Research On The Multi-factor Model For Stock Pricing

For the investors in the securities market share price,the most concered issue is how to price stocks usefully.On the condition that the supply meets the demand,Stock pricing model is to propose the correlations between the price of stocks and factors of financial market indicators.That is to say,a good Stock pricing model is helpful for the investors to evaluate and predict the achievement of the price of the concerned stocks before making a final decision,and then the best investment portfolio is obtained.Following the econometrics arbitrage theory,the dependency relationship between the price of stocks and four factors,including return on equity,exchange rate,the proportion of outstanding shares,and CPI(consumer price index),is studied.Then a stock pricing model is established,which is involved in the four factorsAbove all,taking several stocks in Shenzhen and Shang hai as sample and based on granger testing theory,this paper respectively studied the causal relationship between the price of stocks and several factors including the four factors mentioned above.As a result,the four factors are picked up as the main factors.Upon that conclusion,a stock pricing model,the crucial contribution in the paper, is constructed.Firstly,four types of models for panel data,including the mixed model, fixed effect model,and random effects model,are respectively used to analyse the dozens of major sectors in Shenzhen and Shanghai from January 2004 to December 2008,while Eviews5.0 is a tool.Secondly different analyses are made respectively in the use of raw data,logarithmic data,cross-sectional data,longitudinal data of time with the varieties of time interval.Finally,a hybrid model with the four factors for pricing stocks is coming into being.Empirical Analyses indicate the excellent performances in goodness-of-fit(R~2),parameters testing(T) and regression test(F) etc.