Liquidity Analysis

11.1 What Is Liquidity Risk?

11.2 A Simple Balance Sheet Model of Liquidity
11.3 Balance Sheet Liquidity Characteristics
11.4 Managing Liquidity Risk: Perceptions and Reality
11.5 Measuring Liquidation Cost Risk
11.6 Controlling Liquidity Risk and the Safety Zone
11.7 Diversification of Funding Sources
11.8 The Contingency Planning Process
11.9 Summary: Liquidity Risk and Shareholder Value
Appendix 1lA: Just What Is the "Liquidity Premium" Anyway?
Appendix 1lB: The Cost of Bankruptcy and Liquidation Costs
Appendix 11C: Liquidity in the Banking Industry

233
234 Chapter 11

Chapter 10 analyzed the impact of credit risk, foreign exchange risk, and
stock market risk on the cost of debt and equity for a bank. In every case,
there is a direct link between the level of each type of risk and the bank's
cost of funding, which in tum means that all of these risks have a direct
impact on the risks and techniques that are more traditionally called "asset
and liability management." There is another "traditional" asset and liability
management topic: liquidity risk. It is safe to say that liquidity risk analy
sis is one of the least understood of all ALM areas.
Why is liquidity risk so misunderstood? Liquidity risk is misunder
stood because so many bankers have mistaken liquidity risk for the
disease itself, when in reality it is only the symptom of a more funda
mental problem. Failure to distinguish between the symptom and the dis
ease can get a doctor sued for malpractice. This chapter should protect
bankers against a similar kind of problem.
Liquidity risk includes four basic kinds of analysis:
■ Understanding how to measure the liquidity profile of a bank balance
sheet;
■ Understanding the impacts of the cost of bankruptcy and liquidation
costs;
■ Understanding the liquidity safety zone and the funding implications
of these factors; and,
■ Understanding their impacts on shareholder
value. Each of these topics will be discussed in this
chapter.

11.1 WHAT IS LIQUIDITY RISK?

There is no universally accepted definition of liquidity or liquidity risk.
This is, indeed, one of the reasons why it is so poorly understood: it means
different things to different people. A precise definition of liquidity is not
important; however, understanding its potential consequences and manag
ing it well is critical.
Some of the concepts that are usually expressed on this issue are:
"Liquidity is having access to sufficient cash to maintain a
'business as usual' posture at all times."
"Liquidity is the ability to raise (or borrow) funds at a rea
sonable cost at all times."
Liquidity Analysis 235

"Liquidity buys time to work out problems."

The authors tend to favor the last because it emphasizes the point ex
pressed in the introduction to this chapter. Any difficulty in a bank's li
quidity or funding activities is only the symptom that reflects the existence
of some other, more fundamental problem. The "buying time" perspective
reflects the essential role that liquidity planning plays for a bank. It is
often the only buffer between the bank and those that may be over-react
ing to the "disease," whether the underlying problem is real, but over
stated, or imagined.
No mention has been made about accounting or risk-based capital lev
els. Capital plays a special role as the "liability" with the longest maturity
and as the only "liability" where no regular interest payments can trigger
bankruptcy. Liquidity planning includes more than just capital, since "buy
ing time" can be just as important in the short run as it is in the long run.
How are liquidity and capital linked? A bank will have liquidity problems
in the short run if the market believes that capital has been so impaired
that in the long run the bank may not be able to pay off its liabilities. The
two are linked symptoms of a disease somewhere else on the balance
sheet, usually a loss due to high levels of credit risk or interest rate risk.

11.2 A SIMPLE BALANCE SHEET MODEL OF LIQUIDITY

In this section, a simple theoretical model of balance sheet liquidity is
presented. This type of presentation is extremely effective in communicat
ing to senior management and the board of directors. It is also useful for
establishing liquidity policies and operating limits.
In this approach, all assets are allocated to one of two categories: "liq
uid assets" and "illiquid assets." All liabilities are classified as either "sta
ble liabilities" or "volatile liabilities." Schematically, the balance sheet can
then be presented as shown in Figure 11.1.
The focal point of this framework is the concept of "net liquid assets,"
defined as the difference between liquid assets and volatile liabilities.
(This difference can also be referred to as the "liquidity gap.") When de
fined in this manner, there is a natural tendency to strive for a surplus of
liquid assets over volatile funds. As a result, this definition is preferred
over those that would calculate the ratio of liquid assets to volatile liabili
ties (or to total assets). This differencing, or netting, has the favorable
attribute that an excess of volatile funding over liquid assets is a negative
value, rather than a ratio that would result in a value less than 100%.
236 Chapter 11

Figure 11.1
Balance Sheet Liquidity Model

Volatile

Liquid
Net Liquid
Assets

t Stable

Illiquid

ASSETS LIABILITIES

Psychologically, a 95% ratio of liquid assets to volatile funds may not

sound too bad. But a negative net liquid assets value is usually taken much
more seriously by ALCOs and boards of directors, as well it should.
This concept emphasizes that banks should understand when they are
putting themselves at the risk of not being able to fund their operating cash
needs. Any negative value for net liquid assets may be interpreted as such
a potential circumstance. Simplistically, a negative value here means that
assets readily convertible into cash may not be sufficient to cover the near
term maturities of deposits and debt.
Of course, it is important to establish appropriate criteria for the various
balance sheet components. To use this concept properly, some careful
thought must be given to the precise classifications of balance sheet com
ponents to the "liquid asset" and "stable liability" categories. The next
section will discuss the liquidity characteristics of the main balance sheet
categories. Some rather counterintuitive assessments are encountered!

