claim on a single, risky dividend
payment made at the future date T > 0. The value of this dividend payment is
denoted DT , and it is the date T realization of the geometric Brownian motion
process
dDt /Dt = dt + dz
(15.27)
where and are constants, > 0, and D0 = 1. Note that while the process
in equation (15.27) is observed at each date t [0, T ], only its realization at
date T determines the risky assets single dividend payment, DT .
As with
other endowment economies, it is assumed that the date T dividend payment

is perishable output so that, in equilibrium, it equals aggregate consumption,
CT = DT .
Also, it is assumed that there is a market for risk-free borrowing or lending
where payment occurs with certainty at date T . In other words, individuals can
buy or sell (issue) a zero coupon bond that makes a default-free payment of 1 at
10 Alvaro Sandroni (Sandroni 2000) developed a discrete-time model with similar features
that allows the dierent types of individuals to consume at multiple future dates.
442
date T . This bond is assumed to be in zero net supply, that is, the aggregate
net amount of risk-free lending or borrowing is zero. However, because there
are heterogeneous groups of individuals in the economy, some individuals may
borrow while others will lend.
A.2 Preferences:
All individuals in the economy have identical constant relative risk-aversion
utility defined over their consumption at date T .
dierent groups of representative individuals.
However, there are two
The first group of individuals
are rational traders who have a date 0 endowment equal to one-half of the risky
asset and maximize the expected utility function
E0
"
Cr,T
(15.28)
where Cr,T is the date T consumption of the rational traders and < 1. The
second group of individuals are irrational traders. They also possess a date 0
endowment of one-half of the risky asset but incorrectly believe that the probability measure is dierent from the actual one. Rather than thinking that the
aggregate dividend process is given by (15.27), the irrational traders incorrectly
perceive the dividend process is
z
dDt /Dt = + 2 dt + db
(15.29)
where the irrational traders believe db

z is a Brownian motion whereas, in reality,
db
z = dz dt. The irrationality parameter, , is assumed to be a constant. A
positive value of implies that the irrational individuals are too optimistic about
the risky assets future dividend payment while a negative value of indicates
pessimism regarding the risky assets payo. Hence, rather than believe that

the probability measure is P that is generated by the Brownian motion process
dz, irrational traders believe that the probability measure is generated by db
z
which we refer to as the probability measure Pb.11
Therefore, an irrational
individuals expected utility is
b0
E
"
Cn,T
(15.30)
where Cn,T is the date T consumption of the irrational trader.
15.2.2
Solution Technique
We start by showing that the irrational individuals utility can be re-interpreted

as the state-dependent utility of a rational individual.
Recall from Chapter
10 that as a result of Girsanovs Theorem, a transformation of the type db

z=
dz dt leads to Pb and P being equivalent probability measures and that there
exists a sequence of strictly positive random variables, t , that can transform

one distribution to the other. Specifically, recall that for some date T random
variable, xT , equation (10.27) implies
b0 [xT ] = E0 [ T xT ]/ 0
E
(15.31)
Without loss of generality, let us henceforth simplify the notation by assuming

0 = 1. Then we can apply this result to write the irrational traders expected
utility as
b0
E
"
Cn,T
= E0
"
Cn,T
T
(15.32)
so that T = dPb/dP is the Radon-Nikodym derivative of Pb with respect to P .

Similar to what was shown in Chapter 10 to derive the economys pricing kernel,
11 It should be emphasized that the probability measure P
e is not necessarily the risk-neutral
probability measure. The dividend process is not an asset return process so that is not an
assets expected rate of return and is not a risk premium.
444
t follows the process

d/ = dz
(15.33)
or since and are constants,

1
T = t e 2
2 2
(T t)+(zT zt )
(15.34)
From (15.32) and (15.34) we see that the objective function of the irrational
trader is observationally equivalent to that of a rational trader whose utility is
state dependent. The state variable aecting utility, the Brownian motion zt , is
the same source of uncertainty determining the risky assets dividend payment.
While the ability to transform the behavior of an irrational individual to

that of a rational one may depend on the particular way that irrationality is
modeled, this transformation allows us to use standard methods for determining
the economys equilibrium.
Given the assumption of two dierent groups of
representative individuals, we can solve for an equilibrium where the representative individuals act competitively, taking the price of the risky asset and the
risk-free borrowing or lending rate as given. In addition, because there is only
a single source of uncertainty, that being the risky assets payo, the economy
is dynamically complete.
Given market completeness, let us apply the martingale pricing method introduced in Chapter 13. Each individuals lifetime utility function can be interpreted as of the form (12.54) but with interim utility of consumption equalling
zero and only a utility of terminal bequest being non-zero.
Hence, based on
equation (12.56), the result of each individuals static optimization is that his

terminal marginal utility of consumption is proportional to the pricing kernel:
1
Cr,T
= r MT
(15.35)
1
T Cn,T
= n MT
(15.36)
where r and n are the Lagrange multipliers for the rational and irrational
individuals, respectively. Substituting out for MT , we can write
1
Cr,T = ( T ) 1 Cn,T
(15.37)
where we define r /n . Also note that the individuals terminal consumption must sum to the risky assets dividend payment
Cr,T + Cn,T = DT
(15.38)
Equations (15.37) and (15.38) allow us to write each individuals terminal consumption as
Cr,T =
1
1
1 + ( T ) 1
DT
(15.39)
DT
(15.40)
Substituting (15.39) into (15.37), we also obtain

1
Cn,T =
( T ) 1
1
1 + ( T ) 1
Similar to what was done in Chapter 13, the parameter = r /n is determined by the individuals initial endowments of wealth.
Each individuals
initial wealth is an asset that pays a dividend equal to the individuals terminal
consumption.
To value this wealth, we must determine the form of the sto-
chastic discount factor used to discount consumption. As a prelude, note that
446
the date t price of the zero coupon bond that pays 1 at date T > t is given by
P (t, T ) = Et [MT /Mt ]
(15.41)
In what follows, we deflate all asset prices, including the individuals initial
wealths, by this zero coupon bond price. This is done for analytical convenience,
though it should be noted that using the zero coupon bond as the numeraire is
somewhat dierent from using the value of a money market investment as the
numeraire, as was done in Chapter 10. While the return on the zero coupon
bond over its remaining time to maturity is risk-free, its instantaneous return
will not, in general, be risk-free.
Let us define Wr,0 and Wn,0 as the initial wealths, deflated by the zero coupon
bond price, of the rational and irrational individuals, respectively. They equal
Wr,0
E0 [Cr,T MT ]
E0 [Cr,T MT /M0 ]
=
E0 [MT /M0 ]
E0 [MT ]
h
i
h
i
1
E0 Cr,T
E0 Cr,T Cr,T
/r
h
i =
h
i
1
1
E0 Cr,T
/r
E0 Cr,T
h
i
1
DT
E0 1 + ( T ) 1
h
i1
1
E0 1 + ( T ) 1
DT1
(15.42)
where in the second line of (15.42) we used (15.35) to substitute for MT and then
in the third line we used (15.39) to substitute for Cr,T . A similar derivation
that uses (15.36) and (15.40) leads to
Wn,0
h
i
1
1
DT
E0 ( T ) 1 1 + ( T ) 1
h
=
i1
1
1
1
E0 1 + ( T )
DT
(15.43)
Because it was assumed that the rational and irrational individuals are each

initially endowed with equal one-half shares of the risky asset, then it must be
the case that Wr,0 = Wn,0 . Equating the right hand sides of equations (15.42)
and (15.43) determines the value for . The expectations in these equations can
be computed by noting that T satisfies (15.34) and is lognormally distributed
and that
1 2
DT /Dt = e[ 2 ](T t)+(zT zt )
(15.44)
and is also lognormally distributed.12 It is left as an end-of-chapter exercise to

verify that the value of that solves the equality Wr,0 = Wn,0 is given by
= e
(15.45)
Given this value of , we have now determined the form of the pricing kernel
Define St as the
and can solve for the equilibrium price of the risky asset.
date t < T price of the risky asset deflated by the price of the zero coupon
bond.
Then if we also define T,t T = t e
T 12 2 2 (T t)+(zT zt )
the deflated risky asset price can be written as
Et [DT MT /Mt ]
=
St =
Et [MT /Mt ]
Et
Et
"
1
1
1 + T,t
"
1
1
1 + T,t
DT
DT1
(15.46)
While it is not possible to characterize in closed form the rational and irrational
individuals portfolio policies, we can still derive insights regarding equilibrium
asset pricing.13
15.2.3
Analysis of the Results
12 Recall that it was assumed that D = 1. Note also that powers of / and D , such
0
T
T
0
as DT
, are also lognormally distributed.
13 Kogan, Ross, Wang, and Westerfield show that the individuals demand for the risky
asset, , satisfies the bound || 1 + || (2 ) / (1 ).
448
For the limiting case of there being only rational individuals, that is, = 0,
then T,t = t = 1 and from (15.46) the deflated stock price, Sr,t , is
Sr,t
2
2
Et [DT ]
h
i = Dt e[ ](T t)+ (T t)
1
Et DT
(15.47)
2
2
1
= e[(1) ]T +[(1) 2 ] t+(zt z0 )
A simple application of Its lemma shows that equation (15.47) implies that
the risky assets price follows geometric Brownian motion:
dSr,t /Sr,t = (1 ) 2 dt + dz
(15.48)
Similarly, when all individuals are irrational, the deflated stock price, Sn,t , is
2
2
2(T t)
1
Sn,t = e[(1) ]T +[(1) 2 ] t+(zt z0 ) = Sr,t e
(15.49)
and its rate of return follows the process
dSn,t /Sn,t = (1 ) 2 dt + dz
Note that in (15.49) and (15.50) the eect of is similar to .
(15.50)
When all
individuals are irrational, if is positive, the higher expected dividend growth

acts like lower risk aversion in that individuals find the risky asset, relative
to the zero coupon bond, more attractive.
Equation (15.49) shows that this
greater demand raises the deflated stock price relative to that in an economy
with all rational individuals while equation (15.50) indicates that it also lowers
the stocks equilibrium expected rate of return.
It is also interesting to note that (15.48) and (15.50) indicate that when the
economy is populated by only one type of individual, the volatility of the risky

assets deflated return equals .
In contrast, when both types of individuals
populate the economy, the risky assets volatility, S,t , always exceeds . Applying Its lemma to (15.46), Kogan, Ross, Wang, and Westerfield prove that
the risky assets volatility satisfies the following bounds14
S,t (1 + ||)
(15.51)
The conclusion is that a diversity of beliefs has the eect of raising the equilibrium volatility of the risky asset.
For the special case in which rational and irrational individuals have logarithmic utility, that is, = 0, then (15.46) simplifies to
St
1 + Et [ T ]
Et (1 + T ) DT1
2
= Dt e[ ](T t)
(15.52)
1 + t
1 + t e2 (T t)
1 2
1 2
= e[ 2 ]T 2 (T t)+(zt z0 )
1 + t
1 + t e2 (T t)
For this particular case, the risky assets expected rate of return and variance,
as a function of the distribution of wealth between the rational and irrational
individuals, can be derived explicitly. Define
Wr,t
Wr,t
=
Wr,t + Wn,t
St
(15.53)
as the proportion of total wealth own by the rational individuals. Using (15.42)
and (15.46), we see that when = 0 this ratio equals
14 The proof is given in Appendix B of (Kogan, Ross, Wang, and Westerfield 2006). Below,
we show that this bound is satisfied for the case of individuals with logarithmic utility.
450
t =
"
#
1
1
DT
1 + T,t
Et
Et
"
1
1
1 + T,t
DT
#=
1
1
=
1 + Et [ T ]
1 + t
(15.54)
Viewing St as a function of Dt and t as in the second line of (15.52), Its

lemma can be applied to derive the mean and standard deviation of the risky
assets rate of return. The algebra is lengthy but results in the values
S,t = +
and
"
t
1 + e2 (T t) 1
t 1
S,t = 2S,t (1 t ) S,t
(15.55)
(15.56)
where we have used t = 1/ (1 + t ) to substitute out for t . Note when t = 1

or 0, equations (15.55) and (15.56) are consistent with (15.48) and (15.50) for
the case of = 0.
Kogan, Ross, Wang, and Westerfield use their model to study how terminal
wealth (consumption) is distributed between the rational and irrational individuals as the investment horizon, T , becomes large. The motivation for this
comparative static exercise is the well-known conjecture made by Milton Friedman (Friedman 1953) that irrational traders cannot survive in a competitive
market. The intuition is that when individuals trade based on the wrong beliefs, they will lose money to the rational traders, so that in the long run these
irrational traders will deplete their wealth.
Hence, in the long run, rational
traders should control most of the economys wealth and asset prices should
reflect these rational individuals (correct) beliefs. The implication is that even
when some individuals are irrational, markets should evolve toward long-run
eciency because irrational individuals will be driven to "extinction."

Kogan, Ross, Wang, and Westerfield introduce a definition of what would
constitute the long run dominance of rational individuals and, therefore, the relative extinction of irrational individuals. The relative extinction of an irrational
individual would occur if
lim
Cn,T
T Cr,T
= 0 a.s.
(15.57)
which means that for arbitrarily small the probability of lim
Cn,T
T Cr,T
> equals
zero.15 The relative extinction of a rational individual is defined symmetrically,
and an individual is said to survive relatively in the long run if relative extinction
does not occur.16
For the case of individuals having logarithmic utility, irrational individuals
always suer relative extinction. The proof of this is as follows. Rearranging
(15.37), we have
1
Cn,T
= ( T ) 1
Cr,T
(15.58)
and for the case of = 0, (15.45) implies that = 1. Hence,

Cn,T
Cr,T
= T
(15.59)
1
= e 2
2 2
T +(zT z0 )
Based on the strong Law of Large Numbers for Brownian motions, it can be
15 In
general, a sequence of random variables,

say Xt , is said
to converge to X almost surely

(a.s.) if for arbitrary the probability P lim Xt X > = 0.
16 One
could also define the absolute extinction of the irrational individual. This would
occur if lim Cn,T = 0 almost surely, and an individual is said to survive absolutely in the long
T
run if absolute extinction does not occur. Relative survival is sucient for absolute survival,
but the converse is not true. Similarly, absolute extinction implies relative extinction, but
the converse is not true.
452
shown that for any value of b
lim eaT +b(zT z0 )
0 a<0
=
a>0
(15.60)
where convergence occurs almost surely.17 Since 12 2 2 < 0 in (15.59), we see

that equation (15.57) is proven.
The intuition for this result of why irrational individuals become relatively
extinct is due, in part, to the special properties of logarithmic utility. Note that
the portfolio policy of the logarithmic rational individual is maximize at each
date t the utility
Et [ln Cr,T ] = Et [ln Wr,T ]
(15.61)
This is the same as maximizing the expected continuously compounded return

per unit time:
Et
1
1
ln (Wr,T /Wr,t ) =
[Et [ln (Wr,T )] ln (Wr,t )]
T t
T t
(15.62)
since Wr,t is known at date t and T t > 0. Equivalently from (15.62) the rational log utility individual follows a portfolio policy that maximizes Et [d ln Wr,t ]
at each point in time. The selected portfolio policy is referred to as the "growthoptimum portfolio," because it maximizes the (continuously-compounded) return on wealth.18 Now given that in the model economy there is a single source
of uncertainty aecting portfolio returns, dz, the processes for the rational and
17 See
section 2.9.A of (Karatzas and Shreve 1991).

the standard portfolio choice problem of selecting a portfolio from n risky assets
and an instantaneously risk-free asset, we showed in equation (12.43)
the

12 that
Sof Chapter
growth-optimum portfolio has the risky asset portfolio weights i = n
j=1 ij j r . Note
that this log utility investors portfolio depends only on the current values of the investment
opportunity set and portfolio demands do not reflect a desire to hedge against changes in
investment opportunities.
18 For

irrational individuals wealths can be written as
dWr,t /Wr,t
= r,t dt + r,t dz
(15.63)
dWn,t /Wn,t
= n,t dt + n,t dz
(15.64)
where, in general, the expected rates of returns and volatilities, r,t , n,t , r,t ,
and n,t , are time varying. Applying Its lemma, it is straightforward to show
that the process followed by the log of the ratio of the individuals wealth is
d ln
Wn,t
Wr,t
1
1
n,t 2n,t r,t 2r,t dt + (n,t r,t ) dz
2
2
= Et [d ln Wn,t ] Et [d ln Wr,t ] + (n,t r,t ) dz
(15.65)
Given that the irrational individual chooses a portfolio policy that deviates from
the maximum growth portfolio, we know that Et [d ln Wn,t ] Et [d ln Wr,t ] < 0,
and thus Et [d ln (Wn,t /Wr,t )] < 0, making d ln (Wn,t /Wr,t ) a non-stationary
process whose limit as t is . Hence, the limiting value of Wn,t /Wr,t is
zero, which verifies Friedmans conjecture that irrational individuals lose wealth
to rational ones in the long run.
While irrational individuals lose influence in the long run, as indicated by

equations (15.52), (15.55), and (15.56), their presence may impact the level and
dynamics of asset prices for substantial periods of time prior to becoming "extinct." Moreover, if, as empirical evidence suggests, individuals have constant
relative risk aversion utility with < 0 so that they are more risk averse than
logarithmic utility, it turns out that Friedmans conjecture may not always hold.
To see this, let us compute (15.58) for the general case of = e
454

Cn,T
Cr,T
= ( T ) 1
(15.66)
2
1 2
= e[+ 2 ] 1 T + 1 (zT z0 )
Thus, we see that the limiting behavior of Cn,T /Cr,T is determined by the sign
of the expression + 12 2 or + 12 . Given that < 0, the strong Law

of Large Numbers allows us to conclude
Cn,T
lim
T Cr,T
0
<0
=
0 < < 2
0
2 <
rational trader survives

irrational trader survives
(15.67)
rational trader survives
When the irrational individual is pessimistic ( < 0) or strongly optimistic

( > 2), he becomes relatively extinct in the long run. However, when the
irrational individual is moderately optimistic (0 < < 2), the model has the
opposite implication in that it is the rational individual who becomes relatively
extinct in the long run.
This parametric case is the reverse of Friedmans
conjecture.
The intuition for these results comes from our previous discussion of a log
utility investors choice of the growth optimal portfolio.
When rational indi-
viduals are more risk-averse than log utility ( < 0), their demand for the risky
asset is less than would be chosen by a log utility investor.19 Certeris paribus,
the wealth of these < 0 investors would tend to grow slower than that of
someone with log utility. When < 0, irrationally pessimistic investors would
demand even less of the risky asset than their rational counterparts, which would
move them even further away from the growth optimal portfolio. Hence, in this
case, a rational individuals wealth would tend to grow faster than the wealth
19 For example, recall from Chapter 12s analysis of the standard consumption-portfolio
choice problem when investment opportunities are constant that equation (12.35), =
r
, implies that the demand for the risky asset decreases as risk aversion increases.
(1)2

of the irrational individual, so that the irrational individual does not survive in
the long run.
When the irrational individual is optimistic ( > 0), her demand for the risky
asset will exceed that of rational investor.
When her optimism is moderate,
(0 < < 2), her portfolio demand is closer to the growth optimal portfolio
than is the portfolio demanded by the rational individual. Therefore, in this
case the moderately optimistic individuals wealth grows faster than that of the
rational individual, so that the rational individual suers relative extinction in
the long run. In contrast, when the irrational individual is strongly optimistic
( > 2), her demand for the risky asset is so great that her portfolio choice is
further from the growth optimal portfolio than is the rational individual. For
this case the irrational individuals wealth tends to grow relatively slowly and,
as the pessimistic case, she does not survive in the long run.
The model outlined in this section is clearly a simplification of reality in
that it assumes that individuals gain utility from only terminal, not interim,
consumption.
Interim consumption reduces the growth of wealth, and dif-
ferences between rational and irrational individuals consumption rates could

aect their relative survivability.
The model also assumes that rational and
irrational individuals have the same preferences (levels of risk aversion).
In
general, an individuals portfolio choice, which aects his growth of wealth and
survivability, is determined by risk aversion as well as beliefs. Hence, systematic dierences between rational and irrational investors risk aversions could
influence the models conclusions. In addition, one might expect that irrational
individuals might learn over time of their mistakes since the historical distribution of the dividend process will tend to dier from their beliefs. The eect of
such learning may be that irrationality could diminish with age.20 Lastly, the
20 However, there is empirical psychological evidence (Lord, Ross, and Lepper 1979) showing
that individuals tend to persist too strongly in their initial beliefs after being exposed to
456
model considers only one form of irrationality, namely, systematic optimism or

pessimism. Other forms of irrationality have been identified which presumably
would change the dynamics of wealth and of the equilibrium prices of risky assets.21
Yet, the main conclusions of the model, that irrational investors may
have a significant impact on asset prices and that they may not necessarily
become extinct, are likely to remain robust.
15.3
Summary
There is a growing body of experimental and empirical research documenting

that individuals do not always form beliefs rationally and do not always make
decisions consistent with expected utility theory. Analyzing the asset pricing
implications of such behavior is still at an early stage.
This chapter has at-
tempted to present two of the relatively few general equilibrium models that
incorporate psychological biases or irrationality.
Interestingly, these models
can be solved using essentially the same techniques that previously were used
to derive models of rational, expected utility maximizing individuals.
Both
models embed rationality as a special case, which makes it easy to see how their
behavioral assumptions specifically aect the models results.
As of yet, there appears to be no consensus among financial economists regarding the importance of incorporating aspects of behavioral finance into asset
pricing theories. It is especially unclear whether a behavioral paradigm will be
universally successful in supplanting asset pricing theories built on expected utility. However, it is likely that research exploring the asset pricing implications
of behavioral biases will grow in coming years.
contrary information.
21 Recent models incorporating various forms of irrationality, (Barberis, Shleifer, and Vishny
1998), (Daniel, Hirshleifer, and Subrahmanyam 1998), and (Hong and Stein 1999), have been
constructed to explain the empirical phenomena that stock returns display short run positive
serial correlation (momentum) and the long run negative serial correlation (reversals or mean
reversion).
15.4. EXERCISES
15.4
457
Exercises
1. In the Barberis, Huang, and Santos model, verify that the first order
conditions (15.16) and (15.17) lead to the envelope condition (15.18).
2. In the Barberis, Huang, and Santos model, solve for the price - dividend
ratio, Pt /Dt , for Economy II when utility is standard constant relative
risk aversion, that is,
E0
"
t Ct
t=0
3. In the Kogan, Ross, Wang, and Westerfield model, verify that = e
satisfies the equality Wr,0 = Wn,0 .

