Research On The Correlation And Portfolio Risk Of Exchange Rate Market Based On Mixed Copula Model

With the fast development of globalization,the economic ties among countries are increasingly closed.International bond between the final accounts of the debt due to the change of exchange rate in advance is difficult to grasp,as many countries with floating exchange rate,China's international balance of payments and enterprise economic benefits is also under the influence of foreign exchange risk.The dollar,euro and yen,are the three major currencies in the world,as one of the leaders of the emerging economies,China's financial market develops rapidly,and the interaction between the developed countries is increasingly frequent,therefore,to study the United States,China,Japan correlation with the tail of the euro exchange rate and its risk measures can be of great significance for our correct understanding of the three countries to the euro market,avoiding imported financial risks and promoting the rational development of financial marketsThis article selects USD,CNY and the JPY against the euro EUR as the research object,from the aspects of volatility,yield and the analysis of the tail those three aspects,fully exploring the correlation of money market of China,America and Japan.Finally,we give explanation combined with the corresponding theory.First of all,through descriptive analysis of data,this paper finds that the yields have the characteristics of the "rush" and "fat tails",and reject the null hypothesis of normal distribution.According to AIC criteria we find that using GARCH(1,1)-t model fitting marginal distribution of each yield effect is the best way.We find that the setting is correct through testing the marginal distribution model by K-S inspection for further inspection.Model parameters results suggest that the United States,China,Japan currency yield fluctuations in their early to its variance is larger and the influence of the fluctuation of sustainability is stronger.Secondly,in view of the correlation,this paper uses copulas model as the means to analyze.This article embarks from the survival copulas function,establishing a mixed model which connects a survival copula and its function.Then we describe the related structures of currency market using this mixed model.On the way of choosing a kind of survival Achimedean copula,this paper uses a kind of conditional Kendall's tau.To assure the copula we have selected is the best and the mixed copula has some advantages that other copulas do not have,we analyze and test exchange rates by using Gumbel copula,Clayton copula,Frank copula and other kinds of mixed Copulas.We found that in this paper,the mixed copula model is the best mean to describe the relationship of exchange rate markets.Based on that,we find that there are different dependency structures between different markets.At last,we study the VaR and CVaR value of exchange rates.In order to improve the computation efficiency and reflect copulas theory in calculating the application of portfolio risk,in this paper,a method of use:the Monte Carlo simulations of VaR and CVaR based on copulas,we get the portfolio VaR and CVaR value,and avoid the complex numerical integration as well as the uncorrect assumption.As you can see,the CVaR value is larger than VaR value no matter on which confidence level,suggesting that CVaR is much more conservative than VaR,more can reflect the actual risk investment portfolio.In addition,no matter for which the confidence level of the VaR and CVaR,CNY/EUR-EUR/JPY value is the smallest,so for investors,rather than investment-CNY/EUR USD/EUR and USD/EUREUR/JPY,investment of CNY/EUR-EUR/JPY can give investors with less risk and more earnings.