ALM, FTP & Balance-Sheet

How to read this chapter

Asset Liability Management and Funds Transfer Pricing are where banking becomes joined-up. A bank is not a pile of separate products. It is a living balance sheet made of customer loans, customer deposits, securities, wholesale funding, derivatives, capital, liquidity buffers, payment flows, customer behaviour, market rates, regulatory constraints and management choices. ALM is the discipline that asks whether the shape of that balance sheet is safe, profitable and resilient. FTP is the internal pricing discipline that asks whether each business line is being charged or credited fairly for the funding, liquidity, optionality and interest-rate characteristics it creates.

Read this chapter as the bridge between commercial banking, treasury and markets. A mortgage is not only a customer loan. It has fixed or floating rate behaviour, prepayment risk, liquidity need, funding tenor, capital consumption, FTP cost, interest-rate sensitivity and customer optionality. A current account is not only a deposit balance. It may be operationally stable, rate sensitive, digitally mobile, insured, uninsured, concentrated or relationship driven. A securities portfolio is not only an investment. It may be high-quality liquid assets, repo collateral, duration exposure, accounting volatility and contingency cash. A derivative is not only a hedge. It may create valuation movement, collateral calls, counterparty exposure and accounting complexity. ALM joins these pieces into one management story.

The Basel Committee’s liquidity principles describe liquidity as the ability of a bank to fund increases in assets and meet obligations as they come due without incurring unacceptable losses (Basel liquidity principles). The Liquidity Coverage Ratio promotes short-term resilience through an adequate stock of unencumbered high-quality liquid assets to meet thirty-day stress outflows (Basel LCR standard). The Net Stable Funding Ratio promotes stable funding over a one-year horizon (Basel NSFR standard). The Basel IRRBB framework defines interest-rate risk in the banking book as risk to capital and earnings from adverse interest-rate movements affecting banking book positions (Basel IRRBB framework). ALM sits at the intersection of these ideas.

The practical point is this: a bank can look profitable and still be fragile if its funding is short, deposits are unstable, hedges are misunderstood, securities are long-duration, liquidity is trapped in the wrong entity, or FTP is hiding the real cost of products. ALM and FTP exist to prevent that. They make the bank’s balance-sheet truth visible before stress exposes it.

Learning objectives

By the end of this chapter, you should be able to explain ALM as balance-sheet mismatch management, not merely a committee or reporting function. You should understand FTP as internal economic truth-telling for funding, liquidity, optionality, term structure and product behaviour. You should be able to connect deposits, loans, securities, wholesale funding, derivatives, LCR, NSFR, IRRBB, NII, EVE, ALCO, stress testing, contingency funding and product profitability. You should also be able to translate ALM and FTP into practical calculation logic, requirements, data models, controls, reports and test scenarios for a real banking environment.

1. Why ALM exists

A bank does not naturally have a perfectly matched balance sheet. Customers place deposits that may be contractually repayable on demand but behaviourally stable for years. Borrowers take loans that may amortise, prepay, refinance, default or draw undrawn commitments. Mortgages may be fixed rate while savings balances reprice quickly. Wholesale funding may mature in large blocks. Securities may be liquid but sensitive to rates and spreads. Derivatives may reduce interest-rate risk but create collateral calls. Currencies may not match. ALM exists because these mismatches must be seen, measured, governed and managed.

ALM is therefore not a back-office report pack. It is the discipline of keeping the bank’s balance sheet resilient while allowing the business to serve customers and earn income. It covers liquidity risk, funding strategy, interest-rate risk in the banking book, maturity mismatch, repricing mismatch, currency mismatch, behavioural assumptions, customer optionality, collateral, capital interaction, securities strategy, hedging and product profitability. It connects treasury, finance, risk, product, business lines, technology and executive governance.

A simple five-year fixed-rate loan shows why ALM matters. The customer sees an approved loan and monthly repayment. The business sees margin and relationship value. Treasury asks how the loan is funded. Risk asks how it behaves under rate shocks. Finance asks how income is recognised. FTP asks what internal funding cost applies. ALCO asks whether the bank wants more of this exposure. If the loan is funded by short deposits and market rates rise, the customer rate may be fixed while funding cost rises. If the customer prepays when rates fall, the bank may lose an attractive asset. The product is simple to the customer, but not simple to the balance sheet.

ALM is also an early-warning function. It asks whether asset growth is honestly funded, whether deposit assumptions remain credible, whether wholesale maturities are concentrated, whether currency liquidity is available, whether HQLA is usable, whether hedges behave as expected, whether collateral calls can be met, whether FTP is steering behaviour and whether management decisions are being taken early enough. A strong ALM function studies the balance sheet while management still has choices.

2. The balance sheet as a banking story

The accounting balance sheet has assets, liabilities and equity. Assets include loans, securities, cash, central-bank balances, trading assets and receivables. Liabilities include deposits, wholesale funding, issued debt, repo, derivatives payable, operational payables and other obligations. Equity absorbs loss and supports confidence. ALM reads the same balance sheet through cash-flow timing, repricing, liquidity, optionality and behaviour.

The contractual maturity of an item may differ from its behavioural life. A current account is payable on demand, but part of the balance may behave like stable funding. A term deposit has contractual maturity, but renewal behaviour matters. A mortgage may have a long final maturity, but average life may be shorter because of prepayments. A committed facility may have no drawn balance today, but in stress the customer may draw it. A derivative may have no upfront cash flow, but market movement can create collateral calls. ALM translates accounting items into behavioural and stress cash flows.

The balance sheet also has currency and legal-entity dimensions. A consolidated group can appear comfortable while one currency is tight. A subsidiary may have surplus liquidity that cannot easily move to another entity because of regulation, tax, operational constraints or local stress. A bank may issue foreign-currency debt and swap it back. It may use FX swaps to manage short-term currency liquidity. It may hold HQLA in one currency while outflows arise in another. ALM therefore needs currency-level and entity-level views, not only group totals.

A good ALM story answers what the bank owns, how those assets are funded, when cash flows arrive, when liabilities leave, how rates reset, which deposits are stable, which wholesale maturities create cliffs, which currencies are mismatched, which securities are usable, which hedges exist, which assumptions matter and which decision ALCO must take. If the ALM pack cannot tell that story, it is only a collection of reports.

3. Maturity mismatch

Maturity mismatch means assets and liabilities mature at different times. Banks often transform short-term deposits into longer-term loans. This maturity transformation is central to banking, but it creates liquidity risk. If liabilities leave before assets repay, the bank must refinance, sell assets, repo securities, use central-bank balances, draw committed lines, reduce lending or activate contingency actions. In calm markets, refinancing may be easy. In stress, it may be expensive or unavailable.

A maturity ladder maps expected cash inflows and outflows across time buckets such as overnight, one week, one month, three months, six months, one year and longer horizons. A real ALM ladder should include contractual maturities and behavioural assumptions. Loan repayments, securities maturities, deposit outflows, wholesale funding maturities, repo maturities, derivative cash flows, committed facility drawdowns, operational flows, tax flows, payroll flows and payment settlement flows can all matter. The ladder should be available by significant currency and legal entity.

