Appendix : A Primer on Gap Management - Home - Springer978-1-137-37440-0/1.pdf · Appendix : A...

270

Appendix : A Primer on Gap Management

Earlier in the book we described the importance of cash flow bucketing and analysis in the computation of the net cash balance of a firm. This practice is important for all companies that have an interest in success-fully managing their liquidity, but it is of course vitally important for those in the business of maturity transformation and liquidity provision, i.e., banks and other financial institutions.

As we have also noted, the management of interest rate risk is inti-mately related to the management of funding risk, which is a central part of liquidity management. Gap management, which is a key tool in the analysis and balancing of rate risk, represents an important part of both rate risk management and liquidity analysis. In this brief primer we shall highlight some of the key points associated with gap management, pro-viding several simple examples to help illustrate how it is used in practice. Many of these principles are, of course, applicable in the liquidity analy-sis process, with the caveat that bucketing moves from contractual (or behavioral) maturities to expected cash inflows/outflows.

Interest rate activities

Banks are fundamentally focused on the impact of interest rates on their activities. Indeed, their business strategies are typically centered on their current and expected views of interest rates. The reasons for this are obvious: core banking activities, including investing, lending and fund-ing, are all interest rate sensitive. Current and future net interest income (NII), which is simply interest income less interest expense, is impacted by interest rate changes. And the market value or net worth of a bank is ultimately impacted by NII.

Since interest rates are so fundamental to banking, a brief review of dura-tion is helpful to set the stage for subsequent review of some elemental asset liability management (ALM) measures, including gap analysis, duration gaps and earnings sensitivity. Figure A1.1 summarizes aspects of this framework.

Duration

Duration, a fundamental concept of the interest rate markets that has been in use for many decades, is a standard measure of the linear

Appendix : A Primer on Gap Management 271

price sensitivity of an asset or liability (or off-balance sheet contract) to changes in rates. In fact, we may regard duration as the cash flow weighted maturity of an asset, liability or contract which, except for zero-coupon structures, is shorter than the final maturity. Although there are various formulae for computing duration (e.g., modified, Macaulays, and so on), the basic duration formula is given in (A1.1) as the sum of the time-weighted (t), discounted cash flows (CF), divided by the price (P), or

tDur t cf r p* / (1 ) // (1t * / (1

We shall illustrate the computation in an example below. Ultimately, duration indicates how much the price of a rate sensitive

asset (RSA) or rate sensitive liability (RSL) will increase (decrease) for a small decrease (increase) in rates (e.g., up to several basis points). Since duration only captures the linear relationship over a relatively small rate movement, we need to adjust the overall computation if we want to measure the sensitivity of an RSA or RSL to an instanta-neous large movement in rates (e.g., 25, 50 or more bps); this is done through a convexity adjustment. However, for basic purposes of sim-ple rate change estimates in our discussion, duration is an adequate approximation.

Consider the following example: assume a bank is holding a four-year bond with a 6% annual coupon, that redeems at par. We can compute the bonds duration using the data summarized in Table A1.1.

In this example the bond has a duration of 3.67 (versus a maturity of 4). Any estimate of price sensitivity will thus be a more accurate estimate than a pure examination of maturity. In practical terms the

Interest Rates

FundingInvestingLending

Net Interest Income

Market Value Impact

Gap analysis

Duration gaps

Earnings sens

Duration

MeasuresRate Sensitive Banking Activities

Figure A1.1 Interest rates, rate measures, and banking activities

272 Appendix : A Primer on Gap Management

duration indicates that if rates change by 1 basis point (bp), the price of the bond will change by 0.0367 points. If rates change by 10 bps, the price of the bond will change by 0.367 points. We now have a basic tool by which to estimate rate sensitive asset/liability price changes given rate movements.

Common measures

Let us now consider several common ALM measures, commencing first with gap analysis. Gap analysis, as noted earlier in the book, is a measure of the total rate sensitive assets (RSAs) and rate sensitive liabilities (RSLs) that reprice in a given time period, used to determine whether they are mismatched (recalling the subtle difference between rate gaps and liquidity gaps, where the former is based on repricing and the latter on expected net cash flows). The most granular approach examines mismatches at specific time intervals (gaps or maturity buckets). An extension of this process centers on so-called duration gaps, which is a measure of RSAs and RSLs based on their duration rather than their contractual maturity. A cumula-tive effect based on the duration of the balance sheet and a forecast of rates can be computed to derive the change in market value or net worth of the bank. We can also consider earnings sensitivity, which is an enhanced gap analysis that takes account of embedded options and interest rate forecasts/shocks to reveal sensitivity of bank NII to changing rates.

In basic gap analysis, a bank must first identify its RSAs and RSLs, i.e., all instruments whose price can be impacted by rate changes. An asset or liability is considered rate sensitive if during the selected gap interval:

it matures it represents an interim or partial principal payment it can be repriced

For instance, a 030 day gap might feature the following RSAs and RSLs:

Table A1.1 Sample duration calculation

Year 1 Year 2 Year 3 Year 4 Total

CF 6 6 6 106PV 5.66 5.34 5.04 83.96 100PV/P 0.0566 0.0534 0.0504 0.8396 1.00T * (PV/P) 0.0566 0.1068 0.1511 3.3585Dur 3.673


Maturing instruments or principal payments: If an asset or liability matures within 30 days, the principal amount will be repriced. Any full or partial principal payments within 30 days will be repriced.

Floating rate instruments: If the index contractually changes within 30 days (e.g., 1 week or 1 month EURIBOR), the asset or liability is rate sensitive.

The actual computation of a gap for any particular period t is straightfor-ward and can be given by (A1.2):

t t tGap RSA RSLt ttRSARSA

It should be clear that assets/liabilities that fall outside the gap period are not included in the gap calculation.

Not surprisingly, RSAs and RSLs have a direct impact on NII, since these are the very contracts or instruments that generate interest income or expense. By extension, a bank can also compute a net interest margin (NIM), given simply as NII divided by earning assets.

Naturally, NII (and NIM) are affected by a series of factors, including:

changes in the level of interest rates changes in the composition of assets and liabilities changes in the volume of earning assets and interest-bearing liabilities (and hence the gap) changes in the relationship between the yields on earning assets and rates paid on interest-bearing liabilities

In practice, a bank follows a series of steps in order to understand its sensitivity to rate risk, and how its NII and NIM are likely to fare as rates change. It does so by:

developing an interest rate forecast. selecting a series of maturity buckets or intervals for determining when assets and liabilities will reprice. grouping assets and liabilities into these buckets calculating the gap for each bucket forecasting the change in net interest income given an assumed change in interest rates

Let us consider a simple example to illustrate.


A European bank has the simplified balance sheet shown in Table A1.2:

Under this construct, we can easily determine that

NII = 78.5mm 37.2mm = 41.3mm NIM = 41.3mm / 850mm = 4.86% Gap = 500mm 600mm = 100mm

Let us next assume that short-term rates increase by 100 bps, affecting all RSAs and RSLs as noted in Table A1.3:

Under this scenario,

NII = 83.5mm 43.2mm = 40.3mm NIM = 40.3mm / 850mm = 4.86% Gap = 500mm 600mm = 100mm

Since the bank has been running a negative gap, more liabilities than assets reprice in the higher rate environment, meaning NIM and NII both decline. Not surprisingly, the reverse would occur if short-term rates fell 100 bps.

Finally, let us assume that the mix of assets and liabilities changes so that the gap is smaller, as noted in Table A1.4:

Under this scenario,

Table A1.2 Sample European bank balance sheet

Avg rate mm Avg rate mm

RSAs repricing in 1 year

8% 500 RSLs repricing in 1 year

4% 60

Assets not repricing in 1 year

11% 350 Liabs not repricing in 1 year

6% 220

Non-earning assets 150 Equity 80Total assets 1000 Total Liabs+Equity 1000

Table A1.3 Sample European bank balance sheet 2




5% 600



6% 220



NII = 77.3mm 38mm = 39.3mm NIM = 39.3mm / 850mm = 4.62% Gap = 540mm 560mm = 20mm

Although the gap is smaller meaning rate risk is smaller the banks earning power has declined.

An extension of the gap formula lets a bank quickly determine how interest rate changes affect NII (and NIM); in fact, changes in NII are directly proportional to the size of the gap. This is given in (A1.3):

( )NII r RSA RSL Gap* ( ) *NII rNII * ( ))

Assume, for example, that a bank has RSAs of 500mm and RSLs of 520mm in the gap period. If rates rise by 100bps, the change in NII will be 200,000 (e.g., 20mm * 1%); conversely if rates fall by 100 bps, the change will be 200,000. Naturally, this approach includes simplifying assumptions, including:

parallel rate change to all RSAs and RSLs, instantaneous change at the beginning of each gap period.

These assumptions are not very realistic, but the formula provides for a quick and dirty estimate. Based on this discussion, some rules of thumb can be drawn:

When RSAs and RSLs are equal, we may say that RSA/RSL is 1.0, meaning the bank is perfectly matched. This indicates that for a given increase in rates it will lose on its liabilities and gain on its assets. When RSA < RSL (RSA/RSL is < 1.0), the bank is running a negative gap or is liability sensitive. This also means that

the bank has an asset duration that is shorter than its liability duration, or it has more rate sensitive liabilities than assets. the NII will generally fall (rise) when interest rates rise (fall).

