سود سهام و شتاب قیمت

Stock market evidence shows that momentum profits are lower among dividend-paying firms than their non-paying counterparts due to differences in losers’ returns. Additionally, dividend maintenance is associated with higher returns for losers but not for winners. Finally, buying winners that increased their dividends and shorting losers that decreased their dividends enhances momentum profits. Consistent with the evidence, the behavioral models suggest that investors underreact to the losers’ positive dividend-maintaining news, reducing their return momentum and shrinking the payers’ momentum profit. Also, underreaction to positive news from winners’ dividend-increasing announcements as well as to negative news from losers’ dividend-decreasing announcements explains the higher momentum profits for strategies based on these stocks. The results do not appear consistent with risk-based explanations.

Several studies, starting with Jegadeesh and Titman (1993), document stock return momentum over the medium term. Although it is clear that winners tend to win and losers tend to lose over this horizon, there is substantial debate over the source of the profits and the interpretation of the evidence. New evidence from the markets show that stocks that do not pay dividends generate higher momentum profits than those that do, suggesting managerial dividend policy influences momentum profits. Furthermore, increases in dividends exacerbate the return momentum of winners while dividend cuts increase losers’ momentum, enhancing the profits for portfolios conditioned on past performance and dividend change. These results support the theories, and add to the evidence, that psychological bias underpins momentum profits. The study of the relation between dividend payment and momentum profit is motivated by results in the literature that suggest asymmetry in the dividend information conveyed by winners and losers. Chan et al., 1996 and Chordia and Shivakumar, 2006 report that losers tend to experience declines in earnings while winners tend to display increases in earnings, suggesting dividend maintenance by winners and losers conveys different information. Specifically, dividend maintenance in the face of declining earnings signals that management does not expect the deterioration in earnings to be long-lived, which is positive information for the dividend-maintaining losers. On the other hand, dividend maintenance by winners does not indicate that the increases in their earnings are persistent and, hence, it does not convey good news.1 This points to asymmetry in the dividend news conveyed by winners and losers. Evidence presented in Fuller and Goldstein (2006) supports these contentions. In particular, they find that dividend payers outperform non-payers in down markets (when firms are generally losing), but not in up markets (when firms are generally winning). Consequently, they conclude that dividends are more valuable when firms are losing than when they are winning. This paper examines whether the asymmetry in winners’ and losers’ dividend-maintenance news and dividend change news affect momentum profits. Evidence from the stock markets shows that momentum profits are on average 0.42% per month lower among dividend payers than non-payers. Furthermore, the difference in profits is driven by the higher returns for the dividend-paying losers relative to their non-paying counterparts. Finally, a momentum strategy that buys winners which increased their dividends and shorts losers which reduced their dividend yields an average of 3.08% monthly profit. This is in excess of the approximately 1% profit from the standard momentum strategy, and is unlikely to be swamped by transaction costs since the strategy involves a smaller number of stocks and only dividend payers, which are generally large stocks. The behavioral models for momentum profits (e.g., Daniel et al., 1998, Barberis et al., 1998 and Hong and Stein, 1999) indicate that the differential valuation of dividend-maintaining announcements by winners and losers would lead to different momentum profits for the dividend payers and non-payers. This reasoning results directly from the models’ prediction that stock prices drift in the same direction as the announcement day return.2 That is, if a dividend announcement increases a stock’s price, the post-announcement price will drift upward due to underreaction. The Daniel et al. (1998) model suggests that confirming public information leads to continuing overreaction to private information due to overconfidence and biased self-attribution. They show that in the presence of these psychological biases, a positive non-selective event (Proposition 4) or selective event (Proposition 6) leads to a favorable event-date price change as well as a positive future average price trend.3 Furthermore, the post-event average return will be higher the more the event and the pre-event stock price run-ups are in the opposite direction. In the Barberis et al. (1998) model, conservatism bias leads to slow updates of models in the face of new information, resulting in underreaction. In addition, representativeness bias results in subsequent overreaction to the information. Hong and Stein’s (1999) model assumes that private information diffuses slowly among the “news watchers” over time, and this results in underreaction to private news. The resulting positive serial correlation in returns attracts the attention of momentum traders, whose trading activity results in an eventual overreaction to the news. They argue that their model is also consistent with underreaction to public news since investors invariably use some private information to convert public news (such as dividend announcements) into a judgment about value. Thus, underreaction/overreaction from these behavioral models predicts that a positive (negative) price reaction on an event-date should lead to a positive (negative) future price trend. The behavioral explanations for momentum profits predict that underreaction/overreaction to the good news from dividend maintenance by losers will result in higher post-announcement returns for these stocks relative to those of their non-payer counterparts. In contrast, dividend maintenance by winners does not convey good news, at least not to the extent of losers, and hence their post-announcement price trends should not be significantly higher than those of their non-payer counterparts. That is, if investors underreact to good news, this underreaction should be more pronounced for dividend-maintenance announcements by losers than by winners. Consistent with these predictions, the average dividend-maintaining announcement and post-announcement returns for losers are higher than the corresponding averages for their non-payer counterparts. In contrast, the average dividend-maintaining announcement and post-announcement returns for winners are not significantly higher than the corresponding averages for their non-payer counterparts. Thus, the lower momentum profit among payers is consistent with the behavioral underreaction/overreaction explanation for momentum profits.4 Finally, if dividend maintenance conveys information, changes in dividend payment should provide even stronger signals that can enhance momentum profits. Several studies report that dividend increases convey good news while dividend cuts convey bad news (e.g., Bhattacharya, 1979 and DeAngelo et al., 1992). If investors underreact to dividend news, underreaction to news from changes in dividends can be exploited to enhance momentum profits. Specifically, underreaction to dividend-increasing announcements by winners will increase their return momentum relative to the other winners. Similarly, underreaction to dividend-decreasing announcements by losers will increase their return momentum. Consistent with these predictions, the event-date returns and post-event returns for the winners that increased their dividends are higher than those of the other winners. Similarly, the event-date returns and post-event returns for the losers that decreased their dividends are lower than those of the other losers. This explains the high profits for portfolios that buys winners that increased their dividends and shorts losers that decreased their dividends. There is a separate line of research that suggests that momentum profits represent compensation for risk or trading costs of the strategy. These models indicate that the profits can be explained by β (e.g., Conrad and Kaul, 1998), book-to-market values (e.g., Daniel and Titman, 1999), size (e.g., Lesmond et al., 2004), growth options (Sagi and Seasholes, 2007), industrial effects (e.g., Moskowitz and Grinblatt, 1999), trading costs (e.g., Lesmond et al., 2004 and Korajczyk and Sadka, 2004), and time variation in risk ( Chordia and Shivakumar, 2002, Antoniou et al., 2007 and Li et al., 2008). Overall, it does not appear that momentum profits are solely due to higher risks (e.g., Fama and French, 1996) or to time variation in risk (e.g., Grundy and Martin, 2001). This study reports a link between momentum profit and dividend information, and the relation is robust to size, β, and book-to-market risk factors. Additionally, the dividend effect is unlikely to be confounded by other risk factors. Specifically, if dividend-maintaining news is associated with a particular risk factor, it should have similar effects on winners and losers. Li et al. (2008) suggest that winners incorporate news faster than losers, which could result in different dividend-maintaining announcement effects. However, we see dividend-maintaining announcement effects for losers, which are slower at incorporating news, but not for winners. Thus, the asymmetry in dividend-maintaining announcements by losers and winners is not consistent with different speeds of incorporating the news. The analysis makes at least three contributions to the ongoing debate on the interpretation of the evidence on price momentum: First, it links a policy variable (dividend payment) to momentum profits. In particular, investors underreact to dividend-maintaining announcements by losers and this mitigates the dividend-payers’ momentum profits. Second, an investment strategy that takes long positions in winners that increased their dividends and short positions in losers that decreased their dividends generates profits that are substantially higher than those from the traditional momentum strategy. Finally, the link between dividend policy and momentum profit is more consistent with the behavioral explanations for momentum profits than explanations based on risk. The rest of the paper is organized as follows: Section 2 presents data and evidence on the relation between dividends and momentum profits, Section 3 considers robustness of the results and Section 4 concludes.

