Separate legal terms from behaviour
An overnight account may be withdrawable today, but its paid rate need not move immediately with a policy rate. A floating-rate mortgage may reset only at its next contractual date. Build separate time buckets for asset resets, term funding maturities and assumed changes to administered deposit rates. Mark the estimates for non-maturity deposits explicitly. A stable relationship observed in a period with little competition may change when savers can earn materially more elsewhere.
Run a deliberately small shock
Take €100 million of loans repricing immediately and in full, funded by €100 million of deposits. Assume deposits pass through 40% of a one percentage point rise with no lag. Annualised loan income rises €1 million; deposit cost rises €0.4 million. The resulting €0.6 million is a first-order change in net interest income on this slice alone. A delay in asset resets, migration into term accounts or a different hedge position can shrink or reverse the benefit over the actual reporting period.
Keep earnings and value distinct
A twelve-month interest-income estimate does not measure the change in the present value of all future cash flows. Long fixed-rate assets can lose economic value even if recent net interest income improved. Inspect the two measures under parallel and non-parallel changes in the yield curve. When behavioural assumptions are decisive, show an alternative deposit response rather than presenting a precise figure that mostly reflects the chosen model.
Practical check
State the pass-through assumption and reset lag separately for each deposit group. Reconcile hedges before annualising the income effect.