11.3 BALANCE SHEET LIQUIDITY CHARACTERISTICS

Some balance sheet categories are relatively simple to categorize, while
some are treacherous. The following is a very simplified discussion, in
Liquidity Analysis 237

some cases augmented by utilizing a devil's advocate stand. This is an

extremely serious topic; therefore, all asset and liability managers should
spend some quality time getting acquainted and comfortable with these
types of considerations. Because the consequences of insufficient liquidity
can be so dire, we recommend that when in doub use a conservative
assumption that, if anything, understates net liquid assets.
Many readers will notice that this presentation is very analogous to the
"rate-sensitive assets" and "rate-sensitive liabilities" concept presented in
many discussions of interest rate risk management. In interest rate risk
analyses, the main classification feature is the expected time to the next
repricing or rollover. In liquidity analyses, the main focus is on the effec
tive maturity characteristic. Both techniques require selecting a time
threshold by which items are determined to be sensitive or insensitive,
liquid or illiquid, volatile or stable. For the net liquid assets measure, we
recommend using either a three-month or six-month remaining maturity
criterion. For the following discussion, assume that the threshold selected
is six months.

Asset Accounts
A useful criterion for categorizing the various items on the asset side of
the balance sheet is to ask whether the asset may be liquidated or pledged,
and the bank still maintain a "business as usual" posture to the outside
world. What is "business as usual"? Think of how your customers would
react if your bank suddenly stopped cashing checks, or failed to deliver
good funds at the closing of a mortgage transaction, or reneged on the
requested drawdown of a loan commitment?

Cash and due from banks. Most bankers automatically classify vault
cash, interbank clearing accounts, and required reserve balances at the Fed
as liquid assets. We disagree! These balances are constantly monitored and
decreased to the minimum levels required to carry out normal operations.
No bank leaves material amounts of surplus vault cash, due from balances
or reserve balances. Therefore, these balances usually cannot be substan
tially decreased in the event of a funding problem. By this reasoning, they
are illiquid assets!

Federal funds sold and reverse repo. These are definitely liquid assets. One
possible complication may be correspondent relationships that depend on
your bank for short-term funding. Any balances of this type should be
considered illiquid.
238 Chapter 11

Placements and other money market assets. These are liquid to the
extent they are less than the threshold of six months.

Investment securities. For this analysis, divide the investment portfolio into
I

two important security types: those that can be pledged for repo borrow-
ings in the open markets and those that cannot. Those that are eligible for
repurchase agreement transactions should be entirely classified as liquid
assets, even if they are currently all repoed or otherwise encumbered! (Net
out the repoed securities by classifying their associated repurchase agree
ments as volatile liabilities. Similarly, collateralized borrowings that en
cumber eligible securities should be considered volatile.) For those not
eligible for repo, only those maturing within the six-month threshold may
be considered to be liquid assets.
A similar, yet simpler approach would be to classify all Treasuries and
Agencies as liquid, and all other securities as illiquid. This simplification
is quite useful in conducting peer comparisons where maturities during the
next six months may not be available.

Trading assets. These assets are commonly repoed. As such, they

should be counted as liquid assets. As with Treasury securities in the
investment portfolio, the corresponding repurchase agreements will be
classified as volatile liabilities.

Loans. Our general attitude is that all loans are illiquid. The standard
argu ment against this extreme position is that all balances scheduled to
mature within the six-month liquidity threshold should be considered
liquid. How ever, treating all near-term maturities as liquid would violate
the "business as usual" criterion. Implying that the funds generated from all
scheduled maturities are available for purposes other than lending is
tantamount to shutting down the bank's entire array of loan origination
activities. Com mercial and mortgage loan pipelines are often much
longer than one month in duration. They cannot be shut down on short
notice without se vere disruption. Indeed, they really cannot ever be shut
down without sending shock waves throughout the bank's customer base.
Likewise, consumer lending is being geared more and more toward
cross-selling and relationship building. To cease accepting consumer loan
applications would be just as shocking to this, the bank's most prized cus
tomer base. Our advice is simple: Do not count on decreasing loan origina
tion volumes as a liquidity source.
A special argument is sometimes made about mortgages or some com
mercial loans that qualify as collateral for Federal Reserve discount win-
Liquidity Analysis 239

dow or Federal Home Loan Bank borrowing purposes. However, this ap

proach would violate our "business as usual" criterion. Drawing down
"lender of last resort" credit lines or backup lines would send clear signals
to those entities that something is amiss. This distinction will be expanded
in a later section.
The final special circumstance of loan securitization will be discussed in
Chapter 12.