Table A1.4 Sample European bank balance sheet 3




4% 560



6% 260



When RSA > RSL (RSA/RSL is > 1.0) the bank is running a positive gap or is asset sensitive, indicating that

the bank has an asset duration that is longer than its liability duration, or the bank has more rate sensitive assets than liabilities. the NII will generally rise (fall) when interest rate rises (fall).

The decision to hold a positive or negative gap will depend on the banks expectation of interest rates and its overall business mix and strategy. When rates are expected to rise, a bank should reduce its negative gap or increase its positive gap. When rates are expected to fall, a bank should increase its negative gap or decrease its positive gap. In general, the sum-mary relationships in Table A1.5 may be said to hold true:

In order to reduce rate risk, a bank must narrow its gap meaning it must become less negative or less positive, depending on its current positioning. For instance, if a bank is liability sensitive (e.g., negative gap or RSA/RSL < 1.0), it should attempt to increase RSAs and decrease RSLs and vice-versa. Naturally, to increase rate risk it must widen its gap.

The range of available solutions can be summarized as per Table A1.6:

Table A1.5 Relationships between gap, rates and net interest income

Gap Position

Changes in Interest Rates

Changes in Interest Income

Changes in Interest Expense Change in NII

Positive Increase Increase Increase IncreasePositive Decrease Decrease Decrease DecreaseNegative Increase Increase Increase IncreaseNegative Decrease Decrease Decrease DecreaseZero Increase Increase Increase NoneZero Decrease Decrease Decrease None

Table A1.6 Solution to increase/decrease asset/liability sensitivity

Objective Solutions

Reduce asset sensitivity Buy longer-term securities. Lengthen the maturities of loans. Move from floating-rate loans to term loans.

Increase asset sensitivity Buy short-term securities. Shorten loan maturities. Make more loans on a floating-rate basis.

Reduce liability sensitivity Pay premiums to attract longer-term deposits. Issue long-term subordinated debt.

Increase liability sensitivity Pay premiums to attract short-term deposits. Borrow more via non-core purchased liabilities.


It comes as no surprise that some of these actions will also have an impact on a banks liquidity strategy and position. Essentially, any action that affects the maturity (duration) profile of assets or liabilities will affect the liquidity of the bank. So, if a bank wants to increase its long-term loan growth by funding strictly with short-term deposits, it will be incurring additional liquidity risk. The issue to be considered is whether the explicit spread between the long-term loans and short-term deposits (in a normal positive yield curve environment) compensates sufficiently for the incremental liquidity risk taken. Extending the discussion, we may say that increasing liability sensitivity by boosting short-term depos-its also increases the banks exposure to short-term liability withdrawals (i.e., intensifies potential liquidity risk) in an unstable market; reducing asset sensitivity by making longer term loans may reduce liquidity risk if these loans are term funded, and so on. The main point to stress is that development of a banks strategy for generating NII based on an increase or decrease in its gapping should not be done in isolation it must logi-cally also include a discussion of its strategy for dealing with liquidity issues.

Continuing our discussion, a duration gap can be computed by exam-ining the durations of the RSAs and RSLs (recalling that duration is addi-tive, and it is possible to sum across time and maturity). This is given in (A1.4):

gap RSADur Durgap R RSLRSL Dur* RRSADurR RSLRSL DurR RSA RSA

This formula indicates that the duration gap is equal to the duration of the RSAs, less the duration of the RSLs, adjusted by the weighting of the RSAs and RSLs (or total assets (TA) and total liabilities (TL)). A simple and powerful extension of this formula, shown in (A1.5), allows us to calculate the change in the banks net worth (as a percentage of assets) for a given change in rates.

= NWAssets

Durr

gar pa * ( )+ r

For instance, if a bank estimates its duration gap at 1.7, it would expect its net worth (as a percentage of assets) to decline by 0.17% for every 10bps increase in rates. Naturally, this is to be regarded as a general esti-mate only.


Earnings sensitivity is another tool in understanding ALM and earn-ings. Under this framework a bank can increase the sophistication of its analysis by:

including embedded options that can potentially alter the banks cash flows a process which is ignored by the basic gap analysis described earlier. incorporating interest rates change by different amounts at different times by running multiple what-if scenarios.

Earnings sensitivity generates the potential variation in NII across dif-ferent interest rate environments, based on different assumptions about balance sheet composition, when embedded options will be exercised, the timing of repricings, and so forth. It essentially demonstrates the potential volatility in earnings across each environment; ultimately, the greater the potential variation in earnings (earnings at risk), the greater the amount of risk being assumed by a bank.

The process of creating an earnings sensitivity matrix is based on the following steps:

Step 1: Develop multiple interest rate forecasts and select one as the future evolution of rates; this may span 1, 2, 3, etc., years. Step 2: Compute the enhanced periodic and cumulative gap by making detailed assumptions about when any embedded options might be exercised and how they may change cash flows, assets, liabilities, e.g.,

Option to refinance a loan Call option on a bond in the investment portfolio Depositors option to withdraw funds prior to maturity Cap (maximum) rate on a floating-rate loan

Note: The more detailed the analysis on optionality, the more relevant the results.

Step 3: Apply the rate changes to the enhanced gap measure for each relevant time period to determine the NII. Step 4: Determine the banks NII earnings sensitivity by computing the change in NII for different shocks (e.g., +/ 100, 200 bps, etc.).

These shock scenarios can be applied against any relevant risk limits. Consider the following simple example based on the periodic and cumu-lative option-adjusted gap and a sample interest rate path (most likely versus market implied), as shown in Table A1.7 and Figure A1.2.


The NII results for the next year based on the most likely (ML) inter-est rate scenario can be determined for the stated gap and can then be shocked to extremes of +/ 300 bps to provide an earnings sensitivity range in terms of NII. The process can be repeated for year 2 scenarios and so forth. Any ALCO or Board limits can be included to determine whether warnings or breaches occur. Decisions about the proper periodic or cumulative gap should take account of the construction of the banks balance sheet, paying close attention to liquidity features and the ability to realize cash quickly and cost-effectively.

It is important to remember that measures such as duration, gap analy-sis, duration gaps, earnings sensitivity and the like are simply tools that banks can use to help them gauge rate risk and, by extension, aspects of liquidity risk. These are tools with assumptions, and so must be used care-fully. Naturally, there is no single optimal solution to what a banks gap should be every bank is different. The ultimate answer will be based on a banks interest rate forecast, business strategy and revenue diversifica-tion, funding sources and risk appetite (including both market risk and liquidity risk).

Table A1.7 Sample option-adjusted gap

Total3 Months or less

>36 Months

>612 Months

>13 Years

>35 Years

>510 Years

>1020 Years >20 Years

GAP 20,252 6,249 11,053 43,582 49,200 71,748 51,918 201,000Cumulative 20,252 26,501 15,448 28,134 77,334 149,082 201,00

4.5

4

3.5

3

2.5

21 2 3 4 5

Most Likely Mkt Implied6 7 8 9 10 11 12

Interest rate scenario 1 over next 12 months

Figure A1.2 Sample interest rate scenario

280

Notes

1 Liquidity Risk Defined

1. The specific term liquidation risk is sometimes used to denote the risk of selling an asset at a loss.

2. Although the value of derivative transactions must now be reported on a firms balance sheet and income statement under some accounting systems, the nature of future commitments and obligations under such contracts is still generally oblique.

3. Financial operating risks, such as securities and foreign exchange settlements, delivery of cash against securities, and so forth, comprise a separate category of operational risk that can also impact liquidity risk. For instance, failure by a bank to receive payments on securities that have already been delivered can result in a cash deficit that needs to be funded on a short-term basis.

4. Note that at least one major study has suggested that the interbank funding spread featured two different risk components: at the beginning of the first phase of the crisis (OctoberDecember 2007) credit risk was the dominant factor, while after the Lehman collapse in September 2008 the liquidity risk component moved to the forefront.

5. Unfortunately, the track of record of the public rating agencies in properly evaluat-ing structured credit has been shown to be poor at best. Bad modelling, conflicts of interest and slow reaction times meant that many of the AAA-rated securities backed by mortgages should never have carried such ratings. Furthermore, the record of the agencies in identifying companies with liquidity pressures (leading ultimately to financial distress and either bankruptcy or acquisition) has also been weak, to wit major misses on Penn Central, Drexel, WorldCom, First Executive, Enron, Bear Stearns, Lehman Brothers, among others.

2 Liquidity and Financial Operations

1. Interestingly, half of all US bank failures between 1984 and 1989 occurred within institutions that would have been considered adequately capitalized under the BIS 1988 Capital Accord definitions; capital alone is thus not enough to prevent a firm from failing.

2. We note that sovereign organizations, though not necessarily concerned with value maximization, are often equally interested in minimizing financial problems.

3. Indeed, in an effort to manage greater funding liquidity risks, some securities firms arrange for committed bank line facilities from the banking sector; this gives them some measure of safety in the event the short-term markets prove challenging.

