Theoretical Study Of China's Stock Market Price-Volume Relation With Applications

This dissertation takes a close and deep look at price-volume relation of China's Stock Market and its main subjects are as follows:1.Using Asymmetric Component GARCH model (AC-GARCH) to investigate price-volume relationship in china's stock market and divide volume into expected and unexpected component. The empirical results show the transitory volatility is explained mainly due to unexpected volume, the daily information flow is the source of market volatility replaced by unexpected volume, and this proved the prevailing MDH theory. The volatility persistence is only partially explained by volume, other factors are also too. This is consistent with the emerging markets, but not with the USA. Whose market volatility is explained by volume completely. Volatility in equity markets is asymmetric: the results indicate that the impact of shock from negative returns and increased trading volume on volatility is larger than that from positive returns and decreased trading volume in the same magnitude, respectively after 1997 in China's Stock Markets, especially in Shen Zhen city.2. having introduced dynamic Bivariate Mixture Distribution models (BMD), including standard and modified mixture models, the volume in these models appears endogenous. Parameters of these models are then estimated with MCMC method based on Gibbs sampling. The results show that the BMD models can capture the persistence of return volatility basically, and prove the daily trading volume has informed and noise components. The systematic variation in trading volume is due solely to fluctuations in the informed volume. But the defects in these models exist too, may attributing mainly to strict limitation hypothesis conditions of the models. Next a Generalized bivariate mixture model for stock price volatility and trading volume is used to analyze stock price volatility and trading volume on the China Stock Market. In this model, the traders'sensitivity to new information is traded as an important latent factor. The empirical results based on daily data of individual stocks reveals that the standard mixture model with its assumption that traders'sensitivity to new information is constant over time is clearly rejected against the generalized model. The number of information arrivals as well as the sensetivity to news are important factors accounting for the relationships between price volatility and trading volume. The genaralized mixture model improves obviously the explaination of the behavior of volatility relative to the standard model.3. proposing a generalized mixture of distribution hypothesis (GMDH) and examine whether it explains the relation between ARCH effects and trading volume. The empirical conclusion shows that daily volatility contains a stochastic component that typically accounts for over half of its overall variation. The unexpected component of daily volatility stems from daily information releases and the expected component is driven by past return shocks.4. Linear Granger causality tests conducted to detect information content of volume. The results show that the bidireactional causality is supported between volume and returns. Statistical analysis shows that volume conveys information to the market about the magnitude of price changes and the direction of price changes. This is in accordance with technical analysis of trading strategy.