Maturity mismatch is not automatically bad. Banking earns part of its income from maturity transformation. The risk is unmanaged or mispriced mismatch. A bank that funds long-term illiquid assets with unstable short wholesale money is vulnerable. A bank that ladders funding, diversifies deposits, holds usable liquidity, prices liquidity correctly and stress-tests outflows can take maturity transformation more safely.

FTP should reflect maturity mismatch. A business originating long-term assets should pay the cost of stable funding or term transformation. If FTP charges a five-year loan as if it were funded overnight, the product will look artificially profitable. That encourages growth that may weaken the balance sheet. Truthful maturity pricing is a control, not an accounting detail.

4. Repricing mismatch and IRRBB

Repricing mismatch means assets and liabilities reset interest rates at different times or against different indices. A fixed-rate loan may not reprice for five years. A savings account rate may change tomorrow. A floating-rate loan may reset every three months. A fixed-rate bond portfolio may decline in value when rates rise. A hedge may convert fixed exposure to floating or floating to fixed. Interest-rate risk in the banking book arises because adverse rate movements affect both earnings and economic value.

The Basel IRRBB framework describes IRRBB as current or prospective risk to capital and earnings from adverse movements in interest rates affecting banking book positions. It also identifies gap risk, basis risk and option risk as key sub-types. Gap risk comes from timing differences in rate changes. Basis risk comes from imperfect movement between different interest-rate indices. Option risk comes from explicit or embedded options, including prepayment, early redemption, caps, floors and customer behaviour.

Net interest income sensitivity asks how earnings change over a defined horizon if rates move. If assets reprice faster than liabilities, rising rates may help NII. If liabilities reprice faster, rising rates may hurt NII. But deposit beta matters. If market rates rise by 100 basis points, does the bank pass 20, 40, 70 or 100 basis points to customers? Behaviour matters too. If the bank does not pass enough, will customers leave? NII is not only curve math. It is product behaviour, competition and management action.

Economic value of equity sensitivity asks how the present value of assets, liabilities and off-balance sheet items changes under rate shocks. A long fixed-rate asset loses value when rates rise. A stable non-maturity deposit may have economic value, but that value depends on assumptions. Hedges can reduce or change sensitivity. ALM needs both NII and EVE because the bank is both an earnings business and an economic-value balance sheet.

Supervisory scenarios are not a universal 200-basis-point test

The Basel framework prescribes six EVE shock shapes and two NII parallel-shock scenarios, with currency-dependent calibration and assumptions. The July 2024 recalibration specified implementation by 1 January 2026 at the Basel-standard level. Local effective dates and supervisory constraints must still be checked. A bank should version its jurisdiction, currency, shock parameters and model settings rather than embed an old fixed shock into every report.

Basel’s supervisory EVE outlier threshold compares the maximum decline under the prescribed scenarios with 15 percent of Tier 1 capital. This is a supervisory identification threshold, not a universal bank trading limit, a direct accounting equity loss, or proof that lower exposure is safe. NII, basis, behavioural uncertainty and liquidity effects remain relevant. The versioned SRP31 IRRBB chapter governs the Basel treatment; the bank’s regulator determines its local application.

5. Behavioural assumptions

Behavioural assumptions are the heart and danger of ALM. Many banking products do not behave exactly like contracts. Current accounts may stay for years even though they are repayable on demand. Mortgages may prepay before contractual maturity. Credit-card balances may revolve. Term deposits may roll over. Corporate deposits may be operating balances or rate-seeking treasury balances. Undrawn commitments may remain unused in normal markets but draw quickly in stress.

A behavioural assumption should be evidenced, owned, approved and challenged. Evidence may come from historical data, customer segmentation, product design, rate sensitivity, channel behaviour, stress history, economic cycle, relationship manager insight and competitor behaviour. Ownership means someone is accountable for the methodology. Approval means governance has accepted the assumption. Challenge means risk and treasury can question whether it remains valid.

Deposit beta is a practical example. If policy rates rise, the bank may not immediately pass the full increase to deposit customers. The pass-through depends on product, customer segment, competition, relationship value, digital switching ease and management pricing strategy. Retail operating deposits may behave differently from large corporate treasury deposits. A single beta for every deposit can be misleading.

Prepayment is another practical example. Borrowers may prepay when rates fall, when competitors refinance, when property is sold or when personal circumstances change. Prepayment changes asset duration, NII, EVE and hedge effectiveness. If ALM assumes contractual maturity while customers behave differently, risk measurement becomes weak.

Behavioural assumptions should not be frozen forever. In a rising-rate environment customers become more rate aware. In confidence stress, uninsured or concentrated deposits may leave faster. In digital banking, movement can be faster than old branch-era models suggest. A mature ALM framework reviews assumptions regularly and after market regime changes.

6. FTP as internal truth-telling

Funds Transfer Pricing (FTP) is the internal mechanism used to charge or credit businesses for the funding, liquidity, interest-rate and optionality characteristics of their products. It prevents business lines from treating funding as free. If a business writes a long-term fixed-rate loan, it should be charged for the cost of stable funding and rate transformation. If a business generates stable operational deposits, it may receive internal credit because it provides valuable funding. If a product creates liquidity outflows or drawdown optionality, FTP should reflect that cost.

Good FTP makes product economics and the central treasury residual transparent. It helps the bank price products correctly, compare business profitability fairly and steer the balance sheet toward resilience. Bad FTP creates distorted incentives. A lending business may grow assets that look profitable locally but consume expensive liquidity centrally. A deposit product may look low margin but provide valuable stable funding. A business may sell undrawn commitments without paying for drawdown risk. Treasury then carries hidden cost.

FTP typically includes a base funding curve, term liquidity premium, bank funding spread, currency premium, optionality charge, behavioural maturity, regulatory liquidity cost, sometimes capital-related allocation and product-specific adjustments. The exact methodology is bank-specific. The principle is consistent: internal product economics should reflect the real balance-sheet cost and value.

FTP also supports customer pricing. If wholesale funding cost rises, new loan pricing may need to rise. If stable deposits become more valuable, deposit-gathering businesses should see the value. If liquidity is scarce, products that consume liquidity should become more expensive. FTP turns treasury reality into business behaviour.

7. FTP curves and components

An FTP curve is an internal curve used to price funding across tenors and currencies. It may be built from market rates, swap curves, bank funding spreads, liquidity premiums, term issuance costs and management overlays. It should reflect how the bank can actually fund itself, not an imaginary risk-free curve. A bank normally needs different curves or adjustments by currency because funding markets differ.

The base curve captures the term structure of rates. The bank spread reflects the bank’s own funding cost over the base curve. The liquidity premium reflects the cost of term funding and maintaining liquidity resilience. The optionality charge reflects prepayment, early withdrawal, drawdown or embedded customer options. The behavioural tenor determines how long the product is economically treated. Regulatory cost may reflect LCR, NSFR or other liquidity constraints. These components should be explainable to business users.

FTP curve governance matters. Methodology should be approved by ALCO or the relevant governance body. Curve changes should be documented. Overrides should require approval. Business challenges should follow a defined process. If FTP can be negotiated away casually, it loses control value. If FTP is too opaque, businesses distrust it. The strongest FTP frameworks are rigorous enough for risk and transparent enough to steer behaviour.

FTP should also avoid false precision. A curve with many decimal points can still be wrong if behavioural assumptions are weak. The aim is not decorative mathematics. The aim is a fair internal price that reflects funding, liquidity and interest-rate reality.

