اعتماد بیش از حد مدیر عامل شرکت و سیاست تقسیم سود

We develop a model of the dynamic interaction between CEO overconfidence and dividend policy. The model shows that an overconfident CEO views external financing as costly and hence builds financial slack for future investment needs by lowering the current dividend payout. Consistent with the main prediction, we find that the level of dividend payout is about one-sixth lower in firms managed by CEOs who are more likely to be overconfident. We document that this reduction in dividends associated with CEO overconfidence is greater in firms with lower growth opportunities and lower cash flow. We also show that the magnitude of the positive market reaction to a dividend-increase announcement is higher for firms with greater uncertainty about CEO overconfidence.

The variation in dividend payouts over time and across firms remains one of the major unresolved puzzles in corporate finance, despite an extensive theoretical and empirical literature. In particular, the combined explanatory power of factors like agency problems, asymmetric information, and other market frictions, including taxes, is small compared to the total cross-sectional and time-series variation in dividend choices, leaving much to be explained (Allen and Michaely, 2003 and Brav et al., 2005). We examine an alternative explanation based on differences in managerial beliefs. Following Malmendier and Tate, 2008 and Malmendier and Tate, 2005 and Malmendier et al. (2011), we classify managers as “overconfident” if they overinvest personal funds in their own company.1 While the existing literature has explored the implications of CEO overconfidence for investment, merger, and financing decisions, the implications for dividends remain largely unexplored. The literature on dividend policy generally makes it clear that investment and financing decisions alone do not uniquely determine dividends. So, on the one hand, an overconfident CEO may lower dividends if he or she perceives higher investment needs (Ben-David et al., 2007); on the other hand, the CEO may increase dividends if he or she expects higher cash flows from current investment (Wu and Liu, 2011). Thus, the impact of an overconfident CEO’s beliefs on dividend policy is an open question, yet to be resolved conceptually and empirically. We study this issue by developing a dynamic model of the interaction between dividend policy and CEO overconfidence. We show that an overconfident CEO views future external financing as more costly than internal funds and lowers the current dividend payout to increase the amount of internal capital available for future investment needs. This result is driven by the model’s assumption that an overconfident CEO overestimates the value of new investments. The main testable prediction of the model is that an overconfident CEO pays lower dividends than does a rational CEO. The model also predicts the effect of CEO overconfidence on the dividend payout to be weaker for firms with higher growth opportunities. In addition, the model predicts the stock price response to announcements of dividend changes to be an increasing function of the informativeness of the announcement about CEO overconfidence. We test the model’s predictions using panel data of large US companies over the period, 1980–1994. We employ the measures of CEO overconfidence derived by Malmendier and Tate, 2005 and Malmendier and Tate, 2008. Our results indicate that the level of dividend payout is lower in firms managed by overconfident CEOs. The marginal reduction in dividend payout in firms managed by overconfident CEOs is about one-sixth of the median dividend payout for the firms in our sample.2 This result is robust to alternative measures of CEO overconfidence and to several control variables. We also examine the effect of CEO overconfidence on the relation between dividend policy and growth opportunities, cash flow, and the level of asymmetric information. Consistent with previous evidence, we find a negative relation between growth opportunities and dividend payout. However, the difference in the dividend payout between lower-growth and higher-growth firms is smaller in firms with overconfident CEOs. This finding is consistent with our model’s prediction that CEO overconfidence plays a weaker role in higher-growth firms. We also find that the positive relation between dividend payout and cash flow, documented in previous studies, is stronger in firms with overconfident CEOs. This result suggests that overconfident CEOs may overestimate the ability of current cash flow to predict future cash flow. We find an inverse relation between dividend payout and information asymmetry that does not vary across rational and overconfident CEOs. We conclude with an analysis of the stock market response to announcements of large dividend increases. We estimate a multivariate regression model to investigate the relation between CEO overconfidence and the stock-market response to the dividend-increase announcement. Our results indicate that the magnitude of the positive stock price response is higher for firms with uncertainty about CEO overconfidence than for firms whose CEOs have previously been identified as overconfident. This novel result is consistent with our hypothesis that dividends provide information about CEO overconfidence – dividend increases indicate lower CEO overconfidence – and that this informativeness is higher when there is greater uncertainty about CEO overconfidence. We make three contributions to the dividend policy literature: First, we model and test the relation between managerial overconfidence and dividend policy to show that CEO overconfidence affects dividend policy. Second, to rule out alternative plausible explanations of our results, we develop a series of other predictions that have not been investigated in the literature. Our tests of these predictions strengthen the overconfidence-based interpretation of our results as well as those in related prior work (e.g., Malmendier and Tate, 2005 and Malmendier and Tate, 2008). Specifically, we examine the effect of CEO overconfidence on the relation between dividend policy and cash flow, growth opportunities, and the level of asymmetric information. Third, the findings on the stock market response to announcements of dividend increases by overconfident CEOs indicate that the market recognizes the relation between CEO overconfidence and dividend policy. Our results thus provide a new explanation for the stock market response to announcements of dividend changes. Taken together, our results document a robust effect of CEO overconfidence on dividend policy. Our paper also contributes to the growing literature on behavioral corporate finance. Barberis and Thaler (2003) and Hirshleifer (2001) survey the literature that attempts to explain asset pricing and return patterns based on behavioral characteristics of investors. The literature on behavioral corporate finance that examines the relation between corporate policies and the behavioral characteristics of corporate managers and investors is surveyed by Baker et al. (2007). Hackbarth (2008) shows theoretically that overconfident managers tend to choose higher debt levels. Bernardo and Welch, 2001 and Gervais et al., 2011, and Goel and Thakor (2008) endogenize CEO overconfidence and consider the impact of CEO overconfidence on shareholders. Heaton (2002) examines how managerial optimism affects corporate policies, de Meza and Southey (1996) and Landier and Thesmar (2009) examine financial contracting with optimistic managers, and Bergman and Jenter (2007) link stock option compensation to employee optimism. Bertrand and Schoar (2003) show that differences in style across managers significantly explain corporate decisions and performance. Our study is more closely related to the literature that explores the effect of CEO overconfidence on corporate policies. Malmendier and Tate (2005) document that firms managed by overconfident CEOs exhibit a greater sensitivity of investment spending to internal cash flow. Malmendier and Tate (2008) show that overconfident CEOs are more likely to engage in acquisitions that are value-destroying. Malmendier et al. (2011) argue that overconfident managers perceive their firms to be undervalued and are reluctant to raise funds through costly external sources. They document that the reluctance of overconfident CEOs to raise funds through external sources leads to both a pecking order of financing and debt conservatism. Our results are consistent with the central thesis of this literature that behavioral characteristics of CEOs affect corporate finance policies. The papers that address the relation between managerial overconfidence or overoptimism and dividend policy include Ben-David et al. (2007) and Bouwman (2010). Ben-David et al. (2007) identify miscalibrated Chief Financial Officers based on their forecasts of stock market returns and analyze how this measure is related to corporate policies. Their evidence indicates that firms with miscalibrated CFOs may pay smaller dividends. We focus on CEO overconfidence and get much stronger results with more comprehensive empirical tests. Bouwman (2010) analyzes the stock price reaction to dividend increases by optimistic CEOs. While she does not investigate the impact of CEO overconfidence on dividend policy, she finds that the stock market reacts more positively to dividend increases by optimistic CEOs. Our theoretical analysis provides a more nuanced result that the stock market reaction depends on the extent of uncertainty about CEO overconfidence, and our empirical analysis, which differs from Bouwman (2010), confirms this prediction. Ben-David (2010) surveys the dividend policy literature from a behavioral finance perspective. The paper proceeds as follows. In Section 2, we develop a model of dividend policy and CEO overconfidence. Section 3 describes the data and method. Section 4 presents the empirical results. Section 5 summarizes our findings and discusses the implications of the study.

