The Gold Standard of Bank Asset–Liability Management Services | Interest Rate Gaps, Funding Durations and Balance-Sheet Resilience

A bank lends at long-term fixed rates but funds part of those loans with customer deposits that can be repriced sooner. Interest rates rise. Deposit costs may increase before interest earned on existing loans changes, so profits can be squeezed even while most borrowers continue repaying. A second problem is the change in economic value of assets and liabilities whose cash flows mature at different dates.

Bank asset–liability management (ALM) is the financial service and governance process that examines the bank’s balance sheet as a connected set of earning assets, deposits, borrowings, capital and off-balance arrangements. It helps manage interest-rate mismatches, funding liquidity, currencies and the capacity to sustain customer obligations. The gold standard is accurate cash-flow data, credible behavioural assumptions, independently challenged risk measures and treasury decisions that work under realistic stress.

The Basel IRRBB application guidance effective 1 January 2026 distinguishes effects on net interest income and economic value of equity. This article is distinct from corporate Cash Pooling, an individual Interest Rate Swap trade and institution-wide Financial Risk Stress Testing within the Gold Standard master hub.

ALM at a glance: different risks inside one balance sheet

Two worked ALM sensitivities

These examples are intentionally separate, static and simplified. The duration-based net value reduction is approximately S$1.7 million, not a calculated regulatory EVE measure. Real ALM requires detailed cash flows, options and pricing.

Eighteen chapters: funding, earnings and economic resilience

1. Manage the whole banking book rather than isolated products

Loans and deposits create a joint financial obligation

A bank’s mortgage desk might report profitable loans while the treasury team faces higher deposit funding costs. Each department’s product figures can be correct but the combined balance sheet may be vulnerable to rising interest rates or funding withdrawals. ALM must connect both sides without erasing the economic and legal rights of depositors.

Set the legal entity, perimeter and management purpose

Map loans, securities held for banking purposes, deposits, wholesale borrowings, capital, hedges and relevant off-balance exposures. Preserve the accountable ALM committee and responsible finance and risk functions. A bank cannot claim diversified interest-rate risk based only on the number of loan products it offers.

A balance sheet must have a consistent valuation and legal-entity perimeter. A deposit liability recorded at a foreign branch can be economically connected to local lending without creating unlimited rights to transfer cash across borders. Both the legal and management views should remain available.

2. Interest rate risk in the banking book is not trading-book market risk

A fixed-rate loan can lose economic value when rates rise

The bank might hold a long-term mortgage at amortised cost, so short-term accounting earnings do not show the full change in the value of its future interest cash flows. The interest-rate exposure still affects economic resilience and future funding choices.

Apply the correct banking-book measurement

Basel SRP98 defines IRRBB and distinguishes banking-book economic value from earnings. Record asset classification and actual cash-flow conditions. Do not tell executives that a position is safe from interest rates merely because its financial reporting value does not change every trading day.

A mortgage held at amortised cost can create a real rate exposure without generating a daily trading loss on the published accounts. Executives should see those economic effects alongside the accounting treatment and avoid implying that a lack of immediate profit impact means no value risk.

3. Net interest income and economic value of equity answer different questions

NII tests earnings over a selected future period

Net interest income is interest earned on assets less interest paid on liabilities, incorporating relevant hedging economics under the selected assumptions. It is sensitive to how quickly old loans and deposits reprice, which customers remain and how new business affects the balance sheet.

EVE tests a broader discounted cash-flow effect

Economic value of equity examines the net present value of relevant asset and liability cash flows under alternative interest-rate conditions. Basel treats NII and EVE as complementary but not identical. A bank can improve its first-year earnings while taking a longer-duration economic value loss. Report both measures rather than selecting the one that looks more favourable.

For each IRRBB measure, explain whether management is considering static current positions or a forecast balance sheet that assumes new lending and deposit growth. These approaches can produce materially different outcomes, so a period-to-period comparison needs a stable methodology.

4. Repricing gaps create timing risk even without credit defaults

Deposits can reset before loans

A bank may rely on deposits whose rates can increase quickly while lending at a rate locked for a further year. In that period, funding expenses rise before the loan income adjusts. The resulting loss of margin can occur despite all borrowers and depositors behaving as contractually permitted.

Maintain a dated repricing ladder

Map actual resets, contract maturities, interest floors, expected repayment, and relevant options across rate buckets. Basel describes gap risk in SRP98. The largest headline loan balance is not necessarily the largest first-quarter rate exposure; timing determines the near-term effect.

