Analysis of a Basket of Currencies Portfolio Based on Copula-GARCH Method

This paper presents a new quantitative approach for finding the optimal weights of currency basket, with which the exchange rate risk of the basket is minimized. The article employs t-GARCH and Copula model for better measuring the dependency of currencies in basket. “Maximization by Parts in Likelihood” method is also used to improve the precision of parameters estimation. Our model is applied to daily returns of four currencies in a basket. Taking VaR as the measurement of risk, the empirical result suggests a larger weight should be optimally accorded to the dollar though there is a nonnegligible role for the Euro and Japanese yen.