Asset–liability management (ALM) is a generic term that can mean different things in subtly distinct ways, depending on what type of financial market participant one is. For bankers, the term is used to denote high‐level management of a bank's balance sheet assets and liabilities; in other words, it refers to the process of managing the balance sheet. As such it is a strategy level discipline but at the business line level, it is also a tactical one.

The art of asset and liability management is essentially one of balance sheet risk management and capital management, and although the day‐to‐day activities are run at the desk level, overall direction should be given at the highest level of a banking institution. The risk exposures in a banking environment are multidimensional, and as we have seen they encompass interest‐rate risk, liquidity risk, credit risk, and operational risk.

Traditionally, asset and liability management covered the set of techniques used to manage interest‐rate and liquidity risks; it also dealt with the structure of the bank's balance sheet, which is heavily influenced by funding and regulatory constraints and profitability targets. Interest‐rate risk is one type of market risk. Risks associated with moves in interest rates and levels of liquidity are those that result in adverse fluctuations in earnings levels due to changes in market rates and bank funding costs. By definition, banks' earnings levels are highly sensitive to moves in interest rates and the cost of funds in the wholesale market.

One of the major areas of decision‐making in a bank involves the maturity of assets and liabilities. Typically, longer term interest rates are higher than shorter‐term rates; that is, it is common for the yield curve in the short term (say, 0–3 year range) to be positively sloping. To take advantage of this, banks usually raise a large proportion of their funds from the short‐dated end of the yield curve and lend out these funds for longer maturities at higher rates. The spread between the borrowing and lending rates is the bank's profit. The obvious risk from such a strategy is that the level of short‐term rates rises during the term of the loan, so that when the loan is refinanced the bank makes a lower profit or a net loss. A more critical risk is that funding no longer becomes available when a deposit is rolled over. Managing this risk exposure is the key function of an ALM desk.

Another risk factor is liquidity. From a banking and Treasury point of view, the term liquidity means funding liquidity, or the “nearness” of money. The most liquid asset is cash money. Banks bear several interrelated liquidity risks, including the risk of being unable to pay depositors on demand, an inability to raise funds in the market at reasonable rates, and an insufficient level of funds available with which to make loans. Banks keep only a small portion of their assets in the form of cash, because this earns only a very low return for them, but must make other arrangements to ensure continuous liquidity at all times.

Despite its name, throughout history since the term was first described in the early 1970s, the principal function of the ALM desk has been to manage interest‐rate risk and liquidity risk. This is hardly managing everything to do with the bank's “assets and liabilities”. In some banks the ALM discipline will also encompass the setting of overall policy for credit risk and credit risk management, with tactical level credit policy being set at a lower level within credit committees. But this is rare – in most banks the ALM function (and indeed the asset and liability committee (ALCO)) gets nowhere near credit risk policy and credit risk management. The author considers this not to be best practice, and later on we will provide the rationale why a bank's ALCO oversight remit should extend to credit risk as well, thus living up to the committee's name.

In this chapter we introduce the basic tenets of traditional bank ALM and the key ALM concepts of liquidity, interest‐rate risk, and ALM policy. A subsequent chapter describes modern‐day best practice principles and organisation of the Asset and Liability Committee (ALCO), the paramount balance sheet risk management committee in a bank.

BASIC CONCEPTS

In the era of stable interest rates that preceded the breakdown of the Bretton–Woods agreement, ALM was a more straightforward process, constrained by regulatory restrictions and the saving and borrowing pattern of bank customers. The introduction of the negotiable Certificate of Deposit (CD) by Citibank in the 1960s enabled banks to diversify both their investment and funding sources. With this there developed the concept of the interest margin, which is the spread between the interest earned on assets and that paid on liabilities. This led to the concept of the interest gap and the management of the gap, which is the cornerstone of modern‐day ALM. The increasing volatility of interest rates, and the rise in absolute levels of rates themselves, made gap management a vital part of running the banking book. This development meant that banks could no longer rely permanently on the traditional approach of borrowing short (funding short) to lend long, as a rise in the level of short‐term rates would result in funding losses. The introduction of derivative instruments such as FRAs and swaps in the early 1980s removed the previous uncertainty and allowed banks to continue the traditional approach while hedging against medium‐term uncertainty.

In financial markets, two important strands of risk management are interest‐rate risk and liquidity risk. ALM practice is concerned with managing this risk. Interest‐rate risk exists in two strands. The first strand is the more obvious one: the risk of changes in asset–liability value due to changes in interest rates. Such a change impacts the cash flows of assets and liabilities as well as their present value because financial instruments are valued with reference to market interest rates. The second strand is that associated with optionality, which arises with products such as early‐redeemable loans. The other main type of risk that ALM seeks to manage is liquidity risk, which refers to both the liquidity of markets and an ability to maintain funding of assets and deposit withdrawals, but also to the ease with which assets can be translated into cash. The former is the primary concern of all banks large and small.

ALM is conducted primarily at an overview, balance sheet level. The risk that is managed is an aggregate, group level risk. This makes sense because one could not manage a viable banking business by leaving interest‐rate and liquidity risk management at individual operating levels. We illustrate this in Figure 7.1, which illustrates the cornerstones of traditional ALM. Essentially, interest‐rate risk exposure is managed at the group level by the Treasury desk. The drivers are the different currency interest rates, with each exposure being made up of the net present value (NPV) of cash flow as it changes with changes in interest rates. The discount rate used to calculate NPV is the prevailing market rate for each time bucket in the term structure.

Interest‐rate exposure arises because rates fluctuate from day to day, and continuously over time. The primary risk is that of interest‐rate reset for floating‐rate assets and liabilities. The secondary risk is liquidity risk: unless assets and liabilities are matched by amount and term, assets must be funded on a continuous rolling basis. Equally, the receipt of funds must be placed on a continuous basis. Whether an asset carries a fixed‐ or floating‐rate, reset will determine its exposure to interest‐rate fluctuations. Where an asset is marked at a fixed‐rate, a rise in rates will reduce its NPV and so reduce its value to the bank. This is intuitively easy to grasp, even without recourse to financial arithmetic, because we can see that the asset is now paying a below‐market rate of interest. Or we can think of it as a loss due to opportunity cost foregone, since the assets are earning below what they could earn if they were employed elsewhere in the market. The opposite applies if there is a fall in rates: this causes the NPV of the asset to rise. For assets marked at a floating‐rate of interest, the exposure to fluctuating rates is much less, because the rate receivable on the asset will reset at periodic intervals, which will allow for changes in market rates.