Avoid counting the same funding cost twice

An all-in bank funding curve may already contain the bank credit spread and a term liquidity premium. Adding both components again produces an overstated FTP charge. Conversely, using a risk-free curve alone may omit material funding costs. Methodology should state which costs are embedded in the base curve and which are incremental. Repricing tenor and funding tenor also answer different questions: a five-year floating-rate loan might use short-reset interest-rate pricing plus a longer funding/liquidity component. It should not receive five-year fixed-rate risk treatment merely because its contractual final maturity is five years.

For amortising loans, matched cash-flow transfer pricing may use several curve points or a weighted equivalent rate. Non-maturity deposit replication portfolios spread assumed stable balances across tenors; the model should also show the unstable portion and the effects of changing beta or runoff assumptions. Neither approach authorises the bank to refuse an on-demand depositor’s withdrawal.

8. Practical FTP calculation logic

A practical FTP calculation starts by understanding the product’s real balance-sheet behaviour. The system first identifies product type, currency, legal entity, business line, customer segment, rate type, contractual maturity, repricing frequency, amortisation profile, optionality and liquidity treatment. Without this classification, no FTP calculation can be trusted. A five-year fixed-rate mortgage, a floating-rate corporate loan, an overnight operating deposit, a promotional digital savings balance and an undrawn credit line should not receive the same FTP logic.

For an asset such as a loan, the simplified sequence is: identify expected cash-flow profile, assign behavioural maturity if different from contractual maturity, select the correct currency FTP curve, apply the relevant tenor point or matched-maturity curve, add liquidity premium, add bank funding spread where applicable, add optionality charge where the customer can prepay or draw, add regulatory liquidity cost where material, then calculate the internal charge that flows into product profitability. This is not a universal formula for every bank, but it is the practical operating pattern.

For a liability such as a deposit, the sequence is different. The system identifies deposit type, customer segment, stability, rate sensitivity, insurance or guarantee treatment where relevant, concentration, operational relationship and behavioural life. A stable operational deposit may receive an FTP credit because it provides funding value. A rate-sensitive promotional deposit may receive lower credit because it is more likely to leave or requires high interest expense. A large concentrated corporate deposit may receive careful treatment because losing one client can create a funding gap.

For an undrawn commitment, FTP should recognise optionality. The customer has a right to draw in the future, often exactly when liquidity is more valuable. The FTP engine may apply a drawdown assumption, liquidity conversion factor, term assumption and commitment charge. If the bank treats undrawn lines as free, business profitability is overstated and treasury inherits hidden stress risk.

The calculation should produce explainable output: base curve charge, liquidity premium, optionality component, behavioural adjustment, regulatory/liquidity cost, total FTP charge or credit, effective date, curve version and override status. If business users cannot see why FTP moved, the framework will lose trust. If treasury cannot evidence the calculation, the control will fail audit.

Worked FTP allocation without manufacturing income

Assume a fictional bank has a USD 10 million one-year bullet loan charging 6.00 percent, funded by USD 10 million deposits paying 2.00 percent. For a simplified 360-day actual/360 period, ignore fees, credit losses, capital, operating costs, taxes and external wholesale funding. The loan receives an internal FTP charge of 4.50 percent and the deposit business receives a 3.50 percent FTP credit. These are illustrative bank policy rates, not regulatory prescriptions.

The USD 100,000 treasury residual is allocated income for transformation and centrally managed risks; it is not automatically a separately realised profit from borrowing in the market. If the bank raises external funding or executes a hedge, those real costs affect the whole-bank result. FTP allocation by itself creates no cash inflow from outside the bank.

As a separate asset-pricing example, assume a base curve excluding bank-specific funding/liquidity costs of 3.00 percent, a combined incremental funding/liquidity component of 0.80 percent and an optionality charge of 0.20 percent. Total FTP is 4.00 percent. A 6.00 percent customer rate leaves a 2.00 percent interest spread before other costs, or USD 200,000 over the same assumed year on USD 10 million. Define the accounting and profitability treatment of each component so no cost is charged twice. Curve effective dates, locked versus reset transfer prices and early-prepayment breakage charges need policy and customer-contract review.

9. ALM decision algorithm

ALM is not a software algorithm only, but a real bank does follow a decision sequence. First, collect the balance sheet. Second, classify products by contractual and behavioural features. Third, project cash flows by time bucket, currency and legal entity. Fourth, measure liquidity gaps, LCR, NSFR, NII, EVE, concentration, encumbrance and stress outcomes. Fifth, compare those results with appetite and limits. Sixth, identify drivers and management actions. Seventh, take ALCO decisions and track execution.

This decision sequence matters because ALM can otherwise become report production. A report that says LCR is lower is not enough. The bank needs to know why it is lower. Did deposits run off? Did loan growth consume liquidity? Did HQLA become encumbered? Did wholesale funding mature? Did derivative collateral calls increase? Did currency mismatch widen? Did a source feed fail? The action depends on the cause.

The practical management algorithm is therefore cause-and-action based. If the issue is short-term liquidity pressure, treasury may increase HQLA, raise secured funding, slow outflows or activate contingency funding. If the issue is structural funding, treasury may issue term debt, grow stable deposits or slow long-term asset growth. If the issue is NII sensitivity, ALCO may change deposit pricing, hedge, adjust loan pricing or change product mix. If the issue is EVE sensitivity, ALCO may reduce duration, hedge or change behavioural assumptions. If the issue is FTP distortion, methodology may be revised.

For BA and developers, this means ALM systems should not stop at calculations. They should support driver analysis, drill-down, scenario comparison, threshold alerts, decision notes and action tracking. A bank does not only need the number. It needs the reason, owner and action.

The balance-sheet control loop

The diagram separates internal allocation from external execution. Loans and deposits enter from customer systems; treasury positions enter from deal systems; finance and operations establish actual balances and settlement status. Behavioural models and scenario inputs transform those records into cash-flow, NII and EVE projections. FTP assigns costs and benefits to products. ALCO chooses actions, and dealers execute the authorised funding, securities or hedge trades. Confirmed and settled outcomes return to the source records for the next run.

A projection does not move cash. A committee approval does not execute a hedge. A booked hedge does not prove confirmation or settlement. These distinctions matter when a management pack assumes a funding transaction is available but its trade has been rejected, its collateral has not arrived or its proceeds remain in the wrong legal entity.

10. Liquidity cost allocation

The Basel liquidity principles emphasise the need to allocate liquidity costs, benefits and risks to significant business activities. This is the philosophical foundation behind liquidity transfer pricing. A product that consumes liquidity should carry cost. A product that provides stable liquidity should receive value. This is how the bank prevents individual business lines from optimising themselves while weakening the whole institution.

Liquidity cost allocation should consider product maturity, behavioural stability, customer concentration, currency, collateral, encumbrance, drawdown risk, settlement requirements, intraday liquidity and stress outflows. A stable retail current account is not the same as a large uninsured corporate deposit. A committed facility is not the same as a drawn loan. A security that is unencumbered and repo-eligible is not the same as a pledged security. FTP should reflect these differences.

Liquidity value should not be given blindly for balance growth. A bank may gather deposits quickly through promotional rates, but those balances may leave quickly when rates change. The business may celebrate volume, but ALM may see unstable funding. A good FTP framework rewards quality of funding, not only quantity.