3 Sources of Liquidity

1. It is true, of course, that banks active in the repo and reverse repo market often run matched books so that they do not face an excess of interest rate risk or curve risk. Thus, if a bank allows its reverses to roll off in order to generate additional cash, it must either rebalance its own repurchase agreement position by construct-

Notes 281

ing appropriate interest rate risk hedges or allowing the repos to roll off as well; if it selects the latter it will, of course, face a liquidity constraint of its own as it will have lost a source of funding. However, the securities it receives back from the repo close-outs can then be sold in the marketplace to generate additional cash. Such a process is common when a bank is shrinking its matched book, which often occurs prior to quarter-end financial reporting periods.

2. For instance, in 1996 the Bank of England introduced the Sterling Stock Liquidity Regime (SSLR), which broadened the liquidity requirements for commercial banks with a large retail presence. Under the SSLR banks must supplement the traditional cash inflow/outflow gap/ladder approach (which we discuss in Chapter 8) with a portfolio of liquid sterling assets that can be used to cover any potential loss of wholesale funding. (Note that a separate process, the Mismatch Liquidity Regime, which permits committed funds and a broader array of assets to be used in comput-ing the net liquidity position for each ladder maturity, was introduced in 1999 for non-SSLR banks,. )

3. In fact, certain empirical research (see Schlingemann, Stulz, and Waking, 2001, for example) suggests that a firm is most likely to sell portions of its business that are liquid, or those that are unrelated/immaterial to core operations or poor perform-ers, and far less likely to sell portions that are illiquid.

4. Note that similar local CP markets exist in other countries as well: Canada, for instance, features a relatively active marketplace for unsecured C$-denominated CP.

5. Investor demand for A-2/P-2 and A-3/P-3 issues is very small (less than 5 percent of total outstanding during normal market conditions) and can disappear very quickly, particularly during times of market stress or general credit deterioration, when investors are less eager to speculate on short-term credit spread movements. When this occurs, lower rated issuers with outstanding notes in the market may work with their dealing banks to manage out of the CP market (without causing any disruption in their operations or creating any negative publicity) by tapping alternative facilities, such as bank borrowings. Since the crisis of 20072008, the market for lower-rated CP has essentially become dormant .

6. A small number of CP programs are supported by bank letters of credit or are struc-tured as asset-backed programs, and would thus be considered secured.

7. Even deposits with longer maturities can indirectly affect a banks funding. For example, institutional investors that sell medium-term bank deposits in the sec-ondary market at a discount in order to retrieve their capital can send a negative signal to other low-risk institutional investors, making rollovers of other funds more difficult.

8. It is worth emphasizing that short-term liabilities (regardless of source) have an important role to play in the management of liquidity. In a general sense, we may say that when short-term liabilities (STLs) > short term assets (STAs) a company will be under some degree of liquidity pressure. It must therefore either increase its long-term liabilities (LTLs) to redeem the STLs, or it must convert long-term assets (LTAs) to STAs in order to meet STLs as they come due. Naturally, if STLs < STAs, there is no particular pressure, as enough STAs remain on hand to cover STL obliga-tions. The issue at hand, then, is the dynamism of a companys STLs a company must understand how they behave and whether they can change rapidly in the face of changing markets.

9. Euronote facilities generally permit issuance in the one to ten-year sector (com-plementing the shorter term issuance provided by ECP) and are often backed by bank tender panels, which absorb unsold notes and so provide contingent funding. Tender panel facilities are generally committed or transferable to other banks upon agreement by all parties.

282 Notes

10. Some firms have opted to follow the concept of cash capital as a way of financing long-term assets, particularly those that are illiquid in nature, as well as certain critical accounts, such as haircuts. Though the definition of cash capital can vary, it can include equity, long-term debt, synthetic structures which provide the equivalent of greater than 1 year funding, and so forth.

11. Some securitizations are liability, rather than asset, based and are intended to remove particular types of risks from the liability side of the balance sheet. The concepts are similar, although we shall not consider them in further detail.

4 Funding Liquidity Risk

1. It is worth noting that such reserves may have to be held in low/no-yield assets, which reverts to the issue of risk/return trade-off and enterprise value maximiza-tion.

2. Financial mismanagement has been at the heart of significant corporate problems over the past few decades, including those associated with Enron, Tyco, WorldCom, Swissair, and many others. Many problems ultimately manifested themselves in the form of severe liquidity pressure.

3. Bankers Trust serves as an interesting example of a firm that was plagued by rep-utational problems. It was ultimately subsumed by Deutsche Bank when it was unable to overcome the crisis of confidence. Although not strictly a liquidity prob-lem, the banks well-publicized client derivative lawsuits and settlements (such as Procter & Gamble, Gibson Greetings, and Sandoz) in the early 1990s, coupled with nearly $500 million in losses in Russia in 1998, eroded depositor and investor confidence in the bank and its management. The firm was downgraded on various occasions and was unable to generate a competitive cost of funds as a result of its problems. Deutsche Bank ultimately stepped in to acquire the bank when it was clear that further growth was becoming virtually impossible.

4. For instance, Gatev and Straham (2001) have found that bank assets (securities and bank lines of credit) increase more rapidly when short-term credit spreads (that is, commercial paper less Treasury bills) widen, and that the quantity of assets funded with deposits increases during a crisis period. The study has also found that banks have a comparative advantage over other financiers in extending loans during a crisis, as the yields they pay on flight-to-quality deposits decrease as credit spreads widen. Credit extensions, however, are not granted uniformly; that is, they are not granted to idiosyncratic borrowers who might wish to draw down based on their own credit circumstances, but to systematic borrowers, who draw down based primarily on the availability of market liquidity. Banks can thus lend to highly rated systematic borrowers during times of market stress, suggesting some win-ners exist when exogenous forces are otherwise proving disruptive. The reverse is also shown to be true: when markets regain their balance and depositors withdraw funds to reinvest in higher yield alternatives, banks are no longer flush with cash and thus scale back on the highly rated systematic lending they once engaged in.

5 Asset Liquidity Risk

1. For instance, there is empirical evidence that in some markets futures and the underlying cash assets they reference complement each other, promoting strong liquidity in both sectors Eurodollar futures and Eurodollar deposits are one example of this. In other cases, futures and cash are independent and fraction-alized because they act as true substitutes. In such instances one of the instru-ments might exhibit a great deal more liquidity. In Japan, for example, the 10-year

Notes 283

Japanese Government Bond (JGB) futures contract is far more liquid than the underlying benchmark bond. The same can occur with corporate or sovereign bond issues. If the issuer offers too many different maturities as part of its debt management program, it might fractionalize the market so much that insufficient liquidity builds in any single issue. Alternatively, if it issues only a few maturities, it might fail to generate enough interest among investors and be unable to construct a meaningful yield curve. Balancing these two fungibility forces is not an easy task.

2. Most institutional OTC markets are quote-driven, while many public equity mar-kets are order-driven (or hybrid).

3. Similar spread differentials between on- and off-the-run securities have been observed in other asset classes during a variety of market conditions. For instance, off-the-run asset-backed securities might trade at a spread of +45 bps to the on-the-run asset-backed benchmark during normal market conditions, but as much as +40 bps in poor conditions; the same has been observed for investment grade bonds. Off-the-run high-yield bonds, in turn, might normally trade at +1020 bps to the high yield on-the-run benchmark, and in excess of +5075 bps in poor conditions.

4. A company will only lose ownership of the asset if it defaults on the terms of the collateralized financing transaction: that is, it fails to pay timely principal and interest or breaches a covenant.

6 Liquidity Spirals and Financial Distress

1. There are various academic theories about the underlying causes of bank runs. For instance, Diamond and Dybvig (1983) posit a model where banks are provid-ers of liquidity insurance to depositors who may decide to flee based on random events; Chari and Jagannathan (1988) focus on a model where systemic risks may be derived from idiosyncratic risks that lead to bank runs; Gorton (1988) assumes that bank runs are recession-related and tied directly to corporate and business fail-ures. Despite different models and views, empirical and anecdotal evidence sug-gest that bank runs tend to be driven by institutional, rather than retail, depositors because institutions have better access to information and less insurance coverage. The growing level of interbank and off-balance sheet activity, and the ease by which contagion can spread to other banks in the system, might exacerbate the runs.

2. Although some facilities contain material adverse change clauses giving funding banks the option to opt out if the credit environment has deteriorated enough to pose a financial risk, such clauses are rarely invoked; private restructuring may be a preferred alternative.

3. Credit extensions by banks can accelerate during stressed times as companies draw down facilities (such as during the 1998 Russia/hedge fund crisis, when credit expanded by 30 percent). The main issues center on whether deposits flee or are attracted, and whether banks are willing to assume the additional credit risk. Large banking institutions can often attract deposits, but the same is not necessarily true for small or mid-sized banks.

4. In some national systems regulators limit or prohibit the amount of credit that can be extended to rapidly deteriorating firms, which would exacerbate the problem.

5. It is important to note that liquidity problems are not the only reason a company might enter into a state of financial distress. Difficulties related to earnings, asset quality, overall leverage, or strategy, for example, can be equally damaging and might also induce a state of distress. While these are all important, they are outside the scope of this text.

