The Research And Application Of Financial Risk Contagion Based On Copula Model

The economic globalization and financial liberalization are deepening, the relation of global economic and financial between different countries is becoming close under the background. There is no longer such a country whose economy and finance can't exist from without other country. The information technology is developing quickly at the same time, which makes the market transaction information transfer to the other financial markets that always are thousands miles away almost simultaneously. When a financial market occurs risk crisis, then in a very short period of time, the national or regional financial markets which is closely linked with that market are likely to be infected by the event of infection and risk crisis. And when a country occurred financial crisis, the scope of the financial crisis can spread rapidly by some countries and regions to more countries or regions, causing regional turbulent financial markets and may even escalate into a global economic crisis. Therefore, in this context, to accurately measure the financial risk of infection higher researching on the financial risk of infection has become one of the hot issues of financial academia. Measure the risk of infection is not only able to build a firewall keeping away extreme risk for financial risk management departments to provide policy recommendations, but also to be able to cross-market investors to invest in global to provide help, and thus the subject of this study have an important theoretical and practical significance.Acquiring the return series which reflects the financial market actual running characteristics is one of the fundamental to measure financial risk infection accurate. There are many typical characteristics in financial income general, and the fact that the presence of these typical features will make the financial markets showed a more complex risk characteristics, and will increase the difficulty of risk management. In the study of existing studies measure the financial risk of infection, most scholars have not been able to study under the stylized facts constraints, and then the findings may exist errors. Based on this, wo expand the risk infection research measure under the stylized constraints, first we apply some basic statistical test methods to test what typical facts of financial return series exist, then using ARFIMA-FIAPARCH-SKST model for modeling financial return series, excluding typical facts in financial return series, ultimately obtaining the stand return series which can reflect the financial market actual running characteristic. The empirical results show that the four stock market returns in this study have many typical facts such as correlation, long memory, volatility clustering and leverage effect, through ARFIMA-FIAPARCH-SKST model being used in this paper to obtain a standard yields sequence that does not contain the typical facts, after its integral transformation, then use statistical tests such as KS test methods, test results show gains through standard return serial conversion obedience [0,1] Uniform distribution, so using ARFIMA- FIAPARCH-SKST model for stock market returns modeling is appropriate, which can describe the financial return series important stylized facts accurate.At present, in mostly research scholars concentrate on one thing that the presence or absence of risk infection in the financial market, they doesn't measure financial risk infection intensity. At the same time these research methods in the study of financial markets ignored financial markets' nonlinear and asymmetric tail relationship. The non-linear structure and tail dependence relation are the focus in the risk infection study. The Copula function precisely measure to meet the needs to test financial risk infection. Therefore, this paper, we use Copula methods to measure the financial risk contagion between American stock market and Asian markets during the subprime crisis. We compare different copula models theoretically, chose from SJC-Copula function and mixed Copula function which is made up of Gumbel, Frank and Clayton, and then the better model is used to model Chinese mainland, Hong Kong, Japan and the US stock market during the subprime mortgage crisis, then compare the tail dependence coefficient obtained before the crisis and the tail dependence coefficient obtained after the subprime mortgage crisis, and acquiring the risk infection characters between Asian stock markets and the US stock market. The empirical results show that: After the subprime mortgage crisis, the US stock market and the Chinese mainland stock market was not a direct risk infection, and the Japanese stock market, the Hong Kong stock market and the US stock market have taken place risk infection, the US stock market transmits financial risk to Chinese mainland market through Japan and Hong Kong indirectly in the subprime mortgage crisis.In the context of the deepening financial liberalization, global financial markets gradually form a unified whole, different financial markets influence each other. In the previous text we use Copula function to measure the financial contagion between financial markets, but the model is only able to analyze the risk contagion between two financial markets, we need analyse the risk contagion characteristic between several markets when we build a portfolio of globalization, and therefore we need to further develop the above study, measure the more complex dependence structure between several financial markets and the financial contagion relationship. Current scholars often use Multivariate Copula function, but because of the type of Copula function relatively simple and can't be changed flexibly according to the study, in theory, there are many shortcomings. Vine Copula theory developed on the basis of the Copula function can solve the shortcomings of the above model, vine Copula function has been widely applied in the research field. The applying of vine Copula provides a theoretical basis and technical support for our research. Therefore, we will study the multiple financial markets' risk infection further using vine Copula model. We use Pair Copula decomposition model composed by SJC-Copula function in the study, we chose C breakdown structure. The empirical results show that: After the subprime mortgage crisis, the tail dependence coefficient is smaller between the Chinese mainland stock market and the US stock market to a certain degree, but the tail dependence coefficient of the Chinese mainland and Hong Kong stock market has increased significantly, the tail dependence coefficient in Japan, Hong Kong and the US stock mark, et al.so has increase some degree. Comparing with the bivariate Copula function, we also found that the dependent coefficient is bigger than what we get from bivariate Copula result in several financial markets background, so we think bivariate Copula function may underestimate the relationship between financial markets.