How Banks Reprice and Shrink

The question at the bank level is no longer whether syndicated-loan conditions change, but how a U.S. bank absorbs that change. This chapter follows two responses that can look contradictory until they are measured separately: net interest margins rise while the bank’s assets and loan growth contract. The aim is to read the price, quantity, and timing effects together, with special attention to why banks more exposed to loan-intensive activity respond more strongly.

The price response is higher income under tighter capacity

Net interest margin, or NIM, is a bank-level measure. It is net interest income divided by interest-earning assets, where net interest income is interest income on assets minus interest expenses. That definition matters because NIM is not the same as the spread quoted on one syndicated facility. A bank’s NIM combines the results of many assets and liabilities, and it can reflect both the return the bank earns and the cost it pays for funding. It is a broad accounting measure of intermediation income, not a clean readout of market power alone.

The paper’s Call Reports analysis estimates that a one-standard-deviation increase in exchange-rate volatility is associated with an approximately 10-basis-point annualized increase in NIM for a bank with average exposure. The table’s annualized effects are about 5 basis points at one quarter and roughly 8 to 12 basis points at longer horizons, depending on the horizon and specification. A bank at the 95th percentile of the exposure distribution shows an annualized NIM response about 5 basis points larger than a bank at the 5th percentile. The comparison is important: the result is not simply that banks’ margins move during volatile periods. It is that the response is larger for banks with more loans relative to their interest-earning assets, the paper’s proxy for greater exposure to the syndicated-market pipeline.

“Annualized” is also doing real work in these figures. It describes the rate at which a quarterly change would be expressed over a year; it does not mean that every bank charges an extra 10 basis points on every loan for a full year. Nor is it a forecast for a future episode. The estimate applies to the historical sample of large U.S. commercial banks, under the paper’s exposure design and controls. It is best read as an average-exposure bank’s response to a specified increase in the volatility measure, not as a universal pricing rule.

The decomposition of NIM helps explain why the margin can rise as the balance sheet gets smaller. The paper finds that higher interest income on assets is the main driver. At the two-quarter horizon, the reported annualized increase in interest income on assets is about 15 basis points for an average-exposure bank. Interest expenses rise only modestly at first and, according to the discussion of the longer horizons, do not respond significantly after two quarters. The result is therefore not primarily a story in which banks pay dramatically more for deposits and pass that cost to borrowers.