Empirical analysis of Chinese stock market behavior: evidence from dynamic correlations, herding behavior, and speed of adjustment

Abstract

In its rapid transition to a modern economy, China is undergoing dynamic changes in all of its business sectors and industries, a situation that presents features unique to Chinese culture and China’s economy. This dissertation focuses on several issues related to the Chinese stock market, and these issues are separated into the following three essays.Essay One: Dynamic Correlation Analysis of Chinese StockThis paper examines A-share and B-share market segmentation conditions by employing a dynamic multivariate GARCH model to analyze daily stock-return data for the period 1996 through 2003. Statistics show that stock returns in both A and B shares are positively correlated with the daily change in trading volume or abnormal volume. The evidence reveals that the correlation coefficients between A-share and B-share stock returns are time-varying. Analyzing the dynamic path of the correlation coefficients suggests that the recent increase in correlation coefficients is significantly related to a more liberal policy. There is a substantial spillover effect from the Asian crisis into Chinese stock-return dynamic correlations. The evidence suggests that the time-varying correlation is associated with time-varying risk measured by a daily high-low price differential.Essay Two: Is There Herding Behavior in Chinese Stock Markets? An Examination of Chinese A and B SharesThis essay examines whether herding behavior exists in Chinese A- and B-share markets. By applying the methodology proposed by Chang, Cheng, and Khorana (2000) to examine Chinese stock data, we provide evidence that shows there is herding behavior in both the Shanghai and Shenzhen A-share markets. However, no supportive evidence for herding behavior is found in either B market. Herding behavior in Chinese markets demonstrates similar patterns of asymmetric effects as the market goes up vs. as it goes down, trading volume becomes excessively high vs. excessively low, and volatility becomes excessively high vs. excessively low. There is no concrete evidence in favor of herding behavior across A and B markets, nor across Shanghai and Shenzhen markets. Volatility seems to have more explanatory power than volume in explaining herding behavior.Essay Three: Empirical Analysis of the Speed of Adjustment to Information: Evidence from Chinese Stock MarketsThis essay studies investors’ behavior characterized by different degrees of sophistication involved in Chinese stock markets. By employing a VAR model to examine different speeds of adjustment in response to common information between Chinese A- and B-share markets, we find evidence that domestic investors who mainly invest in A shares adjust to information faster than foreign investors, who can trade only in B shares. By further looking at characteristics of individual firms and market structure, we find evidence that stocks with higher information flows and/or with more prominent status adjust to information faster. The implementation in February 2001 of a more liberal policy of allowing domestic…