An Empirical Study Of The Residual Income Valuation Model

It has been many years that equity valuation is not only important content of financial management, but also one of the basis by which security investors make investment decisions. In current background of vigorously advocating value regression, it is especially important to explore intrinsic value of a company with the help of equity valuation models and statistical methods. Basing on these, this paper takes listed companies, which come from A's stock of Shanghai exchange and Shenzhen exchange from 1995 to 2006, as the research sample, and then examine the residual income valuation model under the linear information dynamics process with many kinds of regression methods, it draws main conclusions as follows.Firstly, by terms of 195 listed companies serves as the sample of this paper, the median of the cost of equity capital is 0.0384 and the mean of that is 0.0475, which are higher than the free-risk rate during 2001 to 2006, and this is consistent with the principle that the risk and the return is symmetrical. Descriptive statistics show that there are more than half of samples whose residual income is negative; what's more, the ratio is still as high as 43.33 percentages though we don't take out the extraordinary items, which means that most of listed companies use their own capital with poor efficiency and they have weak capability of value creating.Secondly, the result, which is from the examination of the linear information dynamics process by equations regression, suggests that residual income and book value of equity are persistent and predictable, and the assumption of persistent operation is met for most of listed companies; however, some companies don't conform to the conservative accounting. Viewing from results of the inter-consistency test, for most of samples, the intrinsic value of equity is statistically lower than market value of that, but it is higher than book value of equity, which shows that price is overvalued, and this is similar with views from domestic papers study stock price bubble.Finally, it is a point of importance that exploring the valuation meaning of variables. Basing on tests of various kinds of models, it can draw conclusions as follows: (1) book value of equity, residual income and earnings per share are significantly positively related to the price adjusted, and these three variables can provide incremental value relevance each other. With regard to the interpretative power of a single variable, earning per share is the highest, followed by residual income, and the lowest is book value of equity, which shows that investors pay more attention to information from the income statement, but ignore the residual income which really represents the value-creation capacity, therefore, the pricing multiplier of income is higher than that of other variables. (2) Results of tests, based on panel data, show that the adjusted R~2 of models become much higher after considering company character and periodical character, which suggest that there are some difference among firm-year data. (3) Looking from the result of expanded tests, the interpretive power of models are also related to sampling year, stock exchanges and the signal of residual income. Besides, the rate of circulating shares, size, the systemic risk of stock, and the property of industry can also provide some marginal interpretive contribution toprice.