An institution can show a small aggregate first-year repricing gap while having a large mismatch in one currency or one maturity bucket. Review more than the net total; offsetting signs can conceal funding fragility or limited cross-market hedging access.

5. A worked first-year income squeeze shows the repricing difference

Assume S$100m loans remain fixed for twelve months

Suppose a fictional bank has S$100 million of loans whose interest cannot reset during the first year, funded in part by S$80 million of deposits that immediately reprice upward by one percentage point. Holding quantities and all other costs fixed, funding interest increases by S$800,000 while loan interest does not initially rise.

Do not confuse later opportunities with immediate cash

After the loan reset, if the entire S$100 million reprices upward by one percentage point while S$80 million funding remains repriced by the same change, an annualised incremental interest margin might be positive S$200,000. The first-year squeeze and later simple spread are different time periods, subject to customer behaviour, pricing and actual loan terms. The example is not a bank forecast.

The illustrative initial S$800,000 earnings squeeze would change if deposits did not pass through the entire interest-rate movement, or customers left when rates rose. Actual deposit pricing and retention assumptions must be tested rather than assumed to follow a simple parallel one-percent move.

6. Basis risk arises even when both sides are called floating

One benchmark can move differently from another

A loan tied to a particular market reference may be funded using deposits priced administratively or linked to a different curve. Even if both reprice each quarter, the two underlying rates may diverge and change the bank’s net interest margin unexpectedly.

Trace the reference and reset conventions

Record index, spread, contractual floor, change notice, fixing date and effective interest period. Do not assume a simple interest-rate swap or a bank’s general asset-liability ratio eliminates this relative movement. Basel SRP98 treats basis risk separately from timing gaps, so controls should preserve that distinction.

Basis risk can worsen when a loan benchmark remains stable but the deposit market becomes competitive. The decision to raise savings rates is sometimes commercial rather than mechanically linked to a reference index, so the bank needs realistic pricing governance in its model.

7. Yield curves need more than an up-or-down shock

Short and long maturities can move differently

A bank’s five-year loans and overnight funding are not equally sensitive to a steepening or flattening yield curve. A parallel rise may be important, but it does not capture every plausible pattern. Currency curves can also behave differently under market stress.

Apply current prescribed shocks where required

The Basel framework includes six interest-rate shock scenarios for EVE and two for NII within its specified IRRBB supervisory approach. Its recalibration dated July 2024 was to be implemented by January 2026, though domestic application depends on the supervisory framework. A scenario tool should use the actual applicable version, not an old screen copied from a 2019 policy.

The recalibrated Basel 2026 shocks use defined currency-specific methodology. Operational teams should retain the correct published scenario version and national implementation, because the same curve movement should not be applied blindly to every currency or maturity in one combined dashboard.

8. Behavioural deposit assumptions can dominate results

Non-maturity deposits have no fixed contractual repricing date

Current and savings account customers can leave funds for many years or withdraw tomorrow. A bank may expect a stable core and model some balances as slow to reprice, but that is an assumption about customer behaviour, not an unconditional funding guarantee.

Challenge stability and price-sensitivity models

Analyse account activity, deposit type, customer concentration, rate competition and stress conditions. Basel SRP98 discusses behavioural treatment of non-maturity deposits. A model that assumes every low-cost deposit remains indefinitely should face robust review, particularly when competitors offer higher interest or customers become concerned about a bank’s condition.

Forecasting stable deposit balances requires evidence about account age, concentration, rate sensitivity and customer type. A new high-yield promotional account is unlikely to behave identically to a longstanding operating account, even if both appear as non-maturity deposits in an accounting report.

9. Borrower prepayment and mortgage options alter duration

A customer may repay early when refinancing becomes attractive

A fixed-rate mortgage’s economic maturity can differ from its legal final term if borrowers have prepayment rights. Rising rates may slow repayments while falling rates speed refinancing, changing asset duration and cash flows just as funding rates move.

Model contractual and behavioural options explicitly

Use permitted prepayment terms, historic behaviour and sensible scenario assumptions. Avoid relying on one static maturity number. A hedge sized against an expected schedule can become mismatched when borrowers change behaviour. The ALM team should know whether it is measuring a contractual obligation or a modelled expectation.

A mortgage model should separate voluntary customer prepayments from contractual scheduled instalments and defaults. When rates fall, borrowers may refinance, reducing asset duration while funding liabilities react differently. The resulting optionality can challenge simple linear duration assumptions.

10. Economic value sensitivity needs realistic duration assumptions

A longer-duration asset typically reacts more to rates

For a rough illustration, a S$50 million fixed-income asset with modified duration four could lose approximately S$2 million of market value from a one-percentage-point rate rise under a first-order approximation. A S$30 million liability with duration one could fall in value by about S$300,000.

