Analysis Of China's Stock Market Volatility Based On ACD-UHF-GARCH Model

Under the condition of market economy, Chinese stock market are more mature than before. At present, the volatility of the stock market research has always been the hot spot of the financial research. Currently, many scholars study on the price and volume of the stock market, but rarely introduce the time factor. This article will both put the transaction volume and time factors in the model, explains the role that duration and trading volume play in the price volatility.Using the high-frequency data to study China's stock market volatility is essentially the application transaction time interval to determine the volatility. High frequency time series contain information of market microstructure and important phenomenon of the information in the day for a long time. In this paper, on the basis of previous studies, using the acquisition transactions and other transaction of Shanghai and Shenzhen trading interval data, choosing ACD model and GARCH model to study the volatility, applying ACD model to calculate the mean and the duration of the transaction conditional expectation, Then apply the resulting parameters to build ultra-high frequency data GARCH model, namely UHF-GARCH model, In the establishment of UHF-GARCH model, this paper not only take into account the duration of the transaction and its conditional expectation, while also taking into account the rate of return on the stock market trading volume and volatility impact.This paper selected the stocks data and index data for empirical research. Choose SPDB and Ping An bank stocks ultra-high frequency data of transaction-to-transaction basis in November and December 2012, index data to select the benchmark Shanghai composite index and Shenzhen component index in 2011 and 2012, two years of high frequency data per minute. Empirical studies show that the features of China's stock market volatility of the trading duration and yield is clustering; Transaction duration and its conditions have an impact on stock market volatility, the larger the duration, the greater the volatility; Trading volumes have more significant impact on volatility, the larger of the trading volume, the greater the volatility is larger. ACD-UHF-GARCH model can well reflect the high frequency nature of volatility clustering; duration and volume have a dynamic effect on the Price volatility.