Long‐run Performance after Stock Splits: 1927 to 1996

Abstract

We measure the postsplit performance of 12,747 stock splits from 1927 to 1996 using two methods to measure abnormal returns: size and book-to-market reference portfolios with bootstrapping, and calendar-time abnormal returns combined with factor models. Between 1927 and 1996, neither method applied to splits 25 percent or larger ¢nds performance signi¢cantly di¡erent from zero. Over selected subperiods, subsamples of 2 -1 splits restricted by book-tomarket availability requirements display positive abnormal returns using some methods. However, these samples show small or negligible abnormal returns using the calendar-time method. Overall, the stock split evidence against market e⁄ciency is neither pervasive nor compelling. THE EFFICIENT-MARKET CONCEPTION of near-perfect capital markets that render only £eeting and nonsystematic gain and loss opportunities to investors has been criticized in recent years by the behavioral ¢nance literature, which o¡ers evidence that stock transactions are often executed (in relation to known events such as stock issues, stock splits, and repurchases) at price levels that imply predictably high or low risk-adjusted returns. If these ¢ndings are factually correct, they pose a challenge to the e⁄cient market hypothesis, which predicts a lack of capital market pro¢t and loss opportunities due to the abilities of investors rapidly and unbiasedly to interpret information according to correct assessments of the underlying economic processes. The behavioral literature attributes its ¢ndings to various investor biases. Supporters of e⁄cient markets de£ate the importance of the behavioral ¢nance ¢ndings by noting that risk-adjustment methods are imperfect, by suggesting that data mining may have occurred, by noting that all the behavioral anomalies taken together suggest an unbiased market at work, and by asking for behavioral models that explain a broader range of evidence (see Fama (1998) for a cogent summary). Haugen (1999Haugen ( , 2002 ably responds from the behavioral camp by pointing out the superior power of capital market phenomena like momentum to predict and explain returns.We believe that certain issues of fact are not yet settled and deserve further examination, one of these being the stock split. Since splits are widely reported and noted, a stock split anomaly would be a particularly £agrant violation of

An Empirical Study of Stock Split Announcements of Select BSE Sectors Using Event Study Methodology

“…They explained this announcement effect was a consequence of an increased payout of the splitting companies. Byun J and Rozeff M (2003) examined the post-split performance of 12,747 stock splits between 1927 and 1996 with statistical tools like time regression analysis and book-to-market reference portfolios. The study finds small or negligible abnormal returns and they conclude that the long-term stock split evidence to market efficiency was neither pervasive nor compelling.…”

Stock Market Reactions to the Announcements and Executions of Stock-Splits and Reverse Stock-Split

“…357-375], Ikenberry and Ramnath [Ikenberry, Ramnath, 2002, pp. 489-526], and Byun and Rozeff [Byun, Rozeff, 2003, pp. 1063-1085.…”

Market Response to Stock Splits in the National Stock Exchange

“…...1 Where, R j,t = Returns on security j on time t P t = Price of the security at time t P t−1 = Price of the security at time t-1 Returns for the event window are computed using the following market model equation R j,t = α j + β j R m + e t ...2 Where, R j,t = The daily return on security j on time t R m = The daily return on the Indian stock market at day t α j and β j = OLS intercept and slope coefficient estimators respectively e t = The error term for security j at day t The expected return for the security j at day t is defined as ER j,t = α j + β j R m ... 7 Where AAR e is AAR of the estimation period for the entire sample and S(AARe) represents the standard deviation of the AAR of the estimation period.…”