An Emperical Study About The Interaction Among The U.S., Japan And China's Exchange Rates Based On The Multivariate GARCH Model

World economy and financial globalization developed so rapidly that the world's major currencies are getting closer, and in the the real exchange rates of China, United States, and Japan also links each other more closely. An impact on the world economy will cause a country's exchange rate fluctuations, and the changes in individual will lead other relevant countries'exchange rates to change through the direct effects between the exchange rates, and the linkage effects brought by oil price volatility and spillover effects. If this volatility continues, it will spread to the whole exchange rate markets, and finally enlarge the volatility of the exchange rate market, and lead to the overall impact on economy. Therefore, different countries can not ignore the relationship between real exchange rates. In this paper, we used the behavior of the real effective exchange rate theory and the diagonal BEKK method to estimate RMB, USD and JPY MGARCH (Multivariate Generalized Autoregression Conditional Heteroskedasticity) model of equilibrium exchange rate. We also discussed the effect between each other. This study is divided into five parts, and the five parts is structured as follows.The first part introduces the research background and significance of this article, summarizes the current situation at home and abroad, and also gives the direction of this paper. The second part descriped, and elaborated the yuan, the dollar, the yen real effective exchange rate change trends from 1994 to 2009, and also summarizes the impact of exchange rate changes on the economy. The third part used MGARCH model to BEER equation of the three exchange rates. In this part, the article first introduced the theory of behavior equilibrium exchange rate BEER, given the equilibrium exchange rate equation; then introduced the theory of MGARCH model; Finally, this article displayed structural ideas and the main equation form of this BEER MGARCH model. MGARCH model is mainly constituted by two parts, the mean equation and conditional variance covariance equations. The mean equation reflects the direct impact of fundamental economy factors, the conditional variance covariance equations reflects the spillover effect of exchange rate fluctuations. Part IV first displayed the empirical results based on MGARCH model, then assessed the effectiveness and stability of the model, and finally gived a detailed analysis of those empirical results for the direct impact of exchange rate macroeconomic fundamentals, as well as the indirect impact between different exchange rates brought by oil price fluctuation and the fluctuation spillover effect. The fifth part is the conclusions and policy recommendations.The results show that, the foundamental Macroeconomy elements which effect RMB, USD and JPY possess their own characteristics. Fluctuations in international crude oil price will also cause linkage between different exchange rates change. In addition, there were significant volatility spillover effects between the three exchange rates.