We speak of risk exposure as being for the group as a whole. This exposure must therefore aggregate the net risk of all the bank's operating business. Even for the simplest banking operation, we can see that this will produce a net mismatch between assets and liabilities, because different business lines will have differing objectives for their individual books. This mismatch will manifest itself in two ways:

• The mismatch between the different terms of assets and liabilities across the term structure;
• The mismatch between the different interest rates that each asset or liability contract has been struck at.

This mismatch is known as the ALM gap. The first type is referred to as the liquidity gap, while the second is known as the interest‐rate gap. We value assets and liabilities at their NPV; hence, we can measure the overall sensitivity of balance sheet NPV to changes in interest rates. As such, then, ALM is an art that encompasses aggregate balance sheet risk management at the group level.

Figure 7.2 shows the aggregate group level ALM profile for a derivatives trading house based in London. There is a slight term mismatch as no assets are deemed to have “overnight” maturity, whereas a significant portion of funding (liabilities) is in the overnight term. One thing we do not know from looking at Figure 7.2 is how this particular institution defines the maturity of its assets. To place them in the relevant maturity buckets, one can adopt one of two approaches, namely:

• Actual duration of the assets;
• “Behavioural” tenor or expected life of the assets;
• “Liquidity duration”, which is the estimated time it would take the firm to dispose of its assets in an enforced or “firesale” situation, such as a withdrawal from business.

Each approach has its adherents. It is up to the individual institution to adopt one method and then adhere consistently to it. The third approach has the disadvantage, however, of being inherently subjective – estimating the time taken to dispose of an asset book is not an exact science and is little more than educated guesswork. Nevertheless, for long‐dated and/or illiquid assets, it is at least a workable method that enables practitioners to work around a specified ALM framework with regard to structuring the liability profile.

LIQUIDITY GAP

There is clearly risk exposure as a result of liquidity mismatches between assets and liabilities. Maturity terms will not match, thereby creating a liquidity gap. The amount of assets and liabilities maturing at any one time will also not match (although overall, by definition, assets must equal liabilities). Liquidity risk is the risk that a bank will not be able to refinance assets as liabilities become due for any reason.¹ To manage this, the bank will hold a portion of its assets in very liquid form.² A surplus of assets over liabilities creates a funding requirement. If there is a surplus of liabilities, the bank will need to find efficient uses for these funds. In either case, the bank has a liquidity gap. This liquidity can be projected over time so that one knows what the situation is each morning, based on net expiring assets and liabilities. The projection will change daily, of course, due to the new business undertaken each day.

We could eliminate liquidity gap risk by matching assets and liabilities across each time bucket. Actually, at the individual loan level this is a popular strategy: if we can invest in an asset paying 5.50% for 3 months and fund this with a 3‐month loan costing 5.00%, we have locked in a 50bp gain that is interest‐rate risk free. However, this is not what banks actually do, as they are in the business of undertaking maturity transformation. Hence, liquidity risk is a key consideration in ALM. A bank with a surplus of long‐term assets over short‐term liabilities will have an ongoing requirement to fund the assets continuously, and there is the ever‐present risk that funds may not be available as and when they are required. The concept of a future funding requirement is itself a driver of interest‐rate risk, because the bank will not know the future interest rates at which it will deal.³ So a key part of ALM involves managing and hedging this forward liquidity risk.

The liquidity gap is the difference, at all future dates, between assets and liabilities of the banking portfolio. Gaps generate liquidity risk. When liabilities exceed assets, there is an excess of funds. An excess does not of course generate liquidity risk, but it does generate interest‐rate risk, because the present value of the book is sensitive to changes in market rates. When assets exceed liabilities, there is a funding deficit and the bank has long‐term commitments that are not currently funded by existing operations. The liquidity risk is that the bank requires funds at a future date to match the assets. The bank is able to remove any liquidity risk by locking in maturities, but of course there is a cost involved as it will be dealing at longer maturities.⁴

Definition and illustration

To reiterate then, the liquidity gap is the difference in maturity between assets and liabilities at each point along the term structure. Because ALM in many banks concerns itself with medium‐term management of risk, this may be a 5‐year horizon. Note from Figure 7.2 how the longest dated time bucket in the ALM profile extended out to only “12 months+”, hence all liabilities longer than 1 year were grouped in one time bucket. This recognises the fact that most liabilities are shorter than 1 year, although a proportion of funding will be longer term – an average of 5 years or so.

For each point along the term structure at which a gap exists, there is (liquidity) gap risk exposure. This is the risk that funds cannot be raised as required, or that the rate payable on these funds is prohibitive.⁵ To manage this risk, a bank must:

• Disperse the funding profile (the liability profile) over more than just a short period of time. For example, it would be excessively risky to concentrate funding in just the overnight to 1‐week time bucket, so a bank will spread the profile across a number of time buckets. Figure 7.3 shows the liability profile for a European multi‐currency asset‐backed commercial paper programme, with liabilities extending from 1 month to 1 year;

  

• Manage expectations such that large‐size funding requirements are diarised well in advance – not planned for times of low liquidity such as the Christmas and New Year period;
• Hold a significant proportion of assets in the form of very liquid instruments such as very‐short‐term cash loans, Treasury bills, and high‐quality short‐term bank certificates of deposit (CDs).

Following these guidelines leads to a reserve of liquidity that can be turned into cash at very short notice in the event of a funding crisis.

The size of the liquidity gap at any one instant is never more than a snapshot in time, because it is constantly changing as new commitments are entered into on both the asset and liability size. For this reason, some writers speak of a “static” gap and a “dynamic” gap, but in practice one recognises that there is only ever a dynamic gap, because the position changes daily. Hence, we will refer only to a liquidity gap.

A further definition is the “marginal” gap, which is the difference between the change in assets and liabilities during a specified time period. This is also known as the “incremental” gap. If the change in assets is greater than the change in liabilities, this is a positive marginal gap, while if the opposite applies it is a negative marginal gap.⁶

We illustrate these values in Table 7.1 and graphically at Figure 7.4. This is a simplified asset–liability profile from a regional European bank, showing gap and marginal gap at each time period. Note that liabilities have been structured to produce an “ALM smile”, which is recognised as following prudent business practice. Generally, no more than 20% of total funding should be in the overnight to 1‐week time bucket – similarly for the 9‐to‐12‐month bucket. The marginal gap is measured as the difference between the change in assets and liabilities from one period to the next.

