Research On Stock Pricing Method

Traditional stock valuation methods mainly include relative valuation and absolute valuation methods. The former includes price-to-earnings (P/E) ratio, price-to-book (P/B) ratio and some other multiples like P/S, EV/EBIDTA. The later includes dividend discount model (DDM), discounted cash flow (DCF) and residual income model (RIV). All of these methods have made great progress both in theory and practical application. However, they also have some imperfections. For instance, the multiples only use a few indicators to evaluate. Furthermore, their application scope is limited, also it needs forecast the future cash flow of the companies and it is difficult.This paper tries to use several indicators to construct a model and evaluate the value of stock. The indicators we choosing mainly include three types:indicator per share, ability of making profit and ability of debt repayment. We use software Eviews 4.0 and Eviews 5.0 do statistic analysis and find that earnings per share (EPS), operation profit margin (OPM) and current ratio (CR) have significance to the stock price. Then we use panel data to do regression and obtain several models. Through unit root test, F test and cointegration test, we obtain the pooled regression model and time and entity fixed effects regression model. Because the p value of intercept term is not so significant in time and entity fixed effect regression model, we finally choose pooled regression model.We choose several companies which have disclosed annual report of 2010 and test the model. We find the error rate is relative less than PE and PB methods. This suggests that the model have a good forecasting ability. Then we use pooled regression model to forecast the stock price of some companies.At last, we discuss the rationality and imperfections of the model. The model mainly uses EPS, OPM and CR as explanatory variables. This has some reasonable. Also the application range of the model is not so wide. It is only used to evaluate the first-tier white horse pharmaceutical stocks. When we evaluate other industries we should construct new models.