When the central bank moves rates, a bank's margin doesn't move with it one-for-one. What matters is which side of the balance sheet reprices first — and how sticky deposit pricing is. Shock the rates below and watch it play out.
Rate shock
Parallel shift applied to everything that reprices within one year, scaled by each item's beta.
+100 bps
−300−200−1000+100+200+300
Baseline NIM
—
—
Shocked NIM
—
—
Δ NII (1-yr)
—
—
1-yr repricing gap
—
beta-adjusted RSA − RSL
The balance sheet
Edit any balance, rate, repricing bucket or beta. Beta = how much of the shock passes into the item's rate (CASA is sticky, so its beta is low). Equity plugs the difference automatically.
Item
Balance (Rs M)
Rate %
Reprices in
Beta %
Income/(Cost)
Repricing gap by bucket
Positive gap = more assets than liabilities reprice in that window → rate rises help. Liabilities shown beta-adjusted.
Positive gap (asset-sensitive)Negative gap (liability-sensitive)
NIM across the whole shock range
What this model does (and doesn't)
Static 1-year earnings-at-risk view: each item repricing inside a year feels the shock for the remainder of the year (bucket midpoints: 0–3m → 10.5/12, 3–6m → 7.5/12, 6–12m → 3/12, over 12m → not at all).
Deposit beta captures pricing stickiness — a 30% beta on CASA means only 30 bps of a 100 bps hike reaches those depositors. Betas are the single biggest judgement call in any real IRRBB model.
Not modelled: balance sheet growth or mix shift, prepayments and early withdrawals, basis risk between benchmark curves, non-parallel shocks, and the EVE (economic value) view. Real ALCO packs cover all of these.
Numbers are illustrative. Nothing here reflects any specific bank.