In M&M’s perfect capital market assumption that the managers and shareholders of a firm do not have conflicts of interests. However, for majority of companies managers and shareholders are separate and therefore they normally have different interests. Managers as agents have the company managing rights delegated by principals, or shareholders. Thus, managers may misbehave in their own interests instead of the shareholders’ interests. Shareholders have to supervise managers’ behaviour even if it inevitably generates costs. The agency costs refer to the implicit costs that generate from the agency problem, namely the managers and shareholders conflict in different interest and the information is asymmetric as only managers are engaged in managing work of the firm’s operation. According to Easterbrook (1984), dividend could reduce the agency costs by cutting the free cash flows in managers’ hands as well as encouraging managers to raise funds in capital markets. It could also be argued that dividend payments help shareholders to limit the managers’ rights to prevent their selfish behaviours. On the other hand, he also argued that dividend may also induce managers to rely upon debts, which makes the firm bear more risks. Jensen (1986) added that managers can benefit from larger size of firms and thus they have incentives to overinvest even if for the projects with negative NPV. If the assumption holds, high dividend could reduce the overinvestment problem so as to increase the firm’s value. Hence, in contrast to M&M theory, free cash flow theory implies that dividend payment is positively related to the firm’s value. Nonetheless, shareholders should consider the trade-off problem brought by the high dividends payment, namely weigh the benefit of reducing agency costs against the costs of dividend taxes and the possibility of bearing more risks for the firm.
Rozeff (1982) constructed a model called “cost minimisation model” to calculate the optimal dividend payout level. In the model the independent variable insider shareholding percentage is negatively correlated to the payout ratio and the number of shareholders is positively correlated to the dividend payout ratio. He found that agency cost reduction effect of dividend is less obviously observed in companies with dispersed ownership or companies that are highly owned by insiders. Dempsey and Laber (1992) supported Rozeff’s model with updated data. An updated study by Holder, Langrehr and Hexter (1998) also showed evidence to support Rozeff’s model by proving the significance of the two variables and their relations with the dividend payout ratio. However, Alli et al. (1993) contended that the ownership dispersion was not a significant factor to affect the dividend’s agency cost reduction effect. Lang and Litzenberger (1989) gave evidence that supported the free cash flow hypothesis by identifying the overinvestment projects first and thereafter quantitatively examining the dividend change announcement’s effect on the firm’s value. Nonetheless, Lie (2000) suggested little evidence was found to prove the free cash flow hypothesis after researching a large number of samples.
Location:Home > Policy Consulting > Policy Analysis > Details
Liu Shangchao:Agency Costs and Free Cash Flow Hypotheses on Dividend Policy
2012-12-27 12:55