Resumen
AbstractThe efficient market hypothesis submits that the expected returns on shares and other financial assets are identical for all the days of the week. Studies of share returns on the New York Stock Exchange have revealed that the expected returns are not identical for the various days of the week. This article examines two hypotheses that have attempted to explain the distribution of returns over different days of the week. The calendar-time hypothesis states that the expected return for Monday is three times the expected return for the other days of the week. The trading-time hypothesis states that the expected return is the same for each day of the week. During the period 1978-1983, the daily returns on shares traded on the JSE were inconsistent with both hypotheses. The average return for Monday was significantly negative while the average return for the other trading days was positive with Wednesday showing the highest return. Evidence is presented to show that Treasury Bills have the same weekend effect as share transactions. An investment strategy based on the observed pattern of share returns over different days of the week is suggested. The implications of the effect of day of the week for tests of market efficiency are examined.