This is especially important in digital banking. Customers can move money quickly. Rate comparison is easier. Online channels can accelerate confidence stress. Historical branch-based assumptions may not fully describe modern behaviour. ALM and FTP must adapt to that reality.

11. LCR in practical ALM

The Liquidity Coverage Ratio promotes short-term resilience by requiring banks to hold sufficient unencumbered high-quality liquid assets to meet net cash outflows over a thirty-day stress scenario. In practical ALM language, LCR asks whether the bank can survive a prescribed short-term liquidity stress using assets that can be converted into cash quickly and with little loss.

LCR affects product appetite. Stable retail deposits may create lower assumed outflows than less stable wholesale funding. Committed facilities can create potential cash outflows. Derivative collateral can create liquidity needs. HQLA securities support the numerator only if they meet eligibility and operational requirements. A business line may not naturally see these effects, so ALM and FTP must translate them into pricing and appetite.

HQLA is not just a label. Assets must be unencumbered and operationally available. A security pledged in repo is not freely available in the same way as an unencumbered security. A security held in the wrong entity or currency may not solve the stress need. A government bond may be high quality but still have market value sensitivity. ALM must look at usability, not only classification.

During stress, banks may use their liquid asset buffer. Basel’s LCR standard recognises use of HQLA in stress rather than mechanical hoarding. But use of the buffer requires governance, communication and recovery actions. ALM should know the drivers before the ratio becomes a problem.

LCR arithmetic and the cash-inflow cap

At the Basel-standard level, LCR = eligible HQLA / stressed net cash outflows, with net outflows calculated as prescribed outflows minus the smaller of eligible inflows and 75 percent of outflows. The normal minimum is 100 percent, subject to the standard’s treatment of use in stress and local application. Basel LCR20 explains the ratio and LCR40.77 defines the inflow cap; the liquidity chapter covers asset eligibility and outflow categories in depth.

If illustrative adjusted eligible HQLA is USD 120 million, weighted outflows USD 100 million and eligible inflows USD 90 million, the cap admits only USD 75 million inflows. Net outflows are USD 25 million and LCR is 480 percent. Ignoring the cap would give 1,200 percent and misstate the metric. Inputs here are already weighted and adjusted; these figures are not raw balance-sheet amounts. A high ratio does not ensure timely cash in a particular currency or payment account.

12. NSFR in practical ALM

The Net Stable Funding Ratio promotes structural funding resilience over a one-year horizon by comparing available stable funding with required stable funding. In simple language, NSFR asks whether longer-term or less liquid assets are supported by sufficiently stable liabilities and capital. It complements LCR. LCR looks at short-term stress. NSFR looks at structural funding.

NSFR matters because product growth consumes stable funding. Long-term loans, securities, off-balance sheet exposures and other assets may require stable funding. Retail deposits, term deposits, wholesale funding, issued debt and capital contribute differently to available stable funding. A business may generate attractive margin, but if the product consumes scarce stable funding, FTP should reflect it.

NSFR discourages excessive reliance on short-term wholesale funding to support longer-term assets. It does not prohibit maturity transformation. It forces the bank to maintain a more resilient structural funding profile. ALCO should see how planned asset growth affects NSFR and whether treasury’s funding plan supports business strategy.

For systems, NSFR depends on product classification, maturity, encumbrance, counterparty type, collateral, off-balance sheet exposure and currency where applicable. A wrong product code or maturity can distort treatment. BA teams should treat NSFR data as part of product architecture, not merely a regulatory reporting field.

NSFR is a weighted requirement

At the Basel-standard level, NSFR = available stable funding (ASF) / required stable funding (RSF), with a normal minimum of 100 percent. The NSF20 calculation and reporting chapter sets the framework. Suppose the applicable classification and weighting have already produced ASF of USD 110 million and RSF of USD 100 million. NSFR is 110 percent. If new assets add USD 15 million weighted RSF with no additional ASF, it falls to 110/115 = 95.65 percent. ALCO must consider funding or balance-sheet actions. The example does not imply a universal weight for that asset type; local categories, remaining maturity, encumbrance and counterparty characteristics determine the input.

13. ALCO as the steering room

Many banks govern ALM through an Asset Liability Committee. An effective ALCO combines reporting with decisions, execution ownership and follow-up. It reviews liquidity, funding, IRRBB, capital interaction, balance-sheet growth, stress results, limit usage, FTP, deposit behaviour, securities strategy, hedge strategy and strategic actions. It should make decisions, not only receive dashboards.

ALCO decisions may include issuing term debt, changing deposit pricing, adjusting loan pricing, revising FTP curves, changing hedge strategy, buying or selling securities, increasing HQLA, reducing wholesale concentration, changing risk appetite, approving contingency actions or slowing balance-sheet growth. These decisions affect customers and business lines. ALCO is not separate from commercial reality.

A good ALCO pack starts with what changed. Which risks increased? Which assumptions changed? Which metrics moved? Which limits are close? Which decisions are needed? Then it supports the story with NII, EVE, liquidity ratios, funding ladder, deposit trends, loan growth, securities portfolio, hedge position, currency liquidity, FTP impact and stress results.

Decision tracking matters. What was approved, why, by whom, for how long and with what expected impact? If ALCO accepts higher duration risk to support income, record it. If it approves a funding plan, track execution. If it changes FTP methodology, show effective date and business impact. Governance without traceability is fragile.

14. Limits, triggers and appetite

ALM needs limits and triggers, not only reports. Common limits include LCR limit, NSFR limit, survival horizon, cumulative liquidity gap, wholesale funding concentration, top depositor concentration, currency mismatch, legal-entity liquidity transfer limits, NII sensitivity, EVE sensitivity, securities duration, HQLA composition, encumbrance, intraday liquidity usage and contingency funding triggers. Limits turn board appetite into daily management discipline.

A trigger is an early-warning point before a hard limit breaks. For example, if LCR falls toward appetite, treasury should know before it breaches. If deposit concentration rises, relationship actions can start early. If wholesale maturities cluster in one month, funding can be issued before markets become difficult. If EVE sensitivity approaches appetite, hedge options can be reviewed before the risk becomes uncomfortable.

Limit breaches should have workflow. The bank should record breach cause, owner, immediate action, approval, expiry, remediation and closure. A breach caused by market movement is different from a breach caused by data error, but both require governance. Repeated temporary approvals should trigger review because they may indicate hidden appetite change.

For systems, limits should be linked to data quality. A green limit based on missing loan feeds, stale market data or incomplete deposit segmentation is not reliable. ALM dashboards should show completeness and freshness alongside utilisation.

15. NII and earnings risk

Net interest income is the difference between interest earned on assets and interest paid on liabilities, adjusted by balance-sheet mix and funding behaviour. For many banks, NII is a major income source. ALM studies how NII changes when rates, volumes, deposit pricing, loan repricing, prepayments and funding costs change.

If assets reprice faster than liabilities, rising rates may increase NII. If liabilities reprice faster, rising rates may reduce margin. But simple labels can mislead. Deposit beta may change. Customers may move balances. Loan demand may fall. Prepayments may slow. Hedges may alter sensitivity. Wholesale spreads may widen. ALM should model scenarios, not rely on slogans.