We model dividend policy as a trade-off between reducing the cost of retaining excess cash and reducing the endogenous cost of external financing for future investment when the CEO acts in the interest of existing shareholders. Overconfident CEOs, who believe that the firm is undervalued and perceive external financing to be more costly relative to rational CEOs, pay lower dividends in order to accumulate greater financial slack for future investment needs. The model yields several testable predictions that we examine empirically. The main testable prediction is that an overconfident CEO pays a lower level of dividends relative to a rational CEO. Our model also predicts the difference in the dividend payout between higher- and lower-growth firms to be smaller in firms managed by overconfident CEOs. Another prediction is that the stock price response to announcements of dividend changes is an increasing function of the uncertainty about CEO overconfidence. We test these predictions and perform related empirical tests, using a panel data of large US companies. We use the measures of CEO overconfidence used in Malmendier and Tate, 2005 and Malmendier and Tate, 2008 and in Malmendier et al. (2011). Consistent with our main prediction, we find that the level of dividend payout is lower in firms managed by overconfident CEOs. The reduction in dividend payout associated with CEO overconfidence is both statistically and economically significant. Next, we document that the difference in the dividend payout between higher-growth and lower-growth firms is smaller for firms with overconfident CEOs. This finding is consistent with the prediction of our model that the reduction in dividend payout caused by CEO overconfidence is smaller in higher-growth firms. We further document that the positive relation between dividend payout and cash flow is stronger in firms with overconfident CEOs. Finally, we analyze market perceptions about the relation between CEO overconfidence and dividend policy by examining the stock price response to announcements of dividend increases by our sample firms. We find that the magnitude of the positive stock price response to announcements of dividend increases is higher in firms in which there is greater uncertainty about the level of CEO’s overconfidence. This finding is consistent with our hypothesis that dividends provide information about CEO overconfidence. Specifically, dividend increases indicate lower CEO overconfidence and that this inference is stronger when there is greater uncertainty about CEO overconfidence. Our empirical evidence collectively suggests that CEO overconfidence has a significant effect on dividend policy.