The Application Of Copula Function In Foreign Exchange Risk Management

At present, the world generally applies the floating exchange rate system and the ex-change rates between the Dollar, the Yen. the Pound and other major currencies are ups and downs from time to time, which lead to the difficulties of settlements in international creditor's rights and debts and generate foreign exchange risk. Our country is among the floating exchange rate system. At the same time foreign exchange rate risk has seriously affected our country's balance of international payments and the economic benefit of en-terprise, especially when our country has rapidly developed nowadays. Therefore, it is vitally significant to manage exchange rate risk. However risk measurement requires a better understanding of the volatility and correlations of asset returns. Traditional meth-ods like multivariate normal distribution are widely used in the aggregation of dependent risks. However, it is proved that such approaches are inflexible and inappropriate.Due to their advantages in measuring correlations copula functions have been widely applied to solve many financial problems, such as risk management, optimal asset alloca-tion and derivative pricing, etc. On the other hand, exchange rate risk, as an important part of market risks, receives a lot of scholars'attention and there are many researches related to it.This paper explores the dependence structure between three exchange rates(EUR/CNY, JPY/CNY and GBP/CNY). In particular, based on survival copula function we propose a dynamic mixed copula approach, which is able to capture the time-varying tail dependence coefficients. First of all, the GARCH(1,1)-t or IGARCH(1,1)-t model is used to examine the marginals, while a couple of copula functions are employed to analyze the joint distribution. In this pairwise analysis, both constant and time-varying conditional dependence parameters are estimated by a two-step maximum likelihood method. It is found that conditional mixed Gumbel copula model always perform better than others in depicting correlations. Based on this model, dynamic tail dependence structure is studied. The findings indicate the existence of asymmetric tail dependence coefficient. Although previous research extensively reports that the lower tail dependence between exchange rates tends to be higher than the upper tail dependence, we find a counterexample where the upper tail dependence is much higher than the lower tail dependence in some short periods. Finally, we discuss the time-varying VaR and CVaR of exchange rate portfolio and find that our model can well capture the most serious period of financial crisis.