4. In the Kogan, Ross, Wang, and Westerfield model, suppose that both representative individuals are rational but have dierent levels of risk aversion. The first type of representative individual maximizes utility of the
form
E0
"
1
Cr,T
1
and the second type of representative individual maximizes utility of the

form
E0
where 1 > 1 > 2 .
"
2
Cn,T
2
Assuming Wr,0 = Wn,0 , solve for the equilibrium
price of the risky asset deflated by the discount bond maturing at date T .
Chapter 16
Asset Pricing with

Dierential Information
The asset pricing models in prior chapters have assumed that individuals have
common information.
Now we will consider situations where individuals can
have dierent private information about an assets future payo or value. Of

particular interest is the influence of private information on a risky assets equilibrium price. We start by analyzing a model developed by Sanford Grossman
(Grossman 1976). It shows how individuals information aects their demands
for an asset, and, via these demands, how private information is contained in the
assets equilibrium price. The model examines two equilibria: a competitive,
but not fully rational, equilibrium; and a fully-revealing rational expectations
equilibrium.
Following this, we examine an extension of the Grossman model that includes
an additional source of uncertainty, namely, shifts in the supply of the risky
asset. A model of this type was developed in a number of studies (Grossman and
Stiglitz 1980), (Hellwig 1980), (Diamond and Verrecchia 1981), and (Grundy
459
460CHAPTER 16. ASSET PRICING WITH DIFFERENTIAL INFORMATION

and McNichols 1989). Importantly, in a rational expectations equilibrium this
additional supply uncertainty makes the equilibrium asset price only partially
reveal the private information of individuals.
We cover one additional model of a risky asset market that also possesses an
equilibrium where private information is partially revealed. It is the seminal
market micro-structure model developed by Albert "Pete" Kyle (Kyle 1985).
This model assumes a market for a particular security in which one agent,
the so-called "insider," has private information and trades with lesser-informed
agents composed of a market maker and "noise" traders. The model solves for
the strategic trading behavior of the insider and market maker and provides a
theoretical framework for determining bid-ask spreads and the market impact
of trades.
The literature on asset pricing in the presence of private information is vast,
and this chapter is meant to provide only a taste of this research area.1 However,
the Grossman and Kyle models are probably the two most common modeling
frameworks in this area of research. Familiarity with these two models provides
a segue to many additional asset pricing results.
16.1
Equilibrium with Private Information
The model by Sanford Grossman (Grossman 1976) that we consider in this section examines how an investors private information about a risky assets future
payo aects her demand for that asset and, in turn, the assets equilibrium
price. In addition, it takes account of the idea that a rational individual can
learn about others private information from the risky assets price, a concept
known as "price discovery." The model is based on the following assumptions.
1 More in-depth coverage of topics in this chapter include books by Maureen OHara
(OHara 1995) and Markus Brunnermeier (Brunnermeier 2001).
16.1. EQUILIBRIUM WITH PRIVATE INFORMATION
16.1.1
461
Grossman Model Assumptions
A.1 Assets
This is a single-period portfolio choice problem. At the beginning of the
period, traders can choose between a risk-free asset, which pays a known endof-period return (one plus the interest rate) of Rf , or a risky asset that has
a beginning-of-period price of P0 per share and has an end-of-period random
payo (price) of Pe1 per share. The unconditional distribution of Pe1 is assumed
to be normally distributed as N(m, 2 ). The aggregate supply of shares of the
but the risk-free asset is in perfectly elastic supply.
risky asset is fixed at X,
A.2 Trader Wealth and Preferences

There are n dierent traders. The ith trader has beginning-of-period wealth
W0i and is assumed to maximize expected utility over end-of-period wealth,
1i . Each trader is assumed to have constant absolute risk aversion (CARA)
W
utility, but traders levels of risk aversion are permitted to dier. Specifically,
the form of the ith traders utility function is assumed to be
1i
1i ) = eai W
, ai > 0.
Ui (W
(16.1)
A.3 Trader Information

At the beginning of the period, the ith trader observes yi , which is a realized
value from the noisy signal of the risky asset end-of-period value
yi = P1 + i
where i N(0, 2i ) and is independent of P1 .
(16.2)
16.1.2
Individuals Asset Demands
Let Xf i be the amount invested in the risk-free asset and Xi be the number of
shares of the risky asset chosen by the ith trader at the beginning of the period.
Thus,
W0i = Xf i + P 0 Xi .
(16.3)
The ith traders wealth accumulation equation can be written as

1i = Rf W0i +
W
i
P1 Rf P 0 Xi .
(16.4)
Denote Ii as the information available to the ith trader at the beginning of the
period. The traders maximization problem is then
i
h
h
1i ) | Ii = max E eai (Rf W0i
max E Ui (W
Xi
Xi
i
[ P1 Rf P 0 ]Xi ) | Ii . (16.5)
1i depends on P1 , it is normally distributed, and, due to the exponential

Since W
form of the utility function, (16.5) is the moment generating function of a normal
random variable. Therefore, as we have seen earlier in the context of meanvariance analysis, the maximization problem is equivalent to
h
i
i
h
1
max E W1i | Ii ai Var W1i | Ii

Xi
2
(16.6)
or
h
i
i
h
1
2
max Xi E P1 | Ii Rf P0 ai Xi Var P1 | Ii
.
Xi
2
(16.7)
The first-order condition with respect to Xi then gives us the optimal number
463
of shares held in the risky asset:

i
h
E P1 | Ii Rf P0
i .
h
Xi =
ai Var P1 | Ii
(16.8)
Equation (16.8) indicates that the demand for the risky asset is increasing in
its expected excess return but declining in its price variance and the investors
risk aversion. Note that the CARA utility assumption results in the investors
demand for the risky asset being independent of wealth.
This simplifies the
derivation of the risky assets equilibrium price.
16.1.3
A Competitive Equilibrium
Now consider an equilibrium in which each trader uses his knowledge of the
unconditional distribution of P1 along with the conditioning information from
his private signal, yi , so that Ii = {yi }. Then using Bayes rule and the fact
that P1 and yi are jointly normally distributed with a squared correlation 2i
2
, the ith traders conditional expected value and variance of P1 are2
2
+ 2i
i
h
E P1 | Ii
Var P1 | Ii
m + 2i (yi m)
(16.9)
2 (1 2i ).
Substituting these into (16.8), we have
Xi =
2 Note
m + 2i (yi m) Rf P0
.
ai 2 (1 2i )
that i is the correlation coecient since
1 , y
cov (P
i )
P y
i
1
(16.10)
t
2 + 2
i
= i .

From the denominator of (16.10) one sees that the individuals demand for the
risky asset is greater the lower is his risk aversion, ai , and the greater is the
precision of his signal (the closer is i to 1, that is, the lower is i ). Now by
aggregating the individual traders risky asset demands for shares and setting
the sum equal to the fixed supply of shares, we can solve for the equilibrium
risky asset price, P0 , that equates supply and demand:
n
X
m + 2 (yi m) Rf P0
(16.11)
# ," n
" n
#
X
1 X m + 2i (yi m)
1
X
=
.
Rf i=1 ai 2 (1 2i )
a 2 (1 2i )
i=1 i
(16.12)
=
=
ai 2 (1 2i )
i=1
X
n
n
X
m + 2i (yi m)
Rf P0
ai 2 (1 2i )
ai 2 (1 2i )
i=1
i=1
or
P0
From (16.12) we see that the price reflects a weighted average of the traders
conditional expectations of the payo of the risky asset. For example the weight
on the ith traders conditional expectation, m + 2i (yi m), is
1
ai 2 (1 2i )
,"
n
X
i=1
1
ai 2 (1 2i )
(16.13)
The more precise (higher i ) is trader is signal or the lower is his risk aversion
(more aggressively he trades), the more that the equilibrium price reflects his
expectations.
16.1.4
A Rational Expectations Equilibrium
The solution for the price, P0 , in equation (16.12) can be interpreted as a competitive equilibrium: each trader uses information on his own signal and, in
465
equilibrium, takes the price of the risky asset as given in deciding on how much
to demand of the risky asset. However, this equilibrium neglects the possibility of individual traders obtaining information about other traders signals
from the equilibrium price itself, what practitioners call price discovery. In
this sense, the previous equilibrium is not a rational expectations equilibrium.
Why?
Suppose traders initially formulate their demands according to equa-
tion (16.10), using only information about their own signals, and the price in
(16.12) results.
Then an individual trader could infer information about the
other traders signals from the formula for P0 in (16.12).
Hence, this trader
would have the incentive to change his or her demand from that initially formulated in (16.10). This implies that equation (16.12) would not be the rational
expectations equilibrium price.
Therefore, to derive a fully rational expectations equilibrium, we need to

allow traders information sets to depend not only on their individual signals,
but on the equilibrium price itself: Ii = {yi , P0 (y)} where y (y1 y2 ... yn ) is
a vector of the traders individual signals and P0 (y) is the rational expectations
equilibrium price.
In equilibrium, the aggregate demand for the shares of the risky asset must
equal the aggregate supply, implying
i
h
n
E P1 | yi , P0 (y) Rf P0 (y)
X
=
.
h
i
X
ai Var P1 | yi , P0 (y)
i=1
(16.14)
Now one can show that a rational expectations equilibrium exists for the case
of the
i s
being independent and having the same variance, that is, 2i = 2 ,
for i = 1, ..., n.

Theorem: There exists a rational expectations equilibrium with P0 (y) given by
P0 (y)
m + 2 (
2 (1 2 ) X
y m)
=
Rf
Rf
," n
#
X 1
a
i=1 i
(16.15)
where y
1X
2
yi and 2
2 .
n i=1
2 + n
Proof: An intuitive outline of the proof is as follows.3 Note that in (16.15)

P0 (y) is a linear function of y with a fixed coecient of 2 /Rf . Therefore, if
a trader observes P0 (y) (and knows the structure of the model, that is, the
other parameters), then he can invert to infer the value of y. Now because
all traders signals were assumed to have equal precision (same 2 ), the average
signal, y, is a sucient statistic for the information contained in all of the
other signals. Further, because of the assumed independence of the signals, the
precision of this average of signals is proportional to the number of traders, n.
Hence, the average signal would have the same precision as a single signal with
variance
2
n .
Now if individual traders demands are given by equation (16.10) but where
yi is replaced with y and i is replaced with , then by aggregating these de as in equation (16.11), we end up with
mands and setting them equal to X
the solution in equation (16.15), which is consistent with our initial assumption
that traders can invert P0 (y) to find y. Hence, P0 (y) in equation (16.15) is the
rational expectations equilibrium price of the risky asset.
Note that the information, y, reflected in the equilibrium price is superior to
any single traders private signal, yi . In fact, since y is a sucient statistic for
all traders information, it makes knowledge of any single signal, yi , redundant.
The equilibrium would be the same if all traders received the same signal, y
N (0,
2
n )
3 See
or if they all decided to share information on their private signals
the original Grossman article (Grossman 1976) for details.
467
among each other before trading commenced.

Therefore, the above equilibrium is a fully-revealing rational expectations
equilibrium. The equilibrium price fully reveals all private information, a condition defined as strong-form market eciency.4
This result has some inter-
esting features in that it shows that prices can aggregate relevant information
to help agents make more ecient investment decisions than would be the case
if they relied solely on their private information and did not attempt to obtain
information from the equilibrium price itself.
However, as shown by Sanford Grossman and Joseph Stiglitz (Grossman
and Stiglitz 1980), this fully revealing equilibrium is not robust to some small
changes in assumptions. Real-world markets are unlikely to be perfectly ecient. For example, suppose each trader needed to pay a tiny cost, c, to obtain
his private signal, yi . With any finite cost of obtaining information, the equilibrium would not exist because each individual receives no additional benefit
from knowing yi given that they can observe y from the price. In other words,
a given individual does not personally benefit from having private (inside) information in a fully-revealing equilibrium. In order for individuals to benefit
from obtaining (costly) information, we need an equilibrium where the price is
only partially revealing. For this to happen, there needs to be one or more additional sources of uncertainty that add noise to individuals signals, so that
other agents cannot infer them perfectly. We now turn to an example of a noisy
rational expectations equilibrium.
16.1.5
A Noisy Rational Expectations Equilibrium
Let us make the following changes to the Grossman models assumptions along
the lines of a model proposed by Bruce Grundy and Maureen McNichols (Grundy
4 This can be compared to semi-strong form market eciency where asset prices need only
reflect all public information.

and McNichols 1989). Suppose that each trader begins the period with a random endowment of the risky asset. Specifically, trader i possesses i shares of
the risky asset so that her initial wealth is W0i = i P0 . The realization of i is
known only to trader i. Across all traders, the endowments, e
i , are indepen-
dently and identically distributed with mean X and variance 2X n. To simplify
the problem, we assume that the number of traders is very large. If we define
e as the per capita supply of the risky asset and let n go to infinity, then by the
X
e is a random variable distributed N X , 2 . Note

Central Limit Theorem X
X
e becomes zero, so
that in the limit as n , the correlation between e
i and X
that trader is observation of her own endowment, e

i , provides no information
e
about the per capita supply, X.
Next, let us modify the type of signal received by each trader to allow for a
common error as well as a trader-specific error. Trader i is assumed to receive

the signal
e + i
yi = P1 +
(16.16)
where
e N (0, 2 ) is the common error independent of P1 and, as before, the
e . Because of
idiosyncratic error i N(0, 2 ) and is independent of P1 and
the infinite number of traders, it is realistic to allow for a common error so that
traders, collectively, would not know the true payo of the risky asset.
Recall from the Grossman model that the rational expectations equilibrium
price in (16.15) was a linear function of y and X. In the current model, the
aggregate supply of the risky asset is not fixed, but random.
However, this
suggests that the equilibrium price will be of the form
e
P0 = 0 + 1 y + 2 X
(16.17)

where now y limn
Pn
i
469
yi /n = P1 +
e.
Although some assumptions dier, trader is demand for the risky asset continues to be of the form (16.8).
Now recall that in a rational expectations
equilibrium, investor is information set includes not only her private information but also the equilibrium price: Ii = {yi , P0 }. Given the assumed structure
e and yi , then investor
in (16.17) and the assumed normal distribution for Pe1 , X,
i optimally forecasts the end-of-period price as the projection

i
h
E Pe1 |Ii = 0 + 1 P0 + 2 yi
(16.18)
where
2
2
2 2
1
1 + + 2 X
=
1 2 + 2
2
1
1 2 + 2 21 2

2 + 2 + 2
2
= m 1 (0 1 m 2 X ) 2 m
(16.19)
If we then average the Xi in (16.8) over all investors, one obtains
0 + ( 1 Rf ) P0 + 2 y
h
i
aVar P1 | Ii )
(16.20)
1 Rf
0
i+
h
i P0 +
h 2
iy
aVar P1 | Ii )
aVar P1 | Ii )
aVar P1 | Ii )
h
where a 1/ limn
1
n
Pn
1
i ai
is the harmonic mean of the investors risk
aversions. Now note that we can re-write equation (16.17) as
X=
0
1
1
+
P0
y
2 2
2
(16.21)
In a rational expectations equilibrium, the relationships between the variables

X, P0 , and y must be consistent with the individual investors expectations.
This implies that the intercepts, and the coecients on P0 and on y, must be
identical in equations (16.20) and (16.21).
By matching the intercepts and
coecients, we obtain three nonlinear equations in the three unknowns 0 , 1 ,

and 2 . Although explicit solutions for 0 , 1 , and 2 cannot be obtained, we
can still interpret some of the characteristics of the equilibrium. To see this,
note that if the coecients on y are equated, one obtains
h 2
i
=
2
aVar P1 | Ii )
(16.22)
Using (16.19) to substitute for 2 and the variance of the projection of P1 on Ii

h
i
to substitute for Var P1 | Ii ) , (16.22) can be re-written as
1
2X
i
= h
2
a 2X (2 + 2 ) + (1 /2 )2 2 2
This is a cubic equation in 1 /2 .
(16.23)
The ratio 1 /2 is a measure of how
aggressively an individual investor responds to his individual private signal,

relative to the average signal, y, reflected in P0 . To see this, note that if one
uses (16.8), (16.20), (16.21), and yi y = i , the individuals demand for the
risky asset can be written as
Xi =
a
ai
1
X i
2
(16.24)
From (16.24) one sees that if there were no information dierences, each investor
would demand a share of the average supply of the risky asset, X, in proportion
to the ratio of the harmonic average of risk aversions to his own risk aversion.
However, unlike the fully revealing equilibrium of the previous section, the individual investor cannot perfectly invert the equilibrium price to find the average
16.2. ASYMMETRIC INFORMATION
471
signal in (16.17) due to the uncertain aggregate supply shift, X. Hence, individual demands do respond to private information as reflected by i . The ratio
1 /2 reflects the simultaneous equation problem faced by the investor in trying
to sort out a shift in supply, X, from a shift in aggregate demand generated by
y. From (16.23) we see that as 2 or 2 , so that investors private
signals become uninformative, then 1 /2 0 and private information has no
eect on demands or the equilibrium price. If, instead, 2 = 0, so that there
is no common error, then (16.23) simplifies to
1
1
=
2
a2
(16.25)
1
a
X
i
ai
ai 2
(16.26)
and (16.24) becomes

Xi =
so that an individuals demand responds to her private signal in direct proportion to the signals precision and indirect proportion to her risk aversion.
16.2
Asymmetric Information, Trading, and Markets
Let us now consider another model with private information that is pertinent
to a security market organized by a market maker. This market maker, who
might be thought of as a specialist on a stock exchange or a security dealer in an
over-the-counter market, sets a risky assets price with the recognition that he
may be trading at that price with a possibly better-informed individual. Albert
"Pete" Kyle (Kyle 1985) developed this model, and it has been widely applied
to study market micro-structure issues. The model is similar to the previous
one in that the equilibrium security price partially reveals the better-informed

individuals private information. Also like the previous model, there is an additional source of uncertainty that prevents a fully-revealing equilibrium, namely,
orders from uninformed noise" or "liquidity" traders that provide camouflage
for the better-informed individuals insider trades. The models results provide
insights regarding the factors aecting bid-ask spreads and the market impact
of trades.
16.2.1
Assumptions of the Kyle Model
The Kyle model is based on the following assumptions.

A.1 Asset Return Distribution
The model is a single period model.5
At the beginning of the period,
agents trade in an asset that has a random end of period liquidation value of
N p0 , 2v .
A.2 Liquidity Traders
Noise traders have needs to trade that are exogenous to the model. It is
assumed that they, as a group, submit a market order to buy u shares of
the asset, where u N 0, 2u . u

and are assumed to be independently
distributed.6
A.3 Better-Informed Traders

The single risk-neutral insider is assumed to have better information than
the other agents. He knows with perfect certainty the realized end of period
value of the risky security (but not u
) and chooses to submit a market order
5 Pete Kyles paper (Kyle 1985) also contains a multi-period continuous time version of
his single period model. Jiang Wang (Wang 1993) has also constructed a continuous time
asset pricing model with asymmetrically-informed investors who have constant absolute risk
aversion utility.
6 Why rational noise traders submit these orders has been modeled by assuming they
have exogenous shocks to their wealth and need to rebalance their portfolio (Spiegel and
Subrahmanyam 1992) or by assuming that they have uncertainty regarding the timing of
their consumption (Gorton and Pennacchi 1993).
473
of size x that maximizes his expected end of period profits.7

A.4 Competitive Market Maker
The single risk-neutral market maker (for example, a New York Stock Exchange specialist) receives the market orders submitted by the noise traders
and the insider, which in total equal u
e+x
e. Importantly, the market maker
cannot distinguish what part of this total order consists of orders made by noise
traders and what part consists of the order of the insider.
(The traders are
anonymous.) The market maker sets the market price, p, and then takes the
position (e
u+x
e) to clear the market. It is assumed that market making is
a perfectly competitive profession, so that the market maker sets the price p
such that, given the total order submitted, his profit at the end of the period is
expected to be zero.
16.2.2
Trading and Pricing Strategies
Since the noise traders order is exogenous, we need only consider the optimal
actions of the market maker and the insider.
The market maker observes only the total order flow, u + x. Given this
information, he must then set the equilibrium market price p that gives him
zero expected profits. Since his end of period profits are (
p) (u + x), this
implies that the price set by the market maker satisfies
p = E [
| u + x]
(16.27)
The information on the total order size is important to the market maker. The
more positive is the total order size, the more likely it is that x is large due to
the insider knowing that is greater than p0 . Thus, the market maker would
7 This assumption can be weakened to the case of the insider having uncertainty over
but having more information on than the other traders. One can also allow the insider to
submit limit orders, that is, orders that are a function of the equilibrium market price (a
demand schedule), as in another model by Pete Kyle (Kyle 1989).

tend to set p higher than otherwise. Similarly, the more negative is u + x, the
more likely it is that x is low because the insider knows is below p0 and is
submitting a sell order.
In this case, the market maker would tend to set p
lower than otherwise. Thus, the pricing rule of the market maker is a function
of x + u, that is, P (x + u).
Since the insider sets x, it is an endogenous variable that depends on .
The insider chooses x to maximize his expected end of period profits,
, given
knowledge of and the way that the market maker behaves in setting the
equilibrium price:
| ] = maxE [( P (x + u)) x | ]
maxE [
x
(16.28)
An equilibrium in this model is a pricing rule chosen by the market maker

and a trading strategy chosen by the insider such that: the insider maximizes
expected profits, given the market makers pricing rule; the market maker sets
the price to earn zero expected profits, given the trading strategy of the insider;
the insider and market maker have rational expectations, that is, the equilibrium
is a fixed-point where an agents actual behavior is that expected by the other.
Insiders Trading Strategy

Suppose the market maker chooses a market price that is a linear function of
the total order flow, P (y) = + y, where y = x + u. We will later argue that
a linear pricing rule is optimal. If this is so, what is the insiders choice of x?
From (16.28) we have
)) x | ] = maxE [( (x + u
)) x | ] (16.29)
maxE [( P (x + u
x
= max ( x) x, since E [
u] = 0
x
475
Thus, the solution to the insiders problem in (16.29) is
x = +
and =
where = 2
1
2 .
(16.30)
Therefore, if the market maker uses a linear pricing-
setting rule, the optimal trading strategy for the insider is a linear trading rule.
Market Makers Pricing Strategy
Next, let us return to the market makers problem of choosing the market price
that, conditional on knowing the total order flow, results in a competitive (zero)
expected profit. Given the assumption that market making is a perfectly competitive profession, a market maker needs to choose the best possible estimate
of E [
| u + x] in setting the price p = E [
| u + x]. What estimate of the mean
of is best? The maximum likelihood estimate of E [
| u + x] is best in the
sense that it attains maximum eciency and is also the minimum variance unbiased estimate.
Note that if the insider follows the optimal trading strategy, which according
to equation (16.30) is x = +
, then from the point of view of the market
maker, and y u
+x = u
+ +
are jointly normally distributed. Because
and y are jointly normal, the maximum likelihood estimate of the mean of
conditional on y is linear in y, that is, E [
| y] is linear in y when they are
jointly normally distributed. Hence, the previously assumed linear pricing rule
is, in fact, optimal in equilibrium. Therefore, the market maker should use the
maximum likelihood estimator, which in the case of and y being normally
distributed is equivalent to the least squares estimator. This is the one that
minimizes
i
h
i
h
y)2
E (
P (y))2 = E (
h
i
= E (
(
u + +
))2
(16.31)
Thus, the optimal pricing rule equals + y where and minimize

i
h
(1 )
u )2
minE (
,
(16.32)
i
h

Recalling the assumptions E [] = p0 , E ( p0 )2 = 2v , E [u] = 0, E u2 =
2u , and E [u] = 0, the above objective function can be written as
2 2
v
min (1 )
,
+ p20 + ( + )2 + 2 2u 2 ( + ) (1 ) p0 (16.33)
The first order conditions with respect to and are
= + p0 (1 )
(16.34)
0 = 2 (1 ) 2v + p20 + 2 ( + ) + 2 2u
2p0 [ ( + ) + (1 )]
(16.35)
Substituting + = p0 (1 ) from (16.34) into (16.35), we see that (16.35)

simplifies to
2v
2 2v + 2u
and =
Substituting in for the definitions = 2
(16.36)
1
2
in (16.34) and (16.35),
477
we have
= p0
1
2
v
u
(16.37)
(16.38)
In summary, the equilibrium price is
p = p0 +
1
2
v
(
u + x)
u
(16.39)
where the equilibrium order submitted by the insider is
x=
16.2.3
u
(
p0 )
v
(16.40)
Analysis of the Results
From (16.40), we see that the greater is the volatility (amount) of noise trading,
u , the larger is the magnitude of the order submitted by the insider for a given
deviation of from its unconditional mean. Hence, the insider trades more
actively on his private information the greater is the camouflage provided by
noise trading.
Greater noise trading makes it more dicult for the market
maker to extract the signal of insider trading from the noise. Note that if
equation (16.40) is substituted into (16.39), one obtains
v
+ 12 (
p0 )
p = p0 + 12 u
u
v
u + p0 +
= 12
u
(16.41)

Thus we see that only one-half of the insiders private information,
1
2
reflected in the equilibrium price, so that the price is not fully revealing.8
, is
To
obtain an equilibrium of incomplete revelation of private information, it was

necessary to have a second source of uncertainty, namely, the amount of noise
trading.
Using (16.40) and (16.41), we can calculate the insiders expected profits:
E [
] = E [x ( p)] = E
u
(
p0 ) 12
v
v
u
p0
u
(16.42)
Conditional on knowing , that is, after learning the realization of at the

beginning of the period, the insider expects profits of
E [
| ] =
1
2
u
( p0 )2
v
(16.43)
Hence, the larger is s deviation from p0 , the larger the expected profit. Unconditional on knowing , that is, before the start of the period, the insider
expects a profit of
E [
] =
1
2
i
u h
E (
p0 )2 = 12 u v
v
(16.44)
which is proportional to the standard deviations of noise traders order and the
end of period value of .
Since, by assumption, the market maker sets the security price in a way that
gives him zero expected profits, the expected profits of the insider equals the
expected losses of the noise traders. In other words, it is the noise traders, not
the market maker, that lose, on average, from the presence of the insider.
From equation (16.39), we see that =
8A
fully-revealing price would be p = .
1 v
2 u
is the amount that the market
16.3. SUMMARY
479
maker raises the price when the total order flow, (u + x), goes up by 1 unit.9
This can be thought of as relating to the securitys bid-ask spread, that is, the
dierence the price for sell orders versus buy orders, though here sell and buy
prices are not fixed but are a function of the order size since the pricing rule is
linear. Moreover, since the amount of order flow necessary to raise the price
by $1 equals 1/ = 2 uv , the model provides a measure of the depth of the
market or market liquidity. The higher is the proportion of noise trading to
the value of insider information,
u
v ,
the deeper or more liquid is the market.
Intuitively, the more noise traders relative to the value of insider information,
the less the market maker needs to adjust the price in response to a given
order, since the likelihood of the order being that of a noise trader, rather
than an insider, is greater. While the greater is the number of noise traders
(that is, the greater is u ), the greater is the expected profits of the insider
(see equation (16.44)) and the greater is the total expected losses of the noise
traders. However, the expected loss per individual noise trader falls with the
greater level of noise trading.10
16.3
Summary
The models considered in this chapter analyze the degree to which private information about an assets future payo or value is reflected in the assets current
price. An investors private information aects an assets price by determining
the investors desired demand (long or short position) for the asset, though the
investors demand also is tempered by risk-aversion. More subtly, we saw that a
rational investor can also learn about the private information of other investors
through the assets price itself, and this price discovery aects the investors
9 It is now common in the market microstructure literature to refer to this measure of order
flow and liquidity as "Kyles lambda."
10 Gary Gorton and George Pennacchi (Gorton and Pennacchi 1993) derive this result by
modeling individual liquidity traders.

