Theoretical And Empirical Study Of A-share And H-share Price Differences

Based on the research achievements of foreign scholars and domestic scholars, this article discussed price differences between A-share and H-share on the basis of selecting the relatively comprehensive explanation variable, studied50listed companies, during July2010to December2012.First of all,this article proved that A-share and H-share market segmentation exist using the Granger causality.And then designed two models to inquire the influencing factors of this price difference, one is multiple cross section model, another is the fixed effect variable coefficient model of panel data. First model using the least squares regression calculation, the second model adopted the residual variance for the weight of each section of the generalized least squares regression calculation. Meanwhile,divided the data into two categories, low premium rate and the high premium rate. After a series of regression testing and correction of process, obtained the following conclusions:1. In the cross section of multivariate regression model, the low premium rate samples and samples of the high premium rate,the variable SIZE to messure asymmetric information is a negative proportional relationship to premium rate, and the results are significant;the variable RV to measure the demand elasticity difference is a negative proportional relationship to premium rate, and the results are significant;the variable EPS to measure the investors concept differences is negative proportion to premium rate, and the results are significant.2. In the fixed effect panel data model with variable coefficients, the measure variable of information asymmetry, an AIE for high premium rate is significantly negative, the influence of sample with expectations, but for the low premium rate the influence of the sample was not significant, while the direction is not clear; Measure the variables of liquidity difference,RL of samples for low premium rate is basic significantly negative, the influence of basic with expectations. For the high premium rate sample all RL is significant, and the influence of basic is negative, and basic with expectations; Measure investor concept differences though the RV for low premium rates, the influence of the sample is significantly negative, basic with expectations, but for the high premium rate sample basic is negative, the influence of basic consistent with the expected, but at the same time, the result was not significant.According to the empirical results, this paper finally puts forward some suggestions, such as strengthening information disclosure, enhancing the H-share liquidity, leading A-share investors to rational investment, speeding up financial innovation, seting up arbitrage mechanism, etc.