	1 week	1 month	3 month	6 month	9–12 month	>12 months	Total
Assets	10	90	460	710	520	100	1,890
Liabilities	100	380	690	410	220	90	1,890
Gap	90	290	230	300	300	10
Marginal	gap	200	60	530	0	290

MANAGING LIQUIDITY

Managing liquidity gaps and the liquidity process is both continuous and dynamic because the ALM profile of a bank changes on a daily basis. Liquidity management is the term used to describe this continuous process of raising and laying off funds, depending on whether one is long or short cash that day.

The basic premise is a simple one: the bank must be “squared off” by the end of each day, which means ensuring the net cash position is zero. Thus, liquidity management is both very short term as well as projected over the long term, because every position put on today creates a funding requirement in the future on its maturity date. The ALM desk must be aware of its future funding or excess cash positions and act accordingly, whether this means raising funds now or hedging forward interest‐rate risk.

The basic case: the funding gap

A funding requirement is dealt with on the day it occurs. The decision on how it will be treated will factor the term that is put on – it also has to allow for any new assets put on that day. As funding is arranged, the gap on that day will be zero. The next day there will be a new funding requirement or a surplus depending on the net position of the book.

This is illustrated in Figure 7.5. Starting from a flat position on the first day (t₀), we observe a gap (the dotted line) on t₁, which is closed by putting on funding to match the asset maturity. The amount of funding to raise and the term for it to run will take into account the future gap as well as that day's banking activities. So, at t₂ we observe a funding excess, which is then laid off. We see at t₃ that invested assets run beyond the maturity of the liabilities at t₂, so we have a funding requirement again at t₃. The decision on the term and amount will be based on the market view of the ALM desk. A matched book approach may well be taken where the desk does not have a strong view or if its view is at odds with market consensus.

There are also external factors to take into account. For instance, the availability of funds in the market may be limited, due to both macro‐level issues and to the bank's own ability to raise funds. The former might be during times of market correction or recession (a “credit crunch”), while the latter might include the bank's credit lines with market counterparties. Moreover, some funds will have been raised in the capital markets and this cash will cover part of the funding requirement. In addition, the ALM desk must consider the cost of the funds it is borrowing – for example, if it thought that interest rates in the short term, or for short‐term periods, were going to fall, it might cover the gap with only short‐term funds so that it can then refinance at expected lower rates. The opposite might be done if the desk thought rates would rise in the near future.

Running a liquidity gap over time, beyond customer requirements, would reflect a particular view of the ALM desk. So, maintaining a consistently underfunded position suggests that interest rates are expected to decline, and so longer term funds can be taken at cost. Maintaining an overfunded gap would imply that the bank thinks rates will be rising, and so longer term funds are locked in now at lower interest rates. Even if the net position is dictated by customer requirements – for example, customers placing more on deposit than they take out in loans – the bank can still manage the resultant gap in the wholesale market.

Generally, excess liabilities at a bank are a rare occurrence and, under most circumstances, such a position is clearly undesirable. This is because the bank will have to achieve target return on capital ratios, and this requires funds to be put to work, so to speak, by acquiring assets. In the case of equity capital it is imperative that these funds are properly employed.⁷ The exact structure of the asset book will depend on the bank's view on interest rates and the yield curve generally. The shape of the yield curve and expectations on this will also influence the structure and tenor of the asset book. The common practice is to spread assets across the term structure with varying maturities. There will also be investments made with a forward start date to lock in rates in the forward curve now. Equally, some investments will be made for very short periods, so that if interest rates rise, when the funds are reinvested they will benefit from the higher rates.

Gap risk and limits

Liquidity gaps are measured by taking the difference between outstanding balances of assets and liabilities over time. At any point a positive gap between assets and liabilities is equivalent to a deficit, and this is measured as a cash amount. The marginal gap is the difference between the changes of assets and liabilities over a given period. A positive marginal gap means that the variation of value of assets exceeds the variation of value of liabilities. As new assets and liabilities are added over time, as part of the ordinary course of business, the gap profile changes.

The gap profile is tabulated or charted (or both) during and at the end of each day as a primary measure of risk. For illustration, a tabulated gap report is shown at Figure 7.1 and is an actual example from a UK banking institution. It shows the assets and liabilities grouped into maturity buckets and the net position for each bucket. It is a snapshot today of the exposure, and hence funding requirement of the bank for future maturity periods.

Table 7.1 is very much a summary figure, because the maturity gaps are very wide. For risk management purposes the buckets would be much narrower, for instance the period between 0 and 12 months might be split into 12 different maturity buckets. An example of a more detailed gap report is shown at Figure 7.7, which is from another UK banking institution. Note that the overall net position is zero, because this is a balance sheet and therefore, not surprisingly, it balances. However, along the maturity buckets or grid points there are net positions that are the gaps that need to be managed.

Limits on a banking book can be set in terms of gap limits. For example, a bank may set a 6‐month gap limit of £10 million. The net position of assets and maturities expiring in 6 months' time could then not exceed £10 million. An example of a gap limit report is shown at Figure 7.6, with the actual net gap positions shown against the gap limits for each maturity. Again, this is an actual limit report from a UK banking institution.

			Time periods
	Total		0–6 months		6–12 months		1–3 years		3–7 years		7+ years
Assets	40,533	6.17%	28,636	6.08%	3,801	6.12%	4,563	6.75%	2,879	6.58%	654	4.47%
Liabilities	40,533	4.31%	30,733	4.04%	3,234	4.61%	3,005	6.29%	2,048	6.54%	1,513	2.21%
Net cumulative positions	0	1.86%	(2,097)		567		1,558		831		(859)
Margin on total assets:	2.58%
Average margin on total assets:	2.53%

The maturity gap can be charted to provide an illustration of net exposure, and an example is shown at Figure 7.9 from yet another UK banking institution. In some firms' reports both the assets and the liabilities are shown for each maturity point, but in our example only the net position is shown. This net position is the gap exposure for that maturity point. A second example, used by the overseas subsidiary of a Middle Eastern commercial bank, which has no funding lines in the inter‐bank market and so does not run short positions, is shown at Figure 7.10, while the gap report for a UK high‐street bank is shown at Figure 7.11. Note the large short gap under the maturity labelled “non‐int”; this stands for non‐interest bearing liabilities and represents the balance of current accounts (cheque or “checking” accounts), which are funds that attract no or very low interest and are in theory very short dated (because they are demand deposits, so may be called at instant notice). In behavioural terms, however, such deposits are treated as “stable” long‐term funding.