equilibrium demands. Indeed, under some circumstances, the assets price may
fully reveal all relevant private information such that any individuals private
information becomes redundant.
Perhaps more realistically, there are non-information-based factors that affect the net supply or demand for an asset. These "noise" factors prevent investors from perfectly inferring the private information signals of others, resulting in an asset price that is less than fully revealing. Noise provides camouflage
for investors with private information, allowing these traders to profit from possessing such information. Their profits come at the expense of liquidity traders
since the greater the likelihood of private information regarding a security, the
larger will be the securitys bid-ask spread. Hence, this theory predicts that a
securitys liquidity is determined by the degree of noise (non-information-based)
trading relative to insider (private-information-based) trading.
16.4
Exercises
1. Show that the maximization problem in (16.7) is equivalent to the maximization problem in (16.5).
2. Show that the results in (16.9) can be derived from Bayes Rule and the
assumption that P1 and yi are normally distributed.
3. Consider a special case of the Grossman model. Traders can choose between holding a risk-free asset, which pays an end-of-period return of Rf ,
or a risky asset that has a beginning of period price of P0 per share and
has an end-of-period payo (price) of Pe1 per share. The unconditional dis
tribution of Pe1 is assumed to be N m, 2 . The risky asset is assumed to
be a derivative security, such as a futures contract, so that its net supply

equals zero.
16.4. EXERCISES
481
There are two dierent traders who maximize expected utility over end-off1i , i = 1, 2. The form of the ith traders utility function
period wealth, W
is
i1i
f1i = eai W
Ui W
, ai > 0.
At the beginning of the period, the ith trader observes yi which is a noisy
signal of the end-of-period value of the risky asset
where
i N
yi = Pe1 + ei
0, 2 and is independent of Pe1 . Note that the variances of
the traders signals are the same. Also assume E [
1 2]
=0.
3.a Suppose each trader does not attempt to infer the other traders information from the equilibrium price, P0 . Solve for each of the traders demands
for the risky asset and the equilibrium price, P0 .
3.b Now suppose each trader does attempt to infer the others signal from the
equilibrium price, P0 . What will be the rational expectations equilibrium
price in this situation? What will be each of the traders equilibrium
demands for the risky asset?
4. In the Kyle (1985) model, replace the original assumption A.3 with the
following new one:
A.3 Better-Informed Traders
The single risk-neutral insider is assumed have better information than
the other agents. He observes a signal of the assets end-of-period value
equal to
s = ve + e
is distributed independently of u
where e
~N 0, 2s , 0 < 2s < 2v , and e
and . The insider does not observe u
, but chooses to submit a market
order of size x that maximizes his expected end-of-period profits.
4.a Suppose that the market makers optimal price setting rule is a linear
function of the order flow
p = + (u + x) .
Write down the expression for the insiders expected profits given this
pricing rule.
4.b Take the first order condition with respect to x and solve for the insiders
optimal trading strategy as a function of the signal and the parameters of
the market makers pricing rule.
Chapter 17
Models of the Term

Structure of Interest Rates
The objective of this chapter is to provide an introduction to the main approaches for modeling the term structure of interest rates and for valuing fixedincome derivatives. It is not meant to be a comprehensive review of this subject,
as the literature on term structure models is voluminous, and many surveys
on this topic, including (Dai and Singleton 2004), (Dai and Singleton 2003),
(Maes 2003), (Piazzesi 2003), (Rebonato 2004), and (Yan 2001) have been written in recent years. This chapter discusses valuation models of bonds and bond
derivatives that are assumed to be default-free, such as Treasury bills, notes,
bonds, and their derivatives.
The following chapter analyzes the eects of
default risk on bond prices.

The chapter is broken into two main sections. In the first we discuss models
that derive the equilibrium bond prices of dierent maturities in terms of particular state variables. One way to think about these models is that the state
variables are the models "input" while the bond values of dierent maturities
465
466
CHAPTER 17. TERM STRUCTURE MODELS
are the models "output." The second section covers models that value fixedincome derivatives, such as interest rate caps and swaptions, in terms a given set
of bond prices. In contrast, these models take the term structure of observed
bond prices as the input and have derivative values as the models output.
17.1
Equilibrium Term Structure Models
Equilibrium term structure models describe the prices (or, equivalently, the
yields) of dierent maturity bonds as functions of one or more state variables or
"factors." The Vasicek (Vasicek 1977) model, introduced in Chapter 9 (see equation 9.28), and the Cox, Ingersoll, and Ross (Cox, Ingersoll, and Ross 1985b)
model, presented in Chapter 13 (see equation 13.51) were examples of singlefactor models. The single factor in the Vasicek model was the instantaneous
maturity interest rate, denoted r (t), which was assumed to follow the OrnsteinUhlenbeck process (9.17). In Cox, Ingersoll, and Rosss one-factor model, the
factor determined the expected returns of the economys production processes,
and, in equilibrium, the instantaneous maturity interest rate was proportional
to this factor and inherited its dynamics. This interest rate followed the square
root process in equation (13.49).
Let us generalize the pricing relationships for default-free zero-coupon bonds
by considering a situation where multiple factors determine their prices. The
derivation of these bond values will utilize our results from earlier chapters. We
start by assuming that there are n state variables, xi , i = 1, ..., n, that follow
the multi-variate diusion process
dx = a (t, x) dt + b (t, x) dz
(17.1)
where x = (x1 ...xn )0 , a (t, x) is a nx1 vector, b (t, x) is a nxn matrix, and
17.1. EQUILIBRIUM TERM STRUCTURE MODELS
467
dz = (dz1 ...dzn )0 is an n1 vector of independent Brownian motion processes so

that dzi dzj = 0 for i 6= j.1 This specification allows for any general correlation
structure for the state variables when b (t, x) has full rank.
Define P (t, T, x) as the date t price of a default-free, zero-coupon bond that
pays 1 at date T . Its lemma gives the process followed by this bonds price.
dP (t, T, x) /P (t, T, x) = p (t, T, x) dt + p (t, T, x)0 dz
(17.2)
where
0
0
p (t, T, x) = a (t, x) Px + Pt + 12 Trace b (t, x) b (t, x) Pxx /P (t, T, x)
(17.3)
and p (t, T, x) is an n 1 vector equal to

p (t, T, x) = b (t, x)0 Px /P (t, T, r)
(17.4)
and where Px is an n 1 vector whose ith element equals the partial derivative
Pxi , Pxx is an nxn matrix whose i, j th element is second order partial derivative
Pxi xj , and Trace[A] is the sum of the diagonal elements of a square matrix A.
Similar to the Black-Scholes hedging argument discussed in Chapter 9 and
applied to derive the Vasicek model, we can form a hedge portfolio of n+1 bonds
having distinctly dierent maturities. By appropriately choosing the portfolio
weights for these n + 1 bonds, the n sources of risk can be hedged so that the
portfolio generates a riskless return. In the absence of arbitrage, this portfolios
return must equal the riskless rate, r (t, x). Making this no-arbitrage restriction
produces the implication that each bonds expected rate of return must satisfy
1 As discussed in Chapter 10, the independence assumption is not important. If there are
correlated sources of risk (Brownian motions), they can be re-defined by a linear transformation to be represented by n orthogonal risk sources.
468
p (t, T, x) = r (t, x) + (t, x) 0 p (t, T, x)
(17.5)
where (t, x) = ( 1 ...n )0 is the n 1 vector of market prices of risks associated

0
with each of the Brownian motions in dz = (dz1 ...dzn ) . By equating (17.5)

to the process for p (t, T, x) given by Its lemma in (17.3), we obtain the
equilibrium partial dierential equation
1
2 Trace
0
0
b (t, x) b (t, x) Pxx + [a (t, x) b (t, x) ] Px rP + Pt = 0 (17.6)
Given functional forms for a (t, x), b (t, x), (t, x), r (t, x), this PDE can be
solved subject to the boundary condition P (T, T, x) = 1. Note that equation
(17.6) depends on the expected changes in the factors under the risk-neutral
measure Q, a (t, x) b (t, x) , rather than the factors expected changes under
the physical measure P , a (t, x).
Hence, to price bonds, one could simply
specify only the factors risk-neutral processes.2 This insight is not surprising
because we saw in Chapter 10 that the Feynman-Kac solution to this PDE is
the risk-neutral pricing equation (10.58).
h R
i
bt e tT r(s,x)ds 1
P (t, T, x) = E
(17.7)
In addition to the pricing relations (17.6) and (17.7), we saw that a third pricing
approach can be based on the pricing kernel that follows the process
dM/M = r (t, x) dt (t, x)0 dz
(17.8)
In this case, pricing can be accomplished under the physical measure based on
2 However, if the factors are observative variables for which data are available, it may be
necessary to specify their physical processes if empirical implementations of the model require
estimates for a (x, t) and b (x, t).

the formula
M (T )
P (t, T, x) = Et 1
M (t)
469
(17.9)
Thus far, we have placed few restrictions on the factors and their relationship
to the short rate, r (t, x), other than to assume that the factors follow the Markov
diusion processes (17.1). Let us next consider some popular parametric forms.
17.1.1
Ane Models
We start with models in which the yields of zero coupon bonds are linear
or "ane" functions of state variables.
This class of models includes those
of Vasicek (Vasicek 1977) and Cox, Ingersoll, and Ross (Cox, Ingersoll, and
Ross 1985b).
Ane models are attractive because they lead to bond price
formulae that are relatively easy to compute and because the parameters of the
state variable processes can often be estimated using relatively straightforward
econometric techniques.
Recall that a zero-coupon bonds continuously compounded yield, Y (t, T, x),
is defined from its price by the relation
P (t, T, x) = eY (t,T,x)(T t)
(17.10)
One popular class of models assumes that zero-coupon bonds continuouslycompounded yields are ane functions of the factors. Defining the time until
maturity as T t, this assumption can be written as
0
Y (t, T, x) = A ( ) + B ( ) x
(17.11)
where A ( ) is a scalar function and B ( ) is an nx1 vector of functions that

do not depend on the factors, x. Because at maturity P (T, T, x) = 1, (17.11)
implies that A (0) = 0 and B (0) is an nx1 vector of zeros. Another implication
470
of (17.11) is that the short rate is also ane in the factors since
A ( ) + B ( )0 x
0
r (t, x) = lim Y (t, T, x) = lim

T t
(17.12)
so that we can write r (t, x) = + 0 x, where = A (0) / is a scalar and

= B (0) / is an nx1 vector of constants.
Under what conditions regarding the factors dynamics would the no-arbitrage,
equilibrium bond yields be ane in the state variables? To answer this, let us
substitute the ane yield assumption of (17.10) and (17.11) into the general
no-arbitrage PDE of (17.6). Doing so, one obtains
0
1
2 B ( )
b (t, x) b (t, x)0 B ( ) [a (t, x) b (t, x) ]0 B ( )
)
+ A(
+
B( )0
x
= + 0 x
(17.13)
Darrell Due and Rui Kan (Due and Kan 1996) characterize sucient conditions for a solution to equation (17.13).
Specifically, two of the conditions
are that the factors risk-neutral instantaneous expected changes and variances
are ane in x.
In other words, if the state variables risk-neutral drifts and
variances are ane in the state variables, so will the equilibrium bond price
yields. These conditions can be written as
a (t, x) b (t, x) = (xx)
(17.14)
p
b (t, x) = s (x)
(17.15)
where x is an nx1 vector of constants, and are nxn matrices of constants,

and s (x) is an nxn diagonal matrix with the ith diagonal term
si (x) = soi + s01i x
(17.16)
471
where soi is a scalar constant and s1i is an nx1 vector of constants. Now, because the state variables covariance matrix equals b (t, x) b (t, x)0 = s (x) 0 ,
additional conditions are needed to ensure that this covariance matrix remains
positive definite for all possible realizations of the state variable, x. Qiang Dai
and Kenneth Singleton (Dai and Singleton 2000) and Darrell Due, Damir Filipovic, and Walter Schachermayer (Due, Filipovic, and Schachermayer 2002)
derive these conditions.3
Given (17.14), (17.15), and (17.16), the partial dierential equation in (17.13)
can be re-written as
0
0
1
2 B ( ) s (x) B ()
[ (xx)]0 B () +
A ( ) B ( )0
+
x
= + 0 x
(17.17)
Note that this equation is linear in the state variables, x. For the equation to
hold for all values of x, the constant terms in the equation must sum to zero
and the terms multiplying each element of x must also sum to zero.
These
conditions imply
A ( )
B ( )
= + (x)0 B ( )
= 0 B ()
1
2
1
2
n
P
i=1
n
P
[0 B ( )]i s0i
(17.18)
[0 B ()]i s1i
(17.19)
i=1
where [0 B ()]i is the ith element of the nx1 vector 0 B ( ). Equations (17.18)
and (17.19) are a system of first order ordinary dierential equations that can
3 These conditions can have important consequences regarding the correlation between the
state variables. For example, if the state variables follow a multivariate Ornstein Uhlenbeck
process, so that the model is a multifactor version of the Vasicek model given in (9.28), (9.29),
and (9.30), then any general correlation structure between the state variables is permitted.
However, if the state variables follow a multivariate square-root process, so that the model
is a multifactor version of the Cox, Ingersoll and Ross model given in (13.51), (13.52), and
(13.53), then the correlation between the state variables must be non-negative.
472
be solved subject to the boundary conditions A (0) = 0 and B (0) = 0. In some

cases, such as a multiple state variable version of the Vasicek model (where s1i
= 0 i), there exist closed-form solutions.
In other cases, fast and accurate
numerical solutions to these ordinary dierential equations can be computed

using techniques such as a Runge Kutta algorithm.
While ane term structure models require that the state variables riskneutral expected changes be ane in the state variables, there is more flexibility
regarding the state variables drifts under the physical measure. Note that the
state variables expected change under the physical measure is
p
a (t, x) = (xx) + s (x)
(17.20)
so that specification of the market prices of risk, , is required to determine

the physical drifts of the state variables.
Qiang Dai and Kenneth Singleton
(Dai and Singleton 2000) study the "completely ane" case where both the
physical and risk-neutral drifts are ane, while Gregory Duee (Duee 2002)
and Jeerson Duarte (Duarte 2004) consider extensions of the physical drifts
that permit nonlinearities.4 Because the means, volatilities, and risk premia of
bond prices estimated from time series data depend on the physical moments
of the state variables, the flexibility in choosing the parametric form for can
allow the model to better fit historical bond price data.
p
and Singleton analyze = s (x) 1 where 1 is an nx1 vector of constants. Duffee considersqthe "essentially ane" modeling of the market price of risk of the form =
p
s (x) 1 + s (x) 2 x, where s (x) is an nxn diagonal matrix whose ith element equals
soi + s01i x
if inf soi + s01i x > 0, zero otherwise, and 2 is an nxn matrix of constants.
This specification allows time variation in the market prices of risk for Gaussian state variables (such as state variables that follow Ornstein-Uhlenbeck processes), allowing their signs
to switch over time. Duarte extend Duees modeling to add a square
q root term. This "semi
p
ane square-root" model takes the form = 1 0 + s (x) 1 + s (x) 2 x where 2 is
an nx1 vector of constants. See, also, work by Patrick Cheridito, Damir Filipovic, and Robert
Kimmel (Cheridito, Filipovic, and Kimmel 2003) for extensions in modeling the market price
of risk for ane models.
4 Dai
17.1.2
473
Quadratic-Gaussian Models
Another class of models assumes that the yields of zero coupon bonds are
quadratic functions of state variables.
This assumption can be expressed as
Y (t, T, x) = A () + B ( )0 x + x0 C ( ) x
(17.21)
where C ( ) is an nxn matrix, and, with no loss of generality, can be assumed to

be symmetric. Similar to our analysis of ane models, since P (T, T, x) = 1, we
must have A (0) = 0, B (0) equal to an nx1 vector of zeros, and C (0) equal to an
nxn matrix of zeros. In addition, the yield on a bond of instantaneous maturity
must be of the form r (t, x) = + 0 x + x0 x where = A (0) / , =
B (0) / , and = C (0) / is an nxn symmetric matrix of constants. Note
that if is a positive semi-definite matrix and 14 0 1 0, then the interest
rate can be restricted from becoming negative.5 Substituting P (t, T, x) =
0
exp A ( ) B () x x0 C ( ) x into the general partial dierential equation
(17.6), we obtain
[B ( ) + 2C ( ) x]0 b (t, x) b (t, x)0 [B ( ) + 2C ( ) x]
Trace b (t, x)0 C () b (t, x)

1
2
[a (t, x) b (t, x) ]0 [B ( ) + 2C ( ) x]
A () B ()0
C ( )
+
x+ x0
x
= + 0 x + x0 x
(17.22)
In addition to yields being quadratic in the state variables, quadratic-gaussian

models then assume that the state variable, x, has a multivariate normal (Gaussian)
5 The
lower bound for r (t) is 14 0 1 , which occurs when x = 12 1 .
474
distribution. Specifically, it is assumed that x follows a multivariate OrnsteinUhlenbeck process:
a (t, x) b (t, x) = (xx)
(17.23)
b (t, x) =
(17.24)
Substituting these assumptions into the partial dierential equation (17.22), one
obtains
1
2
[B ( ) + 2C ( ) x]0 0 [B () + 2C () x]
Trace [0 C ( ) ] [ (xx)]0 [B ( ) + 2C ( ) x]
+
A ( ) B ( )0
C ()
+
x+ x0
x
= + 0 x + x0 x
(17.25)
For this equation to hold for all values of x, it must be the case that the sums of
the equations constant terms, the terms proportional to the elements of x, and
the terms that are products of the elements of x must each equal zero. This
leads to the system of first order ordinary dierential equations
A ( )
= + (x)0 B ( ) 12 B ( )0 0 B () + Trace [0 C ( ) ]
(17.26)
B ( )
C ( )
= 0 B ( ) 2C ( )0 0 B () + 2C ()0 x
(17.27)
= 20 C () 2C ()0 0 C ( )
(17.28)
which are solved subject to the aforementioned boundary conditions, A (0) = 0,

B (0) = 0, and C (0) = 0.
Dong-Hyun Ahn, Robert Dittmar, and Ronald
475
Gallant (Ahn, Dittmar, and Gallant 2002) show that the models of Francis
Longsta (Longsta 1989), David Beaglehole and Mark Tenney (Beaglehole
and Tenney 1992), and George Constantinides (Constantinides 1992) are special cases of quadratic-gaussian models.
They also demonstrate that since
quadratic-gaussian models allow a non-linear relationship between yields and

state variables, these models can outperform ane models in explaining historical bond yield data.
However, quadratic-gaussian models are more dicult to estimate from historical data because, unlike ane models, there is not a one-to-one mapping
between bond yields and the elements of the vector of state variables.
For
example, suppose that at a given point in time we observed bond yields of n

dierent maturities, say Y (t, Ti , x), i = 1, ..., n. Denoting i = Ti t, if yields
0
are ane functions of the state variables then Y (t, Ti , x) i = A ( i ) + B ( i ) x,
i = 1, .., n, represent a set of n linear equations in the n elements of the state

variable x.
Solving these equations for the state variables x1 , x2 , ..., xn ef-
fectively allows one to observe the individual state variables from the observed
yields.
By observing a time-series of these state variables, the parameters of
their physical process could be estimated.

This approach cannot be used when yields are quadratic functions of the
0
state variables since with Y (t, Ti , x) i = A ( i ) + B ( i ) x + x0 C ( i ) x there
is not a one-to-one mapping between yields and state variables x1 , x2 , ..., xn .

There are multiple values of the state variable vector, x, consistent with the set
of yields.6 This diculty requires an dierent approach to inferring the most
likely state variable vector. Ahn, Dittmar, and Gallant use an ecient method
of moments technique that simulates the state variable, x, to estimate the state
variable vector that best fits the data.
6 For
example, if n = 1, there are two state variable roots of the quadratic yield equation.
476
17.1.3
Other Equilibrium Models
Term structure models have been modified to allow state variable processes to
dier from strict diusions. Such models can no longer rely on the Black-Scholes
hedging argument to identify market prices of risk and a risk-neutral pricing
measure.
Because fixed-income markets may not be dynamically complete,
these models need to make additional assumptions regarding the market prices
of risks that cannot be hedged.
A number of researchers, including Chang-Mo Ahn and Howard Thompson (Ahn and Thompson 1988), Sanjiv Das and Silverio Foresi (Das and Foresi
1996), Darrell Due, Jun Pan, and Kenneth Singleton (Due, Pan, and Singleton
2000), Sanjiv Das (Das 2002), and George Chacko and Sanjiv Das (Chacko and
Das 2002), have extended equilibrium models to allow state variables to follow
jump-diusion processes. An interesting application of a model with jumps in
a short-term interest rate is presented by Monika Piazzesi (Piazzesi 2005) who
studies changes in the federal funds rate made by the Federal Reserve.
Other ane equilibrium models have been set in discrete-time, where the
assumed existence of a discrete-time pricing kernel allows one to find solutions
for equilibrium bond prices that have a recursive structure. Examples of models
of this type include work by Tong-Sheng Sun (Sun 1992), David Backus and
Stanley Zin (Backus and Zin 1994), and V. Cvsa and Peter Ritchken (Cvsa and
Ritchken 2001). Term structure models also have been generalized to include
discrete regime shifts in the processes followed by state variables. See work by
Vasant Naik and Moon Hoe Lee (Naik and Lee 1997) and Ravi Bansal and Hao
Zhou (Bansal and Zhou 2002) for models of this type.
Let us now turn to term structure models whose primary purpose is not
to determine the term structure of zero coupon bond prices as a function of
17.2. VALUATION MODELS FOR INTEREST RATE DERIVATIVES
477
state variables. Rather, their objective is to determine the value fixed-income

derivatives as a function of a given term structure of bond prices.
17.2
Valuation Models for Interest Rate Derivatives
Models for valuing bonds and bond derivatives have dierent uses. The equilibrium models of the previous section can provide insights as to the nature
of term structure movements. They allow us to predict how factor dynamics
will aect the prices of bonds of dierent maturities. Equilibrium models may
also be of practical use to bond traders who wish to identify bonds of particular
maturities whose market valuations appear to be over- or under-priced based on
their predicted model prices. Such information could suggest profitable bond
trading strategies.
However, bond prices are modeled for other objectives, such as the pricing
of derivatives whose payos depend on the future prices of bonds or yields.
Equilibrium models may be less than satisfactory for this purpose because it is
bond derivatives, not the underlying bond prices themselves, that one wishes to
value. In this context, one would like to use observed market prices for bonds
as an input into the valuation formulas for derivatives, not model the value of
the underlying bonds themselves.
For such a derivative-pricing exercise, one
would like the model to "fit," or be consistent with, the market prices of the
underlying bonds.
In a discrete-time, binomial model setting, Thomas Ho and Sang Bin Lee
((Ho and Lee 1986)) first introduced the concept of pricing fixed-income derivatives by taking the initial term structure of bond prices as given and then
making assumptions regarding the risk-neutral distribution of future interest
478
rates. This binomial approach was modified for dierent risk-neutral interest
rate distributions by Fischer Black, Emanuel Derman, and William Toy (Black,
Derman, and Toy 1990) and Fischer Black and Piotr Karasinski (Black and
Karasinski 1991). While these binomial models are not covered in this chapter,
we will discuss a similar approach that is set in continuous time. It shares the
characteristics of fitting the currently observed market prices of bonds and of
making particular assumptions regarding the risk-neutral distribution of future
interest rates.
17.2.1
Heath-Jarrow-Morton Models
The approach by David Heath, Robert Jarrow, and Andrew Morton ((Heath,
Jarrow, and Morton 1992)), hereafter referred to as HJM, diers from the previous equilibrium term structure models because it does not begin by specifying
a set of state variables, x, that determines the current term structure of bond
prices. Rather, their approach takes the initial term structure of bond prices as
given (observed) and then specifies how this term structure evolves in the future
in order to value derivatives whose payos depend on future term structures.
Because models of this type do not derive the term structure from more basic
state variables, they cannot provide insights regarding how economic fundamentals determine the maturity structure of zero-coupon bond prices. Instead,
HJM models are used to value fixed-income derivative securities: securities such
as bond and interest rate options whose payos depend on future bond prices
or yields.
An analogy to the HJM approach can be drawn from the risk-neutral valuation of equity options. Recall that in Chapter 10 equation (10.47), we assumed
that the risk-neutral process for the price of a stock, S (t), was geometric Brownian motion, making this price lognormally distributed under the risk-neutral
479
measure. From this assumption, and given the initial price of the stock, S (t),
the Black-Scholes formula for the value of a call option written on this stock
was derived in equations (10.51) and (10.52). Note that we did not attempt to
determine the initial value of the stock in terms of some fundamental state variables, say S (t, x). Rather, the initial stock price, S (t), was taken as given and
an assumption about this stock prices volatility, namely that it was constant
over time, was made.
The HJM approach to valuing fixed-income derivatives is similar but slightly
more complex because it takes as given the entire initial term structure of bond
prices, P (t, T ) T 1, not just a single asset price.
It then assumes risk-
neutral processes for how this initial set of bond prices change over time and
does not attempt to derive these initial prices in terms of state variables, say
P (t, T, x).
However, the way that HJM specify the processes followed by
bond prices is somewhat indirect. They begin by specifying processes for bond
forward rates. A fundamental result of the HJM analysis is to show that, in the
absence of arbitrage, there must be a particular relationship between the drift
and volatility parameters of forward rate processes and that only an assumption
regarding the form of forward rate volatilities is needed for pricing derivatives.
Let us start by defining forward rates. Recall from Chapter 7 that a forward
contract is an agreement between two parties where the long (short) party agrees
to purchase (deliver) an underlying asset in return for paying (receiving) the
forward price.
Consider a forward contract agreed to at date t, where the
contract matures at date T t and the underlying asset is a zero-coupon bond

that matures at date T + where 0.
forward price agreed to by the parties.
Let F (t, T, ) be the equilibrium
Then this contract requires the long
party to pay F (t, T, ) at date T in return for receiving a cashflow of $1 (the

zero-coupon bonds maturity value) at date T + . In the absence of arbitrage,
480
the value of these two cashflows at date t must sum to zero, implying
F (t, T, ) P (t, T ) + P (t, T + ) = 0
(17.29)
so that the equilibrium forward price equals the ratio of the bond prices maturing
at dates T + and T , F (t, T, ) = P (t, T + ) /P (t, T ).
From this forward
price a continuously-compounded forward rate, f (t, T, ), is defined as
ef(t,T, ) F (t, T, ) =
P (t, T + )
P (t, T )
(17.30)
f (t, T, ) = (ln [P (t, T + ) /P (t, T )]) / is the implicit per-period rate of

return (interest rate) that the long party earns by investing $F (t, T, T + ) at
date T and by receiving $1 at date T + .
Now consider the case of such a
forward contract where the underlying bond matures very shortly (e.g., the next
day or instant) after the maturity of the forward contract. This permits us to
define an instantaneous forward rate as
f (t, T ) limf (t, T, ) = lim