All other assets. These are generally considered to be illiquid assets.

Liability Accounts
Again, the liquidity threshold of six months will be an important factor in
these evaluations. However, for retail deposit products, the law of large
numbers and expected portfolio characteristics will also be considered.

Demand and savings deposits. Many banks classify all retail

"nonma turity" products as stable. These are products with no contractual
maturity date. However, experience shows that there will be some
deterioration of a retail deposit base as adverse publicity and downgradings
progress. Sev eral banks have analyzed the core deposit runoff experience of
large re gional banks that ultimately failed. Public regulatory reports can be
examined to determine the extent of retail deposit loss as a function of
rating agency downgrading. The idea would be to reflect the runoff seen at
around the "CID" credit rating stage. The percentages found by deposit
product type would then be considered volatile liabilities.
If this is not practical, use the results from the seasonality analyses de
scribed in Chapter 9. That analysis, which is needed for interest rate risk
analyses as well as for transfer pricing (see Chapter 14), splits demand
deposit balances into core and volatile. The volatile percentage found there
can be applied to the net liquid assets calculation. Under no circumstances
should the bank blindly assume that all retail deposits are stable.

Retail certificates of deposit. Those balances that have more than

six months remaining maturities are considered stable. As with the
demand and savings accounts, it is reasonable to assume that a large
percentage of the near-term maturities will roll over. Indeed, many banks
automatically roll over maturing retail CDs, unless specific instructions are
received oth erwise. If the downgrade and runoff study described earlier is
done, then appropriate assumptions can be established. Without this
data, then as-
240 Chapter 11

sumptions consistent with the core versus volatile split for other retail ac
counts can be utilized.

Commercial demand deposits. These are usually related to long-established

relationships. However, many corporations and businesses have strict
credit rating minimum criteria for their relationship banks. Therefore, as a
bank is downgraded, many strong business relationships may be lost. Also,
corporate treasurers have absolutely no incentive to "stand by their bank"
if their company's ability to conduct financial transactions is threatened.
Therefore, the runoff estimates used for commercial DDA should be sig
nificantly higher than for retail DDA. (This should also be true for the
volatile percentage determined for interest rate risk management purposes.
As above, in the absence of other data, the split used in gap or duration
analyses can be pressed into service here also.)

Large certificates of deposit Only balances beyond the six-month threshold

in remaining maturity can be counted as stable liabilities.

Foreign office time deposits. Balances beyond six months remaining matur
ity are "stable liabilities."

Federal funds purchased and repurchase agreements. As discussed for in

vestment securities, these should all be counted as volatile liabilities.

Long-term debt. Any balances maturing within the six-month limit are
volatile. Also, any debt beyond six months that requires collateral in the
form of repo-eligible investment securities should be considered volatile.

Equity and other liabilities: These are stable liabilities.

This concludes our first liquidity tour of the balance sheet. Of course,
no single definition or equation can be adequate for a subject as diverse
and complex as this. The next several sections will dig through the layers
and nuances of the liquidity management process.

11.4 MANAGING LIQUIDITY RISK: PERCEPTIONS AND

REALITY
The asset and liability manager has to deal with two types of liquidity
risks. One risk is the correct perception by the market that the ALM man-
 Liquidity Analysis 241

ager's bank is in financial trouble. A less common but still possib]e risk is
the incorrect perception by the market that a bank has financial problems.
The bank has to deal with both risks in a way that minimizes the expected
liquidation costs and bankruptcy costs. (A description of the theoretical
effects of liquidation and bankruptcy costs is presented in Appendix 10B.)
These closely related but somewhat different problems will be addressed
in reverse order.
The asset and liability manager is often the first to receive notice that
the market does not think much of his institution. While there is plenty of
sympathy for the underlying validity of the "efficient markets" school of
financial theory, many banking colleagues have had to deal with the very
real consequences of a market that does not want to lend money to their
bank. This phenomenon has been observed in many markets, but New
York and Hong Kong colleagues seem to be the most experienced in it.
The first warning is funding or derivative products counterparties "turning
down the name" of the manager's bank. A more easily observable signal is
the resulting increase in funding costs that the bank faces on a daily basis
relative to both the risk-free rate (U.S. Treasury rates in an American con
text) and to the funding costs faced by peer group banks. As mentioned in
Chapter 10, this means the "implied volatility" of bank assets as viewed
by the market has increased.
What should the asset and liability manager do to deal with the risk of
market misperception? There are four aspects to the ALM manager's du ties:
■ Constant monitoring and reporting of market perceptions to the asset
and liability committee in order to ensure that market signals are not
missed.
■ Ensuring that the financial condition of the bank is always clearly
and accurately communicated to the investor community. Any hint of
"stonewalling" is implicit confirmation of problems and may lead the
market to believe that the situation is worse than it really is.
■ Measuring the liquidation cost risk inherent in the balance sheet.
(The term "liquidity risk" is avoided here to emphasize the fact that
liquidity risk is derived from other risks and that the real concern of
the asset and liability manager is that the bank pays costs to liquidate
assets that could have been avoided.)
■ Structuring the bank's balance sheet so that the bank is in the liquid
ity safety zone.
242 Chapter 11

The first two points cannot be overemphasized, but they are so self-evi
dent it is unnecessary to expand on them. The other two points need to be
addressed in some detail. The emphasis is on managing the expected value
of liquidation costs and postponing their payment for as long as possible.

