Research On The Dependence Of Exchange Rate Based On Copulas

Exchange rate as the relative price between the currencies in two countries, it affects the relative prices of the commodities inevitably. First, from the respective of macroeconomics, the fluctuation of relative prices made by economic globalization will practically influence the foreign exchange market, and then involve, magnify, and spread to other markets.Second, from the respective of microeconomics, the fluctuation of exchange rate will change the condition of international trade, the structure of export commodities and also the upgrading of industrial structures. These bring the uncertainty of income and risk of international trade and more cost of foreign trade, which means that higher requirements of risk management and efficiency from foreign market are necessary. At present, the relationship between foreign market and exchange assets becomes more and more profound, either non-linearity, asymmetrical distribution or tail dependence, which shows that linearanalysis cannot reflect the related information of the foreign market exactly.So here we introduce Copula theory to the analysis related to exchange rate. On the one hand, this text concludes the researches on exchange rate analysis based on Copula; on the other hand, it generally introduces the Copula theory and also the Copula modeling. Then, by the comparative analysis of parametric and non-parametric estimation derived from Copula function, and the selection of optimal Copula function as well, we get the running conditions of these methods by means of Monte Carlo method.After that, by the use of empirical analysis, we get the degree and mode of correlation between RMB exchange rate and EUR, JPY, GBP or HKD. During the construction of marginal distribution function, we use the GPD theory and Empirical Distribution of the Extreme Value theory to fit the marginal distribution function, choose the most fitting Copula function and get the correlation of exchange rates by the jointed distribution based on the mail Copula functions. The result shows that RMB exchange rate is positively correlated but weakly with the four foreign exchange rates, and the correlation between the four nondollar foreign markets is relatively high.At last it comes the conclusion of the whole context, and the prospect of problems raised in the research.