Bank Balance-Sheet Economics and ALM Frameworks

Introduction

A bank’s balance sheet is not merely an accounting statement. It is an economic system that transforms maturities, liquidity, credit risk and interest-rate exposures. Banks typically fund relatively stable or illiquid assets—such as mortgages, corporate loans and securities—with liabilities that may be withdrawn, repriced or rolled over quickly. This transformation creates the economic value of banking, but also exposes the institution to liquidity, interest-rate, solvency and profitability risks.

Asset Liability Management (ALM) is the integrated process through which a bank identifies, measures, controls and prices these exposures. It connects balance-sheet structure with treasury operations, risk appetite, regulatory requirements, capital planning and risk-adjusted performance. Effective ALM therefore requires more than matching asset and liability maturities. It requires understanding customer behaviour, contractual and behavioural cash flows, repricing conventions, embedded options and interactions among risks.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Explain the economic role of bank balance sheets and maturity transformation.
2. Analyse assets and liabilities by liquidity, maturity, repricing and optionality.
3. Apply gap, duration, economic-value and earnings-based ALM measures.
4. Evaluate embedded options and behavioural assumptions in deposits and loans.
5. Connect ALM decisions with liquidity, capital, pricing and risk-adjusted profitability.
6. Critically assess governance, regulatory and technological requirements for ALM.

1. The Economics of Financial Intermediation

1.1 Why banks exist

Financial intermediation reduces information and transaction costs between surplus units (depositors and investors) and deficit units (borrowers). A bank performs several related functions:

• Screening and monitoring: assessing borrowers and monitoring them after lending.
• Liquidity provision: offering deposits and payment services even though loans are less liquid.
• Risk pooling: diversifying individual borrower and depositor risks.
• Maturity transformation: funding longer-term assets with shorter-term or more liquid liabilities.
• Payment and settlement: providing money-like claims transferable at par.
• Delegated monitoring: exploiting economies of scale in credit analysis.

The bank’s balance sheet is the contractual expression of these functions. A simplified accounting identity is:

Assets include cash, central-bank balances, securities, loans and derivatives with positive value. Liabilities include deposits, wholesale funding, debt securities and derivatives with negative value. Equity absorbs losses and provides a buffer against insolvency.

1.2 Intermediation theory and liquidity preference

The Diamond–Dybvig perspective emphasises that depositors value liquidity while borrowers require relatively stable funding. A bank can pool uncertain withdrawal needs: only some depositors withdraw at any time, allowing the bank to invest part of the pool in longer-term assets. This arrangement is efficient when withdrawals are reasonably predictable, but vulnerable to runs when depositors lose confidence.

Liquidity preference also explains why instruments with immediate convertibility generally offer lower yields than illiquid assets. The difference is not pure credit compensation; it may include a liquidity premium. A bank earns a spread by supplying liquidity and bearing the associated risks.

The key insight is that maturity transformation is valuable but fragile. It creates a positive economic margin only if the bank can retain funding, price risk correctly and maintain sufficient liquid resources under stress.

2. Anatomy of the Bank Balance Sheet

2.1 Asset structure

Loan assets generate interest and fee income but are generally illiquid, especially during stress. Securities may be liquid in normal markets but suffer valuation losses or haircuts when yields rise or markets become disorderly.

2.2 Liability and capital structure

A crucial distinction is between contractual maturity and behavioural maturity. A non-maturity deposit may be withdrawable on demand but remain with the bank for years. Conversely, a term deposit may be withdrawn early or not renewed. ALM must model both legal terms and observed customer behaviour.

2.3 Net interest income and the banking margin

Net interest income (NII) is commonly defined as:

Net interest margin (NIM), using average interest-earning assets, is:

The formula is an accounting performance measure, not a complete risk measure. It excludes or incompletely captures credit losses, liquidity costs, capital consumption, market-value changes and operational expenses. A high NIM can coexist with poor risk-adjusted profitability if it is achieved through excessive risk-taking.

3. Maturity Transformation and Liquidity Risk

3.1 The transformation process

Suppose a bank accepts deposits that may be withdrawn promptly and invests part of the proceeds in five-year loans. The bank has transformed short-term claims into longer-term assets. This supports economic investment but creates a funding gap: asset cash flows arrive later than possible liability outflows.

A simplified liquidity gap for a time bucket is:

A negative gap means that expected outflows exceed inflows in that bucket. The bank must cover it with cash, liquid assets, funding capacity or asset sales. Gaps should be assessed cumulatively:

This measure is useful but incomplete. It does not by itself capture asset-sale haircuts, funding concentration, collateral calls, intraday needs or deposit run correlations.