Interpret the approximate net change carefully

The simplified adverse change in net economic value is around S$1.7 million before other banking book components. This ignores convexity, basis differences, options and discount-curve details. Actual EVE calculations require full cash-flow methods and assumptions. Duration approximations explain sensitivity; they are not substitute for the relevant bank’s formal regulatory model.

Duration-based value estimates are first-order approximations and ignore convexity and options. The S$1.7 million example should therefore be presented as a teaching calculation. For real ALM decisions, use verified cash-flow models and disclose what happens when the rate movement is not small.

11. Funding concentration matters beyond the average deposit rate

A small number of customers can move large balances quickly

A bank may rely on a handful of corporate depositors or short-term wholesale borrowing. A low historical funding rate can conceal severe dependence on a few relationships. If those customers withdraw together, market funding may become expensive or unavailable.

Monitor real access to cash and contingency alternatives

Maintain funding source diversification, maturities, collateral eligibility and reliable liquid assets. Link to an approved contingency funding plan and realistic operational settlement arrangements. This is not the same as a capital ratio: a bank can have regulatory capital and still need immediate liquidity to meet deposit withdrawals.

Large depositor concentration also creates an operational question: can the bank identify which accounts are legally accessible and what settlement network is needed for their potential withdrawal? A liquidity amount in a dashboard is not a substitute for a tested payment route.

12. Currency mismatches require separate, traceable risk controls

Funding in one currency can support assets in another

A bank may lend in U.S. dollars while attracting Singapore-dollar deposits, using FX swaps or other arrangements to transform funding. The interest-rate gap interacts with currency basis and the ability to renew a hedge or access foreign-currency cash.

Map both the cash flows and the hedge contracts

Record currency, maturity, counterparties, FX derivative value dates and collateral requirements. Do not collapse foreign-currency balances into one translated SGD amount and conclude funding is usable everywhere. Each legal settlement route matters, particularly if an FX swap matures while the underlying asset remains outstanding.

Foreign-currency funding requires more than a consolidated amount translated into Singapore dollars. Test which currencies can be obtained under stressed swap-market conditions, and whether variation margin or maturity gaps cause immediate cash demands at the wrong subsidiary or branch.

13. Liquidity regulation and IRRBB should inform but not replace each other

Stable funding and interest-rate valuation have different measurements

Liquidity Coverage Ratio and Net Stable Funding Ratio assess specified funding resilience, while IRRBB assesses the effect of rates on economic value and earnings. A bank might meet a liquidity measure but remain sensitive to steep changes in yield curves.

Use separate reporting, connected decisions

The Basel SRP31 describes IRRBB governance, risk measures and supervisory responses. The ALM committee should consider relevant liquidity and capital indicators without representing them as one interchangeable metric. A positive EVE number does not prove enough same-day money is available to customers.

A bank’s liquidity buffer may be substantial while its interest-rate duration remains risky. Decision papers should therefore avoid implying that a successful liquidity stress automatically solves IRRBB; one metric assesses available funding while the other evaluates interest-rate consequences.

14. An interest-rate hedge has a legal and cash settlement cost

A derivative can reduce exposure but add collateral needs

A bank can use swaps to adjust the effective repricing profile of assets and liabilities. The hedge may reduce some rate sensitivity but require margin, create counterparty exposure or have a negative market value on early termination.

Review actual economics beyond a hedge label

Tie every swap to authorised ALM objectives, signed confirmations, reset dates and bank settlements. Compare the hedged banking-book result with the unhedged result under relevant shocks. The Interest Rate Swap Services article explains individual trade administration; bank ALM must judge the combined system and remaining risk.

If an interest-rate swap reduces an asset gap, test the hedge’s actual reset dates and notional after mortgage prepayments. Otherwise the contract may become an unintended stand-alone position, exposing treasury to settlement or collateral requirements that were not part of the initial plan.

15. Capital and earnings cannot be sacrificed to an unapproved strategy

A better-looking ratio can conceal unacceptable concentration

Management can alter product pricing, duration, asset composition or funding mix, but those changes have customer, legal and business effects. A strategy that maximises this quarter’s net interest income may create a long-term economic value loss or leave the bank short of liquidity.

Govern choices at the correct level

The asset–liability committee should receive clear limits, scenario outcomes, funding projections and risk appetite. Document which changes require board approval or supervisor consultation. Evaluate the effects of proposed action across NII, EVE, capital, liquidity and actual customer commitments rather than rewarding one desk’s isolated short-term result.