NII scenarios usually include parallel shocks, curve steepening, curve flattening, basis changes and balance-sheet volume assumptions. They may include management actions such as changing deposit pricing, loan pricing or hedge strategy. The challenge is realism. A scenario that assumes deposit costs never rise may overstate income. A scenario that assumes all customers immediately demand full market rate may be too harsh. Behavioural evidence matters.

For product teams, NII is practical. If a loan margin depends on cheap deposits, the bank must know whether those deposits are truly stable. If a product pays customers aggressively for deposits, the value of the funding may be lower than the headline balance suggests. FTP links these economics.

NII sensitivity with deposit beta and timing stated

In a simplified one-year static-balance-sheet example, USD 100 million floating assets reprice immediately and fully after a +100-basis-point shock. USD 80 million deposits reprice immediately with a 40 percent beta, so their rate rises 40 basis points. Remaining funding is fixed for the horizon. Ignore new business, floors, lags, basis, prepayment and hedge effects. Annual NII change is 100m × 0.0100 − 80m × 0.0040 = +USD 680,000.

If deposit beta becomes 100 percent, the gain is only USD 200,000. If the assets are fixed throughout the horizon while those deposits still reprice at 40 percent beta, the change is −USD 320,000. A beta therefore cannot be interpreted without the asset repricing schedule. If repricing occurs halfway through the year, the relevant period weight changes; do not apply a full-year benefit to a three-month exposure.

Static, constant-balance-sheet and dynamic forecasts need distinct labels. A dynamic commercial forecast may include deposit pricing responses, loan origination and planned hedges, while a supervisory calculation can require different assumptions. Report both the no-action exposure and the management-action case where useful; an unexecuted proposed hedge should not appear as an existing position.

16. EVE and economic-value risk

Economic value of equity measures how the present value of assets, liabilities and off-balance sheet items changes under interest-rate scenarios. While NII focuses on earnings over a defined horizon, EVE focuses on economic value. A long fixed-rate loan loses economic value when rates rise. A stable low-cost deposit may have economic value if it behaves as long-term funding. Hedges can reduce or change EVE sensitivity.

EVE is important because near-term earnings can hide long-term risk. A bank may show strong NII while building a large duration position. It may earn income from long securities while becoming sensitive to rate rises. It may hedge EVE at the cost of current income. ALCO needs to understand the trade-off between earnings protection and economic-value protection.

The Basel IRRBB framework treats earnings and economic value measures as complementary. They share assumptions but answer different questions. NII asks what happens to income. EVE asks what happens to economic value. A bank that manages only one view can miss risk in the other.

EVE requires good cash-flow modelling, discount curves, behavioural assumptions, optionality, prepayment treatment, deposit maturity assumptions, hedge linkage and scenario generation. A wrong behavioural maturity for non-maturity deposits can materially change the result. A missing embedded option can distort duration. EVE is only as good as the assumptions behind it.

EVE sensitivity is not an accounting journal

Use EVE = PV(assets) − PV(liabilities) + PV(off-balance-sheet positions), with a consistent asset/liability sign convention. In a simplified duration approximation, assets have PV USD 120 million and modified duration 4; liabilities have PV USD 110 million and modified duration 1. For a parallel +100-basis-point shock, ignoring convexity, optionality, credit spreads and hedges, asset PV changes by −120m × 4 × 0.01 = −USD 4.8m; liability PV changes by −110m × 1 × 0.01 = −USD 1.1m. EVE change is −4.8m − (−1.1m) = −USD 3.7m. Liability value decreasing offsets part of the asset loss.

This approximation is a teaching check, not a replacement for cash-flow revaluation. With large shocks, nonparallel curves, prepayments or deposit options, duration alone is insufficient. The EVE loss is not automatically booked to profit or loss or OCI: financial-statement effects depend on accounting classification and hedge accounting. Comparing EVE to accounting equity without reconciling scope, discounting and assumptions is misleading.

17. Deposits in ALM and FTP

Deposits are central to ALM because they fund the bank and shape interest-rate risk. But deposits are not one product. Retail current accounts, savings accounts, term deposits, corporate operating balances, public-sector deposits, financial-institution deposits, wealth deposits and promotional digital deposits behave differently. Some are stable. Some are rate sensitive. Some are concentrated. Some are insured. Some are operationally linked to payments. Some are investment balances looking for yield.

ALM segments deposits by behaviour. Stable operating deposits may provide long-term value even if payable on demand. Rate-sensitive deposits may leave quickly when competitors pay more. Large corporate deposits may require relationship-level knowledge. Digital deposits may move faster than legacy assumptions. Term deposits may have contractual maturity but renewal behaviour matters.

FTP credits deposits based on value, not only volume. A stable low-cost operating deposit may receive higher internal credit because it provides valuable funding. A rate-chasing promotional deposit may receive less value because it can leave quickly or requires high interest cost. This is how FTP steers behaviour. Businesses should be rewarded for quality of funding, not only balance size.

Deposit assumptions must be reviewed when market conditions change. In a rising-rate cycle, customers may become more rate aware. In confidence stress, concentrated or uninsured deposits may leave faster. In digital channels, transfers can happen quickly. ALCO should review deposit trends, rate paid, balance migration, concentration and behavioural stability.

18. Loans and customer optionality

Loans create assets, customer relationships and income, but they also create funding need, liquidity risk and interest-rate risk. A fixed-rate loan locks customer pricing but can lose value when rates rise. A floating-rate loan reprices but may create affordability or credit risk when rates rise. A revolving facility may be undrawn today and drawn in stress. A mortgage may prepay when rates fall. A corporate loan may refinance early. These behaviours are optionality.

FTP should charge loans for funding tenor, liquidity cost, optionality, currency, prepayment risk, committed drawdown risk and regulatory cost. A five-year fixed-rate loan should not be priced as if it consumes overnight funding. An undrawn commitment should not be treated as free because the customer may draw under stress. A loan with prepayment rights should reflect the risk that the bank loses an attractive asset when rates fall.

ALM needs loan cash-flow data: contractual maturity, amortisation, rate type, repricing frequency, prepayment assumptions, currency, customer segment, collateral, undrawn commitment, probability of drawdown, credit quality and product type. Missing or wrong fields distort funding and risk measurement.

Product teams should understand FTP before launch. If a product gives customers valuable options, the price should reflect them. Otherwise the business may win volume by giving away balance-sheet value. ALM protects the bank by making optionality visible.

19. Securities portfolio, HQLA and collateral

The securities portfolio plays several ALM roles. It can generate income, provide liquidity, serve as collateral, manage duration, support regulatory liquidity and diversify assets. High-quality liquid assets are central to LCR. But securities also create market risk, accounting effects, repo constraints and collateral management requirements.

A securities portfolio should be divided by purpose. Some securities are immediate liquidity. Some are investment yield. Some are held as collateral. Some are strategic duration positions. Mixing purposes creates confusion. A bond bought for yield may not be suitable as stress liquidity. A bond held as HQLA should be unencumbered, eligible, liquid and operationally available.

Duration matters. A long-duration government bond may be high quality but sensitive to rate rises. If the bank must sell under stress, unrealised loss can become realised. Accounting classification matters because fair-value movement may affect profit and loss or equity depending on treatment. ALCO should understand the income-versus-risk trade-off.

