پیامدهای بازار سرمایه از سبک های افشای داوطلبانه مدیران

This paper studies the capital market consequences of managers establishing an individual forecasting style. Using a manager-firm matched panel dataset, I examine whether and when manager-specific credibility matters. If managers' forecasting styles affect their perceived credibility, then the stock price reaction to forecast news should increase with managers' prior forecasting accuracy. Consistent with this prediction, I find that the stock price reaction to management forecast news is stronger when information uncertainty is high and when the manager has a history of issuing more accurate forecasts, indicating that individual managers benefit from establishing a personal disclosure reputation.

Research examining the credibility of management forecasts has shown that investors' and analysts' responses to management forecasts vary with firms' overall prior forecasting accuracy ( Williams, 1996 and Hutton and Stocken, 2009). These studies do not distinguish between manager- and firm-specific forecasting behavior because under both neoclassical economic and agency theories, managers' individual preferences should not have an effect on corporate decisions. In contrast, several recent studies in the accounting literature find that managers' individual preferences have an effect on firms' voluntary disclosure and financial reporting outcomes ( Bamber et al., 2010, DeJong and Ling, 2010, Dyreng et al., 2010 and Ge et al., 2010). This study extends this line of research and investigates whether and when the stock price reaction to management forecasts news varies with an individual manager's forecasting behavior. 1 If managers' forecasting styles affect their perceived credibility, then the stock price reaction to forecast news should increase with managers' prior forecasting accuracy and this effect is likely to be stronger when managers' individual differences are accentuated. Consistent with investors using managers' prior forecasting behavior to assess the credibility of their current forecasts and managers benefiting from establishing a personal disclosure reputation, I find that the market response to both good and bad news forecasts is stronger for managers with greater prior accuracy when there's higher uncertainty in the information environment.2 Ex ante, it is not clear whether investors should exert effort to understand differences between manager- and firm-specific forecasting behavior if they are only concerned about the firm's overall forecasting accuracy. However, prior research finds that investors and analysts tend to expect better performance from reputable CEOs ( Malmendier and Tate, 2009), indicating that manager-specific attributes do affect the beliefs of market participants. It thus follows that the market response to management forecasts could vary with manager credibility—one important attribute of a manager's overall reputation. Because past forecasting performance is a signal of the manager's forecasting skill and credibility ( Mercer, 2005), market participants should assign greater (less) weight to forecasts issued by managers with higher (lower) prior forecasting accuracy. Moreover, research in psychology and management finds that idiosyncratic personal differences are more likely to affect decision-making processes when individuals are faced with complex situations involving high uncertainty ( Hambrick and Mason, 1984, Caspi and Moffitt, 1993 and Hambrick, 2007). This suggests that there are conditions under which manager-specific effects will play a more important role in determining forecast properties. To address my research question, I follow the methodology introduced by Bertrand and Schoar (2003), which tracks managers across firms over time to identify a set of CEOs and CFOs who are employed by at least two firms during my sample period and also issued management forecasts during their tenure at each firm. This sample selection restriction makes it possible to separate the effects of managers from the stationary firm characteristics (firm fixed effects), time-specific cross-sectional effects (year fixed effects), and time-varying firm characteristics (control variables). Moreover, the advantage of the fixed-effect approach is that it generates parameter estimates of manager- and firm-specific forecast accuracy. Following prior research (Jennings, 1987, Williams, 1996, Rogers and Stocken, 2005 and Hutton and Stocken, 2009), I measure forecast credibility as the stock price reaction to management forecast news. I first examine whether the market reaction is stronger for forecasts issued by managers with higher prior forecasting accuracy, and find that the market response to forecast news is positively associated with manager-specific forecasting accuracy. However, further tests show that this effect is subsumed when I control for firm-specific forecasting records. This result is consistent with Williams (1996) and Hutton and Stocken (2009), who find that security analysts and investors are more responsive to forecast news when firms develop a reputation for issuing more accurate forecasts in the past. While the results discussed above suggest that investors do not distinguish between manager- and firm-specific forecasting styles unconditionally, research in management and psychology suggests that individual differences play a larger role in behavior when uncertainty is high. This suggests that investors should apply Bayesian updating and assign more weight to the highly skilled managers when they are uncertain about the overall precision of the forecast. In this analysis, I use principal components analysis to form two factors that capture different dimensions of uncertainty: information uncertainty and earnings uncertainty. The results show that when information uncertainty is high, the market response to both good and bad news forecasts is stronger for managers with the highest prior forecasting accuracy. I also find that the price reaction to bad news forecasts is stronger for firms with the highest prior forecasting accuracy when information uncertainty is high. However, while the stock price reaction varies with a firm's forecasting record when earnings uncertainty is high, it does not vary with the manager's forecasting record. Taken together, these results suggest that manager-specific credibility matters when individual-specific effects are likely to be accentuated. Although I attempt to distinguish between manager- and firm-specific effects by following a methodology that has been employed in several recent studies, CEOs and CFOs are not assigned to firms randomly. A manager may be perceived as highly accurate because he/she happens to be at a firm with a record of issuing accurate forecasts. If this leads him/her to be hired by another firm that also wishes to initiate such a forecast policy even without hiring a new manager externally, then I would overestimate a high forecast accuracy fixed effect for that manager. Although I explicitly control for firm-specific forecast accuracy in the market reaction tests, the manager fixed effect estimates are still likely to be measured with error, and the results should be interpreted with this caveat in mind. Subject to the limitation discussed above, this paper makes the following contributions: First, this is the first study to document the economic consequences of managers having a style of his/her own. While several studies in the economics, management, finance, and accounting literatures document the existence of individual manager styles, none of these studies examines whether the capital market responds to differences among individual managers' unique styles. My paper extends several recent accounting studies that employ a similar methodology to investigate whether managers have unique styles of their own that are reflected in their earnings forecasts, financial reporting, and tax avoidance choices of the firm for which they are employed at (Bamber et al., 2010, DeJong and Ling, 2010, Dyreng et al., 2010 and Ge et al., 2010).3 I show that the persistence of the manager- and firm-specific forecast accuracy effects differ, which has implications for how investors incorporate managers' and firms' forecasting records into their responses to current management forecasts. Second, to my knowledge, this paper provides the first evidence that managers develop a personal reputation via their disclosure behavior and is consistent with their concerns about establishing a personal reputation for accurate and transparent reporting (Graham et al., 2005). Although prior research has also examined the effect of management reputation on market responses to disclosures, these studies generally assume (implicitly) that a firm's disclosure history is attached to the firm or its management team and do not differentiate between the reputation of a firm and that of an individual manager (Williams, 1996 and Hutton and Stocken, 2009). However, evidence from several recent studies on manager-specific effects suggests that there is reason to believe that it is important to distinguish between the two. Therefore, I identify settings where individual differences are accentuated and provide evidence that investors respond to manager-specific effects when management forecasts are more likely to reflect individual managers' forecasting styles. My third contribution is a response to the Beyer et al.'s (2010) call for research that considers management's tradeoff between high payoffs in the current period and reputation gains for a better understanding of the multi-period interaction between management and the users of corporate disclosures. My evidence is consistent with Stocken (2000), which argues that managers can build a reputation for reporting truthfully. The results presented suggest that managers with high prior forecasting accuracy benefit from a stronger market response to forecast news when information uncertainty is high. The rest of the paper is organized as follows. Section 2 reviews prior literature and develops the hypotheses. Section 3 describes the data and research design. I discuss the empirical results in 4 and 5. Section 6 discusses the robustness tests and Section 7 concludes.