Evidence from the stock markets shows that momentum profits are higher for non-dividend-paying firms than payers. Also, the differential in profits is mainly due to the higher returns for dividend-paying losers compared to their non-paying counterparts. Finally, a strategy that buys winners that increased their dividends and shorts losers that decreased their dividends generates high momentum profits. Both theory and evidence suggest that winners’ dividend-maintaining announcements do not indicate that increases in their earnings are persistent and, hence, the news does not provide any information advantage over their non-payer counterparts. By contrast, the losers’ dividend-maintaining announcements suggest that their managers do not expect deteriorations in their earnings to be persistent, providing positive news that their non-payer counterparts do not provide. As a result, dividend maintenance leads to higher announcement returns for losers relative to their non-payer counterparts while the same announcements do not generate higher returns for winners compared to their non-payer counterparts. The behavioral models indicate that underreaction to the losers’ positive dividend-maintaining news should lead to higher post-announcement price trends for them compared to their non-payer counterparts while the dividend-paying and non-paying winners’ price trends should be similar. Consistent with these predictions, the evidence shows that the post-announcement returns for the dividend-maintaining losers are higher than those for their non-payer counterparts while the post-announcement price trends are similar for the dividend-maintaining and non-paying winners. The higher announcement-day and post-announcement returns for losers relative to their non-payer counterparts and the similar returns for winners irrespective of their dividend-payment status predicts lower momentum profits for payers compared to non-payers, exactly as the evidence indicates. Finally, it is well documented that dividend increases convey good news while dividend cuts convey bad news. If investors underreact to information, the news from dividend increases and cuts can be exploited to enhance momentum profits. In particular, underreaction to the winners’ dividend-increasing news and to the losers’ dividend-decreasing news suggests stronger momentum for these stocks. This predicts that a portfolio that is long in past winners that increased their dividends and short in past losers that decreased their dividends should generate high momentum profits, consistent with the evidence. Furthermore, the profits from this strategy are likely to dominate transaction costs since the strategy involves a small number of generally large firms, only dividend-changing firms. These findings are robust to size, β, and book-to-market values.