National bank window dressing and the call loan market, 1865–1872

•We test for window dressing from 1865 to 1872 in the New York money market.•A change in the reporting law created a natural experiment.•The natural experiment separates window dressing from preferred habitat behavior.•We observe no change in the pattern of interest rates across the change in the law.•The pattern of interest rates provides no evidence for window dressing.

After the American Civil War, market observers attributed increases in interest rates around quarterly reporting dates to window dressing by national banks. Window dressing is a temporary change in portfolio designed to produce a more appealing report to regulators or to the public. This paper tests for increases in interest rates at quarter end under a natural experiment, a change in the reporting law. Using daily data on the call loan interest rate in New York City, we find no evidence of systematic increases in the call loan rate just before the quarterly reporting dates of national banks.