Gaps represent cumulative funding required at all dates. The cumulative funding is not necessarily identical to the new funding required at each period, because the debt issued in previous periods is not necessarily amortised at subsequent periods. The new funding between, for example, months 3 and 4 is not the accumulated deficit between months 2 and 4 because the debt contracted at month 3 is not necessarily amortised at month 4. Marginal gaps may be identified as the new funding required or the new excess funds of the period that should be invested in the market. Note that all the reports are snapshots, at a fixed point in time and the picture is of course a continuously moving one. In practice, the liquidity position of a bank cannot be characterised by one gap at any given date, and the entire gap profile must be used to gauge the extent of the book's profile.

The liquidity book may decide to match its assets with its liabilities. This is known as cash matching and occurs when the time profiles of both assets and liabilities are identical. By following such a course, the bank can lock in the spread between its funding rate and the rate at which it lends cash, and run a guaranteed profit. Under cash matching the liquidity gaps will be zero. Matching the profile of both legs of the book is done at the overall level; that is, cash matching does not mean that deposits should always match loans. This would be difficult as both result from customer demand, although an individual purchase of, say, a CD can be matched with an identical loan. Nevertheless, the bank can elect to match assets and liabilities once the net position is known, and keep the book matched at all times. However, it is highly unusual for a bank to adopt a cash matching strategy.

LIQUIDITY MANAGEMENT

The continuous process of raising new funds or investing surplus funds is known as liquidity management. If we consider that a gap today is funded, thus balancing assets and liabilities and squaring‐off the book, the next day a new deficit or surplus is generated, which also has to be funded. The liquidity management decision must cover the amount required to bridge the gap that exists the following day, as well as position the book across future dates in line with the bank's view on interest rates. Usually, in order to define the maturity structure of debt a target profile of resources is defined. This may be done in several ways. If the objective of ALM is to replicate the asset profile with resources, the new funding should contribute to bringing the resources profile closer to that of the assets, that is, more of a matched book looking forward. This is the lowest‐risk option. Another target profile may be imposed on the bank by liquidity constraints. This may arise if, for example, the bank has a limit on borrowing lines in the market so that it could not raise a certain amount each week or month. For instance, if the maximum that could be raised in 1 week by a bank is £10 million, the maximum period liquidity gap is constrained by that limit. The ALM desk will manage the book in line with the target profile that has been adopted, which requires it to try to reach the required profile over a given time horizon.

Managing the banking book's liquidity is a dynamic process, as loans and deposits are known at any given point, but new business will be taking place continuously and the profile of the book looking forward must be continuously rebalanced to keep it within the target profile. There are several factors that influence this dynamic process, the most important of which are reviewed below.

Interest‐rate risk and source

Interest‐rate risk

Put simply, interest‐rate risk is defined as the potential impact, adverse or otherwise, on the net asset value of a financial institution's balance sheet and earnings resulting from a change in interest rates. Risk exposure exists whenever there is a maturity date mismatch between assets and liabilities, or between principal and interest cash flows. Interest‐rate risk is not necessarily a negative thing; for instance, changes in interest rates that increase the net asset value of a banking institution would be regarded as positive. For this reason, active ALM seeks to position a banking book to gain from changes in rates. The Bank for International Settlements splits interest‐rate risk into two elements: investment risk and income risk. The first risk type is the term for potential risk exposure arising from changes in the market value of fixed interest‐rate cash instruments and off‐balance‐sheet instruments, and is also known as price risk. Investment risk is perhaps best exemplified by the change in value of a plain vanilla bond following a change in interest rates, and from Chapter 2 we know that there is an inverse relationship between changes in rates and the value of such bonds (see Example 2.2). Income risk is the risk of loss of income when there is a non‐synchronous change in deposit and funding rates, and it this risk that is known as gap risk.

ALM covering the formulation of interest‐rate risk policy is usually the responsibility of what is known as the asset‐liability committee or ALCO, which is made up of senior management personnel including the Finance Director and the heads of Treasury and Risk Management. ALCO sets bank policy for balance sheet management and the likely impact on revenue of various scenarios that it considers may occur. The size of ALCO will depend on the complexity of the balance sheet and products traded, and the amount of management information available on individual products and desks.

The process employed by ALCO for ALM will vary according to the particular internal arrangement of the institution. A common procedure involves a monthly presentation to ALCO of the impact of different interest‐rate scenarios on the balance sheet. This presentation may include:

• An analysis of the difference between the actual net interest income (NII) for the previous month and the amount that was forecast at the previous ALCO meeting. This is usually presented as a gap report, broken by maturity buckets and individual products;
• The result of discussion with business unit heads on the basis of the assumptions used in calculating forecasts and impact of interest‐rate changes; scenario analysis usually assumes an unchanging book position between now and 1 month later, which is essentially unrealistic;
• A number of interest‐rate scenarios, based on assumptions of (a) what is expected to happen to the shape and level of the yield curve, and (b) what may happen to the yield curve, for example, extreme scenarios. Essentially, this exercise produces a value for the forecasted NII due to changes in interest rates;
• An update of the latest actual revenue numbers;
• Specific new or one‐off topics may be introduced at ALCO as circumstances dictate; for example, the presentation of the approval process for the introduction of a new product.

Sources of interest‐rate risk

Assets on the balance sheet are affected by absolute changes in interest rates as well as increases in the volatility of interest rates. For instance, fixed‐rate assets will fall in value in the event of a rise in rates, while funding costs will rise. This decreases the margins available. We noted that the way to remove this risk was to lock‐in assets with matching liabilities; however, this is only not always possible, but also sometimes undesirable, as it prevents the ALM manager from taking a view on the yield curve. In a falling interest‐rate environment, deposit‐taking institutions may experience a decline in available funds, requiring new funding sources that may be accessed at less favourable terms. Liabilities are also impacted by a changing interest‐rate environment.