0
ln [P (t, T )]
ln [P (t, T + )] ln [P (t, T )]
=
T
(17.31)
Equation (17.31) is a simple dierential equation that can be solved to obtain
P (t, T ) = e
RT
t
f (t,s)ds
(17.32)
Since this bonds continuously-compounded yield to maturity is defined from

RT
the relation P (t, T ) = eY (t,T )(T t) , we can write Y (t, T ) = T 1t t f (t, s) ds.
Thus, a bonds yield equals the average of the instantaneous forward rates for
horizons out to the bonds maturity. In particular, the yield on an instantaneous

maturity bond is given by r (t) = f (t, t).
481
Because the term structure of instantaneous forward rates, f (t, T ) T t ,

can be determined from the term structure of bond prices, P (t, T ) T t,
or yields, Y (t, T ) T t, specifying the evolution of forward rates over time is
equivalent to specifying the dynamics of bond prices. HJM assume that forward
rates for all horizons are driven by a finite-dimensional Brownian motion:
df (t, T ) = (t, T ) dt + (t, T )0 dz
(17.33)
where (t, T ) is an nx1 vector of volatility functions and dz is an nx1 vector of

independent Brownian motions. Note that since there are an infinite number
of instantaneous forward rates, one for each future horizon, equation (17.33)
represents infinitely many processes that are driven by the same n Brownian
motions.
Importantly, the absence of arbitrage places restrictions on (t, T ) and
(t, T ).
To show this, let us start by deriving the process followed by bond
prices, P (t, T ), implied by the forward rate processes. Note that since ln [P (t, T )]
RT
= t f (t, s) ds, if we dierentiate with respect date t, we find that the process
followed by the log bond price is
RT
d ln [P (t, T )] = f (t, t) dt t df (t, s) ds
(17.34)
RT
= r (t) dt t (t, s) dt + (t, s)0 dz (t) ds

Fubinis Theorem allows us to switch the order of integration:
d ln [P (t, T )] = r (t) dt
RT
t
(t, s) dsdt
RT
t
(t, s) dsdz (t) (17.35)
= r (t) dt I (t, T ) dt I (t, T )0 dz (t)

where we have used the shorthand notation I (t, T )
RT
t
(t, s) ds and I (t, T )
482
RT
t
(t, s) ds to designate these integrals which are known functions as of date
t. Using Its lemma we can derive the bonds rate of return process from the
log process in (17.35).
1
dP (t, T )
= r (t) I (t, T ) + I (t, T )0 I (t, T ) dt I (t, T )0 dz (17.36)
P (t, T )
2
Now recall from (17.5) that the absence of arbitrage requires that the bonds
expected rate of return equal the instantaneous risk-free return plus the product
of the bonds volatilities and the market prices of risk. This is written as
1
r (t) I (t, T ) + I (t, T )0 I (t, T ) = r (t) (t) 0 I (t, T )
2
(17.37)
or
1
0
I (t, T ) = I (t, T ) I (t, T ) + (t) 0 I (t, T )
2
(17.38)
Equation (17.36) and (17.38) shows that the bond price process depends only
on the instantaneous risk-free rate, the volatilities of the forward rates, and the
market prices of risk.
This no-arbitrage condition also has implications for
the risk-neutral process followed by forward rates. If we substitute dz = db

z
(t) dt in (17.33), we obtain
df (t, T ) =
0
0
(t, T ) (t, T ) (t) dt + (t, T ) db
z
=
b (t, T ) dt + (t, T )0 db
z
(17.39)
where
b (t, T ) (t, T ) (t, T )0 (t) is the risk-neutral drift observed at
RT
b (t, s) ds as the
date t for the forward rate at date T . Define
b I (t, T ) t
integral over the drifts of all forward rates from date t to date T . Then using
483
(17.38) we have
b I (t, T ) =
RT
t
b (t, s) ds =
RT
t
(t, s) ds
= I (t, T ) (t) 0 I (t, T )

=
=
or
RT
t
b (t, s) ds =
RT
(t, s)0 ds (t)
1
I (t, T )0 I (t, T ) + (t) 0 I (t, T ) (t) 0 I (t, T )
2
1
I (t, T )0 I (t, T )
(17.40)
2
1
2
R
T
t
0 R
T
(t, s) ds
(t,
s)
ds
.
t
This shows that in the
absence of arbitrage, the risk-neutral drifts of forward rates are completely determined by their volatilities.
Indeed, if we dierentiate
b I (t, T ) with respect
to T to recover
b (t, T ) we obtain
0
0
z
df (t, T ) = (t, T ) I (t, T ) dt + (t, T ) db
R
T
=
(t, T )0 t (t, s) ds dt + (t, T )0 db
z
(17.41)
Equation (17.41) has an important implication, namely, that if we want to

model the risk-neutral dynamics of forward rates in order to price fixed-income
derivatives, we only need specify the form of the forward rates volatility functions.7
One can also use (17.41) to derive the risk-neutral dynamics of the
instantaneous maturity interest rate, r (t) = f (t, t), which is required for discounting risk-neutral payos. Suppose dates are ordered such that 0 t T .
In integrated form (17.41) becomes
f (t, T ) = f (0, T ) +
Rt
(u, T )0 I (u, T ) du +
Rt
(u, T )0 db
z (u)
(17.42)
7 In general, these volatility functions may be stochastic, as they could be specified to

depend on current levels of the forward rates, that is, (t, T, f (t, T )).
484
and for r (t) = f (t, t) this becomes
r (t) = f (0, t) +
Rt
0
0 (u, t) I
(u, t) du +
Rt
0 (u, t)
db
z (u)
(17.43)
Dierentiating with respect to t leads to8

R t (u, t)0 I (u, t)
f (0, t)
dt + (t, t)0 I (t, t) dt + 0
dudt
t
t
R t (u, t)0
+ 0
db
z (u) dt + (t, t)0 db
z
t
Rt
f (0, t)
(u, t)0
0
=
dt + 0 (u, t) (u, t) +
I (u, t) dudt
t
t
R t (u, t)0
db
z (u) dt + (t, t)0 db
z
(17.44)
+ 0
t
dr (t) =
where we have used the fact that I (t, t) = 0 and I (u, t) /t = (u, t).
With these results, one can now value fixed-income derivatives.
As an
example, define C (t) as the current date t price of a European-type contingent

claim that has a payo at date T .
This payo is assumed to depend on the
forward rate curve (equivalently, the term structure of bond prices or yields)
at date T , which we write as C (T, f (T, T + )) where 0. The contingent
claims risk-neutral valuation equation is
h R
i
bt e tT r(s)ds C (T, f (T, T + )) | f (t, t + ) , 0
C (t, f (t, t + )) = E
(17.45)
where the expectation is conditioned on information of the current date t forward

rate curve, f (t, t + ) 0. Equation (17.45) is the risk-neutral expectation
of the claims discounted payo, conditional on information of all currentlyobserved forward rates. In this manner, the contingent claims formula can be
8 Note
that the dynamics of dr are more complicated than simply setting T = t in equation
(17.41) because both arguments of f (t, t) = r (t) are varying simultaneously. Equation (17.44)
(t,u)
is equivalent to dr = df (t, t) + fu
|ut dt.
485
assured of fitting the current term structure of interest rates, since the forward
rate curve, f (t, t + ), is an input. Only for special cases regarding the type of
contingent claim and the assumed forward rate volatilities can the expectation
in (17.45) be computed analytically. In general, it can be computed by a Monte
Carlo simulation of a discrete-time analogue to the continuous-time risk-neutral
forward rate and instantaneous interest rate processes in (17.41) and (17.44).9
Valuing American-type contingent claims using the HJM approach can be
more complicated because, in general, one needs to discretize forward rates to
produce a lattice (e.g., binomial tree) and check the nodes of the lattice to see
if early exercise is optimal.10 However, HJM forward rates will not necessarily
follow Markov processes. From (17.41) and (17.44), one see that if the forward
rate volatility functions are specified to depend on the level of forward rates
themselves, (t, s, f (t, s)), then the evolution of f (t, T ) and r (t) depends on
the entire history of forward rates between two dates such as 0 and t. It will
not be possible to express forward rates as f (0, T, x (0)) and f (t, T, x (t)) where
x (t) is a set of finite state variables.11 Non-Markov processes lead to lattice
structures where the nodes do not re-combine.
This can make computation
extremely time consuming because the number of nodes grows exponentially

(rather than linearly in the case of recombining nodes) with the number of time
steps. Hence, to value American contingent claims using the HJM framework, it
is highly desirable to pick volatility structures that lead to forward rate processes
that are Markov.12 The following is a simple example of an HJM model where
9 An example is presented by Kaushik Amin and Andrew Morton (Amin and Morton 1994).
They value Eurodollar futures and options assuming dierent one-factor (n = 1) specifications for forward rate volatilities.
Their models are nested in the functional form
(t, T ) = [ 0 + 1 (T t)] e(T t) f (t, T ) .
10 Recall that this method was used in Chapter 7 to value an American option.
11 The reason why one may want to assume that forward rate volatilities depend on their
own level is to preclude negative forward rates, a necessary condition if currency is not to
dominate bonds in a nominal term structure model. For example, similar to the square-root
1
model of Cox, Ingersoll, and Ross, one could specify (t, T ) = f (t, T ) 2 .
12 Note, also, that non-Markov short-rate and forward rate process imply that contingent
claims cannot be valued by solving an equilibrium partial dierential equation, such as was
486
forward rate volatilities are deterministic and lead to a Markov process for the
instantaneous interest rate.
Example: Extended Vasicek
Consider the value of a European option maturing at date T where the underlying asset is a zero coupon bond maturing at date T + . It is assumed that
n = 1 and the forward rates volatilities decline exponentially with their time
horizons:
(t, T ) = r e(T t)
(17.46)
where r and are positive constants. From (17.36), this implies that the rate
of return volatility of a zero-coupon bond equals
I (t, T )
RT
t
(t, s) ds =
RT
t
r e(st) ds =
r
1 e(T t)
(17.47)
Note that this volatility function is the same as the Vasicek model of the
term structure given in (9.31).
Hence, the bond prices risk-neutral process
is dP (t, T ) /P (t, T ) = r (t) dt r 1 e(T t) db

z . For option valuation,
it remains to derive the instantaneous-maturity interest rate and its dynamics.

From (17.43) and (17.44) we have:
r (t) = f (0, t)+
R t 2r (tu)
Rt
e
e2(tu) du+ 0 r e(tu) db
z (u) (17.48)
0
done in Chapter 9 in equation (9.27). For conditions on forward rate volatilities that lead
to Markov structures, see the work of Andrew Carverhill (Carverhill 1994), Peter Ritchken
and L. Sankarasubramanian (Ritchken and Sankarasubramanian 1995), and Koji Inui and
Masaaki Kimima (Inui and Kijima 1998).

dr
487
i
Rt h
f (0, t)
dt + 0 2r e2(tu) 2r e(tu) e2(tu) dudt
t
Rt
0 r e(tu) db
z (u) dt + r db
z
(17.49)
Substituting (17.48) into (17.49) and simplifying leads to
dr
Rt
f (0, t)
dt + 0 2r e2(tu) dudt + [f (0, t) r (t)] dt + r db
z
2
1 f (0, t)
+ f (0, t) + r2 1 e2t r (t) dt + r db
=
z
t
= [r (t) r (t)] dt + r db
z
where r (t)
1
f
(17.50)
(0, t) /t + f (0, t) + 2r 1 e2t /2 is the risk-neutral
central tendency of the short-rate process that is a deterministic function of

time.
The process in (17.50) is Markov in that the only stochastic variable
aecting its future distribution is the current level of r (t). However, it diers
from the standard Vasicek model which assumes that the risk-neutral process for
r (t) has a long run mean that is constant.13 By making the central tendency,
r (t), to be a deterministic function of the currently observed forward rate curve,
f (0, t) t 0, the models implied date 0 price of a zero coupon bond, P (0, T ),
will coincide exactly with observed prices.14 This model was proposed by John
Hull and Alan White ((Hull and White 1990), (Hull and White 1993)), and is
referred to as the "Extended Vasicek" model.15
Since, as with the standard Vasicek model, the extended Vasicek model has
13 Recall from equation (10.63) that the unconditional mean of the risk-neutral interest rate
is r + q r /, where r is the mean of the physical process and q is the market price of interest
rate risk.
1
14 It is left as an exercise to verify that when r (t)
f (0, t) /t + f (0, t) +
h
R
i
R
2
T
2
2t
b exp
/ , then P (0, T ) = E
r
(s)
ds
=
exp
T
r 1 e
0
0 f (0, s) ds .
15 Hull and White show that, besides r (t), the parameters (t) and (t) can also be
r
extended to be deterministic functions of time. With these extensions, r (t) remains normally
distributed and analytic solutions to options on discount bonds can be obtained. Making
(t) and r (t) to be time varying allows one to fit other aspects of the term structure, such
as observed volatilities of forward rates.
488
bond return volatilities being a deterministic function of time, the expectation in

(17.45) for the case of a European option has an analytic solution. Alternatively,
the results of Merton (Merton 1973b) given in (9.45) to (9.47) on the pricing of
options when interest rates are random can be applied to derive the solution.
However, instead of Chapter 9s assumption of the underlying asset being an
equity that follows geometric Brownian motion, the underlying asset is a bond
that matures at date T + . For a call option with exercise price X, the boundary
condition is c (T ) = max [P (T, T + ) X, 0]. This leads to the solution
c (t)
where d1 =
P (t, T + ) N (d1 ) P (t, T ) XN (d2 )

R T +
RT
= e t f (t,s)ds N (d1 ) e t f (t,s)ds XN (d2 )
(17.51)
i
h
ln [P (t, T + ) / (P (t, T ) X)] + 12 v (t, T )2 /v (t, T ), d2 = d1
v (t, T ), and where16

v (t, T )2
I (t, u + ) + 2I (t, u) 2 I (t, u + ) I (t, u) du
2
2r
2(T t)
1
e
1 e
3
2
(17.52)
This solution illustrates a general principle of the HJM approach, namely, that
formulas can be derived whose inputs match the initial term structure of bond
prices (P (t, T ) and P (t, T + )) or, equivalently, the initial forward rate curve
(f (t, s) s t).
17.2.2
Market Models
16 Note that when applying Mertons derivation to the case of the underlying asset being a
bond, then , the return correlation between bonds maturing at dates T and T + , equals 1.
This is because there is a single Brownian motion determining the stochastic component of
returns.
489
As shown in the previous section, HJM models begin with a particular specification for instantaneous-maturity, continuously-compounded, forward rates,
and then derivative values are calculated based on these initial forward rates.
However, instantaneous-maturity forward rates are not directly observable, and
in many applications they must be approximated from data on bond yields or
discrete-maturity forward or futures rates that are unavailable at every maturity. A class of models that is a variation on the HJM approach can sometimes
avoid this approximation error and may lead to more simple, analytic solutions
for particular types of derivatives. These models are known as "Market Models"
and are designed to price derivatives whose payos are a function of a discrete
maturity, rather than instantaneous maturity, forward interest rate. Examples
of such derivatives include interest rate caps and floors and swaptions. Let us
illustrate the market model approach by way of these examples.
Example: An Interest Rate Cap

Consider valuing a European option written on a discrete forward rate, such as
one based on the London Interbank Oer Rate (LIBOR). Define L (t, T, ) as the
date t annualized, -period compounded, forward interest rate for borrowing or
lending over the period from future date T to T + .17 In terms of current date t
discount bond prices (P (t, t + )), forward price (F (t, T, )), and continuouslycompounded forward rate (f (t, T, )), this discrete forward rate is defined by
the relation
1
P (t, T + )
= F (t, T, ) = ef (t,T, ) =
P (t, T )
1 + L (t, T, )
(17.53)
17 The convention for LIBOR is to set the compounding interval equal to the underlying
instruments maturity. For example, if = 14 years, then three-month LIBOR is quarterlycompounded. If = 12 years, then six-month LIBOR is semi-annually compounded.
490
Note that when T = t, P (t, t + ) = 1/ [1 + L (t, t, )] defines L (t, t, ) as the

current "spot" -period LIBOR.18 An example of an option written on LIBOR
is a caplet that matures at date T + and is based on the realized spot rate
L (T, T, ). Assuming this caplet has has an exercise cap rate X, its date T +
payo is
c (T + ) = max [L (T, T, ) X, 0]
(17.54)
that is, the option payo at date T + depends on the -period spot LIBOR
at date T .19 Because uncertainty regarding the LIBOR rate is resolved at date
T , which is periods prior to the caplets settlement (payment) date, we can
also write
c (T ) = P (T, T + ) max [L (T, T, ) X, 0]

(17.55)
1
1 X, 0
= P (T, T + ) max
P (T, T + )
1 + X
1 + X
= max 1
, 0 = max 1
,0
P (T, T + )
1 + L (T, T, )
which illustrates that a caplet maturing at date T + is equivalent to a put
option that matures at date T , has an exercise price of 1, and is written on a
zero coupon bond that has a payo of 1 + X at its maturity date of T + .
Similarly, a floorlet, whose date T + payo equals max [X L (T, T, ) , 0],
can be shown to be equivalent to a call option on a zero coupon bond.20
To value a caplet using a market model approach, let us first analyze the
18 This
modeling assumes that LIBOR is the yield on a default-free discount bond. However,
LIBOR is not a fully default-free interest rate, such as a Treasury security rate. It represents
the borrowing rate of a large, generally high credit quality, bank. Typically, the relatively
small amount of default risk is ignored when applying market models to derivatives based on
LIBOR.
19 Caplets are based on a notional principal amount, which here is assumed to be $1. The
value of a caplet having a notional principal of $N is simply N times the value of a caplet
with a notional principal of $1, that is, is payo is N max [L (T, T, ) X, 0].
20 Therefore, the HJM - extended Vasicek solution in (17.51) to (17.52) is a one method for
valuing a floorlet. A straightforward modification of this formula could also value a caplet.
491
dynamics of L (t, T, ). Re-arranging (17.53) gives
L (t, T, ) =
P (t, T )
1
P (t, T + )
(17.56)
We can derive the stochastic process followed by this forward rate in terms of
the bond prices risk neutral process. Note that from (?? along with dz = db
z
z.
(t) dt, we have dP (t, T ) /P (t, T ) = r (t) dt I (t, T )0 db
Applying Its
lemma to (17.56), we obtain

dL (t, T, )
L (t, T, )
I (t, T + )0 [I (t, T + ) I (t, T )] dt
(17.57)
+ [I (t, T + ) I (t, T )]0 db

z
In principle, now we could value a contingent claim written on L (t, T, ) by

calculating the claims discounted expected terminal payo assuming L (t, T, )
follows the process in (17.57).21
However, as will become clear, there is an
alternative probability measure to the one generated by db

z that can be used to
calculate a contingent claims expected payo, and this alternative measure is
analytically more convenient for this particular forward rate application.
To see this, consider the new transformation de
z= db
z + I (t, T + ) dt =
dz+ [ (t) + I (t, T + )] dt. Substituting into (17.57) results in
dL (t, T, )
= [I (t, T + ) I (t, T )]0 de
z
L (t, T, )
(17.58)
so that under the probability measure generated by de

z, the process followed by
L (t, T, ) is a martingale. This probability measure is referred to as the forward
rate measure at date T + . Note that since L (t, T, ) is linear in the bond price
21 Specifically,
RT
bt e t
E
processes.
r(s)ds
if c (t, L (t, T, ))
is the contingent claims value, it could be calculated as
c (T, L (T, T, )) where r (t) and L (t, T, ) are assumed to follow risk-neutral
492
P (t, T ) deflated by P (t, T + ), the forward rate measure at date T + works

by deflating all security prices by the price of the discount bond that matures
at date T + .
This contrasts with the risk-neutral measure where security
prices are deflated by the value of the money market account which follows the
process dB (t) = r (t) B (t) dt.
Not only does L (t, T, ) follow a martingale under the forward measure, but
so does the value of all other securities. To see this, let the date t price of a
contingent claim be given by c (t). In the absence of arbitrage, its price process
is of the form
dc
= [r (t) + (t) 0 c (t)] dt + c (t)0 dz
c
(17.59)
Now define the deflated contingent claims price as C (t) = c (t) /P (t, T + ).
Applying Its lemma gives
dC
C
= [ (t) +I (t, T + )] 0 [c (t) + I (t, T + )] dt
(17.60)
+ [c (t) + I (t, T + )] 0 dz
and making the forward measure transformation de

z = dz + [ (t) + I (t, T + )] dt,
(17.60) becomes the martingale process
dC
= [c (t) + I (t, T + )] 0 de
z
C
(17.61)
et [C (t + )] 0, where E
et [] is the date t expectation under
so that C (t) = E
the forward measure. Now, to show why this transformation can be convenient,
suppose that this contingent claim is the caplet described earlier. This deflated
493
caplets value is given by

et [C (T + )]
C (t) = E
max [L (T, T, ) X, 0]
e
= Et
P (T + , T + )
(17.62)
Noting that C (t) = c (t) /P (t, T + ) and realizing that P (T + , T + ) = 1 ,

we can re-write this as
et [ max [L (T, T, ) X, 0]]
c (t) = P (t, T + ) E
(17.63)
A common practice is to assume that L (T, T, ) is lognormally distributed

under the date T + forward measure.22 This means that [I (t, T + ) I (t, T )]
in 17.58 must be a vector of non-stochastic functions of time that can be calibrated to match observed bond or forward rate volatilities.23
Noting that
L (t, T, ) also has a zero drift, leads to a similar formula first proposed by
Fischer Black (Black 1976) for valuing options on commodity futures:
c (t) = P (t, T + ) [L (t, T, ) N (d1 ) XN (d2 )]
(17.64)
i
h
where d1 = ln (L (t, T, ) /X) + 12 v (t, T )2 /v (t, T ), d2 = d1 v (t, T ), and
v (t, T )2 =
RT
t
|I (s, T + ) I (s, T )| ds
(17.65)
Equation (17.64) is similar to equation (10.57) derived in Chapter 10 for the

