Ebook Dividend and Investment Decisions under Managerial Discretion
In a typical signalling model, dividends are a precise indicator of the future prospects of a firm. Managers are assumed to act in the best interest of existing shareholders - they are indeed usually taken to be a representative of the “average shareholder”. They weigh the costs of dividends to the shareholders and the potential benefits they can draw from separating from low-quality firms, and the optimal dividends will reflect this tradeoff.
It is well known, however, that managerial behavior does not always follow shareholders’ best interest - and that managers have considerable leeway in deciding the actual level of dividends. This is quite likely to lead to important distortions. While 88% of the managers surveyed by Brav, Graham, Harvey and Michaely (2004) say there are negative consequences to reducing dividends, many of them also “tell stories of selling assets, laying off a large number of employees, borrowing heavily, or bypassing positive NPV projects, before slaying the sacred cow by cutting dividends.” Forgoing some valuable investment may be the necessary cost of signaling. We also know, however, that dividends are smoothed (DeAngelo and DeAngelo 2006, Kumar 1988; 90% of the managers in the Brav et al. sample say they smooth dividends) and that dividend changes are only weakly related to future earnings (Benartzi, Michaely and Thaler 1997, DeAngelo, DeAngelo and Skinner 1996). Moreover, it seems that only a tiny minority of managers agree that payout policy is used to make the company look better than competitors, and that payout policy shows that the firms can afford costly signals (Brav et al. 2004). Therefore, while managers are afraid of low dividends, they also do not seem interested in precise signaling.
This paper presents a basic model in which dividends are only a rough indicator of firm quality and managers are self-interested. Firms have different investment opportunities and these opportunities are imperfectly known by investors - or information about them can only be acquired at a cost. This means that firms that pay low dividends because they invest in positive NPV projects face the risk of being pooled with firms that pay the same level of dividends without having similar growth opportunities. Thus firms have to deal with the tradeoff between being perceived as the better firm today or after the results of investment become visible. Managers are the one to decide on dividends and investment, and their interests are not necessarily aligned with those of shareholders or potential investors.
In this framework, higher dividends are still associated with higher average firm quality, but the same level of dividends is associated with firms of different quality. There is usually underinvestment as managers try to boost current share price. The problem is reduced if managers care more about the future, if returns on investment are higher, and the market discount factor is higher.
Interestingly, the model implies that investment is more likely if growth opportunities are more widespread; there is “strength in numbers”. When investors know that many firms have valuable growth opportunities, low dividends will be more easily accepted and share prices will be more favorable to investment. When investment opportunities are less widespread, the lonely firms that have positive NPV projects will find that market rewards are biased against investment. This mechanism goes some way into explaining the well-known phenomenon of “disappearing” (Fama and French 2001) and “reappearing” (Julio and Ikenberry 2004) dividends. It also provides an explanation for the “catering” for investor preferences noted by Baker and Wurgler (2004). Indeed, while “catering” has been challenged by Hoberg and Prabhala (2004), who emphasize the importance of idiosyncratic risk, it may well be that the two competing stories describe the same feature of dividend policy.
Download
PDF Ebook Dividend and Investment Decisions under Managerial Discretion
Posted in :
