CRAR / Capital Adequacy Ratio

Imagine a tightrope walker who carries a safety harness sized not to their body weight alone, but to the actual riskiness of the specific rope they are crossing that day, a thicker, better-anchored harness for a longer, riskier crossing. Capital to Risk-weighted Assets Ratio, or CRAR, works on almost exactly this logic for a bank.

CRAR measures a bank's capital, the money shareholders and certain long-term investors have put in that can absorb losses, against its risk-weighted assets, its loans and investments adjusted for how risky each one actually is, an unsecured personal loan carries a higher risk weight than a loan fully secured by government bonds. The ratio answers a simple but critical question: if a meaningful share of the bank's loans went bad, does it have enough of its own capital cushion to absorb those losses without collapsing or needing an emergency bailout.

India requires banks to maintain CRAR of roughly 11.5% including regulatory buffers, aligned broadly with the global Basel III framework that most major banking systems adopted after the 2008 financial crisis specifically to prevent a repeat of banks operating with dangerously thin capital cushions.

Capital itself is not one uniform pool, it is split into tiers by quality, Tier 1 capital, chiefly equity and retained earnings, is the highest-quality, most loss-absorbing layer, while Tier 2 capital includes instruments that can absorb losses but with somewhat weaker protection. Regulators pay particularly close attention to the Tier 1 ratio specifically, since it represents the capital genuinely available to absorb losses before depositors or the wider system are ever at risk.

Whenever a bank announces a fresh capital raise through a share sale or a bond issuance specifically described as meeting Basel III or CRAR requirements, that capital is being raised precisely to maintain or rebuild this safety cushion, either to support future loan growth or to absorb stress the bank sees coming in its existing loan book.