THE JOURNAL OF FINANCE • VOL. LXIII, NO. 2 • APRIL 2008 Underreaction to Dividend Reductions and Omissions? YI LIU, SAMUEL H. SZEWCZYK, and ZAHER ZANTOUT ∗ ABSTRACT Using a sample of 2,337 cash dividend reduction or omission announcements over the 1927 to 1999 period, this study reports significant negative post-announcement long-term abnormal returns, which last 1 year only. However, this long-term abnormal performance is driven by the post-earnings-announcement drift. After controlling for the earnings performance and the skewness of buy-and-hold abnormal returns, there is no compelling evidence of a post-dividend-reduction or post-dividend-omission price drift. A number of studies present evidence of long-term return anomalies following major corporate events that call into question the informational efficiency of capital markets. In particular, these studies demonstrate that announcement- period abnormal stock returns for various events are not unbiased revisions in investors’ expectations. 1 Fama (1998) reviews this evidence and concludes that these seemingly puzzling findings are the result of chance, or incorrect statistical method, or a misspecified expected return model. 2 In particular, in his review of the paper by Michaely, Thaler, and Womack (1995), who report a stock price underreaction in samples of 561 dividend initiation and 887 dividend omission announcements made over the 1964 to 1988 period, Fama contends that changing the matching criteria might tell a different story, and hence he calls for an out-of-sample test before drawing any inferences about long-term returns following changes in dividend payments. ∗ Yi Liu is assistant professor of Finance at Capital University and at University of North Texas. Samuel H. Szewczyk is associate professor of Finance at Drexel University, and Zaher Zantout is associate professor of Finance at the American University of Sharjah and at Rider University. The authors greatly appreciate the helpful comments of the anonymous referee, Malcolm Baker, Robert Boldin, Richard Bower, De-Wai Chou, Arnie Cowan, Jacqueline Garner, Michael Gombola, Robert Lawson, Edward Nelling, Maria Sanchez, Robert Stambaugh (Editor), and Linghui Tang. The generous financial support of Drexel University, Rider University, and Capital University is also acknowledged. The usual disclaimer applies. 1 Major corporate events for which post-announcement long-term abnormal stock returns are found include: corporate mergers (Agrawal et al. (1992)), proxy contests (Ikenberry and Lakonishok (1993)), spinoffs (Cusatis, Miles, and Woolridge (1993)), initial public offerings (Loughran and Ritter (1995)), seasoned equity offerings (Spiess and Aff leck-Graves (1995)), new exchange listings (Dharan and Ikenberry (1995)), earnings (Bernard and Thomas (1990)), stock splits (Desai and Jain (1997)), share repurchases (Ikenberry, Lakonishok, and Vermaelen (1995)), and dividend initiations and omissions (Michaely, Thaler, and Womack (1995)). 2 He does mention, however, that a couple of anomalies seem to be above suspicion. 987