Abstract

This paper shows that the U.S. equity market reverts the liquidity-driven trading induced
by the payment cycle within a month. The aggregate reversal is robust to transaction costs
and out-of-sample tests as it concentrates on liquid and high-priced stocks and during
expansion periods. The findings lead to a novel interpretation of reversal: the pattern
measures the liquidity not efficiently provided in the market rather than investors’ cognitive
bias or compensation for market-making