Abstract
In this paper we examine the hypothesis that the Monday effect disappeared as a result of arbitrage activity by institutional investor trading empowered by a reduction in transaction costs. Our empirical results argue against this conclusion because of the following: 1) We find that changes in Monday returns are too large to be consistent with arbitrage activity; 2) the shift in the Monday effect occurs primarily in the bull market of the 1990s, much later than the timing of shifts in transaction costs identified in previous research; and 3) when returns are segmented based on magnitude, we find that the patterns in the shifts of weekday returns are inconsistent with profitable arbitrage activity.