3.2 Liquidity buffers and stress testing

A liquidity buffer normally contains cash and assets that can be monetised rapidly with limited loss. A prudent framework tests scenarios such as:

• retail deposit withdrawals;
• loss of unsecured wholesale funding;
• increased collateral requirements;
• market-value haircuts;
• rating downgrades;
• drawdowns of committed credit lines; and
• operational or reputational disruption.

A survival horizon estimates how long the bank can meet stressed net cash outflows without extraordinary central-bank assistance or disorderly asset sales. Stress tests are not forecasts; they are decision tools for determining liquidity limits, contingency actions and buffer size.

3.3 Maturity mismatch is not automatically imprudent

A bank need not eliminate every maturity mismatch. Doing so could prevent productive lending and reduce intermediation efficiency. The appropriate question is whether the mismatch is compensated by stable funding, liquid assets, capital, pricing and credible contingency funding. ALM is therefore an optimisation problem under constraints, not a mechanical exercise in matching every asset with an identical liability.

4. Repricing Risk and Interest-Rate Exposure

4.1 Repricing profiles

Repricing risk arises when assets and liabilities reset their interest rates at different times or according to different benchmarks. A floating-rate loan linked to a policy rate may reprice monthly, while a fixed-rate deposit remains unchanged for one year. Changes in market rates then alter NII.

The traditional rate-sensitive gap is:

where RSA_t is the amount of rate-sensitive assets repricing in bucket t, and RSL_t is the amount of rate-sensitive liabilities repricing in the same bucket. A positive gap generally means that, over the selected horizon and under a parallel rate change, asset income may change more than liability expense. However, the conclusion depends on basis risk, rate floors, caps, administered pricing and customer behaviour.

A simplified change in NII is:

where ∆r_A and ∆r_L are changes in asset and liability rates. If both rates change by the same amount, this becomes:

These relationships assume linear repricing, a parallel shock, unchanged balances and no behavioural response. They are not reliable for large shocks or instruments with options.

Worked Example 1: Repricing gap

A bank has, within a three-month bucket, RSA of $420 million and RSL of $350 million. Assume both reprice fully and immediately, with no volume change. If market rates rise by 1 percentage point:

∆NII ≈ (420 − 350) million × 0.01 = $0.70 million

Thus, quarterly interest income would improve by approximately $0.70 million over the relevant period if the repricing effect applies for the full period. If the shock lasts only half the quarter, the approximate effect would be halved. The calculation does not include basis risk or deposit beta.

4.2 Deposit beta and administered rates

A deposit beta measures the sensitivity of a deposit rate to a market-rate movement. In simplified form:

A beta of 0.60 means that a 1 percentage-point market increase is associated, on average, with a 0.60 percentage-point deposit-rate increase. A beta may differ between products, customer segments and rate cycles. It is an empirical behavioural parameter, not a permanent law.

4.3 Basis risk

Basis risk occurs when two rates are expected to move together but do not. Examples include a loan linked to a policy rate funded by deposits linked to a treasury bill rate, or a loan repricing monthly while its funding reprices quarterly. A simple gap report can show a balanced volume position while the bank remains exposed to the spread between reference rates.

5. Economic Value, Duration and Market-Value Risk

NII analysis focuses on a short horizon. Economic-value analysis considers the present value of all expected future cash flows. For a fixed-income instrument:

where CF_t is cash flow at time t, y is the yield per period, and n is the number of periods. The units of y must correspond to the cash-flow period.

Macaulay duration is:

Modified duration, for annual compounding, is:

For a small parallel yield change:

Duration is a first-order approximation. It assumes a small parallel shift, stable cash flows and no significant convexity or optionality. For larger changes, convexity becomes relevant:

where C is convexity, measured consistently with the duration convention.

In ALM, the bank compares the duration and present value of assets, liabilities and equity. A duration mismatch can cause the economic value of equity to decline when rates change, even when current-period NII initially improves.

6. Embedded Options and Behavioural Modelling

An embedded option is a contractual or behavioural right that changes the timing or amount of cash flows. Important examples include:

• borrower prepayment rights in mortgages;
• loan interest-rate caps and floors;
• deposit early withdrawal rights;
• callable bonds issued by the bank;
• puttable deposits or structured products;
• commitment drawdown options; and
• automatic renewal or attrition behaviour.

Options make cash flows state-dependent. A borrower is more likely to refinance when rates fall; a depositor may move funds when market rates rise. Consequently, the effective maturity and duration of an instrument can differ substantially from its contractual maturity.