22 Assuming
a lognormal distribution for L (t, T, ) is attractive because it prevents this

discrete forward rate from becoming negative, thereby also restricting yields on discount bonds
to be non-negative. Note that if instantaneous-maturity forward rates are assumed to be
lognormally distributed, HJM show that they will be expected to become infinite in finite
time. This is inconsistent with an arbitrage-free bond prices. Fortunately, such an explosion
of rates does not occur when forward rates are discrete (Brace, Gatarek, and Musiela 1997).
23 Note that since (t, t + ) is an integral of instantaneous forward rate volatilities, the
I
lognormallity of I (t, t + ) I (t, T ) puts restrictions on instantaneous forward rates under
an HJM modeling approach. However, we need not focus on this issue for pricing applications
involving a discrete forward rate.
494
case of a call option on a forward or futures price where the underlying is

lognormally-distributed and interest rates are non-stochastic.
An interest rate cap is a portfolio of caplets written on the same -period
LIBOR but maturing at dierent dates T = T1 , T2 , ..., Tn where typically
Tj+1 = Tj + .
Standard practice is to value each individual caplet in the
portfolio in the manner described above, where the caplet maturing at date Tj
is priced using the date Tj + forward measure.
Often, caps are purchased
by issuers of floating-rate bonds whose bond payments coincide with the caplet
maturity dates.
Doing so insures the bond issuer against having to make a
floating coupon rate greater than X (plus a credit spread).
Since a floating
rate bonds coupon rate payable at date T + is most commonly tied to the period LIBOR at date T , caplet payos follow this same structure. Analogous
to a cap, an interest rate floor is a portfolio of floorlets, and can be valued using
the same technique described in this section.
Example: A Swaption
Frequently, a market model approach is applied to value another common interest rate derivative, a swaption. A swaption is an option to become a party
in an interest rate swap at a given future maturity date and at a pre-specified
swap rate.
Let us, then, define the interest rate swap underlying this swap-
tion. A standard (plain vanilla) swap is an agreement between two parties to

exchange fixed interest rate coupon payments for floating interest rate coupon
payments at dates T1 , T2 , ...,Tn+1 where Tj+1 = Tj + and is the maturity
of the LIBOR of the floating rate coupon payments. Thus, if K is the swaps
fixed annualized coupon rate, then at date Tj+1 the fixed-rate payers net payment is [K L (Tj , Tj , )] while that of the floating rate payer is exactly the
495
opposite.24
Note that the swaps series of floating rate payments plus an additional $1
at date Tn+1 can be replicated by starting with $1 at time T0 = T1 and
repeatedly investing this $1 in -maturity LIBOR deposits.25
These are the
same cashflows that one would obtain by investing $1 in a floating rate bond at
date T0 . Similarly, the swaps series of fixed rate payments plus an additional
$1 at date Tn+1 can be replicated by buying a fixed-coupon bond that pays
coupons of K at each swap date and pays a principal of $1 at its maturity
date of Tn+1 . Based on this insight, one can see that the value of a swap to the
floating rate payer is the dierence between a fixed coupon bond having coupon
rate K, and a floating coupon bond having coupons tied to -period LIBOR.
Thus, if t T0 = T1 , then the date t value of the swap to the floating rate
payer is26
K
n+1
P
j=1
P (t, Tj ) + P (t, Tn+1 ) P (t, T0 )
(17.66)
When a standard swap agreement is initiated at time T0 , the fixed rate K is

set such that the value of the swap in (17.66) is zero. This concept of setting
K to make the agreement fair (similar to forward contracts) can be extended to
dates prior to T0 . One can define s0,n (t) as the forward swap rate that makes
the date t value of the swap (starting at date T0 and making n subsequent
exchanges) equal to zero. Setting K = s0,n (t) and equating (17.66) to zero,
24 Recall
that L (Tj , Tj , ) is the spot -period LIBOR at date Tj . Also, as discussed in

the preceeding footnote, this exchange is based on a notional principal of $1. For a notional
principal of $N, all payments are multiplied by N.
25 Thus, $1 invested at time T produces a return of 1 + L (T , T , ) at T . Keeping the
0
0
0
1
cashflow of L (T0 , T0 , ) and re-investing the $1 will then produce a return of 1+ L (T1 , T1 , )
at T2 . Keeping the cashflow of L (T1 , T1 , ) and re-investing the $1 will then produce a
return of 1 + L (T2 , T2 , ) at T3 . This process is repeated until at time Tn+1 a final return
of 1 + L (Tn , Tn , ) is obtained.
26 Notice that P (t, T ) is the date t value of the floating rate bond while the remaining
0
terms are the value of the fixed rate bond.
496
one obtains
s0,n (t) =
=
where B1,n (t)
Pn+1
j=1
P (t, T0 ) P (t, Tn+1 )

P
n+1
j=1 P (t, Tj )
(17.67)
P (t, T0 ) P (t, Tn+1 )

B1,n (t)
P (t, Tj ) is a portfolio of zero coupon bonds that each
pay at the times of the swaps exchanges.

Now a standard swaption is an option to become either a fixed rate payer or
floating rate payer at a fixed swap rate X at a specified future date. Thus, if the
maturity of the swaption is date T0 , at which time the holder of the swaption
has the right, but not the obligation, to become a fixed rate payer (floating rate
receiver), this options payo equals27
c (T0 ) = max [B1,n (T0 ) [s0,n (T0 ) X] , 0]
(17.68)
= max [1 P (T0 , Tn+1 ) B1,n (T0 ) X, 0]

Note from the first line of (17.68) that when the option is in the money, then
B1,n (T0 ) [s0,n (T0 ) X] is the date T0 value of the fixed rate payers savings from
having the swaption relative to entering into a swap at the fair spot rate s0,n (T0 ).
In the second line of (17.68), we have substituted from (17.67) s0,n (T0 ) B1,n (T0 )
= P (T0 , T0 ) P (T0 , Tn+1 ) = 1 P (T0 , Tn+1 ). This illustrates that a swaption
is equivalent to an option on a coupon bond with coupon rate X and exercise
price of 1.
To value this swaption at date t T0 , a convenient approach is to recognize from (17.68) that the swaptions payo is proportional to B1,n (T0 )
Pn+1
j=1 P (T0 , Tj ). This suggests that B1,n (t) is a convenient deflator for valu27 The payo of an option to be a floating rate payer (fixed rate receiver) is
max [B1,n (T0 ) [X s0,n (T0 )] , 0].
497
ing the swap. By normalizing all security prices by B1,n (t), we will value the
swaption using the so-called "forward swap measure."
Similar to valuation under the risk-neutral or forward measure of the previous section, let us define C (t) = c (t) /B1,n (t). Also define dz = dz+
(t) + B1,n (t) dt where B1,n (t) is the date t vector of instantaneous volatil-
ities of the zero coupon bond portfolios value, B1,n (t). Similar to the derivation
in equations (17.59) to (17.61), we have
dC
= c (t) + B1,n (t, T + ) 0 dz
C
(17.69)
so that all deflated asset prices under the forward swap measure follow martingale processes. Thus,
C (t) = E t [C (T0 )]
max [B1,n (T0 ) [s0,n (T0 ) X] , 0]

= Et
B1,n (T0 )
(17.70)
= E t [max [s0,n (T0 ) X, 0]]
Re-written in terms of the undeflated swaptions current value, c (t) = C (t) B1,n (t),
(17.70) becomes
c (t) = B1,n (t) E t [max [s0,n (T0 ) X, 0]]
(17.71)
so that the expected payo under the forward swap measure is discounted by
the current value of a portfolio of zero coupon bonds that mature at the times
of the swaps exchanges.
Importantly, note that s0,n (t) = [P (t, T0 ) P (t, Tn+1 )] /B1,n (T0 ) is the
ratio of the dierence between two security prices deflated by B1,n (t). In the
absence of arbitrage, it must also follow a martingale process under the forward
498
swap measure.
A convenient and commonly made assumption is that this
forward swap rate is lognormally distributed under the forward swap measure:
ds0,n (t)
= s0,n (t) 0 dz
s0,n (t)
(17.72)
so that S0,n (t) is a vector of deterministic functions of time that can be calibrated to match observed forward swap volatilities or zero coupon bond volatilities.28 This assumption results in (17.71) taking a Black-Scholes-type form
c (t) = B1,n (t) [s0,n (t) N (d1 ) XN (d2 )]
(17.73)
i
h
where d1 = ln (s0,n (t) /X) + 12 v (t, T0 )2 /v (t, T0 ), d2 = d1 v (t, T0 ), and
v2 (t, T0 ) =
17.2.3
R T0
t
s0,n (u) 0 s0,n (u) du.
(17.74)
Random Field Models
The term structure models that we have studied thus far have specified a finite
number of Brownian motion processes as the source of uncertainty determining
the evolution of bond prices or forward rates.
For example, the bond price
processes in equilibrium models (equation 17.2) and HJM models (equation

17.36) were driven by an n1 vector of Brownian motions, dz. One implication
of this is that a Black-Scholes hedge portfolio of n dierent maturity bonds can
be used to perfectly replicate the risk of any other maturity bond. As shown
in Chapter 10, in the absence of arbitrage, the fact that any bonds risk can be
hedged with other bonds places restrictions on bonds expected excess rates of
return and results in a unique vector of market prices of risk, (t), associated
28 Applying Its lemma to (17.67) allows one to derive the volatility of s
0,n (t) in terms of
zero coupon bond volatilities.
499
with dz. This implied a bond price process of the form

0
dP (t, T ) /P (t, T ) = [r (t) + (t) 0 p (t, T )] dt + p (t, T ) dz
(17.75)
Moreover, the Black-Scholes hedge, by making the market dynamically complete and by identifying a unique (t) associated with dz, allows us to perform
risk-neutral valuation by the transformation db
z = dz + (t) dt or valuation
using the pricing kernel dM/M = r (t) dt (t)0 dz.
However, the elegance of these models comes with an empirical downside.
The fact that all bond prices depend on the same n 1 vector dz places restrictions on the covariance of bonds rates of return. For example, when n = 1,
the rates of return on all bonds are instantaneously perfectly correlated. While
in these models, the correlation can be made less perfect by increasing n, doing
so introduces more parameters that require estimation.
A related empirical implication of (17.75) or (17.33) is that it restricts the
possible future term structures of bond prices or forward rates. In other words,
starting from the current date t set of bond prices P (t, T ) T > t, an arbitrary
future term structure, P (t + dt, T ) T > t + dt, cannot always be achieved by
any realization of dz.
This is because a given future term structure has an
infinite number of bond prices (each of a dierent maturity), but the finiteness
of dz allows matching this future term structure at only a finite number of
maturity horizons.29 Hence, models based on a finite dz are almost certainly
inconsistent with future observed bond prices and forward rates.
Because of
this, empiricists must assume that data on bond prices (or yields) are observed
with "noise" or that, in the case of HJM-type models, parameters (that the
model assumes to be constant) must be re-calibrated at each observation date
29 For example, consider n = 1. In this case, all bond prices must either rise or fall with a
given realization of dz. This model would not permit a situation where short maturity bond
prices fell but long maturity bond prices rose. The model could produce a realization of dz
that matches long maturity bond prices or short maturity bond prices, but not both.
500
to match the new term structure of forward rates.

Random field models are an attempt to avoid these empirical deficiencies.
Research in this area includes that of David Kennedy (Kennedy 1994), (Kennedy
1997), Robert Goldstein (Goldstein 2000), Pedro Santa-Clara and Didier Sornette (Santa-Clara and Sornette 2001), and Robert Kimmel (Kimmel 2004).
These models specify that each zero coupon bond price, P (t, T ), or each instantaneous forward rate, f (t, T ), is driven by a Brownian motion process that
is unique to the bonds or rates maturity, T . For example, a model of this type
might assume that a bonds risk-neutral process satisfies
zT
dP (t, T ) /P (t, T ) = r (t) dt + p (t, T ) db
T > t
(17.76)
where db
zT (t) is a single Brownian motion process (under the risk-neutral measure) that is unique to the bond that matures at date T .30 The set of Brownian
motions for all zero coupon bonds { zbT (t) }T >t comprise a Brownian "field" or
"sheet." This continuum of Brownian motions has two dimensions: calendar

time, t, and time to maturity, T .
The elements aecting dierent bonds are
linked by an assumed correlation structure:
db
zT1 (t) db
zT2 (t) = (t, T1, T2 ) dt
(17.77)
where (t, T1, T2 ) > 0 is specified to be a particular continuous, dierentiable

function with (t, T, T ) = 1 and
(t,T1 ,T2 )
|T1 =T2
T1
= 0. For example, one simple
specification involving only a single parameter is (t, T1, T2 ) = e
|T1 T2 |
where
is a positive constant.
One can also model the physical process for bond prices corresponding to
30 An alternative way of specifying a random field model is to assume that the
risk-neutral
processes for instantaneous
forward rates are of the form df (t, T ) =
i
h
R
(t,
s)
c
(t,
T,
s)
ds
dt
+
(t,
T
)
dz
(t, T ) T
T where dzT1 dzT2 = c (t, T1 , T2 ) dt. This spect
ification extends the HJM equation (17.41) to a random field driving forward rates.
501
(17.76). If T (t) is the market price of risk associated with db

zT (t), then making
zT + T (t) dt, one obtains
the transformation dzT = db
dP (t, T ) /P (t, T ) = [r (t) + T (t) p (t, T )] r (t) dt + p (t, T ) dzT
T > t
(17.78)
with dzT (t), T > t satisfying the same correlation function as in (17.77). Analogous to the finite-factor pricing kernel process in (17.8), a pricing kernel for
this random field model would be
dM/M = r (t) dt
R
t
[T (t) dzT (t)] dT
(17.79)
so that an integral of the products of market prices of risk and Brownian motions
replaces the usual sum of these products that occur for the finite factor case.31
The benefit of a model like (17.76) and (17.77) is that a realization of the
Brownian field can generate any future term structure of bond prices or forward rates and, hence, be consistent with empirical observation and not require
model re-calibration.
Moreover, with only a few additional parameters, ran-
dom field models can provide a flexible covariance structure among dierent
maturity bonds.
Specifically, unlike finite-dimensional equilibrium models or
HJM models, the covariance matrix of dierent maturity bond returns or forward rates will always be non-singular no matter how many bonds are included.
This could be important when valuing particular fixed-income derivatives where
the underlying is a portfolio of zero coupon bonds, and the correlation between
these bonds aects the overall portfolio volatility.
31 Note, however, that a random field model is not the same as a standard finite factor model
extended to an infinite number of factors. As shown in (17.78), a random field model has
a single Brownian motion driving each bond price or forward rate. A factor model, such as
(17.2) or (17.36), extended to infinite factors would have the same infinite set of Brownian
motions driving each bond price.
502
However, this rich covariance structure requires stronger theoretical assumptions for valuing derivatives compared to finite-dimensional diusion models. A
given bonds return can no longer be perfectly replicated by a portfolio of other
bonds, and, thus, a Black-Scholes hedging argument cannot be used to identify
a unique market price of risk associated with each dzT (t).32
The market for
fixed-income securities is no longer dynamically complete. Hence, one must assume, perhaps due to an underlying preference-based general equilibrium model,
that there exists particular T (t) associated with each dzT (t), or, equivalently,
that a risk-neutral pricing exists.
Random field models can be parameterized by assuming particular functions for bond price or forward rate volatilities.
For example, Pierre Collin-
Dufresne and Robert Goldstein (Collin-Dufresne and Goldstein 2003) propose a

p
stochastic volatility model where, in equation (17.76), p (t, T ) = (t, T ) (t),
where (t, T ) is a deterministic function and where (t) is a volatility factor,
common to all bonds, that follows the square-root process
p
d (t) = (t) dt + (t)db

z
(17.80)
where db
z is a Brownian motion (under the risk-neutral measure) that is assumed to be independent of the Brownian field {db
zT } T > t.
Based on
this parameterization, which is similar to a one factor ane model, they derive
solutions for various interest rate derivatives.33
32 Robert Goldstein (Goldstein 2000) characterizes random field models of the term structure
as being analogous to the APT model ((Ross 1976)). As discussed in Chapter 3, the APT
assumes that a given assets return depends on the risk from a finite number of factors along
with the assets own idiosyncratic risk. Thus, the asset is imperfectly correlated with any
portfolio containing a finite number of other assets. Similarly, in a random field model, a
given bonds return is imperfectly correlated with any portfolio containing a finite number of
other bonds. Taking the analogy a step further, perhaps market prices of risk in a random
field model can be characterized using the notion of asymptotic arbitrage, rather than exact
arbitrage.
33 Robert Kimmel ((Kimmel 2004)) also derives models with stochastic volatility driven by
multiple factors.
503
If, similar to David Kennedy (Kennedy 1994), one makes the more simple
assumption that p (t, T ) and (17.76) and (t, T1, T2 ) in (17.77) are deterministic
functions, then options on bonds, such as caplets and floorets, have a BlackScholes-type valuation formula. For example, suppose as in the HJM-extended
Vasicek case of (17.51) to (17.52) that we value a European call option that
matures at date T , is written on a zero coupon bond that matures at date
T + , and has an exercise price of X.
Similar to (17.63), we can value this
option using the date T forward rate measure:

et [max [p (T, T + ) X, 0]]
c (t) = P (t, T ) E
(17.81)
where p (t, T + ) P (t, T + ) /P (t, T ) is the deflated price of the bond that
matures at date T + . Applying Its lemma to the risk-neutral process for
bond prices in (17.76), we obtain:
dp (t, T + )
p (t, T + )
= p (t, T ) [p (t, T ) (t, T, T + ) p (t, T + )] dt

+ p (t, T + ) db
zT + p (t, T ) db
zT
(17.82)
q
zT + 1 (t, T, T + )2 db
zU,T where
We can re-write db
zT + = (t, T, T + ) db
zT , so that the stochastic

db
zU,T is a Brownian motion uncorrelated with db
q
2
zU,T
component in (17.82) can be written as p (t, T + ) 1 (t, T, T + ) db
+ [p (t, T + ) (t, T, T + ) p (t, T )] db

zT .34 Then making the transformazT + p (t, T ), the process for
tion to the date T forward measure, de
zT = db
34 This re-writing puts the risk-neutral process for p (t, T + ) in the form of our prior analysis
in which the vector of Brownian motions, db
z, was assumed to have independent elements.
This allows us to make the transformation to the forward measure in the same manner as was
done earlier.
504
p (t, T + ) becomes
dp (t, T + )
p (t, T + )
= p (t, T + )
q
1 (t, T, T + )2 db
zU,T
+ [p (t, T + ) (t, T, T + ) p (t, T )] de

zT
= (t, T, ) de
z
(17.83)
where
(t, T, )2 p (t, T + )2 + p (t, T )2 2 (t, T, T + ) p (t, T + ) p (t, T )
(17.84)
Thus, p (t, T + ) is lognormally distributed under the forward rate measure, so
that (17.81) has the Black-Scholes-Merton-type solution
c (t) = P (t, T ) [p (t, T + ) N (d1 ) XN (d2 )]
(17.85)
= P (t, T + ) N (d1 ) P (t, T ) N (d2 )

i
h
2
where d1 = ln (p (t, T + ) /X) + 12 v (t, T ) /v (t, T ), d2 = d1 v (t, T ), and
v (t, T )2 =
RT
t
(u, T, )2 du
(17.86)
and (u, T, ) is defined in (17.84). While this formula is similar to the Vasicekbased ones in (9.45) and (17.51), the volatility function in (17.84) may permit
a relatively more flexible form for matching observed data.
17.3
Summary
This chapter has briefly surveyed some of the important theoretical developments in modeling bond yield curves and valuing fixed income securities. The
17.4. EXERCISES
505
chapters presentation has been in the context of continuous time models and,
to keep its length manageable, many similar models set in discrete time have
been left out.35 Moreover, questions regarding numerical implementation and
parameter estimation for specific models could not be answered in the short
presentations given here.
There is a continuing search for improved ways of describing the term structure of bond prices and of valuing fixed income derivatives.
Researchers in
this field have dierent objectives, and the models that we presented reflect this
diversity.
Much academic research focuses on analyzing equilibrium models
in hopes of better understanding the underlying macro-economic factors that

shape the term structure bond yields. In contrast, practitioner research tends
to focus on finding better ways of valuing and hedging bond-related derivatives.
Researchers eorts in these fields remains strong, and interest in the related subject of valuing defaultable fixed income securities may be growing more quickly.
The next chapter takes up this topic of default-risky bond pricing.
17.4
Exercises
1. Consider the following example of a two-factor term structure model (Jegadeesh

and Pennacchi 1996) (Balduzzi, Das, and Foresi 1998). The instantaneousmaturity interest rate is assumed to follow the physical process
dr(t) = [ (t) r (t)] dt + r dzr

35 Treatments of models set in discrete-time include books by Robert Jarrow (Jarrow 2002),
Bruce Tuckman (Tuckman 2002) and Thomas Ho and Sang Bin Lee (Ho and Lee 2004).
506

and the physical process for the interest rates stochastic "central tendency," (t) satisfies
d (t) = [ (t)] dt + dz
where dzr dz = dt and > 0, r , > 0, > 0, , and are constants.
In addition, define the constant market prices of risk associated with dzr
and dz to be r and .
Re-write this model using the ane model
notation used in this chapter and solve for the equilibrium price of a zero
coupon bond, P (t, T ).
2. Consider the following one-factor quadratic-gaussian model. The single
state variable, x (t), follows the risk-neutral process
z
dx (t) = [x x (t)] dt + x db
and the instantaneous-maturity interest rate is given by r (t, x) = +
x (t) + x (t)2 . Assume , x, , and are positive constants, and that
14 2 / 0 where is also a constant. Solve for the equilibrium price
of a zero coupon bond, P (t, T ).
3. Show that for the extended Vasicek model when r (t) 1 f (0, t) /t +
h
R
i
b exp T r (s) ds
=
f (0, t) + 2r 1 e2t /2 , then P (0, T ) = E
0
R
T
exp 0 f (0, s) ds .
4. Determine the value of an n payment interest rate floor using the LIBOR
market model.
Chapter 18
Models of Credit Risk

To this point, our models of bond pricing assumed that bonds have no risk of default. Therefore, these models are most applicable to valuing default-free bonds
issued by a federal government, which would include Treasury bills, notes, and
bonds.1 However, many debt instruments, including corporate bonds, municipal bonds, and bank loans, have default or credit risk. Valuing defaultable
debt requires an extended modeling approach. We now consider the two primary methods for modeling default risk. The first, suggested in the seminal option pricing paper of Fischer Black and Myron Scholes (Black and Scholes 1973)
and developed by Robert Merton (Merton 1974), Francis Longsta and Eduardo
Schwartz (Longsta and Schwartz 1995), and others, is called the "structural"
approach. This method values a firms debt as an explicit function of the value
of the firms assets and its capital structure.
The second "reduced form" approach more simply assumes default is Poisson
process with a possibly time varying default intensity and default recovery rate.
1 Default can be avoided on government bonds promising a nominal (currency-valued) payment if the government (or its central bank) has the power to print currency. If a federal
government gives up this power, such as governments whose ocial currency is the Euro that
is supplied by the European Central Bank, default on government debt becomes a possibility.
507
508
CHAPTER 18. MODELS OF CREDIT RISK
This method views the exogenously specified default process as the reducedform of a more complicated and complex model of a firms assets and capital
structure. Examples of this approach include work by Robert Jarrow, David
Lando, and Stuart Turnbull (Jarrow, Lando, and Turnbull 1997), Dilip Madan
and Haluk Unal (Madan and Unal 1998), and Darrell Due and Kenneth Singleton (Due and Singleton 1999). This chapter provides an introduction to
the main features of these two methods of modeling default.
18.1
The Structural Approach
This section considers a model similar to that by Robert Merton (Merton 1974).
It specifies the assets, debt, and shareholders equity of a particular firm. Let
A(t) denote the date t value of a firms assets. The firm is assumed to have a
very simple capital structure. In addition to shareholders equity, it has issued
a single zero-coupon bond that promises to pay an amount B at date T > t.
Also let T t be the time until this debt matures. The firm is assumed to
pay dividends to its shareholders at the continuous rate A(t)dt, where is the
firms constant proportion of assets paid in dividends per unit time. The value
of the firms assets are assumed to follow the process
dA/A = ( ) dt + dz
(18.1)
where denotes the instantaneous expected rate of return on the firms assets
and is the constant standard deviation of return on firm assets. Now let
D(t, T ) be the date t market value of the firms debt that is promised the
payment of B at date T . It is assumed that when the debt matures, the firm
pays the promised amount to the debtholders if there is sucient asset value to
do so. If not, the firm defaults (bankruptcy occurs) and the debtholders take
18.1. THE STRUCTURAL APPROACH
509
ownership of all of the firms assets. Hence, the payo to debtholders at date T
can be written as
D (T, T ) = min [B, A (T )]
(18.2)
= B max [0, B A (T )]
From the second line in equation (18.2), we see that the payo to the debtholders
equals the promised payment, B, less the payo on a European put option
written on the firms assets and having exercise price equal to B. Hence, if
we make the usual frictionless market assumptions, then the current market
value of the debt can be derived to equal the present value of the promised
payment less the value of a put option on the dividend-paying assets.2
If we
let P (t, T ) be the current date t price of a default-free zero-coupon bond that
pays $1 at date T and assume that the default-free term structure satisfies the
Vasicek model as specified earlier in (9.28) to (9.30), then using Chapter 9s
results on the pricing of options when interest rates are random, we obtain:
D (t, T ) = P (t, T ) B P (t, T ) BN (h2 ) + e AN (h1 )
(18.3)
= P (t, T ) BN (h2 ) + e AN (h1 )
where h1 = ln e A/ (P (t, T ) B) + 12 v2 /v, h2 = h1 v, and v () is given in
(9.48). Note that if the default-free term structure is assumed to be determin

istic, then we have the usual Black-Scholes value for v = . The promised
yield to maturity on the firms debt, denoted R (t, T ), can be calculated from
(18.3) as R (t, T ) =
as R (t, T )
2 One
ln [B/D (t, T )]. Also, its credit spread can be computed
ln [1/P (t, T )].
needs to assume that the risk of the firms assets, as determined by the dz process, is
a tradeable risk, so that a Black-Scholes hedge involving the firms debt can be constructed.
510
Based on this result, one can also solve for the market value of the firms
shareholders equity, which we denote as E (t). In the absence of taxes and other
transactions costs, the value of investors claims on the firms assets, D (t, T ) +
E (t) must equal the total value of the firms assets, A (t). This allows us to
write
E (t) = A (t) D (t, T )
(18.4)
= A P (t, T ) BN (h2 ) e AN (h1 )
= A 1 e N (h1 ) P (t, T ) BN (h2 )