There are five primary sources of interest‐rate risk inherent in an ALM book, which are described below.

• Gap risk is the risk that revenue and earnings decline as a result of changes in interest rates, due to the difference in the maturity profile of assets, liabilities and off‐balance‐sheet instruments. Another term for gap risk is mismatch risk. An institution with gap risk is exposed to changes in the level of the yield curve, a so‐called parallel shift, or a change in the shape of the yield curve or pivotal shift. Gap risk is measured in terms of short‐term or long‐term risk, which is a function of the impact of rate changes on earnings for a short or long period. Therefore, the maturity profile of the book, and the time to maturity of instruments held on the book, will influence whether the bank is exposed to short‐term or long‐term gap risk.
• Yield curve risk is the risk that non‐parallel or pivotal shifts in the yield curve cause a reduction in NII. The ALM manager will change the structure of the book to take into account their views on the yield curve. For example, a book with a combination of short‐term and long‐term asset‐maturing or liability‐maturity structures⁹ is at risk from a yield curve inversion, sometimes known as a twist in the curve.
• Basis risk arises from the fact that assets are often priced off one interest rate, while funding is priced off another interest rate. Taken one step further, hedge instruments are often linked to a different interest rate to that of the product they are hedging. In the US market, the best example of basis risk is the difference between the Prime rate and Libor. Term loans in the US are often set at Prime, or a relationship to Prime, while bank funding is usually based on the Eurodollar market and linked to Libor. However, the Prime rate is what is known as an “administered” rate and does not change on a daily basis, unlike Libor. While changes in the two rates are positively correlated, they do not change by the same amount, which means that the spread between them changes regularly. This results in the spread earning on a loan product changing over time. Figure 7.12 illustrates the change in spread during 2009–2010.

  

• Another risk for deposit‐taking institutions such as clearing banks is run‐off risk, associated with the non‐interest bearing liabilities (NIBLs) of such banks. The level of interest rates at any one time represents an opportunity cost to depositors who have funds in such facilities. However, in a rising interest‐rate environment, this opportunity cost rises and depositors will withdraw these funds, available at immediate notice, resulting in an outflow of funds for the bank. The funds may be taken out of the banking system completely, for example, for investment in the stock market. This risk is significant and therefore sufficient funds must be maintained at short notice, which is an opportunity cost for the bank itself.
• Many banking products entitle the customer to terminate contractual arrangement ahead of the stated maturity term; this is sometimes referred to as option risk. This is another significant risk as products such as CDs, cheque account balances, and demand deposits can be withdrawn or liquidated at no notice, which is a risk to the level of NII should the option inherent in the products be exercised.

Gap and net interest income

We noted earlier that gap is a measure of the difference in interest‐rate sensitivity of assets and liabilities that revalue at a particular date, expressed as a cash value. Put simply it is:

(9.1)

where A_ir and L_ir are the interest‐rate sensitive assets and interest‐rate sensitive liabilities. When the banking book is described as being positively gapped, and when the book is said to be negatively gapped. The change in NII is given by:

(9.2)

where r is the relevant interest rate used for valuation. The NII of a bank that is positively gapped will increase as interest rates rise, and will decrease as rates decline. This describes a banking book that is asset sensitive; the opposite, when a book is negatively gapped, is known as liability sensitive. The NII of a negatively gapped book will increase when interest rates decline. The value of a book with zero gap is immune to changes in the level of interest rates. The shape of the banking book at any one time is a function of customer demand, the Treasury manager's operating strategy, and view of future interest rates.

Gap analysis is used to measure the difference between interest‐rate‐sensitive assets and liabilities, over specified time periods. Another term for this analysis is periodic gap, and the common expression for each time period is maturity bucket. For a commercial bank the typical maturity buckets are:

• 0–3 months;
• 3–12 months;
• 1–5 years;
• 5 years;

although another common approach is to group assets and liabilities by the buckets or grid points of the Riskmetrics Value‐at‐Risk methodology. Any combination of time periods may be used, however. For instance, certain US commercial banks place assets, liabilities, and off‐balance‐sheet items in terms of known maturities, judgemental maturities, and market‐driven maturities. These are defined as:

• Known maturities: fixed‐rate loans and CDs;
• Judgemental maturities: passbook savings accounts, demand deposits, credit cards, non‐performing loans;
• Market‐driven maturities: option‐based instruments such as mortgages, and other interest‐rate sensitive assets.

The other key measure is cumulative gap, defined as the sum of the individual gaps up to 1 year maturity. Banks traditionally use the cumulative gap to estimate the impact of a change in interest rates on NII.

Assumptions of gap analysis

A number of assumptions are made when using gap analysis, assumptions that may not reflect reality in practice. These include:

• The key assumption that interest‐rate changes manifest themselves as a parallel shift in the yield curve; in practice, changes do not occur as a parallel shift, giving rise to basis risk between short‐term and long‐term assets;
• The expectation that contractual repayment schedules are met; if there is a fall in interest rates, prepayments of loans by borrowers who wish to refinance their loans at lower rates will have an impact on NII. Certain assets and liabilities have option features that are exercised as interest rates change, such as letters of credit and variable rate deposits; early repayment will impact a bank's cash flow;
• That repricing of assets and liabilities takes place in the mid‐point of the time bucket;
• The expectation that all loan payments will occur on schedule; in practice, certain borrowers will repay the loan earlier.

Recognised weaknesses of the gap approach include:

• No incorporation of future growth, or changes in the asset–liability mix;
• No consideration of the time value of money;
• Arbitrary setting of time periods.

Limitations notwithstanding, gap analysis is used extensively. Gup and Brooks (1993, page 59) state the following reasons for the continued popularity of gap analysis:

• It was the first approach introduced to handle interest‐rate risk, and provides reasonable accuracy;
• The data required to perform the analysis is already compiled for the purposes of regulatory reporting;
• The gaps can be calculated using simple spreadsheet software;
• It is easier (and cheaper) to implement than more sophisticated techniques;
• It is straightforward to demonstrate and explain to senior management and shareholders.

Although there are more sophisticated methods available, gap analysis remains in widespread use.

THE ALM DESK

The ALM desk or unit is a specialised business unit that fulfils a range of functions. Its precise remit is a function of the type of activities of the financial institution that it is a part of. Let us consider the main types of activities that are carried out.

