Chinese A And B-share Market Price Discovery And Empirical

If securities issued by the same company are traded in various markets, the price discovery efficiency of these markets may be different. In this paper, several major approaches towards the study of price discovery of assets are adopted to analyze the contribution of A and B shares to the price discovery, respectively. Furthermore, relevant factors which may influence A and B-share market's contribution to the price discovery are analyzed through cross-sectional regression. The main contents include:First of all, Granger causality test result shows that after the openness of B-share market, B shares of more firms in the sample lead A shares than before B-share market's openness and there is a close informational linkage between A and B-share market. In addition, cointegration test is conducted and the empirical evidence indicates that before the B-share market is opened to domestic investors, the two markets are seriously segmented. After the B-share market is opened to domestic investors, the degree to which these two markets are segmented decreases. Then, we use the sample of firms, whose A and B shares are cointegrated, to formulate and estimate the vector error correction model for each firm. The result implies that after the openness of B-share market, B share price of more firms respond to the deviations from the A share price. The extent to which B-share prices respond to A-share prices is more then vice versa.Secondly, four models (Common Factor Contribution Model, Information Share Model, Modified Information Share Model and Kasa Model) are employed to analyze the contribution of A and B shares to the price discovery quantitatively. Before the openness of B-share market, according to common factor contribution methodology, A-share market contributes more to the price discovery. And we find the similar results through the other three models. After the openness of B-share market, all of the four methodologies find that A-share market is still in informational dominance. However, the conclusion about dynamic variation of the contribution of A-share (B-share) market to the price discovery is inconsistent among the four methods: according to common factor contribution methodology, information share model and Kasa model, contribution to the price discovery attributed to A-share market declines with the openness of B-share market, while according to modified information share model, A-share market's contribution to price discovery increases.Finally, the thesis constructs cross-sectional models to examine each relevant factor in order to probe into the main factors that affect the extent to which A-share (B-share) market contributes to price discovery. Our regression results show that duration is the most important factor influencing the contribution to price discovery of A and B-shares. Before the B-share market's openness, A-share (B-share) market's contribution is significantly related to relative systematic risk. After the openness, trading volume ration and relative turnover is significantly related to the contribution of A-share (B-share) market to the price discovery. The coefficients on the illiquid share ration and industry dummies are not significant.