A Statistical Analysis of Global Economies Using Time Varying Copulas

Abstract

The application of time varying copulas has become popular in recent years. Here, we illustrate an application involving stock indices of ten major economies covering all of the six continents. The dependence among them and its variation with respect to time are modeled using ten different copulas. The Gaussian copula is found to give the best fit. Predictions are given in terms of correlations and value at risk.