| Indicator | Latest reading | Source date | Sensitivity per unit | Lag |
|---|---|---|---|---|
| Policy rate (OPR) | 8.75%, unchanged | 22 Jul 2026 | +1.40% / +100bp | 2-3 quarters |
| Private-sector credit growth | +24.4% y/y (banking sector, Q1) | Q1 2026 | +2.60% / +5pp | 2-3 quarters |
| Headline inflation (CCPI) | 7.3% y/y | Jul 2026 | +0.60% / +1pp | 1-2 quarters |
| Rupee (USD/LKR) | ~333 | Aug 2026 | -0.20% / +5% dep. | 1-2 quarters |
| Workers' remittances | Strong through 2026 to date | Jul 2026 | +0.40% / +10% | 1 quarter |
Readings are sourced and dated; sensitivities are analyst priors, not fitted coefficients.
The Q1 2026 financial-sector release is the most important input in the model right now. Banking-sector credit grew 24.4% y/y against 7.9% a year earlier: a step change, not a drift. The credit-to-deposit ratio passed 70% for the first time in three years, which is the mechanical explanation for why rupee and all-currency liquidity coverage fell to 267.9% and 234.7% from 342.4% and 310.6%. Those are still far above the 100% requirement, so the story is deployment, not stress.
Asset quality improved alongside it. The Stage 3 loans ratio fell to 9.4% from 12.7%, with impairment coverage up to 59.5% from 54.1%. Part of that is denominator: a fast-growing book dilutes the ratio. Profit after tax still fell 7.1% y/y on higher operating expenses, which is the line to watch, because cost growth is the one variable inside management's control that is currently going the wrong way.
| Scenario | Macro path | Modelled revenue impact | Read |
|---|---|---|---|
| Hold (base) | OPR 8.75%, credit sustained near 20%+ | +2.1% | Volume carries the line; margin flat. |
| Fuel-led tightening | +100bp, Brent +30%, CCPI +2.5pp | +3.6% | Margin-positive early, demand-negative from Q2. |
| Disinflation and easing | -150bp, CCPI back to 5%, credit +6pp | -0.4% | Volume gain does not offset the repricing. |
Modelled on the MacroLens elasticity matrix v4.2 against an unchanged-macro baseline. Illustrative.
Every reading in Exhibit 1 comes from the primary publisher with its release date. Nothing enters a brief without a traceable page.
Sensitivities, lags and the Exhibit 2 scenario outputs are analyst priors in the published matrix. Fork them and re-run - the priors are in the open.