The Single Index Model
The Single Index Model
The Single Index Model
• Stock return Ri
• 15 .
• …. . error
• ....
• .…..
• …..
• . …. .
• . error
• . . 2
• . . .
• .. -4 Market return 12 Rm
.
• The return on a stock can be written as:
• Ri = ai + iRm
• Where ai = i + ei
• ei = Random Errors
• E(ai) = i
• Thus, Ri = i + iRm + ei
• .
• . . .
• . . . .
• . . . Residual for stock j, ej
• .
• Two Assumptions:
• E(eiej) = 0
• By construction
• Mean of ei = E (ei) = 0
• By Assumption:
• E[ei {Rm – E(Rm)}] = 0
• E(eiej ) = 0 [i.e., Securities only related to through common response to
market].
• By Definitions:
• Variance of ei = ei2
• Variance of Rm = m2
• We can derive the expected return,
standard deviation and covariance as
follows:
• We know,
• Ri = i + iRm + ei
• Ri = [i + iRm + ei]
• 1/n Ri = 1/n [i + iRm + ei ]
• E(Ri) = [i + iE(Rm) + 1/nei]
• E(Ri) = i + iE(Rm)
i2 = i2 m2 + ei2
• We know
• i2 = 1/n [Ri – E(Ri)]2
• i2 = 1/n [i + iRm + ei – i – iE(Rm)]2
• i2 = 1/n [i {Rm – E(Rm)} + ei]2
• i2 = 1/n [i2 {Rm – E(Rm)}2 + 2 i {Rm – E(Rm)}ei + ei2]
• i2 = i2 m2 + 2i 1/n {Rm – E(Rm)}ei + 1/n ei2
• i2 = i2 m2 + ei2 [Since 1/n {Rm – E(Rm)}ei = 0]
ij = ijm2
• ij = 1/n {(Ri – E(Ri)}{Rj – E(Rj)}
• ij = 1/n [i {Rm – E(Rm)} + ei] [j {Rm – E(Rm)} + ej]
• ij = ijm2 + i 1/n {Rm – E(Rm)} ej + j 1/n {Rm –
E(Rm)} ei + 1/n eiej
• ij = ijm2
• Since, the last three terms are zero by assumption
• CHARACTERISTICS OF THE SINGLE-INDEX MODEL
• Portfolio Beta
• βp = Σ Xi βi
• Portfolio Alpha
• p = Σ Xi i
R P X i Ri
Ri i i R m
R P X i i i Rm
R P X i i X i i R m
RP P P Rm
( If R P R m ; P 0; and P 1; The portfolio is the Market Portfolio)
X i i P and X i i P
• The Beta on the market is 1 and stocks
are thought of as being more or less risky
than the market, according to whether the
Beta is larger or smaller than 1
• β=1
• β>1
• β < 1.
β=1
• Ri
• Rm
β > 1 [Aggressive stock]
• Ri
• Rm
β < 1 [Defensive stock]
• Ri
• Rm
• We have shown that we can split the variance of the
return on a security or portfolio into two parts:
N N
X X i X j ij
2
P i
2
i
2
i 1 j 1
i j
X X i X j i j m2
N N
2 2 2 2 2
P i i m ei
i 1 j 1
i j
N N N
X X X i X j i j m2
2
P i
2
i
2 2
m i
2 2
ei
i 1 i 1 j 1
i j
N N
X i X j i j m2 X i2 ei2
2
P
i 1 j 1
N N
X i i X j j m2 X i2 ei2
2
P
i 1 j 1
P2 P2 m2 X i2 ei2
N
• ep2 =
i 1
X 2σ 2
i ei
X
2
P
2
P
2
m i
2 2
ei
1 2
2
P
2
P
2
m
ei
N N
1 2
P P m ei
2 2 2
N
2
• If N is very large, then 1/N = 0 and Residual risk
ei
i.e., unsystematic risk is diversified away.
• Therefore,
2
P
2
P
2
m
P P m
N
P m X i i
i 1
• In an efficient portfolio total portfolio risk is only
systematic risk.
• Since, m is the same, regardless of which stock
we examine, the measure of the contribution of a
security to the risk of a large portfolio is i.
• RA > RB ?
• RA < RB ?
• If equal amount is invested in each securities what is the
portfolio variance P2?
{X-E(X)} {Y-E(Y)}
X Y {X-E(X)} {Y-E(Y)} {X-E(X)}2
E(X) = E(Y) =
0.0553
0.3272
1
n
X X Y Y
0.14458
1.032
1
n
X X
2
0.14001
Ri R m
R i R m 0.3272 (1.03)(0.0553) 0.27
ei = Yi – ( + Xi)
{Y-E(Y)} {Y-E(Y)}2
-0.1708 .02917264
-0.2111 .04456321
0.2028 .04112784
-0.0328 .00107584
-0.1995 .03980025
-0.0169 .00028561
0.2083 .04338889
0.2736 .07485696
-0.0536 .00287296
Σ {Y-E(Y)}2 = .277142
Var of Y = 1/n Σ {Y-E(Y)}2
= (1/9) (.277142) = 0.03079
Relation between r (correlation between individual
security’s return and market return) and unsystematic risk
• σ i2 = β i2 σ m2 + σei2
• 1 = (β i2 σ m2 / σ i2 ) + (σei2/ σ i2 )
• 1 = R2 + (un systematic Risk/Total Risk)
• 1= R2 + .0142 / .03074
• Or, R2 = 1- .4619 = .538
• r = +0.734