A sound behavioural model should:

1. define the population and product scope;
2. identify explanatory variables, such as rates, balances, customer tenure and seasonality;
3. estimate parameters using sufficiently long and representative data;
4. distinguish ordinary periods from stress periods;
5. perform out-of-sample validation and back-testing;
6. apply conservative floors, caps and expert overlays; and
7. document model limitations and governance approvals.

Model risk is material. A deposit assumed to be stable for five years may leave rapidly during a confidence shock. A mortgage prepayment model calibrated during low-rate conditions may fail when refinancing incentives reverse.

7. ALM Measurement Frameworks

7.1 Contractual and behavioural cash-flow analysis

The first layer maps expected cash inflows and outflows into time buckets. It supports liquidity management and identifies refinancing concentrations. Contractual schedules should be supplemented by behavioural assumptions for non-maturity deposits, prepayments, overdrafts and undrawn commitments.

7.2 Earnings-at-risk analysis

Earnings-at-risk estimates the adverse effect of interest-rate scenarios on NII over a chosen horizon. A simplified scenario method is:

The definition must specify whether “worst” means the lowest result across prescribed scenarios or a statistical percentile. Scenario analysis is preferable to a single sensitivity because yield curves can steepen, flatten or twist rather than shift in parallel.

7.3 Economic value of equity

Economic value of equity (EVE) is conceptually:

An interest-rate shock is assessed through:

EVE captures long-term value effects but depends heavily on discount rates, behavioural assumptions and treatment of deposits without contractual maturity. It should complement, not replace, NII analysis.

7.4 Risk-adjusted profitability

A useful management framework adjusts income for expected loss, liquidity cost, operating cost and capital usage. One generic form is:

The exact numerator and denominator must be defined consistently. Expected loss may be expressed as:

where PD is probability of default, LGD is loss given default, and EAD is exposure at default. These quantities must refer to compatible horizons and units. RAROC is not an accounting profit measure; it is a decision tool for comparing returns with the capital required to absorb unexpected loss.

Worked Example 2: Risk-adjusted pricing

A proposed loan generates annual revenue of $1.20 million. Expected loss is $0.30 million, operating cost is $0.20 million, and funding plus liquidity cost is $0.25 million. Economic capital is $5 million.

Risk-adjusted contribution is:

1.20 − 0.30 − 0.20 − 0.25 = $0.45 million

Therefore:

RAROC = 0.45 ÷ 5 = 9%

Management should compare 9% with the bank’s hurdle rate and with returns available from alternative uses of capital. The result is meaningful only if revenue, costs, expected loss and capital are measured on consistent annual bases.

8. An Integrated ALM Governance Framework

ALM is normally governed through a board-approved risk appetite and an Asset Liability Committee (ALCO). A robust framework includes:

• clearly defined responsibilities for the board, ALCO, treasury, risk, finance and business units;
• limits for liquidity gaps, concentration, wholesale funding, interest-rate sensitivity and capital;
• independent risk measurement and challenge;
• daily liquidity monitoring and regular structural ALM reporting;
• documented assumptions for deposits, prepayments and drawdowns;
• stress testing linked to management actions;
• transfer pricing that allocates funding, liquidity and capital costs;
• escalation procedures for limit breaches; and
• internal audit and external supervisory review.

Funds transfer pricing (FTP) assigns the cost or benefit of funding and liquidity to products and business lines. A loan should not appear profitable merely because it receives an unpriced pool of short-term deposits. FTP should reflect tenor, optionality, liquidity value, contingent funding needs and, where appropriate, capital consumption.

9. Illustrative Case: Meridian Bank’s Rising-Rate Problem

Background

Meridian Bank has a large portfolio of five-year fixed-rate mortgages funded by savings deposits and three-month wholesale borrowings. During a low-rate period, the strategy produced strong NIM. The bank’s liquidity buffer was adequate under normal conditions, but its wholesale funding was concentrated among a small number of counterparties.

Problem

Market rates rise by 200 basis points. Wholesale funding reprices quickly, while mortgage yields remain fixed. Depositors begin moving balances into higher-yielding products. At the same time, the market value of the bank’s securities declines, reducing immediately available liquidity.

Analysis

The bank has a negative short-term repricing position because rate-sensitive liabilities exceed rate-sensitive assets. It also has a duration mismatch: fixed-rate mortgage assets have long duration, while deposits and wholesale funding have shorter effective duration. The deposit model understated rate sensitivity by assuming a low beta based on an unusually stable historical period. Liquidity analysis also failed to apply sufficiently severe deposit outflows and securities haircuts.