284 Notes

7 Case Studies in Liquidity Mismanagement

1. In 1984 Continental Illinois, which was heavily reliant on wholesale and interna-tional markets for its funding, was unable to halt interbank deposit withdrawals fuelled by news of problems within the banks emerging market and energy loan portfolios. As liquidity drained away, a consortium of 29 banks and the Federal Deposit Insurance Corporation interceded with a $7.5 billion rescue package that reinjected cash into the bank; this was supplemented by a further $4.5 billion of borrowings from the Federal Reserves discount window. The bailout was consid-ered necessary as the possible disruptions from a large bank failure were deemed to be too great.

2. The reasons for the liquidity problems varied, but the end result was always the same: extreme difficulty accessing funds, resulting in very high funding costs. The Bank of New York suffered a large liquidity deficit in 1985 through operational errors, Salomon Brothers through the Treasury bond auction scan-dal in 1991, and Citibank through massive commercial real estate write-downs in 1991.

3. Although the October 1987 stock market crash was caused by a number of different factors, including excess leverage and speculation, and weakening economic condi-tions, the fall was exacerbated by the use of portfolio insurance, which involved selling assets (such as individual stocks or index futures contracts) into a falling market. The estimated $100 billion-plus of portfolio insurance programs created greater market instability on the downside and converted two-way flows into large one-way flows; the dynamic replication of portfolios via portfolio insurance called for selling more stock or index futures contracts as prices fell, injecting more selling pressure into the market and creating a self-fulfilling downward spiral. During the immediate pre-crash period, some $12 billion of index rebalancing was required, but only $4 billion was executed. By the time the crash was in full motion, the pent-up selling forces were unleashed: when the cash market faced delayed open-ings, the rebalancing programs were shifted to index futures, which exacerbated the fall. Liquidity eventually became so impaired that rebalancing could no longer be undertaken; portfolio insurance techniques were widely criticized in the after-math. Circuit breakers, designed to avoid market free-falls, were eventually insti-tuted and remain in place in various markets to the present time.

4. Following Mexican bank privatizations in 1991 and 1992, the sector entered a phase of rapid asset growth; indiscriminate lending led ultimately to deteriorating asset quality, which strained bank revenues. This was compounded by a very heavy reli-ance on domestic and offshore interbank deposits (63 percent of all funding) rather than more stable retail deposits. When the Mexican government devalued the peso by 56 percent in late 1994, most major banks suffered a rapid, and significant, loss of funds as interbank depositors exited. The ensuing losses left many banks in a weakened state; many smaller institutions were forced to close down or merge, and the government bailed out several larger banks (through the deposit insurance company and central bank).

5. Caprio and Klingebiel (1999) record 112 systemic banking crises in 93 countries between the 1970s and late 1990s, many of them the result of significant liquidity risk difficulties. Importantly, the authors indicate that such crises have impacted developed, as well as developing, nations; countries such as the US, UK, Spain, Finland, and Sweden appear to have been just as susceptible to systemic problems as emerging market systems.

6. Kidder Peabody, Donaldson Lufkin Jenrette, Lehman, Nomura, JP Morgan, Morgan Stanley, Merrill Lynch, Salomon, UBS, Bear Stearns, and Prudential were all lenders to Askin Capital.

Notes 285

7. In January 1995 Orange County sued Merrill for $2.4 billion, claiming it wan-tonly and callously sold the county risky securities in violation of state and fed-eral laws. Orange County claimed ultra vires , saying that it was acting out of its legal scope and that the transactions were unsuitable. County bondholders also sued Merrill and Citron for false disclosure in the July 1994 bond prospectus, and suits were filed against Morgan Stanley, Nomura, and CSFB for their role in grant-ing excessive amounts of leverage. All of the banks initially denied any wrongdo-ing, but eventually settled out of court: Merrill paid $437 million, Morgan Stanley $70 million, CSFB $52 million, and Nomura $48 million. Separately, the courts determined that Citron was to blame for the flawed strategy; Citron pleaded guilty to six felony counts related to making misleading statements in selling securities, falsifying accounting records and redirecting investment funds (not for losing the $1.7 billion).

8. To be sure, many banks were acting in self-interest; this was not a public bailout, obviously, but a reorganization with creditors risking $3.6 billion of their share-holders funds to avoid a catastrophe. The Streets risk exposures were so large that it is likely that many players would have been severely damaged without the bailout.

9. Many major banks gave LTCM the leverage repos, derivatives, and credit facili-ties on very liberal terms, essentially mispricing their risk in order to secure a portion of the funds commission-based business. They also lent without good disclosure (contrary to standard credit operating procedures), opted not to require upfront collateral, and relied too heavily on their own flawed models. In some instances they attempted to replicate LTCMs own positions, injecting the same illiquidity into their own books.

10. For instance, exposure to LTCM caused UBS to lose $690 million, the Bank of Italy $100 million, Credit Suisse $100 million, Dresdner $145 million, Sumitomo Bank $100 million, and so on. Broader market dislocations arising from vola-tility, illiquidity, and forced selling in other asset classes generated additional losses: CSFB reported losses of $1.3 billion in Russia, including nearly $640 mil-lion in rouble forwards purchased via the Moscow Interbank Currency Exchange to hedge rouble exposure; Citibank lost $60 million on Russia and $300 million on its arbitrage positions; Merrill lost $1.5 billion, mostly on spread widening affecting very large, and often illiquid, portfolios of corporate bonds and pre-ferred stock; Goldman reported proprietary trading losses of $650 million across various markets; and so on.

11. SAG featured 72,000 workers, including 21,000 in Switzerland, far more than any other carrier operating similar fleet size and routes, and its employees were among the highest paid in the entire industry. It also incurred additional expenses from its inconvenient dual hub system (Swissair/Zurich, CrossAir/Basel).

12. Readers interested in a broader review of the details of the case from a corporate governance perspective may wish to consult the discussion in Banks (2004a).

13. Including LJM1, LJM2, JEDI 1, JEDI 2, Chewco, and Raptors I-IV. 14. Andersen, Enrons external auditor and architect of the partnerships (for which it

earned many millions of dollars in fees), indicated that it had not accounted for the SPEs correctly; as a result of errors early in the process, the SPEs were not con-solidated on Enrons balance sheet as they should have been, forcing the restate-ments. Many of the SPE transactions between Enron and the LJM1, LJM2, and Chewco partnerships were arranged because the company could not, or would not, do them with third parties; the end-game in all cases appears to have been financial window dressing rather than genuine risk transfer.

15. There was fallout in other sectors as well: external auditor Andersen eventually failed under the weight of criminal obstruction of justice charges, and various

286 Notes

banks and law firms that assisted in the creation of a number of special, and ulti-mately illegal, transactions were fined and sanctioned.

16. http://companyinfo.northernrockassetmanagement.co.uk/downloads/results/ res2006PR_AnnualReportAndAccounts.pdf, p.2.

17. At least some blame for the Northern Rock crisis must also be attributed to the weak regulatory structure in the UK at the time, which spilt duties between the Bank of England and the Financial Services Authority (FSA). Ex-post analysis suggests that the FSA knew about Northern Rocks liquidity problems at an earlier stage but did not act. There is also evidence of lack of coordination and mismanagement of decision-making. Furthermore, the deposit protection scheme was antiquated and at least somewhat flawed, and the regulatory framework lacked any mechanism to deal with the unwinding of troubled banks; much of the regulatory response was ad-hoc in nature. As a result, other troubled banks, including RBS, Bradford and Bingley, and Lloyds TSB, had to be wholly or partly nationalized when they ran into trouble.

18. Banking Reform Protecting Depositors, Discussion Paper, HM Treasury, October 11, 2007.

19. Rather curiously, the Lehman Treasury roadshow material, presented in July 2008 by the bank as it attempted to broaden and strengthen its funding, con-tained an appendix referencing lessons learned from the Bear Stearns liquidity event. Apparently, sufficient lessons were not learned.

20. http://oversight.house.gov/images/stories/Hearings/110th_Congress/ Fuld_Statement.pdf

8 Measuring Liquidity Risk

1. Liquidity ratios are an important aspect of overall financial analysis and bank-ruptcy prediction; empirical research on bankruptcies suggests that the most important financial variables with predictive capabilities include leverage, liquid-ity, profitability, earnings volatility, and company size.

2. Hot money is characterized by a perfectly elastic supply curve, meaning it is extremely sensitive to interest rate levels and changes.

