Review of Quantitative Finance and Accounting, 22: 5–14, 2004
2004 Kluwer Academic Publishers. Manufactured in The Netherlands.
The Reversing Weekend Effect: Evidence from the U.S.
ANTHONY YANXIANG GU
Jones School of Business, State University of New York, 115D South Hall, 1 College Circle, Geneseo, NY 14454,
USA Tel.: (585) 245-5368, Fax: (585) 245-5467
Abstract. The well-known weekend effect has been reversing in Major U.S. indices from late 1980s to late
1990s. The correlation between Monday and Friday returns also exhibited a declining trend, and ﬂuctuated around
zero in the 1990s. A power ratio method is developed to measure consistently the relative contribution of Friday
and Monday returns to the return of the week in each individual year. The revealed dynamics of the anomaly
explains why previous researchers report different or conﬂicting ﬁndings. The anomaly may not be necessarily
related to ﬁrm size.
Keywords: weekend effect, anomaly, reverse
JEL Classiﬁcation: G1
Numerous studies have reported abnormally high average Friday returns and signiﬁcantly
negative average Monday returns in the U.S. stock markets. Pioneer research on the so
called “weekend effect” can be found in Cross (1973), French (1980), Gibbons and Hess
(1981), Hindmarch (1984), Keim and Stambaugh (1984), and Jaffe and Westerﬁeld (1985).
Several researchers explored the possible factors that contribute to the anomaly.
Hindmarch (1984) suggests that institutional trades can partially explain the effect, and
Sias and Starks (1995) believe that institutional traders are the primary drivers of the ef-
fect. Lakonishok and Maberly (1990) and Abraham and Ikenberry (1994) report that share
price does worse on Mondays than on other days of the week, because individual investors
typically sell stocks on Monday. Branch (1974, 2001) suggests that the Monday effect may
be related to weekly cycle in news releases and to weekly pattern in interest rate changes.
Similarly, Steeley (2001) reports that a systematic pattern of market-wide news arrivals
drives the anomaly in the UK stock market. Branch and Echevarria (1991) indicate that
the effect occurs mainly in stocks that do not go ex-dividend on Monday. Schatzberg and
Datta (1992) suggest that some factor unrelated to information arrivals causes the weekend
effect. Coutts and Hayes (1999) indicate that the effect is in part a stock exchange account
settlement effect in major UK indices.
Keim and Stambaugh (1984) report a strong negative relation between Friday returns and
ﬁrm size. Abraham and Ikenberry (1994) notice stronger weekend stock return differences
in small and medium-sized companies. Cross (1973) ﬁnds positive correlation between