11.5 MEASURING LIQUIDATION COST RISK

Unlike the risk-based capital adequacy guidelines, measurement of liquida
tion cost risk can be done with meaning and precision. There are three
elements to this measurement: the costs of liquidation as a function of
time, the liquidity gap for the bank, and liquidity duration. The last con cept
measures the change in the equity value of the bank that would result from
a change in the implied volatility (increased risk) inherent in the bank
balance sheet. Each of these measurement tools will be discussed in order.

Liquidation costs. The major elements of a bank's balance sheet can be

converted to cash with varying degrees of efficiency depending on what
the asset is and how long one has to convert the asset to cash. The costs of
doing so are a function of brokerage and investment banking fees and the
basic bid-offered spread in the market for the asset involved. The fees and
bid-offered spreads are generally dominated by the ease or difficulty of
determining the credit risk of the asset involved. In the case of short-term
government securities, this risk is very literally close to zero. When it
comes to consumer auto loans, it is almost literally impossible for anyone
other than the bank to assess the risk of the assets involved because no
third party has access to enough information to make the credit judgment.
If the bank needs an investor to make a firm bid in a very short period of
time, the bank can get one, but it will be a bid very heavily discounted in
recognition of the lack of information on the risk. As stated in Chapter 12,
there is a serious "moral hazard" risk in the sale of some assets: there is
strong (short-term) incentive for the bank to sell its worst assets when
sorting through its own consumer loan portfolio if the bank feels investors
cannot measure the risk themselves. Investors recognize this risk and dis
count appropriately.
The costs of liquidating various categories of assets will depend on cur
rent market conditions, the institution itself, the time frame for liquidation,
and the asset class. The authors believe most bankers would come up with
a table of liquidation costs very similar to Table 11.1.
Liquidity Analysis 243

Table 11.1
The Cost of Liquidating Typical U.S. Bank Assets
Transactions Costs as a Percentage of Market Value

Time Available for Conversion

One One One One
Asset Hour Dai Week Months
U.S. Treasury Bills 0.031 0.031 0.031 0.000
U.S. Government Bonds 0.063 0.031 0.031 0.031
GNMA and FNMA Securities 0.125 0.063 0.063 0.063
Federal Funds Sold 0.125 0.000 0.000 0.000
Corporate Bonds 0.500 0.250 0.250 0.250
Corporate Loans NA NA 1.000 0.500
Whole Mortgage Loans NA NA NA 1.000
Securitized Consumer Loans NA NA NA 2.000
Other Consumer Loans NA NA NA NA

Note: NA means that liquidation is not available.

It goes without saying that there would be no transactions costs if the

bank had the time to allow assets to mature and collect the underlying
principal. Using charts like this, the bank can calculate the amount of cash
and the amount of liquidation costs realized over a given time period. Un
like the BIS guidelines, these liquidation cost "haircuts" are readily ob
servable to day-to-day market participants.

Liquidity gaps (also known as net liquid assets). Perhaps the best tool
for supplementing measures of liquidation costs is a tool that was of only
limited use in an interest rate risk context. That tool is the "gap" analysis,
as described in Sections 11.2 and 11.3. When using the gap concept for
interest rate risk analysis, assets and liabilities are classified in time cate
gories according to their interest rate reset periods. In the liquidation cost
context, assets and liabilities are classified by maturity. Using this classifi
cation, management can view the amount of cash that is made available
with the passage of time without the payment of any liquidation costs.
244 Chapter 11

Liquidity duration. The discussion of interest rate risk examined the

con cept of duration and interest rate elasticity at length. In that context,
both tools measured the impact on shareholders' equity if the risk-free rates
increased by 1% for all maturities. In the liquidation risk context, the same
tools can be used. In this application, rates would rise not because the
risk-free rates rise, but because the implied volatility of the bank's assets
increases, driving up the marginal cost of funds to the bank through an
increase in the bank's liquidity premium. While the basic calculation
methods remain the same, there are a few differences in application in the
use of liquidation risk elasticity. There are three steps in the process:
I. Calculate the present value of assets and liabilities using bank cost of
funds rates as the discount rates.
This mark-to-market calculation is different from the interest rate risk
case, where the discount rates are the market yields that apply to the
market for that particular asset (Treasury yields are used for Treasury
securities, for example). If the resulting values are below true market
values by more than the cost of liquidating the assets, it is cheaper to
sell the assets than to keep them in a liquidity crisis.
2. Measure the change in the market value of equity from a change in
the bank's cost of funds due to an increase in the risk premium paid
by the bank to raise money.
Like the interest rate risk usage of duration, we are implicitly assum
ing a "parallel shift" in the bank's risk premium. In other words,
assume that the spread over the risk-free rate increases by the same
amount at every maturity. If the LRE (liquidity risk elasticity) is
zero, the bank has zero liquidity risk by this measure. If the LRE is
sharply negative, it will pay the bank to shorten the maturity of its
assets and lengthen the maturity of its liabilities, increasing liquidity,
in order to reduce the amount of liquidity risk.
3. Take management action to control liquidity risk.
These management actions are the subject of the next section.
Clearly, the LRE concept goes a significant step beyond the net liquid
assets, or liquidity gap, approach. The remainder of this chapter will focus
back on the net liquid assets idea. The LRE measure is very useful as a
tool of strategic balance sheet management. It also serves to illustrate that
the analytical concepts and tools described in earlier chapters for interest
rate risk management are also applicable in other risk dimensions.
Liquidity Analysis 245

