The Influence of Trends and Reversals in Earnings Surprises on Stock Return Predictability∗

We find that trends and reversals in past earnings surprises predict future stock returns. These earnings surprise sequences produce a 9.88% risk-adjusted annual return which cannot be attributed to momentum nor post-earnings announcement drift. Our empirical methodology captures the market’s beliefs regarding future returns by estimating the probability associated with different possible return realizations. This probability is able to detect whether the market conditions on past earnings surprises when forming its expected return, and measures the return predictability induced by this conditioning. JEL Classifications: G11, G12, G14 ∗We thank Anthony Bernardo, Michael Brennan, Kalok Chan, Karl Diether, Bruce Grundy, Jeffrey Hales, Bing Han, David Hirshleifer, Dong Hong, Harrison Hong, Andrew Karolyi, Terrance Odean, Scott Richardson, René Stulz, Avanidhar Subrahmanyam, Rossen Valkanov, Scott Weisbenner, and Ingrid Werner for their helpful comments and suggestions, as well as seminar participants at Ohio State University and Singapore Management University. Mitch Warachka is grateful for financial support from the SMU-Wharton Research Center, Singapore Management University. †Ohio State University and Singapore Management University; Address: 606 Fisher Hall, 2100 Neil Ave, Columbus OH, 43210, USA; Email: loh [email protected]. ‡Singapore Management University; Address: #4055 L.K.C. School of Business, 50 Stamford Road, 178899, Singapore; Email: [email protected], Phone: (65) 6828 0249, Fax: (65) 6828 0427. (Corresponding Author).