3. Duration can be measured in a number of different forms, including Macaulays duration and modified duration. For a standard fixed income instrument, these are given as:

2 1

1*

(1 )

*)

( )n)

Dury)

nCy y y2 (1

Dur (P

1 2c nC nM2= * 1 2(1 ) (1 ) (1 )1 2 n n(1 ) 1c c nC nM2+ + + +1 2)) )y y y y(1 ) (1 ) (1 ) (1) (1 ) (1 ) (1

(1 ) (1 ) (1 ) (1 )1 2 n(1 )(1 ) (1 ) (1 ) (1) (1 ) (1) (1 ) (1 ) (1(1 ) (1 ) (1 )) (1 ) (11 2 (1(1 ) (1 ) (1 ) (1) (1 ) (1) (1 ) (1 ) (11 2 (1

100

C 100 100

11

y 100

y+1

11

(1(1 yy(1 )(1 ))=

where C is semi-annual coupon interest, y is the semi-annual yield, n is the number of semi-annual periods, and M is the maturity value of the bond (generally par). These can be supplemented by measures of convexity, or the change in duration for a change in yield, which is important in determining the sensitivity of cash flows to large changes in rates:

http://companyinfo.northernrockassetmanagement.co.uk/downloads/results/res2006PR_AnnualReportAndAccounts.pdf, p.2.http://oversight.house.gov/images/stories/Hearings/110th_Congress/Statement.pdfhttp://oversight.house.gov/images/stories/Hearings/110th_Congress/http://companyinfo.northernrockassetmanagement.co.uk/downloads/results/

Notes 287

21

( 1) ( 1)(1 ) (1 )

n

n)t 2t

t t C n n M( 1) ( 1)1) (Cvx

y y) (1) (122=

1) ((1) (+2( )

t 2= )) (1(12 (12

4. Interestingly, a survey by IOSCO (2002) found that many firms at the turn of the millennium still did not regularly measure commitments or other off-balance sheet contingencies.

5. Note that, while there is no precise definition of the correct maturity buckets to use, and this definition remains at the discretion of individual institutions, many appear to focus on the following buckets as being most relevant: overnight, next day, 7 days, 14 days, 1 month, 2, months, 3 months, 6 months, 1 year, 3 years, 5 years, and 5+ years.

6. The essential maturity transformation function played by banks and other finan-cial institutions necessarily means that contractual short-term cash flows will be large, but may not necessarily be quite as large when considered on a behavioural basis, as those holding the liability cash flows may be quite happy to continue holding them through extensions, rollovers, and so forth.

7. Not surprisingly, banks must be able to adequately model the behaviour of these types of liabilities, so that they can understand how much funding they have under different scenarios. As we might expect, during stressed market situations, the stickiness of liabilities may change, with withdrawals or lack of rollovers accelerating versus some historical norm.

8. See Kyle (1995), for instance. 9. Note that spreads can also be measured in other forms, including the realized

spread, or the differential between the weighted average bidoffer for executed trades over time, and the effective spread, or the actual transaction price incorpo-rating the direction of price movements.

10. It is worth noting that dealing can occur within the full bidoffer spread (such as limit orders and guarantees) so friction may be overstated; in order to avoid this overstatement, a half spread (bidoffer/2, or quoted mid-market point prior to a trade) can be used as a proxy.

11. There is a large body of practical, empirical, and theoretical literature on VAR, which the interested reader may wish to consult; Jorion (1996) serves as an excel-lent primer. In general terms we note that standard parametric VAR (or variance/covariance VAR) relies on use of a multivariate distribution of asset returns and a variance/covariance matrix that is specified by time series returns. Volatilities and correlations may be estimated through historical processes (and may be expo-nentially weighted to give more recent observations greater weight in the sample), or they may be approximated through volatility estimation techniques (such as GARCH). The resulting parametric VAR, which is a multiple of the standard devia-tion of the distribution of portfolio returns, is simple to implement but cannot accommodate non-linear instruments (or can only do so crudely, through the use of delta approximation techniques.)

A second approach, the historical simulation VAR, is useful when the observed distribution of returns cannot be accommodated by a parametric approach. Under this framework, historical returns for assets in the portfolio are sampled, a time series of portfolio returns is simulated, and the appropriate confidence level VAR is determined. This process does not rely on assumptions about the shape of the distribution or estimates of volatility and correlation for individual assets; how-ever, it requires a long history of data (and might be questionable for any data

288 Notes

regime shifts). From a regulatory perspective, the BIS requires banks to perform a VAR computation based on the method of its choosing, adjusted to the 99 percent confidence level and a ten-day liquidation horizon, and scaled by a multiplier of 3 (which represents a safety cushion). The process also requires a regimen of stress testing, backtesting of results, input parameter updates, and appropriate qualita-tive standards and controls. While the aim is constructive and well-intended, it has shortcomings, including some mentioned in this chapter. Sophisticated finan-cial institutions rarely rely on standard VAR as their only means of measuring and managing market risk (and, by extension, aspects of liquidity risk).

12. Readers interested in an excellent overview of alternative LAVAR approaches (e.g., Almgren-Chriss, Jarrow-Subramanian, Bangia, Diebold, Schuermann and Stroughair) may wish to consult Erzegovesi (2002).

13. Note that it is common to define stress testing as an integrated, multivariable test, while scenario analysis is a future state based on the movement of a single vari-able; each scenario analysis calls for holding constant all variables except one.

14. While standard VAR data sets and computational frameworks based on historical data may lack the relevant fat tail characteristics, extreme value theory (EVT), which focuses on the tail of the distribution, tends to provide additional and valu-able information about the low probability/high severity events.

15. It is interesting to consider just how the dominoes toppled during the crisis. Whilst the genesis of the crisis had appeared much earlier, e.g., during the first half of the decade, as loose monetary policy had caused a very significant real estate bubble to develop, the first signs of trouble came in early- to mid-2007. In April 2007, New Century and American Home Mortgage, two large US mortgage originators, reported large losses and eventual closures. In May 2007, UBSs pro-prietary trading internal hedge fund, Dillon Read, shut down after large credit and market losses. In June 2007, two Bear Stearns hedge funds had to be bailed out by the investment bank (consuming precious resources, as would be made clear in early 2008). In August 2007, BNP froze three investment funds, and German bank IKB had to be bailed out by the German banking sector after posting massive losses on its illiquid securities portfolio; another state bank, Sachsen Landesbank, suffered similar losses and required support. By the end of the month, a key fund-ing source, the asset backed commercial paper market, effectively shut down as investors refused to rollover paper any more. This instantly put pressure on all banks active in the securitization market and the structured investment vehicle sector, as they had to bring these assets (many of them illiquid subprime-backed securities) onbalance sheet, funding them through any means possible. Just a month later, the Bank of England supplied Northern Rock with a temporary liquidity facility and, together with HM Treasury, would soon issue a blanket guar-antee on deposits in order to stop the growing deposit withdrawals (as we have discussed earlier in the book). Throughout this period many US and international banks were posting increasing losses as they marked their illiquid asset portfolios to market; some were forced into a deleveraging situation. In November 2007, HSBC was forced to absorb $45 billion in SIV assets. Unfortunately, 2008 was even worse, as illiquidity and market losses intensified. In March 2008, Bear Stearns ran out of cash and was rescued by JP Morgan (with some backing from the US Treasury). Turmoil continued throughout the summer, and in September 2008 events came to a climax: Wachovia and Merrill Lynch, both suffering from con-tinued funds outflows, were forced into shotgun marriages with Wells Fargo and Bank of America, respectively. Shortly thereafter, Lehman Brothers, like Bear, ran out of cash, but was not rescued it was forced into bankruptcy. That triggered panic throughout the markets, causing governments to assemble rescue packages for Fannie Mae and Freddie Mac, AIG and many other US and international banks

Notes 289

(as well as the entire nation of Iceland). The government intervention, through massive liquidity injection and partial nationalizations, calmed the markets and put things on a more stable footing. However, most of 2009 was fragile, with mar-ket conditions far from normal and liquidity in many assets still thin. Not until 2010 did some semblance of normalcy begin to return. So the toppling of the dominoes was extensive, and although credit and market risks featured as initial drivers, liquidity risks came into full force at various occasions during the period. Any successful stress testing exercise demands some blue sky thinking about unimaginable events and connections just as the last crisis demonstrated.

16. Consider, for example, that changing interest rate levels will not have a significant impact on the cash flow portion of the exercise (as the critical time period falls within weeks or months, rather than years) but it will have a rather large effect on the value of assets in the liquidity buffer (particularly those securities that have long durations and are thus very price/rate sensitive).

17. One area of study that attempts to provide better estimates of the tails of statisti-cal distributions is extreme value theory (EVT). EVT estimates probabilities by fitting a model to the tail of a probability function using only extreme event data; a tail index is derived and the thickness of the distribution in extreme regions can be analyzed. EVT is used in catastrophic risk analysis by insurers and reinsurers to capture the low frequency/high severity events to which they are potentially exposed. We might argue that a systemic liquidity crisis, or perhaps a multi-insti-tution liquidity crisis, falls into that category, suggesting that EVT may be useful.

18. One further note bears mentioning for institutions that are forced to dispose of securities held in investment, rather than trading, accounts. Under International Accounting Standards, 35 so-called hold to maturity assets need not be marked to market. However, if these hold to maturity investments are liquidated in support of current liquidity requirements, additional penalties may accrue to the firm, diminishing the cash value that can be extracted from the portfolio.