Repo capacity matters. A security that can be repoed may create cash without outright sale, but haircuts, counterparty capacity, settlement readiness and legal-entity location matter. ALM should track securities by eligibility, haircut, encumbrance, currency, legal entity, custodian and monetisation route. The useful number is usable cash, not only market value.

Three values for the same bond

A USD 10 million market-value bond subject to an illustrative 2 percent repo haircut can raise USD 9.8 million cash before costs, if eligibility, capacity and settlement readiness are satisfied. That potential funding capacity is not another USD 9.8 million asset to add to the bank’s USD 10 million security. Once pledged, the bank must update encumbrance and the funding obligation; it cannot count an unavailable asset twice as freely usable liquidity. The applicable LCR rules determine any regulatory treatment, which is not inferred from this arithmetic.

Accounting carrying value, stress-sale value and collateral borrowing value can all differ. ALCO should see the bridge and the expected time to mobilise the asset. A custodian holding the bond in the wrong account, an unapproved security identifier or a pending settlement can make theoretically eligible collateral unusable at the moment cash is required.

20. Currency and legal-entity constraints

Trapped liquidity is surplus that cannot be moved to the entity or currency needing it within the required time. A banking group may show surplus liquidity at consolidated level while a branch, subsidiary, currency or clearing account is tight. Liquidity cannot always move freely because of regulation, local supervisory expectations, tax, capital rules, settlement timing, documentation, market access or internal risk appetite. ALM must therefore measure liquidity where it is needed, not only where it exists.

Currency mismatch is also critical. A bank may have strong EUR liquidity and weak USD liquidity, or strong domestic-currency deposits and foreign-currency loan growth. FX swaps can transform currency liquidity, but swap markets can become expensive or less available in stress. A consolidated ratio may hide the fact that one currency depends heavily on market access.

Legal-entity ALM matters for resolution and recovery planning. If one subsidiary has HQLA, another entity cannot automatically assume it can use it. ALM reporting should show entity-level liquidity, entity-level funding gaps, currency liquidity, transfer restrictions and intragroup funding dependencies. Senior management needs to know where liquidity can actually be mobilised.

For BA teams, this means every ALM data model must carry legal entity, branch, currency, account location, custodian, nostro, central-bank account, encumbrance and transferability indicators. Without these fields, the system may show liquidity that is not operationally usable.

21. Derivatives and hedging in ALM

Derivatives allow ALM to adjust interest-rate and currency risk without changing underlying customer products immediately. Interest-rate swaps can convert fixed exposure to floating or floating to fixed. Caps and floors can protect against rate extremes. FX swaps can transform currency funding. Cross-currency swaps can manage currency and interest-rate exposures together. But derivatives create valuation, collateral, counterparty, accounting and operational risk.

A hedge should have clear purpose. Is it hedging NII, EVE, a specific loan portfolio, funding issuance, securities duration, deposit behaviour or currency mismatch? The answer affects measurement and accounting. A hedge that reduces one metric may worsen another. A swap may reduce EVE sensitivity but create collateral calls. A hedge may work economically but fail hedge accounting if documentation is weak.

ALCO should review hedge strategy in plain language. What exposure exists? What hedge is proposed? What risk is reduced? What new risk is created? What is the collateral impact? What is the accounting treatment? What happens under stress? Who monitors effectiveness? These questions prevent derivatives from becoming black boxes.

Systems must link hedges to hedged items where needed. Trade capture, valuation, risk, collateral, finance and ALM systems must reconcile. If ALM assumes a hedge exists but risk systems show different terms, management decisions are wrong.

Why a successful hedge can require cash

A bank holds fixed-rate assets and pays fixed/receives floating on a swap to reduce rate-rise EVE exposure. In a rate-rise scenario the fixed assets generally lose economic value while this swap generally gains, all else equal. If rates instead fall, the assets may gain while the pay-fixed swap loses and can require posted VM. The bank does not receive spendable cash merely because its mortgage portfolio’s economic value rose. Model the hedge and hedged item in the same scenario, then separately project collateral calls, coupon dates and realisable cash. Do not state that every economically successful rate-rise hedge must itself post VM in that same simple scenario.

22. Capital, profitability and steering

ALM does not manage liquidity and rates in isolation. Capital and profitability matter. A product may generate margin but consume capital and stable funding. A securities strategy may improve liquidity but reduce return. A deposit campaign may grow balances but increase interest expense. A hedge may reduce risk but cost income. Balance-sheet steering makes these trade-offs explicit.

Return metrics should include funding cost, liquidity cost, capital cost, expected credit loss, operational cost and optionality. A business line that ignores funding and liquidity can overstate profitability. FTP corrects this by charging internal costs. Capital allocation adds another layer. A product that uses scarce capital and stable funding should earn enough return to justify that use.

ALCO and executive management decide the desired balance-sheet shape. Does the bank want more mortgage growth, more corporate lending, more stable deposits, less wholesale reliance, more HQLA, lower duration, higher NII or lower EVE sensitivity? These goals can conflict. ALM provides the numbers and scenarios to make the trade-offs visible.

A strong bank uses FTP and ALM MI continuously. If business growth weakens liquidity, management acts. If deposit mix changes, pricing changes. If rate risk rises, hedging or product strategy changes. Steering is active.

23. Stress testing in ALM

ALM stress testing asks what happens when assumptions fail. What if deposits leave faster than expected? What if wholesale funding closes? What if rates rise or fall sharply? What if the curve steepens or inverts? What if credit spreads widen? What if customers draw committed lines? What if prepayments accelerate or stop? What if FX swap markets become expensive? What if HQLA haircuts increase? What if payment flows create intraday pressure?

Stress testing should join liquidity, NII, EVE, capital, collateral and management actions. A rate shock may reduce securities value, change deposit behaviour and create derivative margin calls. A liquidity stress may force securities sale and crystallise losses. A credit stress may increase drawdowns and reduce funding access. A currency stress may create FX swap dependency. Siloed stress tests miss interactions.

The Basel liquidity principles emphasise stress testing, contingency funding plans and allocating liquidity costs, benefits and risks to business activities. This is directly connected to ALM and FTP. Stress results should feed FTP, funding plan, liquidity buffer, product appetite and ALCO decisions.

Stress tests should produce actions. If the test shows a funding cliff, issue earlier. If it shows deposit concentration, change pricing or relationship plan. If it shows duration risk, hedge or rebalance. If it shows weak HQLA usability, improve collateral positioning. A stress test that changes no decision is weak.

24. Contingency funding and recovery connection

A contingency funding plan defines triggers, escalation, actions, communication, liquidity sources and responsibilities. It should be linked to ALM stress results. If stress testing shows wholesale funding closure as severe, the CFP should include credible alternatives. If deposit outflow is the main risk, relationship, pricing and communication actions should be included. If collateral mobilisation is key, operational readiness should be tested.

Recovery planning is the next level. In severe stress, the bank may need options such as raising capital, selling assets, issuing secured funding, reducing lending, changing pricing, accessing central-bank facilities or restructuring businesses. ALM provides the balance-sheet facts: what assets can be sold, what funding matures, what liquidity is available, which entities are affected and what actions are credible.

The CFP should not be a document nobody uses. It should have stages, owners, dashboards and rehearsals. ALCO should know when it activates. Treasury should know which actions are operationally ready. Business lines should know which products may be slowed or repriced. Operations should know payment and collateral priorities.

