Time Varying Correlation Mixed Normal Garch Model and Currency Futures Hedge Performances

Abstract

In this study, bivariate mixed normal time varying GARCH models are proposed to capture the skewness and kurtosis detected in both conditional and unconditional return distributions. They are compared with different standard bivariate GARCH models in terms of both the percentage variance reduction of the out-of-sample hedged portfolio and the statistical significance test of the performance improvements using Hansen’s (2001) SPA statistics. The models are applied to estimate time varying hedge ratios for the Canadian dollar and UK pound. The out-of-sample evaluation is carried out by comparing the hedged portfolio variances from all models over one to 30 days horizons. The empirical results demonstrate that the standard time varying model significantly outperforms the other competing models at shorter horizons. However, as the hedge horizon is extended to longer than 5 days, it is clearly evident that mixed normal GARCH models are the best at the usual significance level of 5%.