Time-of-month anomaly: reality or mirage?

Abstract

This article asks two questions concerning the robustness of Kohers and Patel's (1999) time-of-month pattern. First, does it appear in other countries, or is it unique to US markets? Second, is it independent of the well-known turn-of-the-month anomaly, or merely a reflection of it? It finds that the time-of-month pattern is largely a mirage in the post-1980 period. For the USA, it disappears when turn-of-month days are removed from the sample. For three out of nine other countries in the sample, some elements of the time-of-month pattern remain once turn-of-month days are removed. However, two other countries exhibit significant effects in the opposite direction. Overall, these patterns do not add up to a consistent and robust international anomaly.