Resumen
The purpose of this paper is to analyze the JPY/USD and the CAD/USD forward exchange premiums by adopting the ARCH/GARCH modeling, given its descriptive and predictive advantages. We estimate a symmetric linear model by taking into account the effect of the mean and the conditional variance in a univariate framework. The estimation results show that the shocks induced by the volatility are very persistent, so that forecasts of the conditional variance converge very slowly to the regular state. This shows the relevance of the heteroscedastic GARCH-in-Mean model in the estimation of the forward premium on the international foreign exchange markets. This result may have several important implications for monetary policy in most developing countries.