From dusk till dawn (and vice versa): Overnight-versus-daytime reversals and feedback trading

Although overnight-versus-daytime return reversals have often been ascribed to the heterogeneous clienteles of the overnight and daytime sessions, there exists no evidence to date on how those clienteles' trading behaviour motivates these reversals. We empirically investigate this issue for the first time by assessing whether these reversals are the result of feedback trading during overnight/daytime hours. Drawing on the S&P 500 ETF for the 1993-2021 period, we find that overnight (daytime) feedback trading largely motivates the expected positive (negative) overnight (daytime) returns; in line with this, days entailing the expected negative overnight-versus -daytime return reversals accommodate stronger feedback trading at the daily (i.e., close-to-close) frequency. Daytime feedback trading is present when the immediately preceding overnight session's returns are positive, while overnight feedback trading reveals a strong Monday-effect. We also show that overnight-versus-daytime variations of feedback trading hold across other large US ETFs.

Automated summary

This study empirically investigates the trading behavior of heterogeneous clienteles during overnight and daytime sessions, and finds that feedback trading largely motivates the expected return reversals. The results show that overnight and daytime feedback trading are influenced by different factors and exhibit variations across different ETFs.