We develop a model of corporate dividend policy based on the idea that management values operating flexibility. By reducing dividends and conserving cash, management increases its flexibility. This improves its ability to invest in projects that it believes are good for the shareholders in the long run but which shareholders would not provide the capital for because they think the projects are value reducing. However, the cost of not paying dividends is a reduction in the current stock price. Management trades off these two aspects of dividends. Flexibility considerations help us understand various dimensions of dividend policy that existing theories do not explain. Our theory generates numerous testable predictions that we confront with the data. The evidence is supportive of the model.

Recent research finds that investors, broadly defined, react to the linguistic tone of quarterly earnings conference calls; there is a positive relation between firms' stock returns and call tone (a measure of “sentiment” related word tabulations). However, this type of soft information can be subtle, context-specific, and difficult to interpret. Moreover, the literature suggests cross-sectional variation in information processing skills among investors. Thus, we test whether sophisticated investors interpret earnings conference call tone differently than investors at large by examining short selling activity and its relation to earnings conference call tone. We find that short sellers target firms with simultaneous high earnings surprise and abnormally high management tone. The combination of positive earnings surprise and unusually positive tone strengthens short sellers' return predictability. This result indicates that short sellers interpret revealed “inflated” call language by managers more completely than naïve investors. The incomplete stock price reaction by naïve investors due to the lack of reliability that they place on this soft information results in overpricing of the stock. However, it also suggests that managers are unable to maintain prolonged overvaluation of their stock by striking an overly optimistic posture in the interactive conference call disclosure forum since short sellers' trades provide additional price discovery.

We examine short selling around dividend announcements and ex-dividend dates. Contrary to our initial expectation, we do not find abnormally high short-selling activity prior to announced dividend decreases, which runs counter to the argument that short sellers have the ability to acquire private information before its public dissemination. However, we find that the common negative relation between current short selling and future daily returns prior to unfavorable dividend announcements is similar to the negative relation during non-event times, suggesting that dividend announcements do not provide unusual trading opportunities for informed traders (Gonedes, 1978, and Benartzi et al., 1997). Around ex-dividend dates, we do find abnormal short selling, which may be explained by the return pattern around ex-dividend days documented by Lakonishok and Vermaelen (1986), who suggest that demand for a particular stock by dividend capture traders drives stock prices above their fundamental value thus providing a profitable trading opportunity for short sellers. Consistent with this conjecture, we find that both the level of short selling and the return predictability of short selling is markedly higher on and after the ex-dividend day than during non-event times.