If an ALM unit has a profit target of zero, it will act as a cost centre with a responsibility to minimise operating costs. This would be consistent with a strategy that emphasises commercial banking as the core business of the firm, and where ALM policy is concerned purely with hedging interest‐rate and liquidity risk.

The next level is where the ALM unit is responsible for minimising the cost of funding. That would allow the unit to maintain an element of exposure to interest‐rate risk, depending on the view that was held as to the future level of interest rates. As we noted above, the core banking activity generates either an excess or shortage of funds. To hedge away all of the excess or shortage, while removing interest‐rate exposure, has an opportunity cost associated with it since it eliminates any potential gain that might arise from movements in market rates. Of course, without a complete hedge, there is an exposure to interest‐rate risk. The ALM desk is responsible for monitoring and managing this risk, and of course is credited with any cost savings in the cost of funds that arise from the exposure. The saving may be measured as the difference between the funding costs of a full hedging policy and the actual policy that the ALM desk adopts. Under this policy, interest‐rate risk limits are set that the ALM desk ensures the bank's operations do not breach.

The final stage of development is to turn the ALM unit into a profit centre, with responsibility for optimising the funding policy within specified limits. The limits may be set as gap limits, Value‐at‐Risk limits or by another measure, such as level of earnings volatility. Under this scenario, the ALM desk is responsible for managing all financial risk.

The final development of the ALM function has resulted in it taking on a more active role. The previous paragraphs described the three stages of development that ALM has undergone, although all three versions are part of the “traditional” approach. Practitioners are now beginning to think of ALM as extending beyond the risk management field, and being responsible for adding value to the net worth of the bank, through proactive positioning of the book and hence the balance sheet. That is, in addition to the traditional function of managing liquidity risk and interest‐rate risk, ALM should be concerned additionally with managing the regulatory capital of the bank and with actively positioning the balance sheet to maximise profit. The latest developments mean that there are now financial institutions that run a much more sophisticated ALM operation than that associated with a traditional banking book.

Let us review the traditional and developed elements of an ALM function.

Generic “traditional” ALM policy for different‐sized banks

ALM discipline as practised traditionally in banks could be summarised as follows:

• The preparation and adoption of a high‐level interest‐rate risk and liquidity risk policy at managing board level; this sets general guidelines on the type and extent of risk exposure that can be taken on by the bank;
• Setting limits on the risk exposure levels of the banking book; this can be by product type, desk, geographic area, and so on, and will be along the maturity spectrum;
• Actively measuring the level of interest‐rate risk exposure at regular, specified intervals;
• Reporting to senior management on general aspects of risk management, risk exposure levels, limit breaches, and so on;
• The monitoring of risk management policies and procedures by an independent “middle office” risk function.

The risk management approach adopted by banks will vary according to their specific markets and appetite for risk. Certain institutions will have their activities set out or proscribed for them under regulatory rules. For instance, building societies in the UK are prohibited from trading in certain instruments under the regulator's guidelines. In this section we present, purely for the purposes of illustration, the ALM policies of three hypothetical banks, called Bank S, Bank M, and Bank L. These are, respectively, a small banking entity with assets of £500 million, a medium‐sized bank with assets of £2.5 billion and a large bank with assets of £10 billion. The following serves to demonstrate the differing approaches that can be taken according to the environment that a financial institution operates in.

ALM policy for Bank S (assets = £500 million)

The aim of the ALM policy for Bank S is to provide guidelines on risk appetite, revenue targets, and rates of return, as well as risk management policy. Areas that may be covered include capital ratios, liquidity, asset mix, rate‐setting policy for loans and deposits, and investment guidelines for the banking portfolio. The key objectives should be:

• To maintain capital ratios at the planned minimum, and to ensure safety of the deposit base;
• To generate a satisfactory revenue stream, both for income purposes and to further protect the deposit base.

The responsibility for overseeing the operations of the bank to ensure that these objectives are achieved is lodged with the ALM Committee. This body monitors the volume and mix of the bank's assets and funding (liabilities), and ensures that this asset mix follows internal guidelines with regard to banking liquidity, capital adequacy, asset base growth targets, risk exposure, and return on capital. The norm is for the committee to meet on a monthly basis; at a minimum the membership of the committee will include the finance director, head of Treasury, and risk manager. For a bank the size of Bank S, the ALM committee membership will possibly be extended to the chief executive, the head of the loans business, and the chief operating officer.

As a matter of course, the committee will wish to discuss and review the following on a regular basis:

• Overall macroeconomic conditions;
• Financial results and key management ratios such as share price analysis and rates of return on capital and equity;
• The house view on the likely direction of short‐term interest rates;
• The current lending strategy and suggestions for changes to this, as well as the current funding strategy;
• Any anticipated changes to the volume and mix of the loan book, and that of the main sources of funding; in addition, the appropriateness or otherwise of alternative sources of funding;
• Suggestions for any alteration to the bank's ALM policy;
• The maturity gap profile and anticipated and suggested changes to it.

The committee will also wish to consider the interest rates offered currently on loans and deposits, and whether these are still appropriate.

Interest‐rate sensitivity is monitored and confirmed as lying within specified parameters; these parameters are regularly reviewed and adjusted if deemed necessary according to changes in the business cycle and economic conditions. Measured using the following ratio:

typical risk levels would be expected to lie between 90 and 120% for the maturity period 0–90 days, and between 80 and 110% for the maturity period over 90 days and less than 365 days.

Put simply, the objective of Bank S would be to remain within specified risk parameters at all times, and to maintain as consistent level of earnings as possible (and one that is immune to changes in the stage of the business cycle).

ALM policy for Bank M (assets = £2.5 billion)

Bank M is our hypothetical “medium‐sized” banking institution. Its ALM policy would be overseen by an Asset–Liability Management Committee or ALCO. Typically, the following members of senior management would be expected to be members of ALCO:

• Deputy chief executive;
• Finance director;
• Head of retail banking;
• Head of corporate banking;
• Head of Treasury;
• Head of risk management;
• Head of internal audit;

together with others such as product specialists who are called to attend as and when required. The finance director will often chair the meeting.

The primary responsibilities of ALCO are detailed below.

