Dividends are the reward to shareholders for supplying capital to the firm. Without the payment of dividends, shares would have no value. Only the promise of future dividends gives value to shares, and therefore under conventional valuation arguments (Williams, 1938) fluctuations in the value of a share are brought about by changes in investors’ expectations about the value of the future dividend stream. Dividends, which are payable in cash (or sometimes optionally in shares – “stock dividends”), are set at the discretion of directors usually on a biannual (UK) or quarterly (USA) basis. The key question is how the directors should set the level of current dividend, or plan future dividend policy, so as to maximize the value of the firm and hence maximize returns to the firm’s owners, the shareholders.
There are two main approaches to explaining the dividend decision. The first starts from the theoretical result that, under perfect market assumptions, both managers and shareholders should be indifferent to the size of the current dividend announcement (Miller and Modigliani, 1961). The various assumptions that lead to this conclusion can be examined and relaxed, leading to a deeper appreciation of the determinants of the dividend level in practice. The second approach starts from the empirical observation that directors have a profound reluctance to decrease the dividend over time (this is certainly true in nominal terms: however, real decreases, such as would be caused by holding the dividend constant, are much more frequent). Investors, while accepting that earnings levels may fluctuate in the short term, seem to be strongly averse to any signal that the underlying level or trend of earnings may be unsustainable. A reduction in dividend, it is argued, has particularly unambiguous information content (but see Taylor, 1979) an d is viewed as strong evidence that managers believe earning levels cannot be maintained, and has a profound effect on the market’s perception of the value of the company. A possible additional consequence is the increased threat of replacement of management through, for example, takeover activity or other means. Consequently, managers may go to exceptional lengths to avoid a reduction in dividend. The corollary is that an increase in dividend should only be made if managers believe the new level is sustainable. These behavioral influences give some predictability to the current dividend based on knowledge of current earnings and past dividends (the “partial adjustment” model of Lintner, 1956). However, attempts to establish a link empirically between current dividend changes and future earnings changes (the signaling hypothesis) have not been convincing despite the plausibility of the theoretical arguments.
Dividend Irrelevancy
Miller and Modigliani (1961) demonstrated that in ideal circumstances the level of a firm’s dividend would not affect the value of the firm, with shareholders being indifferent to an announcement of low or high levels of dividend. The assumptions underpinning this result hinge around the notion that company value depends solely upon the investment opportunities available to it, and that finance for investment is always available. In effect, for a given set of investment opportunities, the firm can raise sufficient capital (internally and externally) to fund both its investment program and its dividend. This result is closely related to the theoretical position established by Miller and Modigliani (1961) that under a similar set of idealized assumptions, company value is un affected by the mix of equity and debt used to finance it.
Of course, the world does not always obey the theoretical assumptions, and many caveats modify the Miller and Modigiliani proposition. From the perspective of shareholders, “irrelevance” implies that they are indifferent between receiving returns as dividends or as capital gains. For a given investment program, a lower dividend implies a greater capital gain and a higher dividend implies a lower capital gain; the overall returns being equivalent in either case. In practice, the tax regime may favor one form of return over another. An investor who is taxed on dividend income, but not on capital gains, will prefer a low dividend policy provided the capital gain reflects the full amount of the retention, and vice versa (short-term capitalization evidence suggesting this is not the case is provided by Elton and Gruber (1970),and a consistent longer-term analysis is found in Auerbach (1979)). For a company, retained earnings may be taxed differently to distributed earnings (as was the case in the UK in the 1950s) leading to different corporate tax bills under different dividend policies. Assuming that this is not the case, the n a company that is acting in the best interests of its shareholders will choose the dividend policy that minimizes the total tax bill of its shareholders. However under a classical tax system such as that operated in the USA, this would seem to suggest that no dividends would be paid at all since the effective rate of tax on capital gains is less than that on income! Even under the comparative neutrality of an imputation tax system as used in the UK, Australasia, and elsewhere, shareholders may not be indifferent between dividend returns and returns taken as capital gains. Mallin’s (1993) evidence of the take up of stock dividends in th e UK shows that on average shareholders elect to take only about 5 percent of dividend in stock form, implying that income returns are generally preferred to capital gain.
From the perspective of a firm’s management, an essential component of the irrelevance view is that investment decisions should not be affected by dividend policy. This amounts to asking two questions. First, is there any discernible evidence that internal investment is affected by dividend levels, and second, is there any evidence that the rates of return generated by employing different forms of finance are different? The answer to the first question depends on whether an investigation i s carried out in cross-section (where following Dhrymes and Kurz (1967) an interactive effect is generally supported) or in time series (where, following Fama (1974), it is not), so this question is not yet resolved. In a situation of market-induced capital rationing, were this situation to exist, it is accepted that investment choices would be heavily influenced by the quantity of retained earnings. Dividend policy would directly impact upon investment policy and the Miller-Modigliani proposition would not apply.
One of the main issues surrounding the second question is the hypothesis that management uses retained earnings inefficiently (Baumol et al. (1970) provide some evidence, although their methodology is problematic). A firm’s ma nagement that pays a low dividend in order to invest retained earnings avoids the costs and the scrutiny that comes with attempting to raise capital in the market. There is a direct financial cost involved in going to the market to raise capital, and indirect costs may be incurred in facilitating the monitoring expected. Avoidance of the discipline of the market leaves management more latitude to enjoy “perquisites” (management perks), with a consequential in crease in agency costs which high dividend levels would avoid.
Conclusion
There is no single theory to explain a firm’s dividend policy or to determine an optimal level for the firm’s dividend. Empirical studies give contradictory evidence, but from a practical viewpoint managers seem to attempt to maintain a particular payout ratio, tempered by a great reluctance to reduce the dividend from last year’s (nominal) level.
Dividend levels as a whole may be paradoxically high. For example, why do many firms incur unnecessary issue and transactions costs by paying dividends and at the same time seek new equity capital – in the case of a rights issue from the same shareholders as receive the dividend? The cross-sectional evidence suggests in addition that high dividend levels may be damaging to a firm’s investment program. Despite these inefficiencies shareholders seem to prefer the cash-in-hand of immediate high dividends (together with the discipline this imposes on management) to the uncertain prom ise of higher dividends in the future.
