Day-of-the-week effect: a meta-analysis

Abstract

This study conducts a meta-analysis on the day-of-the-week effect to shed more light on the replication crisis of this stock market anomaly. The findings confirm that Mondays and Tuesdays provide, on average, lower daily returns. In addition, Wednesdays and Fridays indicate higher returns, with an unexpectedly strong middle-of-the-week effect on Wednesdays. The study highlights the influence of study design on these findings and notes a more substantial effect in the 1980s and 1990s. While differences in empirical methods do not impact the anomaly, index choices affect findings on day-dependent returns. The real estate sector especially stands out with a stronger day-of-the-week effect. However, geographic differences are mostly insignificant except for Oceania. Cultural differences demonstrate a weak but significant effect on abnormal daily returns. From a meta-perspective, outliers remain an essential driver for this stock market anomaly, indicating that study design is not the only factor driving the replication crisis.