1. Introduction
This paper will examine the relationship between the dividend policy and stock price volatility based on the previous studies. Dividend policy is referred as a direction to the dividend paid out. In practice, we can through three aspects to show how the dividend policy is important. Firstly, in term of the clientele effect, the transaction cost and tax of investor position can exert an influence on whether dividend policy gains are preferred which means that dividend policy will exert an important influence on investors’ behaviour. Secondly, according to the signaling effect, dividend paid is the mean by manager to signal the new information to investor. And thirdly on basis of the agency theory, dividend policy would attribute to the conflict between the interests of management and interests of stakeholders.

Through the different dividend policy, firms can use the earning to make scrip dividends, special dividends, share repurchase and non–pecuniary benefits. In 1961, Modigliani and Miller (MM) argue that, given perfect and efficient markets, the pattern of dividend payments by a business have no effect on shareholder wealth. The only way to maximum shareholder wealth is investment with a positive NPV. So depend on M&M theory, Atrill (2000) agrees that to pay a lower dividend will simply be compensated by an increase in share price through reinvestment.

De Angelo and Masulis, Kim (1988) and Miller (1986) supported that dividend paid can greatly reduce tax costs as tax clientele effect. As another theory, agency costs, Jensen and Meckling (1976) argued that the dividend payments reduce the agency cost; the managers would not use the cash to reinvestment.

Meanwhile, Miller and Rock (1985) suggested that the dividend announcement provides information to estimate the firm’s current earnings. In addition, Baskin (1989) proposed that dividend