Based On China's Stock Market Turnover In Arch Models Logarithmic Rate Of Change Of The Empirical Analysis

The relationship between price and trading volume is an important research topic in the research of financial time series. It is regarded as an important thing in the field of theory and practice. Meanwhile, there are a lot of results about this, these results are mainly focus on the rate of return and the relationship between price and trading volume .But there are a few result about the trading volume. So on the base of the former results by using the software Eviews, this paper studies the trading volume of the day, the month, the week (Trading Unit: round lots; Sample interval:1992.1.2------2006.7.31) in the two stock market of our country. It draws thefollowing conclusions:1. When check the ARCH effect of the logarithm vary rate of the trading volume, it finds that there is a sharp higher - order ARCH effect about the day or the week trading volume' logarithm vary rate in our country's stock market. But there is not about the month.2. Modeling with the data, we get ARCH model family. Compared with The TARCH(1,1)-M and the EGARCH(1,1)-M model ,it concludes that this two model are more suit to the logarithm vary rate of the day or week trading volume . Meanwhile, it shows that the information is not symmetrical and the trading volumes have the similar character with the premium in our stock market. And the result shows that the influence of the random impact have higher continuity.3. Using the models to forecast in the short period. The point estimation and the interval estimation are more ideal. It is worthwhile for other people .