Objectives

ALCO is tasked with reviewing the bank's overall funding strategy. Minutes are taken at each meeting, and decisions taken are recorded on the minutes and circulated to attendees and designated key staff. ALCO members are responsible for undertaking a regular review of the following:

• Minutes of the previous meeting;
• The ratio of the interest‐rate‐sensitive assets to liabilities, gap reports, risk reports, and the funding position;
• The bank's view on expected level of interest rates and how the book should be positioned with respect to this view; and related to this, the ALCO view on anticipated funding costs in the short term and medium term;
• Stress testing in the form of “what if?” scenarios to check the effect on the banking book of specified changes in market conditions; and the change in parameters that may be required if there is a change in market conditions or risk tolerance;
• The current interest rates for loans and deposits to ensure that these are in accordance with the overall lending and funding strategy;
• The maturity distribution of the liquidity book (expected to be comprised of T‐bills, CDs, and very short‐dated government bonds); the current liquidity position and the expected position in the short term and medium term.

As ALCO meets on a regular monthly basis, it may not be the case that every aspect of their responsibility is discussed at every meeting; the agenda is set by the chair of the meeting in consultation with committee members. The policies adopted by ALCO should be dynamic and flexible, and capable of adaptation to changes in operating conditions. Any changes will be made on agreement of committee members. Generally, any exceptions to agreed policy can only be with the agreement of the CEO and ALCO itself.

Interest‐rate risk policy

The objective will be to keep earnings volatility resulting from an upward or downward move in interest rates to a minimum. To this end, at each ALCO meeting members will review risk and position reports and discuss these in the light of the risk policy. Generally, the 6‐month and 12‐month A_ir/L_ir cumulative ratio will lie in the range 90–110%. A significant move outside this range will most likely be subject to corrective action. The committee will also consider the results of various scenario analyses on the book, and if these tests indicate a potential earnings impact of greater than, say, 10%, instructions may be given to alter the shape and maturity profile of the book.

Liquidity policy

A primary responsibility of ALCO is to ensure that an adequate level of liquidity is maintained at all times. We define liquidity as:

“…the ability to meet anticipated and unanticipated operating cash needs, loan demand, and deposit withdrawals, without incurring a sustained negative impact on profitability.”

(Gup and Brooks, 1993, page 238)

Generally, a Bank M‐type operation would expect to have a target level for loans to deposits of around 75–85%, and a loans to core deposits ratio of 85–95%. The loan/deposit ratio is reported to ALCO and reviewed on a monthly basis, and a reported figure significantly outside these ranges (say, by 5% or more) will be reviewed and asked to be adjusted to bring it back into line with ALCO policy.

ALM policy for Bank L (assets = £10 billion)

The management policy for ALM at a larger entity will build on that described for a medium‐sized financial institution. If Bank L is a group company, the policy will cover the consolidated balance sheet as well as individual subsidiary balance sheets; the committee will provide direction on the management of assets and liabilities, and the off‐balance‐sheet instruments used to manage interest‐rate and credit risk. A well‐functioning management process will be proactive and concentrate on direction in response to anticipated changes in operating conditions, rather than reactive responses to changes that have already taken place. The primary objectives will be to maximise shareholder value, with target returns on capital of 15–22%.

The responsibility for implementing and overseeing the ALM management policy will reside with ALCO. ALCO will establish the operating guidelines for ALM, and review these guidelines on a periodic basis. The committee will meet on a more frequent basis than would be the case for Bank M, usually on a fortnightly basis. As well as this, it will set policies governing liquidity and funding objectives, investment activities, and interest‐rate risk. It will also oversee the activities of the investment banking division. The head of the ALM desk will prepare the interest‐rate risk sensitivity report and present it to ALCO.

Interest‐rate risk management

ALCO will establish an interest‐rate risk policy that sets direction on acceptable levels of interest‐rate risk. This risk policy is designed to guide management in the evaluation of the impact of interest‐rate risk on the bank's earnings. The extent of risk exposure is a function of the maturity profile of the balance sheet, as well as the frequency of repricing, the level of loan prepayments, and funding costs. Managing interest‐rate risk is, in effect, the adjustment of risk exposure upwards or downwards, which will be in response to ALCO's views on the future direction of interest rates. As part of the risk management process, the committee will monitor the current risk exposure and duration gap, using rate sensitivity analysis and simulation modelling to assess whether the current level of risk is satisfactory.

Measuring interest‐rate risk

Notwithstanding the widespread adoption of Value‐at‐Risk as the key market risk measurement tool, funding books such as repo books continue to use the gap report as a key measure of interest‐rate risk exposure. This enables ALCO to view the risk sensitivity along the maturity structure. Cumulative gap positions and the ratio of assets revaluation to liabilities revaluation are calculated and compared to earnings levels on current the asset–liability position. Generally, the 90‐day, 6‐month, and 1‐year gap positions are the most significant points along the term structure at which interest‐rate risk exposure is calculated. The ratio of gap to earnings assets will be set at the ±15% to ±20% level.

As it is a traditional duration‐based approach, gap reporting is a static measure that measures risk sensitivity at one specific point in time. For this reason, banks combine a Value‐at‐Risk measure as well, or only use Value‐at‐Risk. It is outside the scope of this book to consider Value‐at‐Risk but we cite a useful introductory reference in the Bibliography at the end of this chapter.

Simulation modelling

Simulation modelling is a procedure that measures the potential impact on the banking book, and hence earnings levels, of a user‐specified change in interest rates and/or a change in the shape of the book itself. This process enables senior management to gauge the risk associated with particular strategies. Put simply, the process is:

• Construct a “base” balance sheet and income statement as the starting point (this is derived from the current shape of the banking book, and any changes expected from current growth trends that have been projected forward);
• Assess the impact on the balance sheet of changes under selected scenarios; these might be no change in rates; a 100 basis point and 250 basis point upward parallel shift in the yield curve; a 100 basis point and 250 basis point downward parallel shift; a 25 basis point steepening and flattening of the yield curve, between the 3‐month and the 3‐year maturity points; a combination of a parallel shift with a pivotal shift at a selected point; an increase or decrease in 3‐month T‐bill yield volatility levels; and a 20 basis point change in swap spreads;
• Compare the difference in earnings resulting from any of the scenarios to the anticipated earnings stream under the current environment.

