ALM & IRRBB

IRRBB and Deposit Behaviour: The Governance Questions That Matter

By Jonas (Yonas) Mohamed Osman Abdelghafour · 4 Aug 2026

Behaviour creates the risk profile

Non-maturity deposits have no contractual repricing date, yet they can represent a large share of a bank's funding. Their assumed stability, pass-through and effective maturity materially affect economic value and net interest income sensitivity.

The central governance question is whether observed historical behaviour remains relevant under the current rate, competition and liquidity environment. Long periods of low rates can provide weak evidence for behaviour after rapid repricing.

Segment before estimating

Segmentation should reflect economically meaningful differences such as customer type, product purpose, balance stability, rate sensitivity and digital mobility. Excessive aggregation can hide unstable cohorts; excessive segmentation can create noisy parameters with little data.

A credible approach documents the reason for each segment and tests whether its behaviour is distinct and persistent. Operational changes, promotional campaigns and product migrations should be separated from underlying customer response.

Use multiple forms of evidence

Statistical estimates should be supplemented by cohort analysis, attrition, deposit concentration, pass-through behaviour and stress experience. Challenger assumptions can quantify how results change when customers reprice or withdraw faster than the central estimate.

Back-testing needs care because realised outcomes are affected by the bank's own pricing decisions. Governance should distinguish model error from a deliberate management choice that changed the observed path.

Translate uncertainty into action

Parameter uncertainty should influence risk limits, hedging and buffer decisions. Reporting a single duration or beta without a credible range can create an illusion of control.

Senior management should see the exposure under central and adverse behavioural assumptions, the cost of hedging, and the time required to change pricing or funding. That is the information needed to make an ALM decision.

Primary sources

About the author

Jonas (Yonas) Mohamed Osman Abdelghafour writes about financial risk management, quantitative modelling, actuarial science, banking risk, insurance risk, capital modelling, model validation, climate risk and geopolitical risk. His work focuses on translating complex quantitative and regulatory risk issues into practical frameworks for financial institutions. Author profile.