The robustness of calendar anomalies in daily stock returns

Abstract

This paper investigates the robustness of the major calendar anomalies in stock returns with respect to the choice of return measure, estimation procedure, and time period. For daily returns from 1972 to 1994, the size and statistical significance of the anomalies differ more across return measure than across estimation procedure. For the returns on small-firm stocks, there is robust evidence of weekend effects, pre-holiday effects, and January effects. For the returns on large-firm stocks, calendar anomalies are weaker and essentially disappear after 1986.