ALCO review should assess how changing deposit pricing affects customers as well as bank earnings. A permitted financial decision can still create confusion if notice, effective dates or account terms are implemented inconsistently across the retail bank.

16. Use effective-date reference data rather than historical guesses

The terms of loans and funding can be amended

Customer rates, deposit product conditions and hedges may change after origination. A bank using the latest instrument values for older scenarios can create false backtests or unexplained movements in its reporting.

Preserve a controlled point-in-time balance sheet

Reconcile contract terms, repricing dates, borrower options, issuer identifiers and cash balances for each valuation date. Capture authorised amendments and original versions. Model documentation must specify which legal entities and currencies are included and how cancelled, prepaid or reclassified accounts were treated.

Historic reference data need effective dates for interest floors, margin changes and early repayments. A backtest conducted using today’s terms on last year’s loans can conceal errors in the original risk assessment and create an artificial improvement in forecast accuracy.

17. Backtest assumptions against actual customer behaviour

A model must learn from deposit and mortgage changes

If customers reprice deposits faster or refinance mortgages differently from the model’s estimates, future NII projections can become unreliable. A technically correct calculation is not useful if behavioural parameters have become outdated.

Challenge assumptions, not just software uptime

Compare realised repricing, account retention, prepayment and funding volume with prior forecasts. Investigate deviations and document approved recalibrations. The ALM team should have an independent way to assess vendor models and expert overlays. Material changes need to reach responsible decision makers with the likely capital or earnings consequence.

Independent ALM challenge is most useful when it tests the assumptions on which decisions depend. Focus review on deposit decay, prepayments, benchmark basis and funding access, rather than spending equal effort on every minor rounding difference in a large financial model.

18. The gold standard is resilient banking across time horizons

ALM is not a promise to earn the highest interest spread

It is a process for managing a bank’s obligations when lending, funding and interest-rate conditions interact. No balance sheet can avoid all risks, and every hedging or funding decision introduces trade-offs.

Measure the results that genuinely matter

Audit asset and liability source data, gap ladders, EVE and NII sensitivity, deposit behaviour, currency mismatches, hedge settlements and contingency plans. A gold-standard bank can show how its decisions protect earnings and economic value while preserving safe access to customer banking services through changing markets.

A useful committee action log records the limit or mismatch, approved response, funding or hedging cost, execution date and new measured exposure. Board members should later be able to see whether a genuine financial vulnerability declined, not just whether a committee noted the same warning.

Questions about managing bank assets and liabilities

Are net interest income and economic value of equity the same?

No. NII measures earnings effects over a defined horizon; EVE assesses net present value sensitivity of banking-book cash flows.

Can rising rates hurt a bank even when it still has performing loans?

Yes. Funding may reprice faster than loan income.

Are all customer savings deposits guaranteed stable funding?

No. Behaviour and access rights change with customers and markets.

Do interest-rate swaps eliminate every banking-book exposure?

No. Basis risk, options, counterparty exposure and liquidity can remain.

Does Basel’s 2026 IRRBB shock recalibration mean every local rule is identical?

No. Actual domestic implementation and reporting scope must be checked.

Does a positive capital ratio prove the bank can meet sudden withdrawals?

No. Available liquidity is a separate concern.

An asset–liability committee’s control checklist

• Reconcile banking-book instruments, deposit balances, currency and legal entity scope.
• Distinguish short-term NII from the economic value of equity.
• Validate behavioural deposit and prepayment assumptions with observed outcomes.
• Apply current interest-rate curve scenarios and effective supervisory rules.
• Track funding concentrations, hedge maturities and available foreign-currency cash.
• Connect ALM, liquidity and capital decisions without conflating the ratios.
• Document approved ALCO actions and independently review changed risk after execution.

Current Basel guidance and interconnected eduKate articles

• Basel SRP98: IRRBB Application Guidance Effective January 2026.
• BCBS: 2024 Recalibration of Banking-Book Interest-Rate Shocks, 2026 Implementation.
• Basel SRP31: IRRBB Principles and Supervisory Assessment.
• Basel DIS70: Economic Value and Net Interest Income Disclosures.
• eduKate: Corporate Cash Pooling.
• eduKate: Financial Risk Stress Testing.
• eduKate: Interest Rate Swap Services.
• eduKate: Regulatory Capital Reporting.
• eduKate: Bank Account Switching.

Educational bank risk-management analysis, not a prudential interest-rate risk calculation or an asset-liability committee decision for a specific bank. Actual regulation and contractual exposures determine real requirements.