The Study Of Copula Theory And Its Application In Financial Time Series

Thailand had a serious financial crisis in 90 s,the economic development of Southeast Asian countries declined significantly under the impact of the financial crisis.The United States triggered a crisis in this century,which made the global financial has been widely implicated.The two financial crisis shows that the global economic development is getting closer and closer,and there is a strong correlation between the financial industries.In order to avoid the financial crisis and predict the development of financial market,the research on the correlation between financial markets is getting more and more attention,especially on the correlation degree and the dependence structure of financial time series.The main research contents of the paper are as follows:(1)Dependence on structure and pattern of the Shanghai and Shenzhen stock market is concerned by using the Copula-Kernel method with the Shanghai B and Shenzhen 300 refers to the day's closing price,Nuclear smooth method fits the marginal distribution of the Copula function,Copula function is selected combined with the features of data and the rank correlation coefficients of the Copula function,and the fitting degree is evaluated by the2 L discrete norm.It is found that t-Copula can be used to fit the daily yield sequence of Shanghai and Shenzhen stock market,and the daily return series of Shanghai and Shenzhen stock markets show strong symmetrical tail dependence.(2)The Copula-ARIMA model is constructed to study the correlation structure between China's financial industry and the real estate industry based on the date of our country each quarter of the year GDP in the financial industry and the real estate industry in the quarter value series combined parameter estimation and goodness of fit test.Find Frank Copula function can better describe the correlation between the financial industry and the real estate industry in China,there is a symmetric correlation between the two industries.(3)The exchange rate of RMB against the RMB and the RMB exchange rate is selected,and the relationship between the RMB exchange rate and the RMB exchange rate is studied by using the Copula-MA-ARCH-t model.The AIC criterion is used to determine the condition of the marginal distribution model order and parameter selection,Copula functions are selected by using the 5 commonly used Copula function combined with the rank correlation coefficient and data feature,and these five kinds of Copula function fitting effect diagram are compared by the Q-Q diagram.The results show that: Gumbel Copula can better fit the rank correlation between the dollar exchange rate and the RMB exchange rate from the rank correlation,the t-Copula function can well describe the dollar against the RMB and Hong Kong dollar against the RMB exchange rate,the U.S.dollar against the RMB and Hong Kong dollar against the RMB exchange rate has a strong tail correlation from the fitting results of tail characteristics and the Q-Q diagram.