Chapter 11 · Day 11 — Commercial banks
How a bank is actually judged
Banks need their own framework. Margin, credit quality and capital — in that order — decide the answer.
A bank's product is money. Its revenue is the spread between what it pays depositors and what it charges borrowers, and its main risk is that the borrowers do not repay. Almost none of the general ratios apply.
| Metric | Formula | Reads |
|---|---|---|
| NPL ratio | Non-performing loans ÷ Total loans × 100 | Credit quality — lower is better |
| NIM | Net interest income ÷ Average earning assets × 100 | Core margin |
| ROA | Net profit ÷ Average assets × 100 | Efficiency of the whole balance sheet |
| ROE | Net profit ÷ Average equity × 100 | Return to owners |
| Cost-to-income | Operating expenses ÷ Operating income × 100 | Overhead discipline — lower is better |
| CAR | Capital ÷ Risk-weighted assets × 100 | Cushion against losses (NRB minimum applies) |
| CD ratio | Credit ÷ (Deposits, per NRB definition) | How much of the deposit base is lent |
| Provision coverage | Provisions ÷ Gross NPL × 100 | How much of the bad book is already absorbed |
Gross NPL against net NPL
Gross NPL is the whole non-performing book. Net NPL is what remains after provisions. A bank with 3% gross NPL and 80% provision coverage carries about 0.6% net — materially safer than one with 3% gross and 30% coverage. Always read the two together.
Illustrative comparison — Bank A against Bank B
| Metric | Bank A | Bank B | Stronger |
|---|---|---|---|
| P/B | 1.9× | 1.2× | B on price |
| ROE | 15.2% | 10.4% | A |
| ROA | 1.7% | 1.1% | A |
| Gross NPL | 1.4% | 3.9% | A |
| Provision coverage | 142% | 61% | A |
| NIM | 4.4% | 3.6% | A |
| Cost of funds | 5.9% | 7.2% | A |
| Cost-to-income | 41% | 56% | A |
| Loan growth | 13% | 22% | B on pace |
| Deposit growth | 15% | 12% | A |
| EPS growth | 11% | 4% | A |
Reading it: B is cheaper on P/B and growing loans faster — and those two facts are related to the third. Its NPL is nearly three times A's while its provision coverage is barely half, so future provisions are more likely to hit earnings. Its cost of funds is 130 basis points higher, which is why its margin is thinner. A cheaper multiple on a weaker book is not obviously a bargain.
Why P/B and not P/E for banks
A bank's assets are financial and carried near realisable value, so book value means more than it does for a factory. Earnings, meanwhile, swing on provisioning decisions. P/B against ROE is the standard pairing: a bank earning a high, durable ROE justifies a higher multiple of book.
Saved in this browser only — there is no account to create. Clearing your browser data clears your progress.