Repricing Risk | What Happens When Interest Rates Change Faster Than a Bank’s Balance Sheet

Every financial instrument has a next repricing date

The bank asks when the contractual price of each asset or liability can change.

• A floating-rate corporate loan may reset next month.
• A fixed mortgage may not reset for three years.
• A savings account can be repriced by the bank relatively quickly subject to terms and competition.
• A twelve-month term deposit is contractually fixed until maturity.
• A five-year bond issued by the bank may have a fixed coupon for its life.

The balance sheet is therefore a calendar of future price changes.

A repricing gap is a timing mismatch

Suppose S$60 million of the bank’s assets can reprice within one year while S$80 million of liabilities can reprice within one year. The bank has more rate-sensitive liabilities than rate-sensitive assets in that horizon.

If rates rise, funding costs may increase faster than asset yields. Net interest income can fall.

If the opposite is true—more assets than liabilities reprice quickly—rising rates can initially help income.

This is a simplified gap view. Real banks model many time buckets, behavioural assumptions and optionality.

The same bank can have different gaps at different horizons

A bank can be asset-sensitive over three months, liability-sensitive over one year and roughly matched over five years. There is no single repricing gap that describes the entire balance sheet.

That is why treasury teams build maturity and repricing ladders. They want to know which contractual and behavioural cash flows become sensitive to rates at each future point.

Net interest income sees the short-to-medium-term effect

One way to measure rate risk is to project how net interest income changes under different interest-rate scenarios.

The bank changes assumptions about benchmark rates, deposit pricing, loan resets, prepayments and funding costs, then estimates future interest income and expense.

This reveals earnings sensitivity: how much near-term profitability depends on rates moving in a particular direction.

Economic value sees the longer-duration effect

A second view asks how the present value of future asset and liability cash flows changes when rates move.

Long-term fixed-rate assets generally lose economic value when market rates rise because their old cash flows are discounted at higher rates. Fixed-rate liabilities can also change in economic value.

A bank can therefore show acceptable near-term income while still carrying substantial long-duration economic-value sensitivity.

Deposit behaviour makes liabilities harder to model than their legal terms suggest

A current or savings account may be withdrawable on demand, but many balances remain for years. The contractual maturity is immediate; the behavioural maturity can be much longer.

Banks model how much of these balances are likely to remain and how quickly their interest rates will respond to market changes.

If the model assumes deposits are very stable and slow to reprice but customers suddenly move aggressively toward higher-yielding alternatives, the bank’s liability behaviour can change much faster than expected.

Deposit beta becomes crucial when rates rise

Suppose market rates rise by 2 percentage points. If the bank raises average deposit rates by only 0.5 percentage points, deposit pricing has moved much less than the market. If competition later forces a 1.8 percentage-point increase, funding cost catches up rapidly.

The bank’s initial margin benefit can therefore fade. Repricing risk includes customer response, not only contractual dates.

Fixed-rate mortgages are long assets with behavioural options

A mortgage can be contractually long but behaviourally shorter because borrowers refinance, sell homes or prepay. Those actions depend partly on market rates.

When rates fall, refinancing can accelerate and remove higher-yielding loans from the bank. When rates rise, borrowers can remain longer, leaving the bank with below-market fixed assets.

The effective duration therefore changes in the direction least convenient to the bank unless the exposure is managed.

Loan floors and caps alter repricing

A floating-rate loan with a floor may stop repricing downward once the floor is reached. A cap may stop repricing upward beyond a defined rate.

These contractual nonlinearities mean the sensitivity of the asset changes at different rate levels. The balance sheet does not always respond proportionally.

Basis risk appears when two rates do not move together

A bank can fund itself using one benchmark and lend using another. Even if both instruments float, the two reference rates may not move by the same amount or at the same time.

This creates basis risk. Batch 13 owns that topic in depth, but it belongs here as a boundary: matching “floating with floating” is not necessarily perfect hedging.

Yield-curve risk means different maturities can move differently

Interest rates do not always rise or fall in parallel. Short-term rates can rise while long-term rates stay flat. Long-term rates can fall while short-term rates remain high.

A bank’s assets and liabilities occupy different points on that curve. Repricing risk therefore includes the shape of rate movements, not only one headline policy rate.

A worked miniature: rates rise quickly

Imagine a bank with S$100 million of fixed-rate loans earning 3 per cent. It funds S$70 million with deposits costing 1 per cent and S$20 million with wholesale funding costing 2 per cent.

Market rates rise sharply. Deposits gradually reprice to 2.5 per cent and wholesale funding to 4 per cent. The fixed-rate loans still earn 3 per cent.

The bank’s interest expense rises while asset income barely changes. Margin compresses. If customers also move deposits to competitors, the bank can face a funding-volume problem at the same time.

A worked miniature: rates fall quickly

Now imagine a bank whose loans are mainly floating and reset downward within one month. Its transaction deposits already pay almost zero and cannot fall much further.

Market rates fall by 2 percentage points. Asset yields fall quickly. Deposit costs barely move. Margin compresses from the other direction.

The same concept—different repricing speeds—explains both outcomes.

Hedging can change the repricing map

Banks can use interest-rate swaps, futures and other instruments to offset parts of their rate exposure. A fixed-rate asset can be economically transformed into floating exposure, or vice versa.

But hedging is not a free cancellation. It introduces counterparty, collateral, basis, model and operational risks. The bank must manage the hedge as another financial claim.

Treasury is where customer products become one balance sheet

The mortgage team may care about mortgage pricing. The deposit team may care about attracting savings. The corporate bank may price floating loans. Treasury has to see all those positions together.

This is why asset-liability management exists. A customer product can be sensible in isolation and still create excessive risk when combined with the rest of the bank.

Repricing risk can become strategic risk

If the bank repeatedly prices long fixed-rate assets too cheaply, it can lock in weak economics for years. If it relies on deposits remaining insensitive to competition, it can discover that customers move faster than models assumed.

Interest-rate risk therefore reflects past strategic choices. Today’s treasury problem may have been created by yesterday’s growth targets.

Repricing risk can become credit risk

If the bank passes rising rates through to floating-rate borrowers, debt-service pressure can increase. If it does not pass rates through because assets are fixed, the bank’s margin can weaken.

This is why rate risk cannot be optimised independently. The bank is always choosing a distribution of pressure across itself, its borrowers and its funding providers.

Stress testing is more useful than one rate forecast

No bank can know the exact future path of rates. It can test multiple scenarios: parallel rises, parallel falls, steepening, flattening, sudden shocks and gradual changes.

The bank then asks which balance-sheet assumptions fail under each path. The objective is not to predict the one correct future. It is to remain viable across several plausible futures.

Four misconceptions to remove

Misconception	Better model
“All bank rates move together.”	Assets and liabilities reset on different contractual and behavioural clocks.
“Demand deposits have no maturity risk because they are callable immediately.”	Their behavioural life can be long, but that behaviour can change under stress or competition.
“Floating assets and floating funding perfectly match.”	Different benchmarks and reset dates can create basis and timing risk.
“One forecast of rates is enough.”	Robust banking tests multiple paths because the exact future is unknowable.

A mastery test

1. What is a repricing date?
2. How can more rate-sensitive liabilities than assets hurt income when rates rise?
3. Why can contractual and behavioural deposit maturity differ?
4. How does borrower prepayment alter the effective duration of fixed-rate loans?
5. Why can a bank be exposed even when both its assets and liabilities are floating?

If those answers connect, repricing risk becomes visible for what it is: the risk that the world changes its price of money before the bank’s old promises are ready to change with it.

---