Dividend policy is one of the most important policies concerned with corporate finance for a firm, and it is a decision that balances the use of retained earnings between the firm’s reinvestment and the payment to shareholders. It determines shareholders’currentreturn and thefirm’sfuture prospect,and italso affectsbenefit allocation between different groups of shareholders. Therefore, dividend policy is an important issue relating to the valuation of a firm. In many of the classic principle-agent relationships regarding dividend policy of a firm, asymmetric information is one of the most important frameworks such as the Miller and Rock (1985), and John and Williams (1985). However, they did not allow investors agree to disagree. On the basis of that there are heterogeneous beliefs existing among divergent investors, the model constructed in this paper is not under asymmetric information but a heterogeneous framework. I will determine and interpret optimal dividend policy based on notonly by managers’,butalso investors’viewpoints.
Classic literatures assume that investors have the same beliefs given the same information. However, the common beliefs assumption is suitable for traditional and matured industries because available information is rich and a large amount of experience has been accumulated. Conversely, in this paper the common beliefs assumption is not allowed even given the same information because people agree to disagree due to facing a completely new industry, i.e. biotechnology.
There have been some studies regarding diversity of opinions between different parties. Manove and Padilla (1998) and DeMarzo et al. (1998) have ever considered models where agents have different priors initially and updated in a Bayesian manner.
However, in both cases agents are irrational in that they do not use all the information available to them in an optimal way. In my analysis, all agents are fully rational.
Allen and Gale (1999) argued that common priors assumption is appropriate when information is plentiful, a large amount of experience has been accumulated, and posterior beliefs have converged. It can be argued that common priors assumption is not appropriate when considering new industries such as biotechnology and new technologies such as personal computers. They offered a model to compare the effectiveness of financial markets and financial intermediaries in financing new industries and technologies in the presence of diversity of opinions. The results showed that financial markets tend to be superior when there is significant diversity of opinions. However, it is complex and too many parameters were used to be practical.
In our model, heterogeneous beliefs of investors are directly given. A simple, clear, and general production/investment function is utilized to directly express real estimation on the future earnings of a firm. This function will help to explore heterogeneity deeper. More concrete results are therefore obtained.
The common priors assumption is not appropriate when considering new industries. Harris and Raviv (1993) proposed a model allowing for differences in prior beliefs. Kandel and Pearson (1995) provided empirical evidence that trading around earnings announcements is due to differences in priors. However, their models could not illustrate a concrete industry or a new technology. In this paper, we will take biotechnology as an example. When a new medicine is innovated, there will be heterogeneous beliefs on its future prospect given the same investment.
Investors agree to disagree because the amount of data available based on actual experience with this new product is nonexistent or small.
Chen (2003) proposed a mathematical model to study dividend policy. She focused on how differences of opinions create incentives for the investors to trade.
From the viewpoint of risk, she concluded that an optimistic investor perceives more risk and hence behaves like trend chasers. Although she has ever found that high dividend payout ratio mitigates the divergence in opinions among traders, how the optimal dividend policy is determined and changed according to heterogeneous beliefs is not yet discussed. In this paper, not only does heterogeneous beliefs to result in changing optimal dividend policy be studied but the degree of diversity of opinions be
also considered.
Grullon et al. (2002) proposed maturity hypothesis to examine the relation between dividend changes and risk changes. As firms mature, their investment opportunities set shrinks, resulting in a decline in their future profitability. The decline in investment opportunities generates an increase in free cash flows, leading to an increase in dividends. In our model, an increase or decrease in dividend is according to the ratio of pessimistic to optimistic investors and heterogeneous beliefs of investors.
Bhattacharya (1979), Miller and Rock (1985), and John and Williams (1985) proposed that firms adjust dividends to signal their prospects. A rise in dividend typically signals that the firm will do better, and a decrease suggests that it will do worse. In this research, however, dividend policy is not determined by a firm’s character but investors’ viewpoints. As the beliefs of investors including both optimisticand pessimisticoneschange,afirm’soptimaldividend policy willbehence adjusted. Low dividend policy is appropriate as in investors’ beliefs are simultaneously increased. We interpret this novel result not based on the traditional signaling theories, but on heterogeneity. Specifically, the ratio of pessimistic to optimistic investors and the degree of divergent beliefs could change the optimal dividend policy.
This research may provide a useful reference for researchers attempting to develop a theoretical model regarding optimal dividend policy under a framework classified by heterogeneous beliefs of investors. In addition, dividend policy in a new industry such as biotechnology will be also illustrated to enrich this study.