Day of the Week Effects and Asset Returns

The number of empirical studies using daily stock returns is rapidly increasing.' Researchers generally assume that the distribution of stock returns is identical for all days of the week-a convenient statistical assumption but not a necessary condition of market equilibrium. Nevertheless, there are reasons to suspect that the distribution of returns may vary according to the day of the week; the most obvious is the impact of weekends on Monday's return. Since Monday's return is calculated over three instead of one calender day, the mean and variance may well be higher on Monday compared with any other daily return (perhaps three times as large). Previous research has examined the daily distribution of stock returns for a Monday effect. While Fama (1965) does not compare daily mean returns, he does report that Monday's variance is about 20o greater than other daily returns. With A traditional distributional assumption regarding the returns on a financial asset specifies that the expected returns are identical for all days of the week. Contrary to this plausible assumption, this paper discovers that the expected returns on common stocks and treasury bills are not constant across days of the week. The most notable evidence is for Monday's returns where the mean is unusually low or even negative. Several explanations of the results are investigated, but none proves satisfactory. Aside from documenting significant day of the week effects, the implications of the results for tests of market efficiency are examined. While market-adjusted returns continue to exhibit day of the week effects, these effects are no longer concentrated on Monday. * During the last year we have had many conversations with our colleagues throughout the professions. We are particularly thankful to the participants at sessions of the Western and American Finance Associations where preliminary versions of this paper were presented. Special thanks are due to Gene Fama, Gailen Hite, Jon Ingersoll, Robert Jarrow, Merton Miller, Clifford Smith and Rex Thompson. The expert research assistance of Frank Rabinovitch and YoungSup Yun is gratefully acknowledged. 1. The list of papers using daily stock returns is too long to name. Recent examples include Charest (1978), Dodd (1980), Banz (1981), and Reinganum (1981).