Repricing Gap

A repricing gap is the difference between the amount of a bank's assets and the amount of its liabilities whose interest rates will reset, or which will mature, within a given period. It is one of the oldest and simplest measures of interest rate risk in banking. Treasury and risk teams use it to estimate how a change in market rates would affect earnings, because items that reprice soon take on new rates quickly while items that reprice later keep their old rates.

The measure is a standard tool of asset liability management and is one of the methods banks use to assess interest rate risk in the banking book. It looks at the timing of rate changes, which sets it apart from liquidity gap analysis, which looks at the timing of cash flows, and from the duration gap, which looks at changes in value.

A bank sorts its interest-bearing assets and liabilities into time bands, such as up to one month, one to three months, three to twelve months, one to five years and over five years. Each item is placed in the band in which it next reprices. A floating-rate loan that resets every three months goes in the three-month band whatever its final maturity. A five-year fixed-rate loan goes in the five-year band, although scheduled repayments of principal are slotted in as they fall due. Off-balance-sheet positions such as an interest rate swap are included as a pair of offsetting entries.

For each band, rate-sensitive liabilities are subtracted from rate-sensitive assets to give the periodic gap. Adding the bands together from the shortest onwards gives the cumulative gap, and the one-year cumulative gap is the figure most often quoted.

The gap for a time band and its estimated effect on earnings are calculated as:

• $\text{RSA}$ is rate-sensitive assets, those repricing within the band.
• $\text{RSL}$ is rate-sensitive liabilities repricing within the band.
• $\Delta r$ is the change in interest rates, written as a decimal.
• $\Delta \text{NII}$ is the change in annual net interest income.

Suppose a bank has 400 million of assets and 550 million of liabilities repricing within one year. Its one-year gap is 400 − 550 = −150 million. If rates rise by one percentage point (0.01), annual net interest income changes by roughly −150 million × 0.01 = −1.5 million. The bank is liability-sensitive: its funding costs rise faster than its asset yields. A bank with a positive gap is asset-sensitive and gains when rates rise.

The gap report is reviewed by the asset and liability committee, which sets limits on the size of the gap in each band, often as a share of total assets or capital. If the gap is outside its limit, the bank can change the mix of fixed and floating lending, lengthen or shorten its funding, or hedge with swaps. The same report feeds measures of earnings at risk and supports the supervisory view of the bank's economic value of equity.

Because it is easy to build and to explain, the repricing gap remains common in smaller banks and as a first check in larger ones, alongside simulation models.

The repricing gap has well-known weaknesses:

• Basis risk. It assumes all rates move by the same amount, but deposit rates, loan rates and market benchmarks often move by different amounts; see basis risk and deposit beta.
• Timing within bands. Items in the same band are treated as repricing at the same moment, which hides mismatches inside the band.
• Customer behaviour. Borrowers may repay early and depositors may withdraw or switch products, so contractual repricing dates can mislead. Current and savings accounts with no fixed maturity need a behavioural assumption.
• Earnings only. The gap says little about the effect of rate changes on the long-term value of the balance sheet, and it ignores yield curve risk from non-parallel shifts.

For these reasons the Basel Committee's standards on interest rate risk in the banking book expect banks to use both earnings-based and value-based measures, with the gap report as one input among several.

• Basel Committee on Banking Supervision