Approach and solution

ALCO implements a coordinated response:

1. It raises the liquidity buffer and diversifies wholesale maturities.
2. It uses interest-rate swaps, subject to hedge-accounting and counterparty controls, to reduce excessive duration exposure.
3. It revises deposit beta and attrition assumptions using segmented behavioural data.
4. It introduces FTP charges for long-term fixed-rate lending and contingent liquidity usage.
5. It slows origination of underpriced fixed-rate mortgages and promotes products with repricing features.
6. It conducts reverse stress tests to identify the combination of withdrawals, funding closure and market losses that would breach risk appetite.

Outcome

NIM initially declines because the bank prices liquidity and duration more realistically. However, risk-adjusted profitability improves, funding concentration falls and the bank becomes less vulnerable to a confidence shock. The case illustrates that short-term accounting performance can conceal structural risk.

Learning points

• Contractual maturity alone is insufficient.
• NII and EVE can give different signals and must be used together.
• Behavioural assumptions should be stress-sensitive.
• Hedging does not remove liquidity or basis risk.
• Product pricing must include funding, liquidity and capital costs.

10. Regulatory, Ethical and Technological Considerations

Regulators generally expect banks to maintain adequate liquidity, capital, interest-rate risk controls, stress-testing systems, governance and reliable data. Specific rules vary by jurisdiction, but common themes include high-quality liquid asset buffers, stable funding expectations, supervisory limits and disclosure.

Ethically, ALM decisions affect depositors, borrowers and financial stability. A bank should not rely on opaque pricing, misleading guarantees or models that systematically disadvantage customer groups. Incentives should reward sustainable risk-adjusted returns rather than short-term volume or NIM alone.

Technology enables near-real-time cash-flow aggregation, behavioural segmentation, scenario analytics and model monitoring. Artificial intelligence may improve deposit-flow forecasting, but it introduces explainability, data-quality, privacy, cybersecurity and model-drift risks. AI outputs should support—not replace—expert judgement, validation and governance. A model that predicts normal deposit behaviour accurately may still fail under an unprecedented run.

11. Limitations and Best Practices

No single ALM metric captures all balance-sheet risks. Gap analysis is intuitive but ignores rate magnitude and duration. Duration analysis captures value sensitivity but can fail with options and non-parallel shocks. NII analysis supports budgeting but may overlook long-term value destruction. Stress testing depends on scenario quality and management credibility.

Best practice is therefore to:

• combine contractual, behavioural, earnings and economic-value views;
• use multiple interest-rate and liquidity scenarios;
• analyse basis risk, optionality and concentration explicitly;
• validate behavioural models and challenge optimistic assumptions;
• integrate ALM with capital, liquidity and recovery planning;
• use transparent FTP and risk-adjusted performance measures;
• maintain reliable, timely and reconciled data; and
• test whether identified risks can actually be mitigated in stressed markets.

Chapter Summary

Bank balance sheets create value by performing intermediation, liquidity provision and maturity transformation. The same structure exposes banks to liquidity, repricing, basis, duration, credit and solvency risks. ALM analyses these risks through contractual and behavioural cash flows, repricing gaps, NII sensitivity, duration, EVE, stress testing and risk-adjusted profitability.

Maturity transformation is economically useful but cannot be judged through maturity matching alone. Deposits, loans and securities contain behavioural features and embedded options that alter effective cash flows. Accordingly, models must be validated, stressed and governed. NII and EVE provide complementary short- and long-term perspectives, while RAROC connects risk-taking with capital allocation and pricing. Effective ALM requires coordinated governance through ALCO, credible liquidity buffers, sound FTP, independent challenge and integration with regulatory and strategic planning.

Key Terms

Asset Liability Management: The integrated management of balance-sheet risks, returns and funding structure.

Maturity transformation: Funding longer-term or less liquid assets with shorter-term or more liquid liabilities.

Financial intermediation: Channeling funds between savers and borrowers.

Liquidity preference: The preference for assets or claims that can be converted into cash quickly and reliably.

Rate-sensitive asset: An asset whose interest rate or cash flow resets within a specified period.

Rate-sensitive liability: A liability whose funding cost resets within a specified period.

Repricing gap: Rate-sensitive assets minus rate-sensitive liabilities in a time bucket.

Basis risk: Risk that related reference rates do not move together.

Deposit beta: Sensitivity of a deposit rate to a market-rate change.

Duration: Sensitivity of an instrument’s value to changes in yield.

Convexity: The curvature of the price–yield relationship.

Economic value of equity: Present value of assets less present value of liabilities.

Embedded option: A contractual or behavioural right that changes expected cash flows.

