Since the dividend irrelevance theory of Miller and Modigliani (1961), academics and practitioners still have little understanding of the managerial incentives underpinning dividend policy. Black (1976) observed, “The harder we look at the dividend picture, the more it seems like a puzzle, with pieces that just don’t fit together.”
This thesis aims to shed additional light on the dividend puzzle. Accordingly, two theoretical models have been developed to help explain why firms pay dividends or repurchase their own shares. The models consider the case in which the managers of a high-quality firm (firm H) and a low-quality firm (firm L) choose to use corporate cash flows to pay dividends, repurchase shares, or invest in a real project from which they can earn private benefits. I focus on the case in which firm H has a positive NPV project whereas firm L has a negative NPV project.
In the first model, developed in spirit of Isagawa (2000), I show that paying dividends is a dominated strategy for firm H, regardless of the managerial weight parameter. If the manager is myopic, firm L will choose to repurchase shares at the detriment of existing shareholders. If the manager is farsighted, on the other hand, firm L will choose to pay dividends. I also consider the case in which investors are irrational in that they do not update their beliefs upon observing one firm repurchasing shares while the other firm paying dividends. The model shows that, in inefficient market, firm L will not mimic given that firm H repurchases shares since it cannot obtain any benefit from doing so.
In the second model, built on Fairchild and Zhang’s (2005) work, in which the managerial payout decisions depend on the relative magnitudes of dividend and repurchase catering premia, I demonstrate that a myopic manager of firm H may pass up a positive NPV project in order to cater to investor demand for dividends or share repurchases (an adverse selection problem). In addition, I show that the agency cost of free cash flow can be mitigated if the dividend-catering premium is sufficiently high. That is, firm L’s manager will have a strong incentive to return excess cash to shareholders rather than invest it in a negative NPV project.
Then, I investigate dividend changes in Thailand over the period 2002-2005. To test the signalling and free cash flow hypotheses, I first analyse profitability changes around dividend changes and benchmark them with control firms, and examine the relation between dividend changes and the past and future profitability. Consistent with Benartzi et al.’s (1997) evidence in the U.S., dividend changes in Thailand do not signal future profitability but rather the past performance. Then, I examine the determinants of dividend changes and firm’s decision to change dividends. I also investigate the short-run and long-run stock price performance of dividend-changing firms, and the relation between announcement returns and hypothesised independent variables. Finally, I examine firms’ investment behaviour following dividend changes. The results do not support the view that dividend changes signal future profitability. Overall, the findings are broadly consistent with the free cash flow hypothesis rather than the signalling hypothesis.
Additionally, I provide preliminary evidence on open-market share repurchases (OMRs) in Thailand over the period December 2001 to January 2007. I find that stock prices react positively to OMR announcements and continue to increase in the longer term, suggesting that stock market underreacts to the signal conveyed by the managers of repurchasing firms. Comparing the actual repurchase cost with the costs of benchmark portfolios, I find that the actual repurchase cost is the lowest. This finding suggests that the managers of repurchasing firms have substantial ability to time the market.
|Date of Award||1 Sept 2009|
|Supervisor||Richard Fairchild (Supervisor)|
- free cash flows
- share repurchases