Shareholders equity is similar to a call option on the firms assets in the
sense that at the debts maturity date, equity holders receive the payment
max [A (T ) B, 0].
Shareholders limited liability gives them the option of
receiving the firms residual value when it is positive. However, shareholders

equity diers from the standard European call option if the firm pays dividends
prior to the debts maturity. As is reflected in the first term in the last line of
(18.4), the firms shareholders, unlike the holders of standard options, receive
these dividends.
Chapter 12 of Robert Mertons book (Merton 1992) gives an in-depth analysis of the comparative statics properties of the debt and equity formulas similar
to equations (18.3) and (18.4), as well as the firms credit spread. Note that an
equity formula such as (18.4) can be useful because for firms that have publiclytraded shareholders equity, observation of the firms market value of equity and
its volatility can be used to infer the market value and volatility of the firms
assets. The market value and volatility of the firms assets can then be used
as inputs into (18.3) so that the firms default risky debt can then be valued.
Such an exercise based on the Merton model has been done by the credit-rating
18.1. THE STRUCTURAL APPROACH
511
firm Moodys KMV to forecast corporate defaults.3

The Merton models assumption that the firm has a single issue of zerocoupon debt is unrealistic, since it is commonly the case that firms have multiple
coupon-paying debt issues with dierent maturities and dierent seniorities in
the event of default. Modeling multiple debt issues and determining the point
at which an asset deficiency triggers default is a complex task.4
In response,
some research has taken a dierent tack by assuming that when the firms assets
hits a lower boundary, default is triggered. This default boundary is presumed
to bear a monotonic relation to the firms total outstanding debt.
With the
initial value of the firms assets exceeding this boundary, determining future
default amounts to computing the first passage time of the assets through this
boundary.
Francis Longsta and Eduardo Schwartz (Longsta and Schwartz 1995) develop such a model following the earlier work of Fischer Black and John Cox
(Black and Cox 1976). They assume a default boundary that is constant over
time and, when assets sink to the level of this boundary, bondholders are assumed to recover an exogenously given proportion of their bonds face values.
This contrasts with the Merton model where, in the case of default, bondholders
recover A (T ), the stochastic value of firm assets at the bonds maturity date,
which results in a loss of B A (T ). In the Longsta-Schwartz model, possible
default occurs at a stochastic date, say , defined by the first (passage) time
that A ( ) = k, where k is the pre-determined default boundary. Bondholders
are assumed to recover P ( , T ) B, where < 1 is the recovery rate equaling
a proportion of the market value of an otherwise equivalent default-free bond,
3 For a description of the KMV application of the Merton model for forecasting defaults,
see (Crosbie and Bohn 2002). Alan Marcus and Israel Shaked (Marcus and Shaked 1984)
apply the Merton model to analyzing the default risk of commercial banks.
4 A study by Edward Jones, Scott Mason, and Eric Rosenfeld (Jones, Mason, and Rosenfeld
1984) is an example.
512
P ( , T ) B.5
This exogenous recovery rate, , is permitted to dier for bonds
of with dierent maturity and seniority characteristics and might be estimated

from the historical recovery rates of dierent types of bonds.
Pierre Collin-Dufresne and Robert Goldstein (Collin-Dufresne and Goldstein
2001) modify the Longsta-Schwartz model to permit a firms default boundary
to be stochastic. Motivated by the tendency of firms to target their leverage ratios by partially adjusting their debt and equity over time, Collin-Dufresne and
Goldstein permit the ratio of firm assets to firm debt (the default-boundary) to
follow a mean reverting process with default triggered when this ratio declines to
unity.6 Chunsheng Zhou (Zhou 2001) modifies the Longsta Schwartz model in
another direction by allowing the firms assets to follow a mixed jump-diusion
process.
While these "first passage time" models seek to provide more real-
ism than the more simple Merton model, they come at the cost of requiring
numerical, rather than closed-form, solutions.7
For firms with complicated debt structures, these first passage time models
simplify the determination of default by assuming it occurs when a firms assets
sink to a specified boundary. The interaction between default and the level and
timing of particular promised bond payments are not directly modeled, except
in so far as they aect the specification of the default boundary. In the next
section, we consider the reduced form approach which goes a step further by
not directly modeling either the firms assets or its overall debt level.
5 P ( , T ) B is the market value of a zero-coupon bond paying the face value of B at date
T . However, Longsta and Schwartz do not limit their analysis to defaultable zero coupon
bonds. Indeed, they value both fixed and floating coupon bonds assuming a Vasicek model
of the term structure. Hence, in general, recovery equals a fixed proportion, , of the market
value of an otherwise equivalent default-free (fixed or floating rate) bond.
6 More precisely, they assume that the risk-neutral process for the log of the ratio of firm
debt to assets, say l (t) = ln [k (t) /A (t)], follows an Ornstein-Uhlenbeck process. For an
example of a model displaying mean-reverting leverage in the context of commercial bank
defaults, see (Pennacchi 2005).
7 An exception is the closed-form solutions obtained by Stijn Claessens and George Pennacchi (Claessens and Pennacchi 1996) who model default-risky sovereign debt such as Brady
bonds.
18.2. THE REDUCED-FORM APPROACH
18.2
513
The Reduced-Form Approach
With the reduced form method, default need not be tied directly to the dynamics
of a firms assets and liabilities.
As a result, this approach may provide less
insight regarding the economic fundamentals of default.
However, because
reduced form models generate default based on an exogenous Poisson process,

they may better capture the eects on default of additional unobserved factors
and provide richer dynamics for the term structure of credit spreads.8 Reduced
form modeling also can be convenient because, as will be shown, defaultable
bonds are valued using techniques similar to those used to value default-free
bonds.
To illustrate reduced form modeling, we begin by analyzing a defaultable
zero-coupon bond and, later, generalize the results to multiple-payment bonds.
As in the previous section, let D (t, T ) be the date t value of a default-risky
zero-coupon bond that promises to pay B at its maturity date of T . However,
unlike previous sections structural models where default is necessarily linked
to the dynamics of the firms capital structure, we now assume that a possible
default event depends on an exogenous process that may or may not depend
directly on the firms capital structure. Default for a particular firms bond is
modeled as a Poisson process with a time-varying default intensity. Conditional
on default having not occurred prior to date t, the instantaneous probability of
8 In most structural models, (Zhou 2001) is a notable exception, a firms assets are assumed
to follow a diusion process which has a continuous sample path. An implication of this is
that default becomes highly unlikely for short horizons if the firm currently has a substantial
dierence between assets and liabilities. Hence, these models generate very small credit
spreads for the short-maturity debt of credit-worthy corporation, counter to empirical evidence
that finds more significant spreads. Small spreads occur because default over a short horizon
cannot come as a sudden surprise. This is not the case with reduced form models where
sudden default is always possible due its Poisson nature. Hence, these models can more
easily match the significant credit spreads on short-term corporate debt. Darrell Due and
David Lando (Due and Lando 2001) present a structural model where investors have less
(accounting) information regarding the value of a firms assets than do the firms insiders.
Hence, like the jump-diusion model of Chunsheng Zhou (Zhou 2001), investors valuation of
the firms assets can take discrete jumps when inside information is revealed. This model
generates a Poisson default intensity equivalent to a particular reduced form model.
514
default during the interval (t, t + dt) is denoted (t) dt where (t) is the physical
default intensity or hazard rate and is assumed to be non-negative.9
Note
from this definition, the physical probability of the firm not defaulting (that is,
surviving) over the time interval from t to , where t < T , is
R
Et e t (u)du
18.2.1
(18.5)
A Zero-Recovery Bond
To determine D (t, T ), an assumption must be made regarding the payo received by bondholders should the bond default.
We begin by assuming that
bondholders recover nothing should the bond default and, later, we generalize
this assumption to permit a possible non-zero recovery value. With zero recovery, the bondholders date T payo is either D (T, T ) = B if there is no default
or D (T, T ) = 0 if default has occurred over the interval from t to T .
Ap-
plying risk-neutral pricing, the date t value of the zero-recovery bond, denoted
DZ (t, T ), can be written as
h R
i
bt e tT r(u)du D (T, T )
DZ (t, T ) = E
(18.6)
bt [] is the
where r (t) is the date t instantaneous default-free interest rate, and E
date t risk-neutral expectations operator. To compute this expression, we need
to determine the expression for D (T, T ) in terms of the risk-neutral default

intensity, rather than the physical default intensity.
The risk-neutral default
intensity will account for the market price risk associated with the Poisson
9 Recall that in Chapter 11 we modeled jumps in asset prices as following a Poisson process
with jump intensity . Here, a one-time default follows a Poisson process, and its intensity
is explicitly time-varying.
515
arrival of a default event.

To understand the role of default risk, suppose that both the default-free
term structure and the firms default intensity depend on a set of n state variables, xi , i = 1, ..., n, that follow the multi-variate Markov diusion process10
dx = a (t, x) dt + b (t, x) dz
(18.7)
0
where x = (x1 ...xn ) , a (t, x) is a nx1 vector, b (t, x) is a nxn matrix, and
0
dz = (dz1 ...dzn ) is an n 1 vector of independent Brownian motion processes
so that dzi dzj = 0 for i 6= j. As in the previous chapter, x (t) includes macroeconomic factors that aect the default-free term structure, but it now also
includes firm-specific factors that aect the likelihood of default for the particular firm. Similar to (17.8), the stochastic discount factor for pricing the firms
default risky bond will be of the form
dM/M = r (t, x) dt (t, x)0 dz (t, x) [dq (t, x) dt]
(18.8)
where (t, x) is an nx1 vector of the market prices of risk associated with
the elements of dz and (t, x) is the market price of risk associated with the
actual default event which occurs when the Poisson process q (t) jumps from
0 (the no default state) to 1 (the absorbing default state) at which time dq =
1.11
b (t, x), is then given by

b (t, x) =
The risk-neutral default intensity,
[1 (t, x)] (t, x).
Note that in this modeling context, default is "doubly-
10 For concreteness our presentation assumes an equilibrium Markov state variable environment. However, much of our results on reduced form pricing of defaultable bonds carry over
to a non-Markov, no-arbitrage context, such as the Heath-Jarrow-Morton framework. See
(Due and Singleton 1999).
11 Recall from the discussion in Chapter 11 that jumps in an assets value, as would occur
when a bond defaults, cannot always be hedged. Thus, in general it may not be possible to
determine (t, x) based on a no arbitrage restriction. This market price of default risk may
need to be determined from an equilibrium model of investor preferences.
516
stochastic." Default depends on the Brownian motion vector dz that drives x

b (t, x), changes over time, but
and determines how the likelihood of default,
it also depends the Poisson process dq which determines the arrival of default.
Hence, default risk reflects two types of risk-premia, (t, x) and (t, x).
Based on the calculation of survival probability in (18.5), the value of the
zero-recovery defaultable bond is
R
RT
h R
i
b
b
tT r(u)du t (u)du
]du B (18.9)
b
bt e tT [r(u)+(u)
DZ (t, T ) = Et e
e
B =E
Equation (18.9) shows that valuing this zero-recovery defaultable bond is similar
b (u)
to valuing a default free bond except that we use the discount rate of r (u)+
b (t, x), and
rather than just r (u). Given specific functional forms for r (t, x),
the risk-neutral state variable process (specifications of (18.7) and (t, x)), the
expression in (18.9) can be computed.
18.2.2
Specifying Recovery Values
The value of a bond that has a possibly non-negative recovery value in the
event of default equals the value in (18.9) plus the present value of the amount
recovered in default. Suppose that if the bond defaults at date where t <
T , bondholders recover an amount w (, x) at date . Now note the the
risk-neutral probability density of defaulting at time is
e
R
t
b
(u)du
b
( )
(18.10)
h R
i
b ( ) is discounted by exp
b (u) du because default at date
In (18.10),
t
is conditioned on not having defaulted previously. Therefore, the present value

of recovery in the event of default, DR (t, T ) is computed by integrating the
expected discounted value of recovery over all possible default dates from t to
517
T:
bt
DR (t, T ) = E
bt
= E
"Z
"Z
r(u)du
w () e
R
t
b
[r(u)+(u)
]du b
b
(u)du
b
( ) d
( ) w ( ) d
#
(18.11)
Putting this together with (18.9) gives the bonds total value, D (t, T ) =
DZ (t, T ) + DR (t, T ), as
"
Z
R
b ]ds
tT [r(s)+(s)
b
D (t, T ) = Et e
B+
R
t
b
[r(s)+(s)
]ds
b ( ) w ( ) d
(18.12)
Recovery Proportional to Par Value
Let us consider some particular specifications for w (, x).
One assumption
used by several researchers is that bondholders recover at the default date a

proportion of the bonds face or par value, that is, w (, x) = ( , x) B where
(, x) is usually assumed to be a constant, say .12
be written as
DR (t, T ) = B
where
In this case (18.11) can
k (t, ) d
(18.13)
h R
i
b
]du
b ()
bt e t [r(u)+(u)
k (t, ) E
(18.14)
b (u, x) are ane functions of x

has a closed-form solution when r (u, x) and
and the vector x in (18.7) has a risk-neutral process that is also ane.13 In this
case, the recovery value in (18.13) can be computed by numerical integration of
12 Work by Darrell Due ((Due 1998)), David Lando((Langer 1998)), and Dilip Madan and
Haluk Unal((Madan and Unal 1998)) make this assumption. As reported by Gregory Duee
(Duee 1999), the recovery rate, , estimated by Moodys for senior unsecured bondholders
is approxmately 44 percent.
13 This is shown in (Due, Pan, and Singleton 2000).
518
k (t, ) over the interval from t to T .
Recovery Proportional to Par Value, Payable at Maturity

An alternative recovery assumption is that if default occurs at date , the bondholders recover a proportion (, x) of the bonds face value, B, payable at the
maturity date T .14 This is equivalent to assuming that the bondholders recover
a proportion ( , x) of the market value of a default-free discount bond paying
B at date T , that is, w ( , x) = ( , x) P ( , T ) B.
Under this assumption
(18.11) becomes
bt
DR (t, T ) = E
bt
= E
"Z
"Z
"
bt e
= E
RT
t
b
[r(u)+(u)
]du
b ( ) ( , x) e
b
(u)du
b
r(u)du
( ) ( , x) e
R
t
b
(u)du
b
RT
t
RT
r(u)du
r(u)du
Bd
#
Bd
( ) ( , x) d B
(18.15)
For the specific case of ( , x) = , a constant, this expression can be simplified

i
h R
RT
b
b ( ) d is the total risk-neutral
(u) du
by noting that the term t exp t
probability of default for the period from date t to the maturity date T . Therei
h R
T b
(u) du , that is, one minus the probability of
fore it must equal 1 exp t
surviving over the same period. Making this substitution and using (18.9) we
have
R
RT
b
(u)du
tT r(u)du
b
t
1e
B
DR (t, T ) = Et e
h RT
i
RT
b
]du B
bt e t r(u)du e t [r(u)+(u)
= E
= BP (t, T ) DZ (t, T )
(18.16)
14 This specification has been studied by Robert Jarrow and Stuart Turnbull ((Jarrow and
Turnbull 1995)) and David Lando ((Langer 1998)).
519
Therefore, the total value of the bond is
D (t, T ) = DZ (t, T ) + DR (t, T ) = 1 DZ (t, T ) + BP (t, T )
(18.17)
Hence, this recovery assumption amounts to requiring only a solution for the
value of a zero-recovery bond.
Recovery Proportional to Market Value
Let us consider one additional recovery assumption analyzed by Darrell Due
and Kenneth Singleton (Due and Singleton 1999).
When default occurs,
bondholders are assumed to recover a proportion of what was the bonds market
value just prior to default.
This is equivalent to assuming that the bonds
market value jumps downward at the default date , suering a proportional

loss of L ( , x). Specifically, at default D ( , T ) jumps to
D + , T = w ( , x) = D , T [1 L ( , x)]
(18.18)
By specifying a proportional loss in value at the time of default, the bonds

dynamics become similar to the jump-diusion model of asset prices presented
in Chapter 11. Treating the defaultable bond as a contingent claim and applying
Its lemma, its process prior to default is similar to equation (11.7):
dD (t, T ) /D (t, T ) = (D kD ) dt + 0D dz + dqD
(18.19)
where D and the nx1 vector D are given by the usual Its lemma expressions
similar to (11.8) and (11.9).
From (11.6) and (18.18), we have that when a
jump occurs dqD = [D ( + , T ) D ( , T )] /D ( , T ) = L (, x), so that we

can write dqD = L (t, x) dq. Also, from (11.10), kD , the expected jump size,
is given by kD ( ) E [D ( + , T ) D ( , T )] /D ( , T ) = L ( , x), so
520
that the drift term in (18.19) becomes D + (t, x) L (t, x).

Now since under the risk-neutral measure, the defaultable bonds total expected rate of return, D , equals the instantaneous maturity default-free rate,
r (t), we can write the bonds risk-neutral process prior to default as
b (t, x) L
b (t, x) dq (18.20)
b (t, x) dt + 0D db
zL
dD (t, T ) /D (t, T ) = r (t, x) +
b (t, x) is the risk-neutral expected proportional loss given default.15 The

where L
intuition of (18.20) is that because the bond has a risk-neutral expected loss
b (t, x), and the risk-neutral instantaneous probability of degiven default of L
b (t, x), when the bond does not default it must earn an
fault (dq = 1) is
b (t, x) L
b (t, x) to make its unconditional risk-neutral
excess expected return of
expected return equal r (t). Based on a derivation similar to that used to obtain
(11.16) and (17.6), one can show that the defaultable bonds value satisfies the
equilibrium partial dierential equation
1
2 Trace
a (t, x)0 Dx R (t, x) D + Dt = 0

b (t, x) b (t, x)0 Dxx + b
(18.21)
where Dx denotes the nx1 vector of first derivatives of D (t, x) with respect to
each of the factors and, similarly, Dxx is the nxn matrix of second order mixed
partial derivatives. In addition, b
a (t, x) = a (t, x) b (t, x) is the risk-neutral
b (t, x) L
b (t, x) is the
drift of the factor process (18.7) and R (t, x) r (t, x) +
defaultable bonds risk-neutral drift in the process (18.20). Now note that should
15 As with the risk-neutral default intensity,

b (t, x), there may be a market price of recovery
b (t, x) that distinguishes it from the physical expected loss at default,
risk associated with L
L (t, x). This market price of recovery risk cannot, in general, be determined from a no
arbitrage restriction because recovery risk may be unhedgeable. Most commonly, modelers
simply posit functional forms for risk-neutral variables in order to derive formulas for defaultable bond values. Dierences between risk-neutral default intensities and losses at default
and their physical counterparts might be inferred based on the market prices of defaultable
bonds and historical (physical) default and recovery rates.
521
the bond reach the maturity date, T , without defaulting, then D (T, T ) = B
which determines the boundary condition for (18.21).
The PDE (18.21) is
in the form of a PDE for a standard contingent claim except that R (t, x) has
replaced r (t, x) in the standard PDE. This insight allows us to write the PDEs
Feynman-Kac solution as:16
h R
i
bt e tT R(u,x)du B
D (t, T ) = E
(18.22)
b (t, x) L
b (t, x) can be viewed as the "default-adjusted"
where R (t, x) r (t, x)+
discount rate.
b (t, x) L
b (t, x) has the interpretation
The product s (t, x)
as the "credit spread" on an instantaneous-maturity defaultable bond. Since
b (t, x) and L
b (t, x) are not individually identified in (18.22), when implementing
this formula we can simply specify a single functional form for s (t, x).
18.2.3
Examples
Because default intensities and/or credit spreads must be non-negative, a popular stochastic process for modeling these variables is the mean-reverting, squareroot process used in the term structure model of John Cox, Jonathan Ingersoll,
and Stephen Ross (Cox, Ingersoll, and Ross 1985b).
To take a very simple
a (t, x) =
example, suppose that x = (x1 x2 ) is a two-dimensional vector, b
0
(1 (x1 x1 ) 2 (x2 x2 )) , and b (t, x) is a diagonal matrix with first and
second diagonal elements of 1 x1 and 2 x2 , respectively. If one assumes
b (t, x) = x2 (t), this has the implication that the defaultr (t, x) = x1 (t) and
free term structure and the risk-neutral default intensity are independent. Arguably, this is unrealistic since empirical work has found a negative correlation
16 Recall from Chapter 10 that (10.14) was shown to be the Feynman-Kac solution to the
Black-Scholes PDE (10.7). See Darell Due and Kenneth Singleton (Due and Singleton
1999) for an alternative derivation of (18.22) that does not involve specification of factors or
the bonds PDE.
522
between default-free interest rates and the likelihood of corporate defaults.17

b (t, x) while restricting
Allowing for non-zero correlation between r (t, x) and
each to be positive is certainly feasible but comes at the cost of requiring numerical, rather than closed-form, solutions for defaultable bond values.18
Hence,
for simplicity of presentation, we maintain the independence assumption.
With r (t, x) = x1 (t) and denoting x1 = r, we obtain the Cox, Ingersoll, and
Ross formula for the value of a default-free discount bond:19
P (t, T ) = A1 ( ) eB1 ( )r(t)
(18.23)
where
21 e(1 +1 ) 2
A1 ( )
(1 + 1 ) (e1 1) + 21
B1 ()
and 1
p
21 + 2 21 .
21 r/21
2 e1 1
(1 + 1 ) (e1 1) + 2 1
(18.24)
(18.25)
b (t, x) = x2 (t) and denoting x2 = , then

Also with
b (t) we can write

based on (18.9) and the assumed independence of r (t) and
17 This evidence is presented in work by Gregory Duee ((Duee 1999)) and Pierre CollinDufresne and Bruno Solnik ((Collin-Dufresne and Solnik 2001)).
18 For models with more flexible correlation structures that require numerical solutions, see
examples given by Darrell Due and Kenneth Singleton (Due and Singleton 1999). Some
b (t) (or s (t) =
b (t) L (t)) be positive by
research has dropped the restriction that r (t) and
assuming these variables follow multi-variate ane Gaussian processes. This permits general
correlation between default-free interest rates and default intensities as well as closed-form
solutions for defaultable bonds. The model in work by C.V.N. Krishnan, Peter Ritchken, and
James Thomson ((Krishnan, Ritchken, and Thomson 2004)) is an example of this.
19 The formula in (18.23) to (18.25) is the same as (13.51) to (13.53) except that it is written
in terms of the parameters of the risk-neutral, rather than physical, process for r (t). Hence,
relative to
our earlier notation, 1 = + where the market price of interest rate risk equals
(t) = r/ 1 .
523
the value of the zero-recovery bond as

h R
i
b
]ds B
bt e tT [r(s)+(s)
DZ (t, T ) = E
h R
i h RT b
i
bt e tT r(s)ds E
bt e t (s)ds B
= E
= P (t, T ) V (t, T ) B
(18.26)
where
b
V (t, T ) = A2 () eB2 ( )(t)
(18.27)
and where A2 ( ) is the same as A1 ( ) in (18.24) and B2 () is the same as

B1 ( ) in (18.25) except that 2 replaces 1 , 2 replaces 1 , replaces r, and
p
2 22 + 222 replaces 1 .
If we assume that recovery is a fixed proportion, , of par value, payable at
maturity, then based on (18.17) the value of the defaultable bond equals
1 DZ (t, T ) + BP (t, T )
= + 1 V (t, T ) P (t, T ) B
D (t, T ) =
(18.28)
V (t, T ) in (18.27) is analogous to a bond price in the standard Cox, Ingersoll,

b (t)
and Ross term structure model, and as such it will be inversely related to
b (t) is strictly positive, which can be ensured

and strictly less than 1 whenever
when 22 22 .
Thus, (18.28) confirms that the defaultable bonds value
declines as its risk-neutral default intensity rises.

