The Study Based On The Extreme Value Correlation Of Chinese Stock Market

In the recent years, with the regulation in the stock market of our country, people's daily life are influenced more and more by the stock market. Influenced by the information of market, the stock price may surge hugely and course huge lost during the investors. Based on this situation, how to expect the profits and lost of the holding assets becomes the focus for the investors.The huge fluctuation of the financial assets could be generalized as: the corresponding data have surpassed the fixed threshold or been to the maximum during the certain time, and course the crisis. The classical statistical extreme value themes concern about the maximum/minimum in the random variable series. Extreme Value Theory and methods can deal with the tail behavior for return series. Concerning about the fat tail in the financial risk, it could measure the financial risk underlying the huge fluctuation more effectively.In the financial market, the extreme value correlation of the two or more than two financial assets always become the focal point in the research. The traditional methods of linear correlation methods have certain defects in the theory and empirical research. A new theory and method Copula in financial field have made us find new directions. Using Copula theory, from the marginal functions and Copula function between them, we can get joint distribution which is more close to the practice data and build the risk management model more effectively. It has important meaning in the theory to recognize the financial market correctly and avert the financial risk effectively.Using the Extreme Value Theory and Copula theory, I have tried to model and analyze the stock markets in our country. The main research results and conclusions are as follows:(1), It generalized the two main distributions in the Extreme Value theory, especially the Generalized Extreme Value model. With the data from Shanghai Stock Market, based on the GEV model, it did the data analysis, parameter estimations, statistical test and empirical results analysis. Based on the two statistical tools given by Longin, it gave the probability of exceedance of an extreme price movement and its associated waiting time period.(2), It generalized the definition, character and classification of Copula. Based on the four different Copula functions, with the daily returns of Shanghai and Shenzhen stock market, it gave the parameter estimations of relevant Copula and the best Copula functions with the stock market in our country.(3), It introduced the multivariate Extreme Value distribution and Copula functions of the extreme value distribution, applied the Logistic model to analyze the correlation of the Shanghai and Shenzhen Stock market.