The goal of this paper is to examine whether market responses to management forecasts vary with managers' forecasting credibility, where credibility is captured by manager-specific prior forecasting accuracy. Following several recent studies that adopt an empirical design by Bertrand and Schoar (2003), I separate firms' overall forecasting records into a firm- and a manager-specific component. I first show that the firm- and manager-specific estimates are differentially associated with forecasting accuracy in an out-of-sample test. I then examine the market response to forecasts and show that the effect of the manager's forecasting behavior is subsumed once the firm's forecasting record is controlled for. However, I also find that investors respond more strongly to forecasts issued by managers with the highest prior forecasting accuracy when information uncertainty is high. This finding suggests that managers do benefit from establishing a strong forecasting record and is also consistent with managers' concerns about establishing a reputation for accurate reporting. This paper extends recent studies (Bamber et al., 2010, DeJong and Ling, 2010, Dyreng et al., 2010 and Ge et al., 2010) that provide evidence of significant manager effects on an array of firms' accounting and financial reporting choices and is the first to document the capital market consequences to managers for establishing an individual forecasting style. While prior and concurrent research (Williams, 1996 and Hutton and Stocken, 2009) has only examined the effect of management reputation at the firm level, I provide evidence that individual manager styles also matter to capital market participants. However, it is important to note that because CEOs and CFOs are not assigned to firms randomly, the manager fixed-effect estimates may also be capturing a firm's change in disclosure policy, rather than the manager's active influence on the firm's forecasting decision. Although I explicitly control for firm-specific forecast accuracy in the market reaction tests, the manager fixed-effect estimates are still likely to be measured with error, and the results should be interpreted with this caveat in mind. Future studies can examine whether the labor market for top executives is also efficient in recognizing differences in managers' disclosure styles. Are individual managers' forecasting styles correlated with their other financial reporting or operating styles? Moreover, are managers with higher forecasting accuracy more likely to be employed because they've signaled their talent? These are some questions related to this line of research that can lead to further insights.