A strong ALM function connects normal steering with stress playbooks. The same data used in calm markets should support crisis decisions. If the bank uses one set of numbers in ALCO and another in crisis, confusion follows.

25. Intraday liquidity and payments connection

Intraday liquidity is often missed in simple ALM explanations, but it is critical in real banks. A bank may be solvent and have enough end-of-day liquidity, yet still face pressure during the day if payments, settlements, securities movements, CLS obligations, clearing systems, margin calls or correspondent banking flows require liquidity at specific times. Payment timing matters.

Treasury and ALM need visibility of major payment flows, settlement cut-offs, clearing obligations, collateral movements, nostro balances, central-bank account balances and intraday credit usage. A corporate payment run, securities settlement cycle, margin call or FX settlement can create time-specific liquidity need. Liquidity available at 5 PM may not solve a payment obligation at 10 AM.

Intraday liquidity also affects operational resilience. If a payment system outage delays incoming flows but outgoing obligations continue, liquidity pressure can rise. If collateral movements fail, margin obligations may remain. If a nostro is funded late, payments may queue. ALM should connect with payment operations and treasury dealers, not only finance balances.

For BA teams, intraday liquidity requirements include timestamped balances, expected inflows, expected outflows, payment queue, settlement system, currency, nostro account, counterparty, cut-off, collateral movement, overdraft usage and exception workflow. This is where banking operations and ALM meet directly.

26. ALM run cycle and operating calendar

Another practical topic often missed is the run cycle. ALM is not only monthly ALCO. A real bank has daily liquidity monitoring, intraday liquidity monitoring, weekly funding updates, monthly ALCO packs, monthly or quarterly FTP updates, periodic behavioural assumption reviews, regulatory reporting cycles, stress testing cycles and annual framework reviews. Each cycle has owners, cut-offs, data feeds, sign-offs and escalation steps.

The daily run may focus on liquidity position, cash ladder, large flows, market funding, HQLA, collateral, central-bank account, nostro funding and limit usage. The monthly run may focus on NII, EVE, FTP, balance-sheet trends, funding plan and ALCO decisions. The quarterly run may review assumptions, stress testing, model performance and regulatory packs. The annual run may review policy, appetite, FTP methodology and model validation.

Cut-off discipline matters. If loan data arrives after the ALM batch, the report may miss new assets. If deposit data is stale, liquidity assumptions are wrong. If market curves are updated after FTP calculation, profitability can be inconsistent. If finance reconciliations happen after ALCO, management may be looking at unverified numbers. The operating calendar should define when data arrives, when calculations run, when exceptions are resolved and when reports are signed off.

For systems, run-cycle controls should include batch status, source completeness, reconciliation status, market data timestamp, assumption version, FTP curve version, scenario version, report version and sign-off evidence. A practical ALM platform should make the run visible, not hidden inside batch jobs.

27. Data model for ALM and FTP

ALM data comes from many systems: core banking, loan platforms, deposit systems, treasury systems, securities systems, derivatives platforms, collateral systems, payment systems, general ledger, customer master, market data, risk systems and finance. The data model must capture product type, balance, currency, legal entity, customer segment, contractual maturity, behavioural maturity, repricing date, rate type, index, spread, amortisation, optionality, collateral, undrawn commitment, account type, deposit stability segment, FTP curve, liquidity category, accounting classification and hedge linkage.

Data lineage is critical. A number in an ALCO pack should be traceable to source. If loan balances differ from finance, reconcile. If deposit segmentation differs from product systems, explain. If FTP charges differ from profitability reports, investigate. If securities HQLA differs from treasury inventory, resolve. ALM is credible only when data is credible.

Common defects include missing maturity, wrong rate type, incorrect repricing frequency, stale customer segment, wrong currency, duplicate account, missing commitment, wrong product code, broken hedge link, manual override without approval and late source feed. Each defect can change risk and economics. BA requirements should include validation, exception reporting, owner and remediation workflow.

FTP data must link customer product economics to treasury cost. A loan should receive the right curve, tenor, liquidity premium and optionality charge. A deposit should receive the right behavioural value. A commitment should receive drawdown cost. If mapping is wrong, business profitability is wrong.

28. Systems architecture

An ALM and FTP architecture typically includes source systems, data pipelines, data warehouse, ALM engine, FTP engine, market data, behavioural models, stress engine, reporting layer, finance reconciliation, risk reporting and governance workflow. Some banks use vendor ALM platforms. Others build internal engines. The architecture matters less than the controls: complete data, explainable assumptions, consistent calculations, reconciliation and audit trail.

The ALM engine projects cash flows, repricing, liquidity gaps, NII, EVE and stress. The FTP engine assigns internal funding charges or credits. Market data provides curves and rates. Behavioural models provide assumptions for non-maturity deposits, prepayments and drawdowns. Reporting presents outputs to ALCO, treasury, risk and business lines. Finance reconciles balances. Model governance validates assumptions.

Architecture should support scenario management. Users need to run base case, rate shocks, liquidity stress, deposit runoff, loan growth, funding plan, hedge strategy and FTP changes. Scenarios should be versioned. Assumptions should be stored. Results should be reproducible. A management decision based on a scenario should be traceable later.

Access control matters. Not everyone should change assumptions. Manual overrides should require approval. Model changes should be documented. FTP curve changes should have governance. ALM systems influence pricing and risk appetite, so weak access control can create financial impact.

Finance, risk and operations reconcile different objects

Finance reconciles product principal, accrued interest, carrying values and official ledger balances at a defined cut-off. ALM separately reconciles projected cash flows and repricing dates to the relevant contractual records, then explains behavioural overlays. The sum of projected future cash flows should not be forced to equal current carrying value: interest, discounting, impairments, fees and timing make them different objects.

Front office supplies executed funding and hedge trades. Middle office challenges scenarios, validates market inputs and limits, and monitors sensitivity. Back office proves confirmation, settlement, collateral and account balances. An ALM data store should retain both transaction state and projected treatment. If settlement fails, treasury updates usable liquidity even when finance still recognises a booked receivable. Reconcile FTP allocations across business-unit profitability and the central treasury residual so internal charges cancel on consolidation. Retain cut-off, source counts, balance totals, exception owners and rerun version as audit evidence.

29. Controls and audit evidence

ALM and FTP controls must be evidenced. A behavioural assumption should show methodology, data, owner, approval and review date. An FTP curve should show source data, calculation date, approval and effective date. A stress scenario should show assumptions and results. A hedge should show purpose, linkage, accounting treatment and monitoring. A limit breach should show cause, approval and remediation.

Internal audit may review ALM governance, liquidity risk management, IRRBB measurement, FTP methodology, data quality, model validation, ALCO minutes, stress testing, contingency funding, hedge governance and regulatory reporting. The question is not only whether a policy exists. The question is whether the control operates and whether evidence proves it.

Model risk applies to ALM assumptions. Deposit models, prepayment models, drawdown assumptions, NII models, EVE models and FTP models influence management decisions. They need validation proportionate to materiality and complexity. A simple rule can still be material if it drives billions of balance-sheet treatment.

A mature bank treats findings as improvement. Repeated ALM findings usually point to unclear ownership, weak data, manual workarounds, assumption governance gaps or poor reconciliation. The remediation should fix root cause, not only the report page.