19. The haircuts applied to fixed assets can be more challenging to determine because they are based on assets that are unique and often lack a ready market of buyers. An industrial company with an unencumbered factory valued at $500 million (after depreciation) may not receive the full value in a disposal or pledging sce-nario; how much it receives will depend on both the perceived worth of the fac-tory to a lender or buyer and the time horizon during which transaction execution must occur. Engineers and auditors must perform due diligence on the estimated sales value of the factory, and investment bankers might then approach commer-cial banks and/or competitors to determine whether they would be willing to lend or buy at the independently assessed value. The process is time-consuming and far from transparent, meaning an ex-ante haircut value can be difficult to ascertain. The time dimension again features in the equation, with rapid disposal or pledg-ing creating a larger discount. As with any illiquid asset, it is reasonable to assume that in most instances the haircut will be considerable, 2575 percent of book value, and possibly even more.

9 Controlling Liquidity Risk

1. These are often driven by expectations regarding key financial indicators such as economic growth, interest rates, term structure movements, foreign exchange rates, consumer confidence, and inflation.

2. In the banking sector, for example, the BIS has specifically noted in its Liquidity Risk discussion paper from 2008 (Principle 4) that a bank should incorporate liquidity costs, benefits and risks in product pricing, performance measurement and new product approvals.

290 Notes

3. Consider, for instance, that in the United States a bank holding company (BHC) cannot accept deposits or make use of the Federal Reserve discount window, and must therefore rely very heavily on the interbank market for liquidity. In addi-tion, a BHC faces considerable constraints regarding upstreaming of funds and divi-dends from main bank units, putting even greater pressures on funding (e.g., credit from a main bank unit to a BHC must be collateralized, a main bank unit cannot repay a BHCs obligations, dividends can only be moved upstream as long as the main banks capital ratios remain sound). Similar restrictions exist in the insur-ance industry and in other regulated/unregulated industrial sectors. All of these relationships must be well understood before liquidity can be properly controlled, hence the importance of a legal entity and consolidated focus.

4. To give just one example, the US Office of the Comptroller of the Currency (OCC) believes that a funding concentration exists when a single decision or single market factor could lead to a large withdrawal of funds. The OCC has also indicated that US banks must be very cautious about over-reliance on the interbank funding market, even if activity is spread out among a large number of banking institutions.

5. In some cases, institutions can create products that allow better control of what might otherwise be an uncertain maturity; for instance, a liability may include an early exercise penalty that effectively dissuades presentation prior to the original contract maturity.

6. Even in instances when a MAC is not publicly triggered, there is the possibility that bankers will press the borrowing company to take certain actions or enact certain reforms that will add pressure to its financial position.

7. It is interesting to note that at least one banking regulator has proposed the estab-lishment of one week and one month cash flow gap limits built atop multi-week dynamic stress factors that are applied to actual and contingent inflows and out-flows. Larger stress factors are used when currency mismatches exist. This approach can be viewed as a matrix of hybrid multi-currency cash flow gap limits that have been recalibrated for the effects of stress testing.

8. Consider the case of US bank Wachovia, which suffered greatly during the financial crisis. Specifically, in the aftermath of the Lehman collapse, the bank lost approxi-mately $6.8 billion in deposits and commercial paper. Though the bank still held a liquidity buffer of more than $100 billion, it had done a poor job of communi-cating effectively with its financing providers; concerns about the banks mort-gage portfolio and lack of clear answers led to growing nervousness. Shortly after the silent run on the deposits, counterparties demanded repayment of a further $50 billion of credit. Over the following weekend, US regulators determined that Wachovia stood to lose a further $115 billion of funding from retail and corporate deposits (versus $100 billion in coverage). Given this liquidity squeeze, regulators approved an 11 th hour acquisition of Wachovia by beleaguered Citibank; the bid was trumped at the last minute by the much stronger Wells Fargo, which essen-tially saved Wachovia from collapse. While we may never know what might have happened had Wachovia been more forthright and constructive in its dialogue and relationships with funders, it is an important lesson to bear in mind. Loss of confi-dence can occur quickly and can be damaging.

9. In some instances regulations or accounting conventions require firms to comply with specific liquidity-related disclosures. For instance, in the regulatory commu-nity, the US Securities and Exchange Commission, commenting on Management Discussion and Analysis of Financial Condition and Results, has indicated:

In determining required or appropriate disclosure, companies should evalu-ate separately their ability to meet upcoming cash requirements over both the short-term and long-term. Merely stating that a company has adequate resources is insufficient unless no additional more detailed or nuanced information is

Notes 291

material. One starting point for a companys discussion and analysis of cash requirements is a tabular disclosure of contractual obligations, supplemented with additional information that is material to an understanding of a companys cash requirement. Companies should address, where material, the difficulties involved in assessing the effect of amount and timing of uncertain events, such as loss contingencies, on cash requirements and liquidity. Any such discussion should be specific to the circumstances and informative.

(SEC, 2003) Within the accounting sector, the International Account Standards Board, ref-

erencing disclosures related to financial instruments, has noted that an entity shall disclose (a) a maturity analysis that shows for financial liabilities the remain-ing contract maturities; and (b) a description of how it manages the liquidity risk inherent in (a). Other accounting systems have requirements of their own.

10 Liquidity Crisis Management

1. Reductions in net working capital are very industry dependent. In some instances working capital is relatively high, e.g., construction (2025% of assets) while in others it is comparatively low, e.g., oil and gas, mining (10%). Those with higher working capital certainly have greater flexibility in reducing as needed; common methods include collecting receivables, reducing inventories, and managing paya-bles more conservatively.

11 New Regulatory Initiatives

1. Level 1 assets are limited to: Coins and banknotes Central bank reserves (including required reserves) to the extent that the central bank policies allow them to be drawn down in times of stress Marketable securities representing claims on or guaranteed by sovereigns, central banks, PSEs, the Bank for International Settlements, the International Monetary Fund, the European Central Bank and European Community, or mul-tilateral development banks, as long as they have a 0% risk weight under Basel II, traded in a deep market with a proven track record of liquidity under all conditions, and not issued by a financial institution Domestic (local currency) sovereign or central bank debt for non-0% risk weighted sovereigns.

Level 2A assets include: Marketable securities representing claims on or guaranteed by sovereigns, central banks, PSEs or multilateral development banks that satisfy all of the following conditions: assigned a 20% risk weight under the Basel II Standardized Approach for credit risk; traded in large, deep and active repo or cash markets character-ized by a low level of concentration; have a proven record as a reliable source of liquidity in the markets (repo or sale) even during stressed market conditions (i.e., maximum decline of price not exceeding 10% or increase in haircut not exceeding 10 percentage points over a 30-day period during a relevant period of significant liquidity stress); and not an obligation of a financial institution or any of its affiliated entities. Corporate debt securities (including commercial paper) and covered bonds that satisfy all of the following conditions: in the case of corporate debt securities: not issued by a financial institution or any of its affiliated entities; in the case

292 Notes

of covered bonds: not issued by the bank itself or any of its affiliated entities; and are plain vanilla in nature, with no complex features or subordination. Additionally, they must

have a long-term credit rating from a recognized external credit assessment institution (ECAI) of at least AA-21 or in the absence of a long-term rating, a short-term rating equivalent in quality to the long-term rating; or (ii) do not have a credit assessment by a recognized ECAI but are internally rated as having a probability of default (PD) corresponding to a credit rating of at least AA-;

be traded in large, deep and active repo or cash markets characterized by a low level of concentration; and

have a proven record as a reliable source of liquidity in the markets (repo or sale) even during stressed market conditions: i.e., maximum decline of price or increase in haircut over a 30-day period during a relevant period of significant liquidity stress not exceeding 10%.

Level 2B assets include:

Residential mortgage backed securities (RMBS) that satisfy all of the following conditions may be included in Level 2B, subject to a 25% haircut:

not issued by, and the underlying assets have not been originated by, the bank itself or any of its affiliated entities;

have a long-term credit rating from a recognized ECAI of AA or higher, or in the absence of a long-term rating, a short-term rating equivalent in quality to the long-term rating;

traded in large, deep and active repo or cash markets characterized by a low level of concentration; have a proven record as a reliable source of liquidity in the markets (repo or

sale) even during stressed market conditions, i.e., a maximum decline of price not exceeding 20%;

the underlying asset pool is restricted to residential mortgages and cannot contain structured products;

the underlying mortgages are full recourse loans (i.e., in the case of fore-closure the mortgage owner remains liable for any shortfall in sales pro-ceeds from the property) and have a maximum loan-to-value ratio (LTV) of 80% on average at issuance; and

the securitizations are subject to risk retention regulations which require issuers to retain an interest in the assets they securities.

Corporate debt securities (including commercial paper) that satisfy all of the following conditions may be included in Level 2B, subject to a 50% haircut:

not issued by a financial institution or any of its affiliated entities; either (i) have a long-term credit rating from a recognized ECAI between

A+ and BBB- or in the absence of a long term rating, a short-term rating equivalent in quality to the long-term rating; or (ii) do not have a credit assessment by a recognized ECAI and are internally rated as having a PD corresponding to a credit rating of between A+ and BBB-;

traded in large, deep and active repo or cash markets characterized by a low level of concentration; and

have a proven record as a reliable source of liquidity in the markets (repo or sale) even during stressed market conditions, i.e., a maximum decline of price not exceeding 20% or increase in haircut over a 30-day period not exceeding 20 percentage points during a relevant period of significant liquidity stress.

