37 Pages Posted: 20 Mar 2005
Date Written: July 26, 2004
This paper introduces seasonality into a model of expected stock returns. We confirm previous findings that there is no evidence for cross-sectional variation in expected stock returns when we restrict the means to be constant throughout the year. Yet, we show there is substantial variation when considering each month of the year separately. Applying a seasonal structure we estimate an annualized standard deviation of 13.8%. There is strong evidence stocks have distinct expected returns in January, February, ... December. The estimated seasonal variation in expected returns is positive in every calendar month and especially high during October, December, and January. This structure is independent of industry, size, and earnings announcements. These results support the inclusion of seasonal structure into asset-pricing models.
Suggested Citation: Suggested Citation
Heston, Steven L. and Sadka, Ronnie, Seasonality in the Cross-Section of Expected Stock Returns (July 26, 2004). AFA 2006 Boston Meetings Paper. Available at SSRN: https://ssrn.com/abstract=687022 or http://dx.doi.org/10.2139/ssrn.687022