Bank margin: separate deposit price from funding mix
A bank’s net interest margin can weaken without a rise in the rate paid on any individual deposit product. Customers may move from noninterest deposits into higher-cost products. Separate that funding-mix change from repricing, then compare both with the yield and mix of earning assets.
Rebuild the numerator and denominator
The Federal Reserve defines NIM as interest income less interest expense divided by average interest-earning assets. A quarterly calculation needs consistent annualization. Check whether the issuer uses a tax-equivalent adjustment, daily averages or another disclosed basis. Those differences matter when reconciling a published margin.
Do not subtract a headline deposit rate directly from a headline loan rate and call the answer NIM. Loans are only part of earning assets, while deposits can be noninterest-bearing or supplemented by wholesale funding. The asset and liability amounts need not match category by category.
Hypothetical funding migration
Assume annual average earning assets of 1,000 million yield 5%, generating interest income of 50. Funding comprises 300 of noninterest deposits, 600 of deposits priced at 2%, and 100 of equity, with no other interest-bearing liabilities. Interest expense is 12, giving NIM of (50 − 12) / 1,000 = 3.8%.
Now assume 100 migrates from noninterest deposits to the same 2% deposit product. The product’s rate is unchanged, but interest expense rises to 14 and NIM falls to 3.6%. A 20 basis-point decline has come entirely from mix in this simplified example.
| Funding worksheet | Initial average | After migration | Annual cost change |
|---|---|---|---|
| Noninterest deposits | 300 | 200 | 0 |
| Interest-bearing deposits | 600 | 700 | +2 |
| Deposit product rate | 2% | 2% | No repricing |
| Interest expense | 12 | 14 | +2 |
| Net interest margin | 3.8% | 3.6% | −20 basis points |
If the 700 interest-bearing balance subsequently reprices to 2.5%, expense increases another 3.5. Keep this price effect separate from the earlier migration. These sequential bridges depend on the chosen ordering; state the ordering or show an interaction term when both balances and rates change simultaneously.
Inspect asset-side offsets
Create a companion table for loans, securities, interest-bearing cash and other earning assets. Record average balance, yield and repricing timing. A floating-rate loan book may adjust sooner than a fixed-rate securities portfolio. A larger cash allocation may reduce the blended asset yield even if every individual loan rate remains unchanged.
Funding stability also needs evidence beyond margin. A high NIM could reflect higher-risk assets rather than a better deposit franchise. Compare deposit concentration, wholesale dependence and loan performance separately. No single spread proves that the earnings mechanism is durable.
Make the follow-up observable
For the next quarter, compare average noninterest deposits with ending balances and explain large differences. A late-quarter deposit move may have little effect on the current average but a larger effect in the following period. Where only ending figures are available, label the estimated bridge rather than presenting it as an exact reconciliation.
The completed worksheet should distinguish asset yield, deposit pricing, deposit mix, wholesale funding and unexplained residual. It is an operating evidence tool, not a forecast of rates or bank solvency. Avoid importing one jurisdiction’s deposit classifications into another bank without checking its reporting framework.
FAQ
Is deposit beta the same as deposit mix?
No. Beta compares a deposit-rate change with a reference-rate change. Mix concerns the balances held in differently priced products.
Can ending deposits replace average deposits?
Only as an explicitly approximate input. Large movements within a quarter can make an ending-balance calculation materially different from the reported margin.