Notes 293

Common equity shares that satisfy all of the following conditions may be included in Level 2B, subject to a 50% haircut:

not issued by a financial institution or any of its affiliated entities; exchange traded and centrally cleared; a constituent of the major stock index in the home jurisdiction or where

the liquidity risk is taken, as decided by the supervisor in the jurisdiction where the index is located;

denominated in the domestic currency of a banks home jurisdiction or in the currency of the jurisdiction where a banks liquidity risk is taken;

traded in large, deep and active repo or cash markets characterized by a low level of concentration; and

have a proven record as a reliable source of liquidity in the markets (repo or sale) even during stressed market conditions, i.e., a maximum decline of share price not exceeding 40% or increase in haircut not exceeding 40 percentage points over a 30-day period during a relevant period of sig-nificant liquidity.

2. Another definition of counterbalancing capacity is given as the liquid asset portfolio plus new short-term debt plus new long-term debt.

3. The rules distinguish between Individual Liquidity Assessment Standards (ILAS) and non-ILAS entities. ILAS include banks, building societies and investment companies, while non-ILAS include small banks, limited license operations and limited activity operations.

12 Summary: Toward Active Liquidity Risk Management

1. Consider, for example, that UK securities firms were historically not directly sub-ject to the liquidity rules imposed by the Bank of England on other domestic bank-ing institutions, even though they may have had considerable liquidity exposures. That has changed with FSAs BIPRU 12.

2. Elimination of too big to fail has been a priority for a number of national regula-tors, who seek to avoid a repetition of the bailouts they were party to in 20082009. While various initiatives have been put forth, including shrinking bank activities and granting regulators defined authorities in a wind down or resolution phase, there does not seem to be any widespread belief that the too big to fail concept, with its attendant call on taxpayers, has actually been resolved.

3. The process should also involve ancillary and indirect forms of support, such as temporarily relaxing collateral, reserve, and solvency requirements for other insti-tutions in the system, and sterilizing any system-wide liquidity injections in order to avoid inflationary pressures and currency weakness. In general, more extreme government actions, such as interest rate cuts, capital controls, or deposit freezing should be avoided, as these can send a potentially damaging signal to the market-place at large, and actually affect liquidity.

294

Selected References

American Academy of Actuaries (AAA) (2000) Report of the American Academy of Actuaries Life Liquidity Working Group, Washington, DC: AAA.

American Bankers Association (ABA) (2011) Dodd Frank Act Implementation: Executive Overview, Washington DC: ABA.

Bangia, A., Diebold, F., Schuermann, T., and Stroughair, J. (1999) Modeling Liquidity Risk with Implications for Trading Market Risk Management, Working Paper, University of Pennsylvania.

Banks, E. (2004a) Corporate Governance , Basingstoke, UK: Palgrave Macmillan. Banks, E. (2004b) The Credit Risk of Complex Derivatives , 3rd edn, Basingstoke, UK:

Palgrave Macmillan. Banks, E. (2006) The Failure of Wall Street , NY: Palgrave Macmillan. Banks, E. (2010) Risk and Financial Catastrophe , UK: Palgrave Macmillan. Banks, E. (2011) See No Evil , Basingstoke, UK: Palgrave Macmillan. Banks, E. and Dunn, R. (2003) Practical Risk Management , London: John Wiley. Barrett, A. (2003) Cash: Burn, Baby, Burn, Business Week , April 28. Basel Committee on Banking Supervision (BIS) (1992) A Framework for Measuring

and Managing Liquidity , Basel: BIS. Basel Committee on Banking Supervision (BIS) (2008) Principles for Sound Liquidity

Risk Management , Basel: BIS. Basel Committee on Banking Supervision (BIS) (2010) Basel III: International

Framework for Liquidity Risk Management Standards and Monitoring , Basel: BIS. Basel Committee on Banking Supervision (BIS) (2013) Basel III: The LCR and

Liquidity Risk Monitoring Tools , Basel: BIS. BIS (1999) Market Liquidity: Research Findings and Selected Policy Implications, Basel:

BIS. BIS (2000) Sound Practices for Managing Liquidity in Banking Organizations , Basel:

BIS. Berkowitz, J. (2000) Incorporating Liquidity Risk into VAR models, Working

Paper, University of California at Irvine. Bernstein, P. (1987) Liquidity, Stock Markets, and Market-Makers, Financial

Management , Summer, pp. 5462. Borio, C., (2009) Ten Propositions about Liquidity Crises, BIS Working Paper

293, Basel: BIS. Brockman, P. and Chung, D. (2001) Managerial Timing and Corporate Liquidity:

Evidence from Actual Share Repurchases, Journal of Financial Economics , Vol. 61, No. 3 (September).

Brown, K., Fiedler, R., and Moloney, J. (2001) Liquidity In a Dry Climate, E-Risk Portal , New York: E-Risk.

Buraschi, A. (2002) Liquidity Risk and Specialness, Journal of Financial Economics , Vol. 64, No. 2 (May).

Canadian Institute of Actuaries (CIA) (1996) Liquidity Risk Measurement, Subcommittee on Liquidity Risk Measurement, Ottawa, Canada.

Caprio, G. and Klingebiel, D. (1999) Episodes of Systemic and Borderline Financial Crises, World Bank Monograph, Washington, DC: World Bank.

Selected References 295

Chari, V. and Jagannathan, R. (1988) Banking Panics, Information, and Rational Expectations Equilibrium, Journal of Finance , Vol. 43.

Committee of European Bank Supervisors (CEBS) (2009) Consultation Paper on Liquidity Buffers and Survival Periods, Brussels: CEBS.

Committee of European Bank Supervisors (CEBS) (2010) EU Banks Liquidity Stress Testing and Contingency Funding Plans, Brussels: CEBS.

Covitz, D. and Downing, C. (2002) Insolvency Or Liquidity Squeeze? Federal Reserve Bank Bulletin, Washington, DC: Federal Reserve Bank (FRB).

Coyle, B. (2000) Cash Flow Forecasting and Liquidity , Chicago: Glenbrook. Davis, E. (1999) a Reappraisal of Market Liquidity Risk in the Light of the

Russia/LTCM Global Securities Market Crisis, Presentation at Basel Banking Supervisors Conference.

Decker, P. (2000) The Changing Character of Liquidity and Liquidity Risk Management, Chicago: Federal Reserve Bank of Chicago.

Desocio, A. (2011) The Interbank Market after the Financial Turmoil, Bank of Italy Working Paper 819, Rome: Bank of Italy.

Diamond, D. and Dybvig, P. (1983) Bank Runs, Deposit Insurance, and Liquidity, Journal of Political Econom y, Vol. 91 (June).

Diamond, D. and Rajan, R. (1999) Liquidity Risk, Liquidity Creation and Financial Fragility: a Theory of Banking, Working Paper, National Bureau of Economic Research.

Drehmann, M. and Nikolaou, K. (2010) Funding Liquidity Risk: Definitions and Measurement, BIS Working Paper 3164, Basel: BIS.

Dudley, W. (2009) More Lessons from the Crisis, Federal Reserve Bank of New York speech, New York: FRB.

Ernst and Young (2012) Bank Risk Management Survey: Liquidity Management , New York: E&Y.

Erzegovesi, L. (2002) VaR and Liquidity Risk, Working Paper, University of Trento, Italy.

European Commission (2009) Directive 2009/138 of the European Parliament on Insurance and Reinsurance (Solvency II), Brussels: EC.

Federal Deposit Insurance Corporation (FDIC) (1998) DOS Manual of Examination Policies: Liquidity and Funds Management, Washington, DC: FDIC.

Federal Reserve Bank of San Francisco (2008), What Is Liquidity Risk, Economic Letter 200833, San Francisco: FRBSF.

Fernandez, F. (1999) Liquidity Risk: New Approaches to Measurement and Monitoring, Working Paper, Securities Industry Association.

Financial Services Authority (FSA) (2003) Liquidity Risk in the Integrated Prudential Sourcebook: A Quantitative Framework , London: FSA.

Financial Services Authority (FSA) (2008) Strengthening Liquidity Standards, London: FSA.

Financial Services Authority (FSA) (2010) Liquidity Regime and Non-ILAS BIPRU Firms , London: FSA.

Gatev, E. and Straham, P. (2001) Banks Advantage in Hedging Liquidity Risk: Theory and Evidence from the Commercial Paper Market, Working Paper, University of Pennsylvania.

Gorton, G. (1988) Banking Panics, and Business Cycles, Oxford Economic Papers , Vol. 40.

Gorton, G. and Huang, K. (2002) Liquidity, Efficiency, and Bank Bailouts, Working Paper, University of Pennsylvania.

