Study On Risk Contagion Across Markets Based On Smooth-Transition Mixed-Copula

With the speeding up of the financial integration process in China, the financial market dependence structure is becoming more and more complex, the dependent patterns are often nonlinear and asymmetric. The relationship between financial markets is increasing, the spread of financial risk is more diverse, and the effect is more far-reaching. Under this background, risk contagion and volatility spillover issue has become the focus of research, the lack of measurement of the dependence between the financial markets in the extreme cases, may underestimate the market risk. Therefore, how to effectively measure the dependence structure and risk contagion effects is a problem to be solved now.Dependency metrics is the prerequisite for analysis of financial problems, the traditional econometric model is only applicable for descript the linear correlation, in solving complex, non-linear, asymmetric dependence of the financial market, a new measure method is needed. As a tool for multivariate statistics and dependency analysis, Copula function has advantages in describing the dependence relationship between financial variables, but a single Copula function can only measure single structure. In view of the complexity of the dependence structure of financial time series, it is necessary to construct a more flexible Copula function, which is a mixed Copula function. The mixed Copula function not only covers the excellent properties of a single Copula function, but also can adjust the dependence pattern according to the change of weights. In order to maximize Gumbel Copula and Clayton Copula’s advantages in describe the tail dependence, and overcome the defects of the single Copula function in the characterization of asymmetric tail dependence, this paper takes the Clayton Copual, Gumbel Copula and their survival functions as the basis to construct three different Mixed-Copula. The combination of Clayton Copula and Gumbel Copula(CCG), Clayton Copula and its survival functions(CCSC), Gumbel Copula and its survival function (CGSG).Based on the mixed Copula function, use the multi-regime smooth transition model to dynamic of Copula parameter, a new time variant Copula model has constructed to study the asymmetric evolutionary trend and risk contagion of the tail dependence between the three stock markets in China. The advantage of this model is that it can adjust the position of the inflection point, it is to say that the model is able to capture the structural change of the dependence.Taking Shanghai Composite Index, Shenzhen composite index, Hang Seng Index and Taiwan weighted index as samples, using multi-regime smooth transition mixed Copula model to analyze the varying trend of the tail dependence between two stock markets and the risk contagion effect. The evidence shows that:the change of the tail dependence between the different stock market is not same, the position and degree of the structural changes are different; the evolution trend of the same composition in the upper and lower tail dependence is also different, except to the Shanghai stock market and Shenzhen stock market, others tail dependence exist asymmetry, especially at the beginning of the sample; in the total sample period, several events such as the introduce of QDII and QFII system, the share split reform, the subprime mortgage crisis, QFII and RQFII amount of expansion, etc. has produce a different degree of impact to the dependence between stock market. However there are some common phenomenon, the tail dependence rises sharply in the extreme crisis, which makes the risk contagion across market, it needs the supervision authorities to pay more attention to the Risk Spillover Effect among the mainland stock market, Hongkong stock market and Taiwan stock market, to prevent the occurrence of the chain effect of financial contagion in the stock market. In addition, because the tail dependence between the stock market has obvious dynamic characteristics, and the dependence change is more intense during the crisis, investors should pay close attention to the changes of the dependence during the crisis, and adjust the combination to reduce investment risk.