11.6 CONTROLLING LIQUIDITY RISK AND THE SAFETY ZONE

Once management has measured the amount of liquidity risk, the practical
question is simple: what should be done about it? The true objective of
liquidity risk management is the same when the bank is really in financial
difficulties and when the market just thinks it is. The objective is simple:
to buy time. This was asserted in Section 11.1. However, now we can
more fully explain why this is important.
"Buying time" reduces liquidation costs because, as Table 11.1 shows,
the more time management has to sell assets, the smaller will be the trans
action costs that management pays in the process. This is particularly im
portant when the market's perception of financial difficulties is incorrect.
Buying time allows management to prove its case that nothing is wrong. in
addition to allowing the treasury staff time to generate cash from less liq
uid assets at the smallest possible liquidation costs. As Hong Kong banks
have proven over and over again, the best remedy for a run on deposits is
to pay every depositor who wants his money until depositors realize there
is no need to make a withdrawal in the first place.
When the institution truly is in difficulty, buying time reduces liquida
tion costs to a minimum and increases the possibility that the bank's fun
damental problem resolves itself. One concrete example is the U.S.
savings and loan crisis. If a savings and loan institution funded a 30-year
fixed rate loan with 10-year fixed rate money, the institution might be
bankrupt on a mark-to-market basis if rates rose dramatically after one
year. Fortunately for the savings and loan association, however, it would
have nine years in which it could wait (and hope) for a drop in rates to
save the institution. As stated in the interest rate risk sections, it is better
not to depend on good fortune for solvency. Nonetheless, a conservative
posture from a liquidity risk point of view provides a second line of de
fense against many problems. Even postponing the inevitable sometimes
has value.
What is the liquidity risk counterpart of the safety zone that was dis
cussed in an interest rate context? The boundaries of the liquidity risk
safety zone are fairly wide. One boundary of the safety zone is a balance
sheet where the liquidity gap is positive at every maturity threshold. That
is, the cash flow from liquid assets is always greater than the amount of
volatile liabilities coming due. This means the bank will never have to pay
liquidation costs unless the bank is truly bankrupt. Even if it is bankrupt,
liquidation costs will be minimized.
The other boundary of the liquidity risk safety zone is less obvious and
much more dependent on the institutional setting in which the bank oper-
246 Chapter 11

ates. The bank should always be able to generate enough cash on its own
to survive a market misperception of a crisis long enough for either (a) the
government to come to the aid of the bank or (b) the crisis to end natu
rally, whichever is shorter. In the United States, most large banks would
want to be able to survive based on their own resources for at least two
weeks before they are forced to borrow from the Federal Reserve. In addi
tion, they would want subsequent cash flow to be strong enough to pay
down Federal Reserve borrowings quickly enough to avoid lasting govern
ment control of the institution. In the case of a more laissez-faire market
like Hong Kong, the bank's financial posture would need to be much more
liquid, since it is much less likely that a rescuer would be as benevolent at
the U.S. government generally has been to bank shareholders.
These two extremes bracket the asset and liability management structure
that defines the safety zone:
■ On the "most liquid" side of the safety zone, the bank can meet all
maturing volatile liabilities with the proceeds of liquid assets.
■ On the "least liquid" side of the safety zone, the bank can generate
only enough cash to stay afloat until the lifeguard arrives. This would
be one to two to four weeks in a U.S. context and more in less be nign
environments.

11.7 DIVERSIFICATION OF FUNDING SOURCES

Apart from structuring the maturity of assets and liabilities, the bank has to
emphasize the diversification of funding supplies in the same way that
assets are diversified in order to minimize credit risk. The funding sources
for each institution should be managed so that there is minimal impact if
any one institution changes its mind concerning the credit risk of the bank.
If any one institution provides more than 5% of the funding to the bank,
that is cause for concern, particularly if the maturities involved are short
term. While diversification of funding sources is a holy grail of treasury
managers, diversity may have a price: What is a banker to do if the biggest
supplier of funds consistently offers them at a price below market? There
is no easy answer to this question, but we offer one anyway:
The asset and liability manager should seek to retain
the cheapest funding on a risk-adjusted basis for the
bank, regardless of maturity and regardless of source.
Liquidity Analysis 247

ff the interest rate risk characteristics are the only problem with the
funds offered, the bank can change the interest rate risk characteristics
using swaps, options, and futures. If the bank's closest correspondent bank
offers fed funds at 0.25% below market, the bank should take the funds
even if it pushes that source of funds beyond the risk limit (although this
transaction must be approved in advance on a case-by-case basis by the
Asset and Liability Committee or another management authority). The
proper usage of such funds is limited; the bank should lay off the funds
only in such a way that the liquidity risk of the bank is not increased. How
can this be done? In this particular example, the funds should be invested
in fed funds sold at going market rates. From a liquidity gap (net liquid
assets) perspective, there is no additional risk. From a liquidation cost
point of view, there would be no cost of liquidating the asset position.
When should attractive funding be turned down? Only in those cases
when there is no "zero liquidity risk" investment alternative for placing the
funds at a positive spread.