A slightly dierent defaultable bond formula can be obtained when recovery
b (t, x) L
b (t, x) = x2
is assumed to be proportional to market value and s (t, x)
524
with the notation x2 = s. In this case, (18.22) becomes

h R
i
bt e tT [r(u)+s(u)]du B
D (t, T ) = E
h R
i h RT
i
bt e tT r(u)du E
bt e t s(u)du B
= E
(18.29)
S (t, T ) = A2 ( ) eB2 ( )s(t)
(18.30)
= P (t, T ) S (t, T ) B
where
and where A2 ( ) is the same as A1 ( ) in (18.24) and B2 ( ) is the same as

B1 ( ) in (18.25) except that 2 replaces 1 , 2 replaces 1 , s replaces r, and
p
2 22 + 222 replaces 1 . This defaultable bond is priced similar to a
default-free bond except that the instantaneous maturity interest rate, R (t) =
r (t) + s (t) is now the sum of two non-negative square root processes. Hence
the defaultable bond is inversely related to s (t) and can be strictly less than
the default-free bond as s (t) can always be positive when 22 s 22 .
Valuing the defaultable coupon bond of a particular issuer (e.g., corporation) is straightforward given the preceding analysis of defaultable zero-coupon
bonds. Suppose that the issuers coupon bond promises n cashflows, with the
ith promised cashflow being equal to ci and being paid at date Ti > t. Then
the value of this coupon bond in terms of our zero-coupon bond formulas is
n
P
i=1
D (t, Ti )
ci
B
(18.31)
Our results can also be applied to valuing credit derivatives. A credit default
swap is a popular credit derivative that typically has the following structure.
One party, the protection buyer, makes periodic payments until the contracts
maturity date as long as a particular issuer, bond, or loan does not default.
525
The other party, the protection seller, receives these payments in return for
paying the dierence between the bond or loans par value and its recovery
value if default occurs prior to the maturity of the swap contract. At the initial
agreement date of this swap contract, the periodic payments are set such that
the initial contract has a zero market value.
We can use our previous analysis to value each side of this swap. Let the
contract specify equal period payments of c at future dates t + , t + 2, ...,
t + n.20 Then recognizing that these payments are contingent on default not
occurring and that they have zero value following a possible default event, their
market value equals
n
c P
DZ (t, t + i)
B i=1
(18.32)
where DZ (t, T ) is the value of the zero-recovery bond given in (18.9).
If we
let w (, x) be the recovery value of the defaultable bond (or loan) underlying
the swap contract, then assuming this bonds maturity date is T t + n, the
value of the swap protection can be computed similar to (18.11) as
bt
E
"Z
t+n
R
t
b
[r(u)+(u)
]du
b () [B w ( )] d
(18.33)
The protection sellers payment in the event of default, B w ( ), is often

simplified by assuming recovery is a fixed proportion of par value, that is, B
w ( ) = B B = B 1 . For this special case, (18.33) becomes
B 1
t+n
k (t, ) d
(18.34)
where k (t, ) is defined in (18.14). For given assumptions regarding the funcb (t, x), and w (t, x), and the state variables x, the value
tion forms of r (t, x),
20 A period of = one-half year is common since these payments often coincide with an
underlying coupon bond making semi-annual payments.
526
of the swap payments, c, can be determined that equates (18.32) to (18.33).

A general issue that arises when implementing the reduced-form approach to
b (t), s (t), or w (t)
valuing risky debt is determining the proper current values
that may not be directly observable.
One or more of these default variables
might be inferred by setting the actual market prices of one or more of an issuers
b (t),
bonds to their theoretical formulas. Then, based on the implied values of
s (t), or w (t), one can determine whether a given bond of the same issuer is overor under-priced relative to other bonds.
Alternatively, these implied default
variables could be used to set the price of a new bond of the same issuer or a
credit derivative (such as a default swap) written on the issuers bonds.
18.3
Summary
Research on credit risk has grown significantly in recent years, generating and
generated by greater interest in credit risk management and credit derivatives.
This chapter introduced the two main branches of modeling defaultable bond
values.
The structural approach derives default based on the interaction be-
tween a firms assets and liabilities.
Potentially, it can improve our under-
standing between capital structure and corporate bond prices.
In contrast,
the reduced form method abstracts from specific aspects of a firms financial
structure, but it can permit a more flexible modeling of default probabilities
and may provide a better fit to the prices of an issuers bonds.
While, due to space constraints, this chapter has been limited to models of
corporate defaults, the credit risk literature also encompasses additional topics
such as consumer credit risk and the credit risk of (securitized) portfolios of
loans and bonds.
Interest by both academics and practitioners in the broad
field of credit risk will undoubtedly continue.
18.4. EXERCISES
18.4
527
Exercises
1. Consider the example given in the structural approach to modeling default risk. Maintain the assumptions made in the chapter but now suppose
that a third party guarantees the firms debtholders that if the firm defaults, the debtholders will receive their promised payment of B. In other
words, this third-party guarantor will make a payment to the debtholders
equal to the dierence between the promised payment and the firms assets if default occurs. (Banks often provide such a guarantee in the form
of a letter of credit. Insurance companies often provide such a guarantee
in the form of bond insurance.)
What would be the fair value of this bond insurance at the initial date, t?
In other words, what would be the competitive bond insurance premium
to be charged at date t?
2. Consider a Merton (1974)-type structural model of credit risk. A firm

is assumed to have shareholders equity and two zero-coupon bonds that
both mature at date T . The first bond is senior debt and promises to pay
B1 at maturity date T while the second bond is junior (or subordinated)
debt and promises to pay B2 at maturity date T . Let A (t), D1 (t), and
D2 (t) be the date t values of the firms assets, senior debt, and junior
debt, respectively. Then the maturity value of the bonds are
B1
if A (T ) B1
D1 (T ) =
A (T ) otherwise
528
B2
D2 (T ) =
A (T ) B1
if A (T ) B1 + B2
if B1 + B2 > A (T ) B1
otherwise
The firm is assumed to pay no dividends to its shareholders, and the value
of shareholders equity at date T , E (T ), is assumed to be
A (T ) (B1 + B2 ) if A (T ) B1 + B2
E (T ) =
0
otherwise
Assume that the value of the firms assets follows the process
dA/A = dt + dz
where denotes the instantaneous expected rate of return on the firms

assets and is the constant standard deviation of return on firm assets. In
addition, the continuously-compounded risk-free interest rate is assumed
to be the constant r. Let the current date be t, and define the time until
the debt matures as T t.
2.a Give a formula for the current, date t, value of shareholders equity, E (t).
2.b Give a formula for the current, date t, value of the senior debt, D1 (t).
2.c Using the results from 2.a and 2.b, give a formula for the current, date t,
value of the junior debt, D2 (t).
3. Consider a portfolio of m dierent defaultable bonds (or loans), where
the ith bond has a default intensity of i (t, x) where x is a vector of state
variables that follows the multi-variate diusion process in (18.7). Assume
that the only source of correlation between the bonds defaults is through
18.4. EXERCISES
529
their default intensities. Suppose that the maturity dates for the bonds
all exceed date T > t. Write down the expression for the probability that
none of the bonds in the portfolio defaults over the period from date t to
date T .
4. Consider the standard (plain vanilla) swap contract described in the previous chapter. In equation (17.67) it was shown that under the assumption
that each partys payments were default free, the equilibrium swap rate
agreed to at the initiation of the contract, date T0 , equals
s0,n (T0 ) =
1 P (T0 , Tn+1 )
P
n+1
j=1 P (T0 , Tj )
where for this contract fixed interest rate coupon payments are exchanged
for floating interest rate coupon payments at the dates T1 , T2 , ...,Tn+1 with
Tj+1 = Tj + and is the maturity of the LIBOR of the floating rate
coupon payments.
This swap rate formula is valid when neither of the
parties have credit risk. Suppose, instead, that they both have the same
credit risk, and it is equivalent to the credit risk reflected in LIBOR interest
rates.21 Moreover, assume a reduced form model of default with recovery
proportional to market value, so that the value of a LIBOR discount bond
promising $1 at maturity date Tj is given by (18.22):
RT
j
bT0 e T0 R(u,x)du
D (T0 , Tj ) = E
where the default-adjusted instantaneous discount rate R (t, x) r (t, x)+
b (t, x) L
b (t, x) is assumed to be the same for both parties. Assume that if
default occurs at some date < Tn+1 , the counterparty whose position is
"in-the-money" (whose position has positive value) suers a proportional

21 Recall
that LIBOR reflects the level of default risk for a large international bank.
530

loss of L ( , x) in that position. Show that under these assumptions, the
equilibrium swap rate is
s0,n (T0 ) =
1 D (T0 , Tn+1 )
P
n+1
j=1 D (T0 , Tj )
Bibliography
Abel, A. B. (1990): Asset Prices Under Habit Formation and Catching Up
with the Joneses, American Economic Review, 80, 3842.
Ahn, C.-M., and H. E. Thompson (1988): Jump-Diusion Processes and
the Term Structure of Interest Rates, Journal of Finance, 43, 155174.
Ahn, D.-H., R. F. Dittmar, and A. R. Gallant (2002): Quadratic Term
Structure Models: Theory and Evidence, Review of Financial Studies, 15,
243288.
Amin, K. I., and A. J. Morton (1994): Implied Volatility Functions in
Arbitrage-Free Term Structure Models, Journal of Financial Economics,
35, 141180.
Andersen, T. G., T. Bollerslev, F. X. Diebold, and H. Ebens (2001):
The Distribution of Realized Stock Return Volatility, Journal of Financial
Economics, 61, 4376.
Anderson, R. W., and J.-P. Danthine (1981): Cross Hedging, Journal
of Political Economy, 89(6), 11821196.
Ang, A., and M. Piazzesi (2003): A No-Arbitrage Vector Autoregression
of Term Structure Dynamics with Macroeconomic and Latent Variables,
Journal of Monetary Economics, 50(4), 745787.
567
568
BIBLIOGRAPHY
Arrow, K. J. (1953): The Role of Securities in the Optimal Allocation of

Risk-bearing, conomtric.
(1964): The Role of Securities in the Optimal Allocation of Riskbearing, Review of Economics Studies, 31(2), 9196.
(1971): Essays in the Theory of Risk-Bearing. North-Holland, Amsterdam.
Attanasio, O. P., and G. Weber (1993): Consumption Growth, the Interest Rate, and Aggregation, Review of Economic Studies, 60(3), 631649.
Backus, D., and S. Zin (1994): Reverse Engineering the Yield Curve,
NBER Working Paper No. 4676.
Bakshi, G. S., C. Cao, and Z. Chen (1997): Empirical Performance of
Alternative Option Pricing Models, Journal of Finance, 52(5), 20032049.
Bakshi, G. S., and Z. Chen (1996): Inflation, Asset Prices, and the Term
Structure of Interest Rates in Monetary Economies, Review of Financial
Studies, 9(1), 241275.
Balduzzi, P., S. Das, and S. Foresi (1998): The Central Tendency: A
Second Factor in Bond Yields, Review of Economics and Statistics, 80(1),
6272.
Bansal, R., and H. Zhou (2002): The Term Structure of Interest Rates with
Regime Shifts, Journal of Finance, 57, 19972043.
Barberis, N., M. Huang, and J. Santos (2001): Prospect Theory and
Asset Prices, Quarterly Journal of Economics, 116, 153.
Barberis, N., A. Shleifer, and R. Vishny (1998): A Model of Investor
Sentiment, Journal of Financial Economics, 49(3), 307343.
BIBLIOGRAPHY
569
Barberis, N., and R. Thaler (2002): A Survey of Behavioral Finance,

NBER Working Paper No. 9222.
Barsky, R. B., F. T. Juster, M. S. Kimball, and M. D. Shapiro (1997):
Preference Parameters and Behavioral Heterogeneity: An Experimental Approach in the Health and Retirement Study, Quarterly Journal of Economics,
112(2), 537579.
Basel Committee on Banking Supervision (2005): International Convergeance of Capital Measurement and Capital Standards: A Revised Framework, Discussion paper, Bank for International Settlements, Basel.
Bates, D. S. (1991): The Crash of 87 - Was It Expected? The Evidence
from Options Markets, Journal of Finance, 46, 10091044.
(1996): Jumps and Stochastic Volatility: Exchange Rate Processes
Implicit in Deutsche Mark Options, Review of Financial Studies, 9, 69107.
(2002): Empirical Option Pricing: A Retrospection, Journal of
Econometrics, 116, 387404.
Beaglehole, D., and M. Tenney (1992): A Nonlinear Equilibrium Model
of the Term Structure of Interest Rates: Corrections and Additions, Journal
of Financial Economics, 32, 345454.
Beaudry, P., and E. van Wincoop (1996): The Intertemporal Elasticity of
Substitution: An Exploration Using a U.S. Panel of State Data, Economica,
63(251), 495512.
Bellman, R. (1957): Dynamic Programming. Princeton University Press,
Princeton, N.J.
Bernoulli, D. (1954): Exposition of a New Theory on the Measurement of
Risk, Econometrica, 22(1), 2336.
570
BIBLIOGRAPHY
Bernstein, P. L. (1992): Capital Ideas: The Improbable Origins of Modern

Wall Street. Free Press.
Bhamra, H. S., and R. Uppal (2003): The Role of Risk Aversion and Intertemporal Substitution in Dynamic Consumption-Portfolio Choice with Recursive Utility, London Business School working paper.
Black, F. (1972): Capital Market Equilibrium with Restricted Borrowing,
Journal of Business, 45, 444455.
(1976): The Pricing of Commodity Contracts, Journal of Financial
Black, F., and J. C. Cox (1976): Valuing Corporate Securities: Some Eects
of Bond Indenture Provisions, Journal of Finance, 31, 351367.
Black, F., E. Derman, and W. Toy (1990): A One-Factor Model of Interest
Rates and Its Application to Treasury Bond Options, Financial Analysts
Journal, 46(1), 3339.
Black, F., and P. Karasinski (1991): Bond and Option Pricing When
Short Rates Are Lognormal, Financial Analysts Journal, 47(4), 5259.
Black, F., and M. Scholes (1973): The Pricing of Options and Corporate
Liabilities, Journal of Political Economy, 81, 637659.
Blanchard, O. J. (1979): Speculative Bubbles, Crashes and Rational Expectations, Economics Letters, 3, 387389.
Bodie, Z., R. C. Merton, and W. Samuelson (1992): Labor Supply Flexibility and Portfolio Choice in a Life-cycle Model, Journal of Economic Dynamics and Control.
BIBLIOGRAPHY
571
Bollerslev, T., R. Y. Chou, and K. F. Kroner (1992): ARCH Modeling in Finance: A Review of Theory and Empirical Evidence, Journal of
Econometrics, 52, 559.
Brace, A., D. Gatarek, and M. Musiela (1997): The Market Model of
Interest Rate Dynamics, Mathematical Finance, 7(2), 127147.
Breeden, D. T. (1979): An Intertemporal Asset Pricing Model with Stochastic Consumption and Investment Opportunities, Journal of Financial
Brunnermeier, M. K. (2001): Asset Pricing under Asymmetric Information:
Bubbles, Crashes, Technical Analysis, and Herding. Oxford University Press,
Oxford.
Campbell, J. (1999): Asset Prices, Consumption, and the Business Cycle,
in Handbook of Macroeconomics, ed. by J. B. Taylor, and M. Woodford. Amsterdam: North Holland.
Campbell, J. Y. (2000): Asset Pricing at the Millenium, Journal of Finance.
Campbell, J. Y., and J. H. Cochrane (1999): By Force of Habit: A Consumption Based Explanation of Aggregate Stock Market Behavior, Journal
Campbell, J. Y., A. Lo, and C. MacKinlay (1997): The Econometrics of
Financial Markets. Princeton University Press, Princeton.
Campbell, J. Y., and N. G. Mankiw (1989): Consumption, Income, and
Interest Rates: Reinterpreting the Time Series Evidence, in National Bureau
of Economic Research Macroeconomics Annual, ed. by O. J. Blanchard, and
S. Fischer, vol. 4. MIT Press, Cambridge, MA.
572
BIBLIOGRAPHY
Campbell, J. Y., and L. M. Viceira (2002): Strategic Asset Allocation:

Portfolio Choice for Long-Term Investors. Oxford University Press, Oxford.
Carhart, M. M. (1997): On Persistence in Mutual Fund Performance, Journal of Finance, 52.
Carrier, G. F., and C. E. Pearson (1976): Partial Dierential Equations:
Theory and Technique. Academic Press, New York.
Carverhill, A. (1994): When Is the Short Rate Markovian?, Mathematical
Finance, 4, 305312.
Cecchetti, S. G., P.-S. Lam, and N. C. Mark (1993): The Equity Premium and the Risk-Free Rate: Matching the Moments, Journal of Monetary
(1994): Testing Volatility Restrictions on Intertemporal Marginal
Rates of Substitution Implied by Euler Equations and Asset Returns, Journal of Finance, 49, 123152.
Chacko, G., and S. Das (2002): Pricing Interest Rate Derivatives: A General
Approach, Review of Financial Studies, 15, 195241.
Chapman, D. A. (1998): Habit Formation and Aggregate Consumption,
Econometrica, 66, 12231230.
Chen, N.-F., R. Roll, and S. A. Ross (1986): Economic Forces and the
Stock Market, Journal of Business, 59(3), 383403.
Cheridito, P., D. Filipovic, and R. Kimmel (2003): Market Price of Risk
Specifications for Ane Models: Theory and Evidence, Princeton University
Working Paper, December.
BIBLIOGRAPHY
573
Churchill, R. V., and J. W. Brown (1978): Fourier Series and Boundary

Value Problems. McGraw-Hill, New York.
Claessens, S., and G. G. Pennacchi (1996): Estimating the Likelihood
of Mexican Default from the Market Prices of Brady Bonds, Journal of
Financial and Quantitative Analysis, 31(1), 109126.
Collin-Dufresne, P., and R. S. Goldstein (2001): Do Credit Spreads
Reflect Stationary Leverage Ratios?, Journal of Finance, 56, 19291958.
(2003): Generalizing the Ane Framework to HJM and Random
Field Models, Carnegie-Mellon University and Washington University working paper.
Collin-Dufresne, P., and B. Solnik (2001): On the Term Structure of
Default Premia in the Swap and LIBOR Markets, Journal of Finance, 56,
10951115.
Connor, G., and R. A. Korajczyk (1995): The Arbitrage Pricing Theory
and Multifactor Models of Asset Returns, in Finance, Handbooks in Operations Research and Management Science, ed. by R. A. Jarrow, V. Maksimovic,
and W. T. Ziemba, vol. 4, chap. 23. North Holland, Amsterdam.
Constantinides, G. M. (1986): Capital Market Equilibrium with Transaction Costs, Journal of Political Economy, 94, 842862.
(1990): Habit Formation: A Resolution of the Equity Premium Puzzle, Journal of Political Economy, 98(3), 519543.
(1992): A Theory of the Nominal Term Structure of Interest Rates,
Review of Financial Studies, 5, 531552.
Coval, J. D., and T. Shumway (2003): Do Behavioral Biases Aect Prices?,
Journal of Finance, forthcoming.
574
BIBLIOGRAPHY
Cox, D. R. (1955): Some Statistical Methods Connected with Series of

Events, Journal of the Royal Statistical Society Series B, 17(2), 129164.
Cox, J., and S. A. Ross (1976): The Valuation of Options for Alternative
Stochastic Processes, Journal of Financial Economics, 3, 145166.
Cox, J., S. A. Ross, and M. Rubinstein (1979): Option Pricing: A Simplified Approach, Journal of Financial Economics, 7, 229263.
Cox, J. C., and C.-F. Huang (1989): Optimal Consumption and Portfolio
Policies When Asset Prices Follow a Diusion Process, Journal of Economic
Theory, 49, 3383.
Cox, J. C., J. E. Ingersoll, and S. A. Ross (1981): The Relation Between
Forward Prices and Futures Prices, Journal of Financial Economics, 9, 321
346.
(1985a): An Intertemporal General Equilibrium Model of Asset
Prices, Econometrica, 53, 363384.
Cox, J. C., J. E. Ingersoll, Jr., and S. A. Ross (1985b): A Theory of
the Term Structure of Interest Rates, Econometrica, 53, 385408.
Crosbie, P. J., and J. R. Bohn (2002): Modeling Default Risk. KMV, LLC,
San Francisco, CA.
Cvsa, V., and P. H. Ritchken (2001): Pricing Claims Under Garch-Level
Dependent Interest Rate Processes, Management Science, 47, 16931711.
Dai, Q., A. Le, and K. Singleton (2006): Discrete-time Dynamic Term
Structure Models with Generalized Market Prices of Risk, UNC, NYU, and
Stanford University working paper.
BIBLIOGRAPHY
575
Dai, Q., and K. Singleton (2000): Specification Analysis of Ane Term

Structure Models, Journal of Finance, 55, 19431978.
(2003): Term Structure Modeling in Theory and Reality, Review of
Financial Studies, 16, 631678.
(2004): Fixed-Income Pricing, in Handbook of Economics and Finance, ed. by G. Constantinides, M. Harris, and R. Stulz. North-Holland,
Amsterdam.
Daniel, K. D., D. Hirshleifer, and A. Subrahmanyam (1998): Investor
Psychology and Security Market Under- and Over-reactions, Journal of Finance, 53(6), 18391886.
Daniel, K. D., D. Hirshleifer, and S. H. Teoh (2001): Investor Psychology in Capital Markets: Evidence and Policy Implications, Northwestern
University Working Paper.
Das, S. R. (2002): The Surprise Element: Jumps in Interest Rates, Journal
of Econometrics, 106, 2765.
Das, S. R., and S. Foresi (1996): Exact Solutions for Bond and Option
Prices with Systematic Jump Risk, Review of Derivatives Research, 1, 724.
Debreu, G. (1959): Theory of Value. Cowles Foundation Monograph 17, Yale
University Press, New Haven, CT.
DeLong, J. B., A. Shleifer, L. Summers, and R. J. Waldmann (1991):
The Survival of Noise Traders in Financial Markets, Journal of Business,
64(1), 120.
Detemple, J., and F. Zapatero (1991): Asset Prices in an Exchange Economy with Habit Formation, Econometrica, 59, 16331658.
576
BIBLIOGRAPHY
Diamond, D. W., and R. E. Verrecchia (1981): Information Aggregation

in a Noisy Rational Expectations Economy, Journal of Financial Economics,
9(3), 221235.
Diba, B. T., and H. I. Grossman (1988): The Theory of Rational Bubbles
in Stock Prices, Economic Journal, 98, 746754.
Diebold, F., M. Piazzesi, and G. Rudebusch (2005): Modeling Bond
Yields in Finance and Macroeconomics, American Economic Review, 95(2),
415420.
Dimson, E., P. Marsh, and M. Staunton (2002): Triumph of the Optimists: 101 Years of Global Investment Returns. Princeton University Press,
Princeton, N.J.
Duarte, J. (2004): Evaluating an Alternative Risk Preference in Ane Term
Structure Models, Review of Financial Studies, 17, 379404.
Duesenberry, J. S. (1949): Income, Saving, and the Theory of Consumer
Behavior. Harvard University Press, Cambridge, MA.
Duffee, G. (1999): Estimating the Price of Default Risk, Review of Financial Studies, 12, 197226.
(2002): Term Premia and Interest Rate Forecasts in Ane Models,
Journal of Finance, 57, 405443.
Duffie, D. (1998): Defaultable Term Structure Models with Fractional Recovery of Par, Stanford University Working Paper.
Duffie, D., and L. Epstein (1992a): Asset Pricing with Stochastic Dierential Utility, Review of Financial Studies, 5, 411436.
(1992b): Stochastic Dierential Utility, Econometrica, 60, 353394.
BIBLIOGRAPHY
Duffie, D., D. Filipovic, and W. Schachermayer (2002):
577
Ane
Processes and Applications in Finance, Annals of Applied Probability, 13,

9841053.
Duffie, D., W. Fleming, M. Soner, and T. Zariphopoulou (1997):
Hedging in Incomplete Markets with HARA Utility, Journal of Economic
Dynamics and Control, 21, 753782.
Duffie, D., and R. Kan (1996): A Yield-Factor Model of Interest Rates,
Mathematical Finance, 6, 379406.
Duffie, D., and D. Lando (2001): Term Structures of Credit Spreads with
Incomplete Accounting Information, Econometrica, 69, 633664.
Duffie, D., J. Pan, and K. Singleton (2000): Transform Analysis and
Asset Pricing for Ane Jump-Diusions, Econometrica, 68, 13431376.
Duffie, D., and K. Singleton (1999): Modeling Term Structures of Defaultable Bonds, Review of Financial Studies, 12, 687720.
Dumas, B., and E. Luciano (1991): An Exact Solution to a Dynamic Portfolio Choice Problem under Transactions Costs, Journal of Finance, 46,
577596.
Epstein, L., and S. Zin (1989): Substitution, Risk Aversion, and the Temporal Behavior of Consumption Growth and Asset Returns I: A Theoretical
Framework, Econometrica, 57, 937969.
Fama, E. F., and K. R. French (1988): Permanent and Temporary Components of Stock Prices, Journal of Political Economy, 96(2), 246273.
(1993): Common risk factors in the returns on stocks and bonds,
Journal of Financial Economics, 33, 356.
578
BIBLIOGRAPHY
Ferson, W. E., and G. M. Constantinides (1991): Habit Persistence and