296 Selected References

HM Treasury (1999) UK Economic Glossary , London: HM Treasury. International Association of Insurance Supervisors (IAIS) (2000) Solvency

Subcommittee Issues Paper, Basel: IAIS. International Institute of Finance (2009) Reform in the Financial Services

Industry, Washington DC: IIF. International Organization of Securities Commissioners (IOSCO) (2002) Sound

Practices for the Management of Liquidity Risk at Securities Firms , Geneva: IOSCO. Jameson, R. (2001) Whos Afraid of Liquidity Risk? E-Risk Portal , New York:

E-Risk. Jarrow, R. and Subramanian, A. (1997) Mopping Up Liquidity, Risk , December

10. Jorion, P. (1996) Value at Risk , Burr Ridge, IL: Irwin. Kyle, A. (1995) Continuous Auctions and Insider Trading, Econometrica , Vol. 53,

No. 6. Locke, P. and Sarkar, A. (1996) Volatility and Liquidity in Futures Markets,

Federal Reserve Bank of New York Paper, New York: FRB. Lott, F. and Marcucci, J. (2003) Corporate Liquidity Demand in the US, Working

Paper, Bank of Italy and University of California San Diego. Lucas, R. and Stokey, N. (2011) Liquidity Crisis: Understanding Sources and

Limiting Consequences, Federal Reserve Bank of Minneapolis Economic Policy Paper 113, Minneapolis: FRBM.

Lucas, Y. (2009) The Impact of Financial Crises on Transfer Systems, Paris: Banque de France.

Matz, L. (2011) Liquidity Risk Measurement and Management: Basel III and Beyond , New York: Xlibris Corp.

Matz, L. and Neu, P. (2006) Liquidity Risk: Measurement and Management, Singapore: John Wiley.

Muranaga, J. and Ohsawa, M. (1999) Measurement of Liquidity Risk in the Context of Market Risk Calculation, Bank of Japan Bulletin , Tokyo: Bank of Japan.

Nikolaou, K. (2009) Liquidity Risk Concepts, ECB Working Paper 1008, Brussels: ECB.

Northern Rock (2006) Annual Report, and Accounts, Newcastle upon Tyne. Office of the Comptroller of the Currency (OCC) (2001) Liquidity: Comptrollers

Handbook , Washington, DC: OCC. Office of the Superintendent of Financial Institutions Canada (OSFI) (1995)

Liquidity Guideline , Ottawa, Canada: OSFI. Schlingemann, F., Stulz, R., and Waking, R. (2001) Divestitures and the Liquidity

of the Market for Corporate Assets, Working Paper, National Bureau of Economic Research.

Securities and Exchange Commission (SEC) (2003) Management Discussion and Analysis of Financial Condition and Results: Interpretative Guidance, Washington, DC: SEC.

Sender, H. and Singer, J. (2003) a Betting Man and His Funds Hard Fall, Wall Street Journal , April 9.

Sprinzen, S. and Petit, M. (2002) Framework for Analyzing Corporate Liquidity Under Stress, Standard and Poors CreditWeek , New York: Standard and Poors.

Taleb, N. (1996) Dynamic Hedging, New York: John Wiley.

297

active liquidity risk management, future of 25969

macro role 2639avoiding fragmentation 266conducting regular inspections 265harmonizing accounting

treatment 2667minimizing costs 266providing selective lender of last

resort support 2678reinforcing proper capital

allocation 267micro role: best practices 2603

creating a sound governance framework 2601

developing and implementing a crisis management process 263

implementing proper measures and reporting 2612

performing ongoing reviews 265using tactical controls 262

asset liquidity risk 5, 93107effects of 1067exogenous considerations 968nature of asset liquidity risk

problems 1016excessive concentrations 1023insufficient collateral 1056lack of asset marketability 99101lack of unencumbered assets 1012misvalued assets 1035

sources of 96asset liquidity, sources of 4955

Bank for International Settlements (BIS) 16

Basel III 24352BIS, see Bank for International

Settlements

case studies 12252Askin Capital 1269Drexel Burnham Lambert 1236Enron 1404General American 1356Lehman Brothers 14852Long Term Capital Management 1314

Northern Rock 1448Orange County 12931Swissair 13640

cash flow gaps 1627Committee of European Banking

Supervisors Liquidity Consultation 2534

conduit 16contagion 1201controlling liquidity risk 188226

governance structure 18890liquidity risk controls 196218

asset liquidity controls 197204collateral/pledging limits 2024liquidity and fixed asset

limits 198202funding liquidity controls 20410

committed facility limits 20710diversified funding limits 2057

joint liquidity controls 21012balance sheet target limits

21112cash flow gap limits 21011hybrid ratio limits 212

off balance sheet controls 21214forward commitments and

contingencies 21314other safeguards 21519

defensive interval 21516external relationship

management 21819mark and model

verification 21617penalties 21718reserves 216

liquidity risk management 1914financial and human

resources 1924risk plan 1912

liquidity risk monitoring 21926asset and funding portfolios 220forward balance sheet 221general indicators 2223management goals 2245off balance sheet commitments and

contingencies 221stress scenarios 2212

Index

298 Index

controlling liquidity risk continuedtechnical capabilities 2256

management duties 1949corporate liquidity ratios 160counterbalancing capacity 251credit risk 4, 13

defensive interval 21516defensive measures 2327definition of liquidity risk 37depth 168Dodd Frank Act 2523duration 2702

endogenous liquidity 5, 910, 43exogenous liquidity 5, 910, 47, 968

financial distress 1191201financial institution liquidity ratios 161financial instrument liquidity

measures 16873financial risk 3Financial Services Authority BIPRU 12,

2546funding liquidity risk 4, 7792

effects of 912exogenous considerations 857nature of funding problems 8791

commitment withdrawals 8990excessive concentrations 901lack of market access 889rollover problems 878

sources of 7885mismanagement 84negative perceptions/market

actions 845unfavorable legal/regular

actions 834unpredictable cash flows 7883

funding liquidity, sources of 5564

gap analysis 2712gap management 2709

common measures 2729duration 2702interest rate activities 270

haircut 501high quality liquid assets (HQLA) 50,

2446HQLA, see High quality liquid assets

joint asset/funding liquidity risk 5, 10813

liquid assets 4953liquidity adjusted value at risk 1703liquidity committee 18890liquidity contingency risk 5liquidity coverage ratio 161, 2446liquidity crisis management 22741

defensive measures 2327communication 2367recentralization 2326

asset management 235funding management 2325risk hedging 236

disaster recovery 239ex-ante market access 2312invoking and terminating the

program 2379scope and focus 22731testing the plan 23941

liquidity mismatch risk 5liquidity ratios 15762liquidity risk

controlling 188226measuring 15587monitoring 21926policies 196pricing 1934

liquidity risk and financial operations 2447

endogenous versus exogenous liquidity 437

external requirements 2931financial imperatives 289general approaches to liquidity

management 278liquidity operating requirements

247liquidity profiles by industry 3342

capital intensive companies 412financial institutions 3440municipalities and sovereigns 42non-financial service

companies 401liquidity risk/return tradeoff

313liquidity risk controls 196218liquidity risk defined 323

definition of 37liquidity risk and financial

crises 1422liquidity, risk and the

corporation 713market risk, credit risk and liquidity

risk 1314

Index 299

liquidity risk management, active 25969

liquidity, sources of 4873liquidity spirals and financial

distress 10821contagion 1201financial distress 1191201joint asset and funding risks 10813

causes 11213problems 10911

liquidity spirals 11319causes 11519

debt investors and banks 11618management reaction 119rating agencies 11819s

problems 11315liquidity warehouse 53, 199

market risk 4, 13, 93measuring liquidity risk 15587

common liquidity measures 15586cash flow gaps 1627financial instrument liquidity

measures 16873liquidity ratios 15762stress tests 17387

asset disposals, secured funding and haircuts 17982

cash flows 1789collateral 1845covenants and

terminations 1834currency exposures 185event risks and joint

scenarios 1856liquidation horizon 178market parameters 1767stress test framework 1756unsecured funding 1823

net funding requirement 1635, 211net interest income 273net interest margin 273net stable funding ratio 161, 24650new regulatory initiatives 24258

Basel III 24352monitoring 2502new liquidity measures 24350

liquidity coverage ratio 2446net stable funding ratio 24650

Committee of European Banking Supervisors Liquidity Consultation 2534

Dodd Frank Act 2523Financial Services Authority BIPRU 12,

2546solvency II and other insurance

regulations 2568

off balance sheet liquidity, sources of 6471

operating risk 34

rate sensitive assets 1667, 271rate sensitive liabilities 1667, 271rating agencies 301repurchase agreement 16resiliency 1689risk appetite 1923

securitization 17, 656shadow banking 16silent bank run 17, 146SIV, see Structured investment vehiclesolvency 25solvency II and other insurance

regulations 2568sources of liquidity 4873

amalgamating sources of liquidity 713

sources of asset liquidity 4955fixed assets and intangibles 545

fixed assets 545intangibles 55

liquid assets 4953cash and marketable

securities 4952inventories 523receivables 52

sources of funding liquidity 5564equity capital 634medium/long-term funding

marketsfunding agreements and

guaranteed investment contracts 603

loans 623long-term bonds 612medium term notes and

Euronotes 60short-term funding markets 5760

commercial paper and euro commercial paper 57

deposits and

Appendix : A Primer on Gap Management - Home - Springer978-1-137-37440-0/1.pdf · Appendix : A...

Documents

Transcript of Appendix : A Primer on Gap Management - Home - Springer978-1-137-37440-0/1.pdf · Appendix : A...