Generally, the committee will have set guidelines about the significance of simulation results; for example, there may be a rule that a 100 basis point change in interest rates should not impact NII by more than 10%. If results indicate such an impact, ALCO will determine if the current risk strategy is satisfactory or whether adjustments are necessary.

CRITIQUE OF THE TRADITIONAL APPROACH TO ALM

Traditionally, the main approach of ALM concentrated on interest sensitivity and net present value sensitivity of a bank's loan–deposit book. The usual interest sensitivity report is the maturity gap report, which we reviewed briefly earlier. The maturity gap report is not perfect, however, and can be said to have the following drawbacks:

• The repricing intervals chosen for gap analysis are ultimately arbitrary, and there may be significant mismatches within a repricing interval. For instance, a common repricing interval chosen is the 1‐year gap and the 1–3‐year gap; there are (albeit extreme) circumstances when mismatches would go undetected by the model. Consider a banking book that is composed solely of liabilities that reprice in 1 month's time, and an equal cash value of assets that reprice in 11 months' time. The 1‐year gap of the book (assuming no other positions) would be zero, implying no risk to net interest income. In fact, under our scenario the net interest income is significantly at risk from a rise in interest rates;
• Maturity gap models assume that interest rates change by a uniform magnitude and direction. For any given change in the general level of interest rates, however, it is more realistic for different maturity interest rates to change by different amounts, what is known as a non‐parallel shift;
• Maturity gap models assume that principal cash flows do not change when interest rates change. Therefore, it is not possible effectively to incorporate the impact of options embedded in certain financial instruments. Instruments such as mortgage‐backed bonds and convertibles do not fall accurately into a gap analysis, as only their first‐order risk exposure is captured;
• Notwithstanding these drawbacks, the gap model is widely used as it is easily understood in the commercial banking and mortgage industry, and its application does not require a knowledge of sophisticated financial modelling techniques.

But perhaps the biggest drawback of the traditional approach to ALM in banks is that it is reactive in nature. The department(s) that are responsible for managing ALM risk in banks are not responsible for originating most of the assets and liabilities that are on the balance sheet at any time. In effect, Treasury is given a balance sheet to manage, after the fact, for interest‐rate and liquidity risk purposes. This makes strategic planning less of a practical exercise and often more of just a budgeting one.

STRATEGIC ALM

The twenty‐first century approach to ALM in banks must, given the operating environment post‐2008, be more proactive in nature rather than reactive as it has been traditionally. Strategic ALM is a single, integrated aggregate approach to balance sheet management that ties in asset origination with the liabilities raising. It attempts to break down “silos” in the organisation, in that asset type is relevant and appropriate to funding type and source (and vice versa).

Strategic ALM addresses a three‐dimensional optimisation problem that we summarise as follows, in line with the requirements of the three key stakeholders of a bank:

• The liquidity and regulatory requirement aspects (Regulator);
• The NII/NIM aspects (Shareholder);
• The customer franchise aspects (Customer);

and must be a high‐level, strategic discipline driven from the top down. It is by nature proactive and not reactive.

A key operating model decision is whether the ALM function should be a profit centre or a cost centre.

• If an ALM unit has a profit target of zero, it will act as a cost centre with a responsibility to minimise operating costs. This would be consistent with a strategy that emphasises commercial banking as the core business of the firm, and where ALM policy is concerned purely with hedging interest‐rate and liquidity risk;
• The next stage is where the ALM unit is responsible for minimising the cost of funding. Any saving is measured as the difference between the funding costs of a full hedging policy and the actual policy that the ALM desk adopts. Under this policy, interest‐rate risk limits are set that the ALM desk ensures the bank's operations do not breach;
• The final stage of development is to turn the ALM unit into a profit centre, with responsibility for optimising the funding policy within specified limits.

The op model decision is a key part of business development.

For a bank's funding structure to be assessed on an aggregate balance sheet approach, it must measure the quality and adequacy of the funding structure (liabilities) alongside the capital and asset side of the balance sheet. This gives a more holistic picture of the robustness and resilience of the funding model under stress. The robustness of funding is almost as much a function of the liquidity, maturity, and product type of the asset base as it is of the type and composition of the liabilities.

The following are key considerations of the ALM function:

• Liquid assets vs illiquid assets share;
• How much illiquid assets are funded by unstable and/or short‐term liabilities;
• Breakdown of liabilities:
  • Retail deposits: stable and less stable;
  • Wholesale funding: secured, senior unsecured;
  • Capital: subordinated/hybrid; equity.

On the liability side, the ALCO will look at:

• Debt buybacks, especially of expensive instruments issued post‐crash;
• Wider investor base;
• Private placement programme;
• Fit‐for‐purpose allocation of liquidity costs to business lines (FTP);
• Design and use of adequate stress testing policy and scenarios;
• Adequate risk management of intra‐day liquidity risk;
• Strong public disclosure to promote market discipline.

On the asset side, strategic action could include:

• Increasing liquid assets as a share of the balance sheet (although liquidity and ROE concerns must be balanced);
• De‐linking the bank – sovereign risk exposure connection;
  • The LCR doesn't have to be all sovereign debt;
• Avoiding lower loan origination standards as the cycle moves into the bull market phase;
• Addressing asset quality problems:
  • Ring‐fenced NPLs and impaired loans? (A sort of “non‐core” part of the balance sheet that indicates you are addressing the problem and looking at disposal);
• Review the bank's operating model. Retail/wholesale mix? Franchise viability? Comparative advantage?;
• Limit asset encumbrance: this contradicts pressure for secured funding.

CONCLUSIONS

Asset‐liability management is a discipline that is as old as banking itself, even if it was only articulated in formal terms fairy recently. The traditional approach to it was essentially reactive in nature, however, with business lines originating assets and liabilities that were then “risk managed” for interest‐rate and liquidity risk by Treasury and Risk (and in some cases Finance) departments. As we noted in the chapter, this is not fit for purpose in the twenty‐first century. In a bank, the balance sheet is everything, because the very survival of the institution depends on the balance sheet being robust and long‐term viable. So for an ALM function to be able to carry out its function effectively, banks need to change practices in two areas:

• Bring credit risk policy under the aegis of the ALM function; and
• Move from a reactive to a more proactive ALM process whereby the origination of balance sheet assets and liabilities is more integrated and also under ALCO direction.

We term this more proactive approach Strategic ALM.