Durability in Aggregate Consumption; Empirical Tests, Journal of Financial
Friedman, M. (1953): The Case for Flexible Exchange Rates, in Essays
in Positive Economics, ed. by M. Friedman. University of Chicago Press,
Chicago, IL.
Goldstein, R. (2000): The Term Structure of Interest Rates as a Random
Field, Review of Financial Studies, 13, 365384.
Gorton, G. B., and G. G. Pennacchi (1993): Security Baskets and IndexLinked Securities, Journal of Business, 66, 127.
Grossman, S. J. (1976): On the Eciency of Competitive Stoch Markets
Where Traders Have Diverse Information, Journal of Finance, 31, 573585.
Grossman, S. J., and J. E. Stiglitz (1980): On the impossibility of informationally ecient markets, American Economic Review, 70(3), 393408.
Grundy, B. D., and M. McNichols (1989): Trade and Revelation of Information through Prices and Direct Disclosure, Review of Financial Studies,
2, 495526.
Hall, R. E. (1988): Intertemporal Substitution and Consumption, Journal
Hansen, L. P., and R. Jagannathan (1991): Implications of Security Market Data for Models of Dynamic Economies, Journal of Political Economy,
99(2), 225262.
Hansen, L. P., and K. J. Singleton (1983): Stochastic Consumption, Risk
Aversion, and the Temporal Behavior of Asset Returns, Journal of Political
Economy, 91, 249265.
BIBLIOGRAPHY
579
Harrison, J. M., and D. M. Kreps (1979): Martingales and Arbitrage in

Multiperiod Securities Markets, Journal of Economic Theory, 20, 381408.
Harrison, J. M., and S. Pliska (1981): Martingales and Stochastic Integrals and in Theory of Continuous Trading, Stochastic Processes and their
Applications, 11, 215260.
Harvey, C. R., and A. Siddique (2000): Conditional Skewness in Asset
Pricing Tests, Journal Finance, 55, 12631295.
He, H., and N. D. Pearson (1991): Consumption and Portfolio Policies with
Incomplete Markets and Short-Sale Constraints: The Infinite Dimensional
Case, Journal of Economic Theory, 54, 259304.
Heath, D., R. A. Jarrow, and A. Morton (1992): Bond Pricing and the
Term Structure of Interest Rates: A New Methodology for Contingent Claims
Valuation, Econometrica, 60(1), 77105.
Heaton, J. C., and D. J. Lucas (2000): Portfolio Choice and Asset Prices:
The Importance of Entrepreneurial Risk, Journal of Finance, 55(3), 1163
1198.
Hellwig, M. F. (1980): On the Aggregation of Information in Competitive
Markets, Journal of Economic Theory, 22, 477498.
Heston, S. (1993): A Closed-Form Solution for Options with Stochastic
Volatility with Applications to Bond and Currency Options, Review of Financial Studies, 6(2), 327343.
Hindy, A., and C.-F. Huang (1993): Optimal Consumption and Portfolio
Rules with Durability and Local Substitution, Econometrica, 61, 85121.
Hirshleifer, D. (2001): Investor Psychology and Asset Pricing, Journal of
Finance, 64(4), 15331597.
580
BIBLIOGRAPHY
Ho, T., and S. B. Lee (1986): Term Structure Movements and Pricing Interest Rate Contingent Claims, Journal of Finance, 41, 10111028.
(2004): The Oxford Guide to Financial Modeling: Applications for
Capital Markets, Corporate Finance, Risk Management, and Financial Institutions. Oxford University Press, New York.
Hong, H., and J. C. Stein (1999): A Unified Theory of Underreaction,
Momentum Trading and Overreaction in Asset Markets, Journal of Finance,
54(6), 21432184.
Huang, J.-z., and M. Huang (2003): How Much of the Corporate-Treasury
Yield Spread Is Due to Credit Risk?, Penn State University, NYU, and
Stanford Business School Working Paper.
Huberman, G. (1982): A Simple Approach to Arbitrage Pricing Theory,
Journal of Economic Theory, 28, 183191.
Hull, J., and A. White (1990): Pricing Interest Rate Derivative Securities,
Review of Financial Studies, 3, 573592.
(1993): One-Factor Interest-Rate Models and the Valuation of
Interest-Rate Derivative Securities, Journal of Financial and Quantitative
Analysis, 28, 235254.
Hull, J. C. (2000): Options, Futures, and other Derivative Securities. PrenticeHall, Inc., Upper Saddle River, NJ, fourth edn.
Ingersoll, J. E. (1987): Theory of Financial Decision Making. Rowman &
Littlefield, Totowa, NJ.
Inui, K., and M. Kijima (1998): A Markovian Framework in Heath-JarrowMorton Models, Journal of Financial and Quantitative Analysis, 33, 423
440.
BIBLIOGRAPHY
581
It, K. (1944): Stochastic Integral, Proceedings of the Imperial Academy of

Tokyo, 20, 519524.
(1951): On Stochastic Dierential Equations, Memoirs of the American Mathematical Society, 4, 151.
Jagannathan, R., and E. McGrattan (1995): The CAPM Debate, Federal Reserve Bank of Minneapolis Quarterly Review, 19(4), 217.
Jagannathan, R., and Z. Wang (1996): The CAPM is Alive and Well,
Journal of Finance, 51(1), 353.
Jarrow, R. A. (2002): Modelling Fixed Income Securities and Interest Rate
Options. McGraw Hill, New York, second edn.
Jarrow, R. A., D. Lando, and S. M. Turnbull (1997): A Markov Model
for the Term Structure of Credit Risk Spreads, Review of Financial Studies,
10(2), 481523.
Jarrow, R. A., and G. S. Oldfield (1981): Forward Contracts and Futures
Contracts, Journal of Financial Economics, 9, 373382.
Jarrow, R. A., and S. M. Turnbull (1995): Pricing Derivatives on Financial Securities subject to Credit Risk, Journal of Finance, 50(1), 5386.
Jegadeesh, N., and G. Pennacchi (1996): The Behavior of Interest Rates
Implied by the Term Structure of Eurodollar Futures, Journal of Money,
Credit and Banking, 28, 426446.
Jegadeesh, N., and S. Titman (1993): Returns to buying winners and
selling losers: Implications for stock market eciency, Journal of Finance,
48(1), 6591.
582
BIBLIOGRAPHY
Jones, E., S. Mason, and E. Rosenfeld (1984): Contingent Claims Analysis of Corporate Capital Structures: An Empirical Investigation, Journal of
Finance, 39, 611627.
Jones, E. P. (1984): Option Arbitrage and Strategy with Large Price
Changes, Journal of Financial Economics, 13, 91113.
Judd, K. L. (1998): Numerical Methods in Economics. MIT Press, Cambridge,
MA.
Kahneman, D., and A. Tversky (1979): Prospect Theory: An analysis of
decision under risk, Econometrica, 47, 263291.
Karatzas, I., J. Lehoczky, and S. E. Shreve (1987): Optimal Portfolio
and Consumption Decisions for a "Small Investor" on a Finite Horizon,
SIAM Journal of Control and Optimization, 25, 15571586.
Karatzas, I., and S. E. Shreve (1991): Brownian Motion and Stochastic
Calculus. Springer-Verlag, New York.
Karlin, S., and H. M. Taylor (1975): A First Course in Stochastic
Processes. Academic Press, New York, NY.
(1981): A Second Course in Stochastic Processes. Academic Press,
New York, NY.
Kennedy, D. (1994): The Term Structure of Interest Rates as a Gaussian
Random Field, Mathematical Finance, 4, 247258.
(1997): Characterizing Gaussian Models of the Term Structure of
Interest Rates, Mathematical Finance, 7, 107118.
Kimmel, R. L. (2004): Modeling the Term Structure of Interest Rates: A
New Approach, Journal of Financial Economics, 72, 143183.
BIBLIOGRAPHY
583
Kindleberger, C. P. (2001): Manias, Panics, and Crashes: A History of

Financial Crises. John Wiley and Sons, New York.
Kogan, L., S. Ross, J. Wang, and M. Westerfield (2006): The Price
Impact and Survival of Irrational Traders, Journal of Finance, 61, 195,229.
Kraus, A., and R. Litzenberger (1976): Skewness Preference and the
Valuation of Risk Assets, Journal of Finance, 31, 10851100.
Kreps, D., and E. Porteus (1978): Temporal Resolution of Uncertainty and
Dynamic Choice Theory, Econometrica, 46, 185200.
Kreps, D. M. (1990): A Course in Microeconomic Theory. Princeton University Press, Princeton, NJ.
Krishnan, C., P. Ritchken, and J. Thomson (2004): Monitoring and
Controlling Bank Risk: Does Risky Debt Help?, Journal of Finance, 60,
343377.
Kyle, A. S. (1985): Continuous Auctions and Insider Trading, Econometrica, 53, 13151335.
(1989): Informed Speculation with Imperfect Competition, Review
of Economic Studies, 56, 317356.
Lakonishok, J., A. Shleifer, and R. W. Vishny (1994): Contrarian investment, extrapolation and risk, Journal of Finance, 49, 15411578.
Lando, D. (1998): Cox Processes and Credit-Risky Securities, Review of
Derivatives Research, 2, 99120.
Langeteig, T. C. (1980): A Multivariate Model of the Term Structure,
Journal of Finance, 25, 7197.
584
BIBLIOGRAPHY
Leippold, M., and L. Wu (2002): Asset Pricing under the Quadratic Class,
Journal of Financial and Quantitative Analysis, 37(2), 271295.
Lettau, M., and S. Ludvigson (2001): Resurrecting the (C)CAPM: A
Cross-Sectional Test When Risk Premia Are Time-Varying, Journal of Political Economy, 109, 12381287.
Lettau, M., and H. Uhlig (2000): Can Habit Formation Be Reconciled with
Business Cycle Facts?, Review of Economic Dynamics, 3, 7999.
Lintner, J. (1965): The Valuation of Risky Assets and the Selection of Risky
Investments in Stock Portfolios and Capital Budgets, Review of Economics
and Statistics, 47, 1337.
Litterman, R., and J. Scheinkman (1988): Common Factors Aecting
Bond Returns, Journal of Fixed Income, 1, 5461.
Lo, A. W. (1988): Maximum Likelihood Estimation of Generalized Ito
Processes with Discretely Sampled Data, Econometric Theory, 4, 231247.
Longstaff, F. (1989): A Nonlinear General Equilibrium Model of the Term
Structure of Interest Rates, Journal of Financial Economics, 23, 195224.
Longstaff, F., and E. Schwartz (1995): A Simple Approach to Valuing
Risky Fixed and Floating Rate Debt, Journal of Finance, 50, 789819.
Lord, C. G., L. Ross, and M. R. Lepper (1979): Biased assimilation and
attitude polarization: the eects of prior theories on subsequently considered
evidence, Journal of Personality and Social Psychology, 37(11), 20982109.
Lucas, R. E. (1972): Expectations and the Neutrality of Money, Journal of
Economic Theory, 4, 103124.
BIBLIOGRAPHY
585
Lucas, Jr., R. E. (1976): Econometric Policy Evaluation: A Critique, in

The Phillips Curve and Labor Markets, vol. 1 of Carnegie Rochester Conference Series on Public Policy, pp. 1946. North-Holland, Amsterdam.
(1978): Asset Prices in an Exchange Economy, Econometrica, 46,
142614.
(1987): Models of Business Cycles. Basil-Blackwell, New York.
Machina, M. J. (1987): Choice Under Uncertainty: Problems Solved and
Unsolved, Journal of Economic Perspectives, 1(1), 121154.
Madan, D., and H. Unal (1998): Pricing the Risks of Default, Review of
Derivatives Research, 2, 121160.
Maes, K. (2003): Modeling the Term Structure of Interest Rates: Where Do
We Stand?, KULeuven University and University of Amsterdam working
paper.
Marcus, A. J., and I. Shaked (1984): The Valuation of FDIC Deposit
Insurance Using Option-Pricing Estimates, Journal of Money, Credit and
Banking, 16, 446460.
Markowitz, H. M. (1952): Portfolio Selection, Journal of Finance, 7(1),
7791.
Marshall, A. (1920): Principles of Economics: An Introductory Volume.
Macmillan, London.
Mas-Colell, A., M. D. Whinston, and J. R. Green (1995): Microeconomic Theory. Oxford University Press, first edn.
McDonald, R. L. (2002): Derivatives Markets. Pearson Addison Wesley,
Reading, MA, first edn.
586
BIBLIOGRAPHY
Mehra, R., and E. Prescott (1985): The Equity Premium: A Puzzle,

Journal of Monetary Economics, 15, 145161.
Menzly, L., T. Santos, and P. Veronesi (2001): Habit Formation and
the Cross Section of Stock Returns, Center for Research in Security Prices
Working paper no.534, University of Chicago.
Merton, R. C. (1969): Lifetime Portfolio Selection under Uncertainty: The
Continuous-Time Case, Review of Economics and Statistics, 51(3), 24757.
(1971): Optimum Consumption and Portfolio Rules in a ContinuousTime Model, Journal of Economic Theory, 3(4), 373413.
(1972): An Analytical Derivation of the Ecient Portfolio Frontier,
Journal of Financial and Quantitative Analysis, 7, 18511872.
(1973a): An Intertemporal Capital Asset Pricing Model, Econometrica, 41(5), 867887.
(1973b): Theory of Rational Option Pricing, Bell Journal of Economics and Management Science, 4, 141183.
(1974): On the Pricing of Corporate Debt: The Risk Structure of
Interest Rates, Journal of Finance, 29, 449470.
(1976): Option Pricing when the Underlying Stock Returns are Discontinuous, Journal of Financial Economics, 3, 125144.
(1980): On Estimating the Expected Return on the Market: An
Exploratory Investigation, Journal of Financial Economics, 8, 323361.
(1992): Continuous Time Finance. Basil Blackwell, Cambridge, MA.
Mossin, J. (1966): Equilibrium in a Capital Asset Market, Econometrica,
34, 768783.
BIBLIOGRAPHY
587
(1968): Optimal Multiperiod Portfolio Policies, Journal of Business,

41(2), 215229.
Muth, J. F. (1961): Rational Expectations and the Theory of Price Movements, Econometrica, 29(3), 315335.
Myers, S. C. (1968): A Time-State Preference Model of Security Valuation,
Journal of Financial and Quantitative Analysis, 3(1), 134.
Naik, V., and M. Lee (1990): General Equilibrium Pricing of Options and the
Market Portfolio with Discontinuous Returns, Review of Financial Studies,
3, 493521.
Naik, V., and M. H. Lee (1997): Yield Curve Dynamics with Discrete
Shifts in Economic Regimes: Theory and Estimation, University of British
Columbia working paper.
Neftci, S. N. (1996): Introduction to the Mathematics of Financial Derivatives. Academic PressThe Free Press, San Diego, CA.
Obstfeld, M. (1994): Risk-Taking, Global Diversification, and Growth,
American Economic Review, 84, 13101329.
OHara, M. (1995): Market Microstructure Theory. Blackwell, Oxford.
Pennacchi, G. G. (1991): Identifying the Dynamics of Real Interest Rates
and Inflation: Evidence Using Survey Data, Review of Financial Studies, 4,
5386.
(2005): Risk-Based Capital Standards, Deposit Insurance, and Procyclicality, Journal of Financial Intermediation, 14, 432465.
588
BIBLIOGRAPHY
Piazzesi, M. (2005a): Ane Term Structure Models, in Handbook of Financial Econometrics, ed. by Y. Ait-Sahalia, and L. Hansen. Elsevier, Amsterdam.
(2005b): Bond Yields and the Federal Reserve, Journal of Political
Economy, 113, 311344.
Pliska, S. (1986): A Stochastic Calculus Model of Continuous Trading: Optimal Portfolios, Mathematics of Operations Research, 11, 371382.
Poterba, J. M., and L. H. Summers (1988): Mean Reversion in Stock
Returns: Evidence and Implications, Journal of Financial Economics, 22(1),
2759.
Pratt, J. W. (1964): Risk-Aversion in the Small and in the Large, Econometrica, 32(1), 12236.
Rabin, M., and R. H. Thaler (2001): Anomalies: Risk Aversion, Journal
of Economic Perspectives, 15(1), 219232.
Ramsey, F. P. (1928): A Mathematical Theory of Saving, Economic Journal, 38(4), 543559.
Rebonato, R. (2004): Interest-Rate Term Structure Pricing Models: A Review, Proceedings of the Royal Society of London, 460, 667728.
Richards, A. J. (1997): Winner-Loser Reversals in National Stock Market
Indices: Can They be Explained?, Journal of Finance, 52(5), 21292144.
Ritchken, P., and L. Sankarasubramanian (1995): Volatility Structures
of Forward Rates and the Dynamics of the Term Structure, Mathematical
Finance, 5, 5572.
BIBLIOGRAPHY
589
Ritter, J. R. (1991): The Long-Run Performance of Initial Public Oerings,

Journal of Finance, 46(1), 327.
Rogers, C. G., and D. Talay (1997): Numerical Methods in Finance. Cambridge University Press, Cambridge, UK.
Roll, R. W. (1977): A Critique of the Asset Pricing Theorys Tests, Journal
of Financial Economics, 4, 129176.
Ross, S. A. (1976): The Arbitrage Theory of Capital Asset Pricing, Journal
of Economic Theory, 13, 341360.
Runggaldier, W. (2003): Jump-Diusion Models, in Finance, Handbook of
Heavy Tailed Distributions in Finance, ed. by S. T. Rachev, vol. 1, chap. 5,
pp. 169209. Elsevier, Amsterdam.
Ryder, H. E., and G. M. Heal (1973): Optimum Growth with Intertemporally Dependent Preferences, Review of Economic Studies, 40, 143.
Samuelson, P. A. (1965): Rational Theory of Warrant Pricing, Industrial
Management Review, 6, 1331.
(1969): Lifetime Portfolio Selection by Dynamic Stochastic Programming, Review of Economics and Statistics, 51(3), 239246.
Sandroni, A. (2000): Do markets favor agents able to make accurate predictions?, Econometrica, 68, 13031342.
Santa-Clara, P., and D. Sornette (2001): The Dynamics of the Forward Interest Rate Curve with Stochastic String Shocks, Review of Financial
Studies, 14(1), 149185.
Santos, M. S., and M. Woodford (1997): Rational Asset Pricing Bubbles,
Econometrica, 65(1), 1957.
590
BIBLIOGRAPHY
Scheinkman, J. A., and W. Xiong (2003): Overconfidence and Speculative

Bubbles, Journal of Political Economy, 111(6), 11831219.
Schroder, M., and C. Skiadas (2002): An Isomorphism Between Asset
Pricing Models With and Without Linear Habit Formation, Review of Financial Studies, 15(4), 11891221.
Sharpe, W. F. (1964): Capital Asset Prices: A Theory of Market Equilibrium
Under Conditions of Risk, Journal of Finance, 19, 425442.
Shiller, R. J. (1982): Consumption, Asset Markets, and Macroeconomic
Fluctuations, Carnegie Rochester Conference Series on Public Policy, 17,
203238.
Shimko, D. C. (1992): Finance in Continuous Time: A Primer. Kolb Publishing, Miami, FL.
Shleifer, A., and R. Vishny (1997): The Limits to Arbitrage, Journal of
Finance, 52(1), 3555.
Siegel, J. J., and R. H. Thaler (1997): Anomalies: The Equity Premium
Puzzle, Journal of Economic Perspectives, 11, 191200.
Spiegel, M., and A. Subrahmanyam (1992): Informed Speculation and
Hedging in a Noncompetitive Securities Market, Review of Financial Studies,
5, 307329.
Sun, T.-S. (1992): Real and Nominal Interest Rates: A Discrete-Time Model
and Its Continuous-Time Limit, Review of Financial Studies, 5, 581611.
Sundaresan, S. M. (1989): Intertemporally Dependent Preferences and the
Volatility of Consumption and Wealth, Review of Financial Studies, 2(1),
7389.
BIBLIOGRAPHY
591
Thaler, R., and E. J. Johnson (1990): Gambling with the House Money
and Trying to Break Even: The Eects of Prior Outcomes on Risky Choice,
Management Science, 36, 643660.
Tirole, J. (1982): On the Possibility of Speculation Under Rational Expectations, Econometrica, 50(5), 11631181.
Tobin, J. (1958): Liquidity Preference as Behavior Towards Risk, Review of
Economic Studies, 67, 6586.
Treynor, J. L. (1961): Toward a Theory of Market Value of Risky Assets,
unpublished manuscript.
Tuckman, B. (2002): Fixed Income Securities: Tools for Todays Markets.
John Wiley and Sons, Inc., New York, second edn.
Varian, H. R. (1992): Microeconomic Analysis. W. W. Norton & Company,
third edn.
Vasicek, O. A. (1977): An Equilibrium Characterization of the Term Structure, Journal of Financial Economics, 5(2), 177188.
von Neumann, J., and O. Morgenstern (1944): Theory of Games and
Economic Behavior. Princeton University Press, Princeton, NJ.
Wachter, J. A. (2002): Portfolio and Consumption Decisions under MeanReverting Returns: An Exact Solution for Complete Markets, Journal of
Financial and Quantitative Analysis, 37(1), 6391.
Wang, J. (1993): A Model of Intertemporal Asset Prices Under Asymmetric
Information, Review of Economic Studies, 60, 249282.
Weil, P. (1989): The Equity Premium Puzzle and the Risk-Free Rate Puzzle,
Journal of Monetary Economics, 24, 401421.
592
BIBLIOGRAPHY
Welch, I. (2000): Views of Financial Economists on the Equity Premium and

on Professional Controversies, Journal of Business, 73(4), 501537.
Yan, H. (2001): Dynamic Models of the Term Structure, Financial Analysts
Journal, 57(4), 6076.
Zhou, C. (2001): The Term Structure of Credit Spreads with Jump Risk,
Journal of Banking and Finance, 25, 20152040.

Theory of Asset Pricing - Pennacchi

Uploaded by

Copyright:

Available Formats

Theory of Asset Pricing - Pennacchi

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Theory of Asset Pricing - Pennacchi

Uploaded by

Copyright:

Available Formats

Expected Utility and Risk

Economists refer to these two bases

of prices as investor "tastes" and the economys "technologies" for generating

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

These findings have motivated a search for improved models

Theoretical innovations both within and outside the

Preferences when Returns are Uncertain

Economists typically analyze the price of a good or service by modeling the

The main distinction between assets is

Later chapters will generalize the situation

to consider an individuals choices over multiple periods among assets paying

measure of the attractiveness of an asset is the average or expected value of

Suppose an asset oers a single random payo at a particular

1.1. PREFERENCES WHEN RETURNS ARE UNCERTAIN

assets expected payos?

This valuation concept was the prevailing wisdom

He did it by posing the following problem that

became known as the St. Petersberg Paradox:

Peter tosses a coin and continues to do so until it should land "heads"

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

where Ui is the utility associated with payo xi . Moreover, he hypothesized

1.1. PREFERENCES WHEN RETURNS ARE UNCERTAIN

dierent lottery is characterized by another set of probabilities, for example,

P = {p1 , ..., pn }. Let , , and denote preference and indierence between

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

primary area of microeconomics analyzes a consumers optimal choice of multiple goods

1.1. PREFERENCES WHEN RETURNS ARE UNCERTAIN

However, experimental evidence finds

some systematic violations of this independence axiom, making it a questionable

Machina (Machina 1987) summa-

rizes violations of the independence axiom and reviews alternative approaches

In spite of these deficiencies, von Neumann -

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

proach to modeling investor preferences, though research exploring alternative

Therefore, it may be that

ei ej even if the monetary payment corresponding to xi was less than that

1.1. PREFERENCES WHEN RETURNS ARE UNCERTAIN

p1 e1 + ... + pi1 ei1 + pi [Ui en + (1 Ui )e1 ]

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

pi Ui . Thus, we see that lottery P is equivalent to a

compound lottery consisting of a probability of obtaining elementary lottery

is, Ui = U (xi ) = xi . This is a monotonically increasing, concave function of

1.1. PREFERENCES WHEN RETURNS ARE UNCERTAIN

x, which, here, is assumed to be simply a monetary amount (in units of ducats).

Hence we can conclude that the St.

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

Risk Aversion and Risk Premia

As mentioned in the previous section, Daniel Bernoulli proposed that utility

By risk averse we mean that

11 When the random payo, x

1.2. RISK AVERSION AND RISK PREMIA

which implies 1 /2 = (1 p) /p, or, solving for p, p = 2 / (1 2 ).

course since 0 < p < 1, 1 and 2 are of opposite signs.

if it is not accepted, expected utility is given by E [U (W )] = U (W ). Thus, an

CHAPTER 1. EXPECTED UTILITY AND RISK AVERSION

since p1 + (1 p)2 = 0 by the assumption that the lottery is fair. Re-writing

U (W + p1 + (1 p)2 ) > pU (W + 1 ) + (1 p)U (W + 2 )

which is the definition of U being a concave function. A function is concave

which is the desired result.