30. BA and testing scenarios

A strong ALM test pack begins with a new fixed-rate loan. The loan should flow from source system into ALM with correct balance, maturity, rate, currency, repayment schedule, product type, customer segment, FTP curve and liquidity treatment. The ALM engine should project cash flows, apply FTP, calculate NII and EVE sensitivity, and reconcile balance to finance.

Test a savings account deposit. Contractual maturity may be on demand, but behavioural maturity and repricing assumptions should apply. Change deposit rate and test NII. Apply deposit runoff stress and test liquidity. Change customer segment and test FTP credit. Test large corporate deposits separately from retail deposits. This proves behaviour matters.

Test a committed facility. The drawn balance may be zero, but the undrawn amount should appear in liquidity stress and FTP optionality cost. Draw part of the facility and test loan balance, funding need, LCR impact and profitability. If the system ignores undrawn commitments, ALM risk is understated.

Test a securities purchase. The security should feed balance, maturity, coupon, duration, HQLA category, encumbrance, accounting classification and liquidity stress. Pledge the security in repo and test encumbrance. Sell the security and test cash and accounting. Apply a rate shock and test valuation and EVE impact.

Test a hedge. Book an interest-rate swap linked to a loan portfolio. Confirm that ALM receives hedge cash flows, risk receives sensitivities, collateral system receives exposure and finance receives valuation. Terminate the hedge and test all downstream updates. A hedge that exists in one system only is not a real control.

Test FTP. Create a five-year fixed-rate loan and verify the five-year funding curve, liquidity premium and optionality charge. Create an overnight deposit and verify behavioural value. Change FTP curve effective date and confirm old and new bookings are treated correctly. Test manual FTP override and approval. Test profitability report reconciliation.

Test stress. Apply deposit runoff, rate shock, wholesale funding closure, collateral call, FX swap widening and committed facility drawdown. Confirm that liquidity, NII, EVE, FTP and ALCO reports show the result. Store scenario assumptions and reproduce the result. Stress testing without reproducibility is weak.

Test legal-entity and currency liquidity. Create surplus liquidity in one entity and deficit in another. Verify that consolidated liquidity does not hide entity stress. Create a USD funding gap and EUR surplus. Verify that FX swap dependency is shown. Test transfer restrictions and trapped liquidity flags.

Test ALM run-cycle controls. Break one source feed, delay market curves, change FTP curve version, rerun the batch and verify status, exception, sign-off and report version. This proves the bank can trust the process, not only the calculation.

31. Practitioner casebook

A bank grows mortgage lending aggressively because customer demand is strong. The business sees volume and customer margin. ALM shows the mortgages are long fixed-rate assets funded partly by short deposits and wholesale funding. EVE sensitivity rises. FTP is updated to charge longer-term funding and optionality. The business complains that profitability has fallen. ALCO explains that profitability without funding cost is not real profitability. The lesson is that FTP protects the bank from false margin.

A digital savings product gathers large balances quickly through promotional rates. Finance celebrates deposit growth. ALM segments the balances and finds they are rate sensitive and concentrated in new customers with limited operating relationship. LCR may improve temporarily, but behavioural stability is uncertain. FTP gives lower value than stable operating deposits. The lesson is that balance volume is not the same as funding quality.

A securities portfolio earns attractive yield by extending duration. Rates rise. The securities show unrealised losses and EVE sensitivity worsens. Treasury argues the bonds are high quality. Risk argues they are long duration. ALCO must decide whether income was worth the risk. The lesson is that HQLA quality does not remove rate risk.

A corporate facility is undrawn for years. During market stress, several clients draw at once. Liquidity outflows rise just when wholesale funding is expensive. FTP had not charged enough for undrawn optionality. The lesson is that commitments are liquidity promises and must be priced.

A hedge reduces EVE sensitivity but creates collateral calls under rate shock. Treasury must fund margin. The hedge is economically useful, but liquidity stress increases. The lesson is that ALM and liquidity must analyse derivatives together.

A payment operations issue delays major incoming flows while outgoing clearing obligations continue. End-of-day liquidity looks comfortable, but intraday liquidity becomes tight. Treasury funds the nostro late and payments queue. The lesson is that ALM must understand timing, not only balances.

A non-maturity deposit model assumes long behavioural life based on old data. After sustained rate rises, customers move funds faster than expected. NII and EVE assumptions become unreliable. The model is recalibrated and ALCO reduces reliance on that deposit segment. The lesson is that behavioural assumptions must evolve with market regime.

A group has excess liquidity in one subsidiary and a shortage in another. The consolidated ALCO pack looks comfortable, but local transfer restrictions mean liquidity cannot move quickly. The entity under pressure must raise local funding. The lesson is that ALM must measure usable liquidity, not theoretical group liquidity.

A business challenges FTP because a product appears uncompetitive. Treasury decomposes the FTP charge into base curve, liquidity premium, term funding spread and prepayment cost. The business understands that the product is not cheap to fund and adjusts customer pricing. The lesson is that transparent FTP builds trust even when the answer is uncomfortable.

32. Source references used for this chapter

• BIS Basel Committee - Principles for Sound Liquidity Risk Management and Supervision - Basel source for liquidity risk governance, liquidity cost allocation, stress testing, contingency funding, intraday liquidity and collateral principles.
• BIS Basel Committee - Liquidity Coverage Ratio - Primary Basel source for LCR, HQLA and short-term liquidity resilience.
• BIS Basel Committee - Net Stable Funding Ratio - Primary Basel source for NSFR and stable structural funding.
• BIS Basel Framework - SRP31 Interest rate risk in the banking book - Current Basel framework source defining IRRBB, NII, EVE, gap risk, basis risk, option risk and supervisory expectations.
• BIS Basel Framework - SRP98 Application guidance on IRRBB - Basel application guidance explaining IRRBB measurement techniques, earnings measures, economic value measures and behavioural optionality.
• BIS Basel Framework - LCR30 High-quality liquid assets - Basel framework source explaining HQLA characteristics and operational requirements.
• BIS Disclosure Requirements - IRRBB - Basel disclosure source for IRRBB policy and quantitative NII/EVE shock information.
• BIS FSI - Basel III liquidity monitoring tools - BIS source explaining liquidity monitoring tools beyond LCR and NSFR.

Final takeaways

ALM asks whether the bank can live with the shape of its balance sheet. FTP asks whether each business is being charged or credited fairly for the shape it creates. Treasury executes funding and liquidity actions. Risk challenges assumptions. Finance connects numbers to official books. Product teams feel the pricing impact. Senior management owns appetite. When these teams work together, balance-sheet steering becomes real.

The learner should leave this chapter with a practical instinct. When you see a loan, ask how it is funded, how it reprices, how it behaves, what liquidity it consumes and what FTP it receives. When you see a deposit, ask whether it is stable, rate sensitive, concentrated and valuable. When you see securities, ask whether they are income assets, HQLA, collateral or duration risk. When you see a hedge, ask what risk it reduces and what new risk it creates. When you see FTP, ask whether it tells the truth.

That is ALM, FTP and balance-sheet steering inside a real bank. It is not a monthly report ritual. It is the management discipline that connects products, treasury, liquidity, rates, capital, profitability, payment timing, customer behaviour, legal-entity reality and executive decisions into one controlled banking story.