11.8 THE CONTINGENCY PLANNING PROCESS

No asset and liability management process is complete without a detailed
contingency plan that shows a step-by-step process for who does what
when the bank's (or more realistically, bank holding company's) funding
is cut off. Most bankers will have close colleagues who have been forced
to put such plans into action, either due to natural calamities like fires or
earthquakes or due to adverse credit risk at their own institution. In times
of crisis, there is no time to talk about who should do what-or worse yet,
who should have done what. The detailed contingency plan makes the cal
culations of previous sections a practical reality.
The contingency plan is an exercise in "critical path" planning in the
language of an industrial company. Many participants have to take action
simultaneously for each part of the plan to be completed in proper se
quence.
In an American context, almost every bank or bank holding company
contingency plan will specify the following:
1. Senior executive (and two back-ups) responsible for contacting the
local Federal Reserve Bank president and the chairman of the Federal
Reserve Board in Washington D.C.
2. Senior executive responsible for contacting presidents of lead banks
in commercial paper or other back-up lines of credit.
248 Chapter 11

3. Senior executive responsible for contacting senior counterparts at one

or two investment banking firms.
4. Decision process for activation of back-up lines of credit.
5. Decision process for sale of government securities and other liquid
asset portfolios.
6. Decision process for borrowing from Federal Reserve Banlc.
7. Precise instructions on asset sales over these timeframes:
■ One day
■ Two days
■ One week
■ One month
■ Six months
8. Reallocation of systems and consumer lending human resources to
speed asset sales.
9. Lines of authority for sale of subsidiaries or assets to be securitized.
The contingency plan should demonstrate in a concrete way that the
bank has sufficient internal cash flow to survive with no outside assistance
for the minimum time period specified by the bank's senior management.
Typically, the liquidity gap is a very important tool in the process of de
veloping the liquidity plan.
As mentioned in Section 11.4, the asset and liability manager's task
over the long run is to structure the balance sheet so that the costs of asset
liquidation are minimized once a crisis has occurred. Once a liquidity plan
has been put into action, however, speed of execution becomes the primary
objective. There will be no time and the bank will have no negotiating
power to pinch pennies on liquidation costs, lawyer's fees, and invesbnent
banlcing fees. The best way to manage those fees is to position the bank's
liquidity risk position such that the contingency plan never needs to be put
into action.
A few major banks have gone so far as to declare a "liquidity crisis" as
a real-world drill to ensure that bank staff is prepared to execute the con
tingency plan. This drill is an expensive one, but it is one that may be
worthwhile for those banks who are most aggressively taking liquidity
risk.
Liquidity Analysis 249

11.9 SUMMARY: LIQUIDITY RISK AND SHAREHOLDER VALUE

There is no free lunch when it comes to liquidity risk, just like the interest
rate risk case. The bank cannot create shareholder value by taldng liquidity
risk. The bank can only destroy value from talcing liquidity risk outside of
the safety zone. This value destruction is due to the potential costs of
bankruptcy, which shareholders and debt holders cannot avoid through ar
bitrage actions for their own account.
The bank will maximize shareholder value by optimizing the expected
liquidation costs and bankruptcy costs compared to the economic incen
tives for taking risk, like the tax deductions available for the costs of debt.
One of the tools for controlling this risk is the securitization of balance
sheet assets, the subject of Chapter 12.
250 Chapter 11

APPENDIX 11A: JUST WHAT IS THE "LIQUIDITY PREMIUM"