Behavioural maturity: Expected economic maturity based on customer behaviour rather than legal terms.

Earnings-at-risk: Potential adverse effect of scenarios on earnings, commonly NII.

Funds transfer pricing: Internal allocation of funding, liquidity and related costs.

Expected loss: Probability of default multiplied by loss given default and exposure at default.

RAROC: Risk-adjusted return divided by economic capital.

ALCO: Committee responsible for balance-sheet and market-risk oversight.

Review Questions

Short Answer Questions

1. Why is maturity transformation central to banking?
2. Distinguish contractual maturity from behavioural maturity.
3. What does a positive repricing gap indicate under a parallel rate increase?
4. Why are NII and EVE complementary ALM measures?
5. Define basis risk and provide one banking example.

Long Answer Questions

1. Explain the economic functions of banks and show how they generate both value and vulnerability.
2. Compare gap analysis, duration analysis and stress testing as ALM tools.
3. Discuss the importance of embedded options in mortgage and deposit portfolios.
4. Explain how FTP and RAROC can improve balance-sheet pricing and capital allocation.
5. Design a governance framework for an ALCO managing liquidity and interest-rate risk.

Analytical / Application Questions

1. A bank has RSA of $600 million and RSL of $750 million in a repricing bucket. Estimate the effect of a 1% parallel rate rise on the relevant-period NII, assuming full repricing and unchanged balances. Interpret the result and identify two limitations.
2. A bank’s fixed-rate assets have materially longer duration than its liabilities. Analyse the likely effect of a sharp rate increase on NII and EVE, and recommend two possible responses.
3. Develop a stress scenario combining deposit attrition, wholesale-funding closure and securities haircuts. Explain the data, assumptions, limits and management actions required.

Multiple Choice Questions with Answers

Question 1

Which economic function most directly describes a bank funding long-term loans with short-term deposits?

A. Maturity transformation

B. Capital conservation

C. Currency sterilisation

D. Tax arbitrage

Correct Answer: A. Maturity transformation

Question 2

A bank has RSA of $420 million and RSL of $350 million. If both reprice fully and rates rise by 1 percentage point, the approximate change in the relevant-period NII is:

A. $0.35 million decrease

B. $0.70 million increase

C. $4.20 million increase

D. $7.70 million decrease

Correct Answer: B. $0.70 million increase

Question 3

Which measure is defined conceptually as PV(Assets) minus PV(Liabilities)?

A. Net interest margin

B. Liquidity coverage ratio

C. Economic value of equity

D. Deposit beta

Correct Answer: C. Economic value of equity

Question 4

A deposit beta of 0.60 implies that a 1 percentage-point increase in the reference market rate is associated, on average, with:

A. A 0.06 percentage-point deposit-rate increase

B. A 0.60 percentage-point deposit-rate increase

C. A 1.60 percentage-point deposit-rate increase

D. No deposit-rate change

Correct Answer: B. A 0.60 percentage-point deposit-rate increase

Question 5

Which risk arises when a loan references one benchmark while its funding references another?

A. Basis risk

B. Settlement risk

C. Legal risk

D. Concentration risk

Correct Answer: A. Basis risk

Question 6

Which instrument feature is most likely to make effective mortgage duration change when interest rates fall?

A. Borrower prepayment right

B. Fixed accounting classification

C. Cash reserve requirement

D. Loan documentation date

Correct Answer: A. Borrower prepayment right

Question 7

Using the simplified formula, if annual risk-adjusted contribution is $0.45 million and economic capital is $5 million, RAROC is:

A. 5%

B. 9%

C. 11.1%

D. 45%

Correct Answer: B. 9%

Question 8

Which statement best describes a limitation of a simple repricing gap?

A. It fully captures customer options

B. It assumes all assets are risk-free

C. It may ignore basis risk and behavioural responses

D. It measures only credit losses

Correct Answer: C. It may ignore basis risk and behavioural responses

Question 9

What is the primary purpose of funds transfer pricing?

A. To eliminate all liquidity risk

B. To allocate funding and liquidity costs to products and businesses

C. To replace regulatory capital

D. To guarantee a positive NIM

Correct Answer: B. To allocate funding and liquidity costs to products and businesses

Question 10

Why should ALM use both NII sensitivity and EVE analysis?

A. They are identical measures with different names

B. NII is short-horizon earnings-focused, while EVE captures longer-term present-value effects

C. EVE measures only operational losses

D. NII excludes all interest-rate effects

Correct Answer: B. NII is short-horizon earnings-focused, while EVE captures longer-term present-value effects