The Research Of Dependence Structure Between The Stock Price And The Trading Volume Of Hushen Stock Market Based On Mixed Copula-Vec Model

This thesis researches on the dependency of the valorem amount between stock market. It takes the daily closing price and daily trading volume of Shanghai composite index and Shenzhen component index as research objects. Further, the paper also analyzes the dependency structure between the price and the trading volume by modeling mixed Copula-VEC.In chapter one, we not only elaborate the background and significance of our study but also overview the present situation of domestic and overseas researches.Chapter two provides an overall idea for modeling in this paper, including the ways of constructing VAR and VEC model. What’s more, it also introduces the definition and the property of Copula function together with Sklar theorem and the method of constructing copula model. In addition, this chapter describes the key steps of modeling and the advantages of modeling mixed Copula-VEC.In chapter three, we present the VEC model of stock price and daily trading volume of Shanghai stock market and Shenzhen stock market, choosing the closing prices and trading volume of Shenzhen and Shanghai stock market (from 30th July,2011 to 30th July,2012) as research objects; processing the statistics to obtain four groups of Logarithmic Sequence and testing the stationary and cointegration, on the basis of the statistic analysis, to construct VEC(12) and to pass through the statistical test; using the model to impulse response function and decompose variance and thus, drawing a conclusion that the dynamic relationship of the stock price and the trading volume is non-symmetrical and that the price has a greater influence on the trading volume compared with the influence which the trading volume has on the price. The result is that the dynamic relationship of the valorem amount between Shanghai stock market and Shenzhen stock market is quite symmetrical.In chapter four, we take residuals which have no autocorrelation as research objects and estimate the marginal distribution of the valorem amount of Shanghai stock market and Shenzhen stock market by using non-parametric kernel density method.Basing on chapter three and chapter four, we establish mixed Copula-VEC model. Firstly, we select Clayton, Gumbel, Frank Copula function and the linear combination of the three functions mentioned before. We model Copula-VEC respectively and do parametic estimation and test them. Comparing four models, we finally decide on mixed Copula-VEC model (5.7) because it is the most suitable one for studying the dependency of the valorem amount between Shanghai stock market and Shenzhen stock market. At last, we draw a conclusion from the mixed Copula-VEC model(5.7) that the relationship of valorem amount between Shanghai stock market and Shenzhen stock market is not only positive dependent but also negative dependent on each other. Besides, the sequences of them have significant non-symmetric tailed relevance. The detailed results are that the stock price and the trading volume of the low tailed relevant coefficient is 0.0245 and the up tailed relevant coefficient is 0.3075 in the Shanghai stock market; the stock price and the trading volume of the low tailed relevant coefficient is 0.0062 and the up tailed relevant coefficient is 0.4294 in the Shenzhen stock market.Chapter VI is aimed at summarizing results of the whole paper and making a prospect of further researches.