ANYWAY?
(This appendix is a continuation of Appendix l0C.)
Chapter 10 introduced the risk-premium that most banks must pay in
funding the bank, referred to as the spread over Treasuries. This risk pre
mium truly is the liquidity premium. What many bankers call the liquidity
premium has a different definition. Let's say the bank decides to issue a
floating rate certificate of deposit with a two-year maturity that pays inter
est at the risk-free rate plus a premium set by the market. We have already
seen what this risk premium is in the one-year case. What many bankers
call the liquidity premium is the difference between what the bank will
have to pay on its floating rate two-year certificate of deposit and the cost
of its own one-year cost of funding. In reality, this "liquidity premium" is
simply a reflection that the risk of loss over a long time period is greater
in the real world than it is over a short time horizon. We can see this in
the following example.
Let us assume that the one-year Treasury rate is expected to stay con
stant at the equivalent of 6% per year. We also assume that the market
expects the bank's probability of default in year two is twice as high as it
is in year one. What would be the spread to the risk-free rate for a two year
floating rate certificate of deposit for the bank? The answers are con tained
in Table 1lA.l.
To put the discussion in real-world terms, let's call the bank's one-year
certificate of deposit cost LIBOR flat. If the bank's one-year probability of
default is 0.10% (leading to a two-year probability of default of 0.20%),
the bank can only issue a two-year floating certificate of deposit at 6.16%,
a 5 basis point spread over LIBOR and a 16 basis point spread over the
risk-free rate. This five basis point premium for two-year funds with a
one-year floating rate is what many bankers call the "liquidity premium."
In Table 11A. l that LIBOR is only 11 basis points over the risk-free rate
for a one-year maturity. If the probability of default is 50 basis points in
the first year and 1% in the second, the cost of the two-year floating cer
tificate of deposit is 26 basis points over LIBOR, a much higher liquidity
prenuum.
What happens if the probability of default is the same in year one as it
is in year two? Then there is no "liquidity" premium in the informal usage
of the term, but there is a risk or liquidity premium over the risk-free rate
that is the same as that of a one-year certificate of deposit.
Liquidity Analysis 251

Table 11A.1
Tw Year Floating Rate Break-Even Funding Costs

Year 1 Year 2 Risk Premium Over:

Probability Probabiltiy Break-Even Risk-Free One-Year
of Default of Default Two-Year Cou on Rate Rate
0.01% 0.02% 6.02% 0.02% 0.01%
0.10% 0.20% 6.16% 0.16% 0.05%
0.20% 0.40% 6.32% 0.32% 0.10%
0.30% 0.60% 6.47% 0.47% 0.16%
0.50% 1.00% 6.79% 0.79% 0.26%
1.00% 2.00% 7.60% 1.60% 0.52%
2.00% 4.00% 9.23% 3.23% 1.07%
3.00% 6.00% 10.92% 4.92% 1.64%
5.00% 10.00% 14.43% 8.43% 2.85%

The break-even coupon can be calculated using the formula below

when: (a) "default" means a total loss of interest and principal and (b)
investors are risk neutral. We assume P1 is the probability of default in
year one and p2 is the probability of default in year two. Just to simplify
the notation, we assume the risk-free rate rr is constant.

With a little bit of algebra, this reduces to the formula for the coupon
on a one-year certificate of deposit given above. Note that in this formula,
we are assuming interest rates are given as decimals and have not been
divided by 100 to convert from percentage figures.

APPENDIX 118: THE COST OF BANKRUPTCY AND

LIQUIDATION COSTS
This appendix introduces some important "real world" considerations into
the theoretical discussion from Appendix 11A. While this could be done in
252 Chapter 11

a precise way analytically, we will restrict ourselves to a logical argument

rather than a mathematical one. The added consideration is a simple one
that bankers and taxpayers have learned over and over again in the United
States in the 1980s: Bankruptcy has a cost in and of itself. Debt holders
lose more than just the difference between the true market value of assets
and the value of principal plus accrued interest. They only receive a pay out
after the enormous costs of the bankruptcy proceedings themselves. In
addition to court costs, legal costs, accountant's fees, and investment bank
ing fees, there is the cost and inconvenience of waiting for payment be
yond the scheduled maturity date. Even more important is the huge price
paid, both financially and emotionally, by customers and employees.
What is the impact of these costs from a practical point of view? Until
now, we have made the very defensible Modigiliani and Miller assumption
that one cannot change the value of the bank by calling debt equity and
equity debt. When bankruptcy has a cost, this powerful assumption is no
longer true. As the possibility of bankruptcy increases, the expected value
of incurring the costs of bankruptcy increases. This is a dead-weight loss
to both debt and equity holders, not just an amount of money that passes
from debt holder to equity holder (or vice-versa) without changing the
aggregate value of the bank.
Because of the cost of bankruptcy, there is an optimal capital structure
and liability maturity structure that results from the tradeoff from the in
creased tax deductions on additional debt versus the increased expected
value of bankruptcy costs. The volatility of asset values is perhaps the
single most important determinant of the expected value of bankruptcy
costs, in addition to the impact volatility has on debt borrowing costs.
There is one more important real world consideration in managing li
quidity risk. Many of the assets held by the bank cannot be converted to
cash costlessly; there is a transaction cost in getting rid of them in order to
make a cash payment. The amount of these costs is generally a function of
how much time one has to convert them to cash, a phenomenon of which
anyone who has had to sell his house in a hurry is painfully aware. If
bankruptcy has already occurred, these liquidation costs are less important
except for their impact on the time it takes to liquidate a bankrupt com
pany in the most economical way. In the case of a solvent institution that
suddenly has to liquidate assets unexpectedly, however, they become very
important.
Liquidity Analysis 253

APPENDIX 11C: LIQUIDITY IN THE BANKING INDUSTRY

Tables 11C.1 and 11C.2 derive approximate net liquid assets to total assets
ratios for all commercial banks by size range. The two tables pertain to
1985 and 1991 data to illustrate recent trends in balance sheet liquidity.
One characteristic feature of these results is that liquidity ratios decrease
with increasing bank size. The other striking feature is that the liquidity
ratio for every bank size range has improved between 1985 and 1991.
254 Chapter 11

Table 11C.1
Liquidity In the Banking lndustry-1985 Data
(Values are Percentages of Total Assets)

Under $300MM Over Ten

$300MM to $5 Bn $5Bn• Largest
ASSETS:
Federal Funds Sold 5.55 5.11 3.17 3.71
Placements 2.86 4.02 7.85 7.75
Loans 54.21 59.73 62.96 62.99
Commercial 20.70 27.63 40.12 46.20
Real Estate 20.67 17.96 12.50 11.08
Consumer 12.84 14.14 10.34 5.71
Securities 27.17 18.30 9.41 4.19
Treasuries 3.82 7.95 4.06 1.94
Agencies 14.55 3.12 1.34 0.45
Munis 8.11 6.05 3.59 1.70
Other 0.69 1.18 0.42 0.10
Trading Account 0.03 0.29 1.24 3.55
Total Earning Assets 89.82 87.45 84.63 82.19

LIABILITIES:
DDA 15.39 18.63 16.29 11.84
Interest Checking 8.02 5.35 3.10 1.47
Savings (Including MMDA) 21.05 20.70 13.09 9.76
Small CDs 31.33 18.44 10.03 4.34
Large CDs 11.63 13.58 12.26 8.21
Foreign Time 0.00 2.66 14.45 33.23
Federal Funds Purchased 1.60 8.59 13.17 7.75
Other Borrowings 0.63 2.58 5.00 6.44
Total Deposits and Borrowings 89.65 90.53 87.39 83.04

cont
Liquidity Analysis 255

Under $300MM Over Ten

$300MM to $5 Bn $5Bn• Largest

Liquid Assets 26.81 20.49 17.66 17.40

Federal Funds Sold 5.55 5.11 3.17 3.71
Placements 2.86 4.02 7.85 7.75
Treasuries and Agencies 18.37 11.07 5.40 2.39
Trading Account 0.03 0.29 1.24 3.55
Volatile Liabilities 13.23 24.83 39.88 49.19
Large CDs 11.63 13.58 12.26 8.21
Foreign CDs 0.00 2.66 14.45 33.23
Federal Funds Purchased 1.60 8.59 13.17 7.75
NET LIQUID ASSETS 13.58 -4.34 -22.22 -31.79

*Excluding 10 largest banks.

Source: "Recent Developments Affecting the Profitability and Practices of Commercial

Banks." Federal Reserve Bulletin (July 1992) pp. 459-483.
256 Chapter 11

Table 11C.2
Liquidity In the Banking lndustry-1991 Data
(Values are Percentages of Total Assets)

Under $300MM Over Ten

$300MM to $5 Bn $5Bn* Largest
ASSETS:
Federal Funds Sold 5.52 4.65 4.59 2.80
Placements 1.53 1.80 4.28 4.13
Loans 54.22 61.09 60.89 63.03
Commercial 15.25 19.53 28.52 34.74
Real Estate 28.82 27.24 21.40 21.21
Consumer 10.15 14.32 10.97 7.08
Securities 30.03 20.82 16.17 9.23
Treasuries 9.30 5.88 3.73 1.31
Agencies 13.84 9.54 8.41 3.41
Munis 4.32 2.69 1.76 0.76
Other 2.57 2.71 2.27 3.75
Trading Account 0.06 0.56 0.94 6.02
Total Earning Assets 91.36 88.92 86.87 85.21

LIABILITIES:
DDA 11.95 13.97 13.97 10.33
Interest Checking 11.07 8.02 5.59 2.96
Savings (Including MMDA) 19.23 20.69 17.87 13.56
Small CDs 35.39 24.92 17.73 5.78
Large CDs 10.01 10.32 11.17 6.23
Foreign Time 0.00 1.57 6.55 21.n
Federal Funds Purchased 1.33 7.26 10.24 6.57
Other Borrowings 0.52 3.04 5.42 9.88
Total Deposits and Borrowings 89.50 89.79 88.54 83.08

cont.
Liquidity Analysis 257

Under $300MM Over Ten

$300MM to $5 Bn $5Bn* Largest
Liquid Assets 30.25 22.43 21.95 17.67
Federal Funds Sold 5.52 4.65 4.59 2.80
Placements 1.53 1.80 4.28 4.13
Treasuries and Agencies 23.14 15.42 12.14 4.72
Trading Account 0.06 0.56 0.94 6.02
Volatile Liabilities 11.34 19.15 27.96 40.57
Large CDs 10.01 10.32 11.17 6.23
Foreign CDs 0.00 1.57 6.55 27.77
Federal Funds Purchased 1.33 7.26 10.24 6.57
NET LIQUID ASSETS 18.91 3.28 -6.01 -22.90

*Excluding 1O largest banks.

Source: "Recent Developments Affecting the Profitability and Practices of Commercial

Banks." Federal Reserve Bulletin (July 1992) pp. 459-483.