The use of conventional augmented CAPM specification in estimating the exchange rate exposure may result in less reliable estimates for, at least, two reasons. First, it does not take into account a few important stylized facts associated with financial time series. Second, one cannot estimate the total impact of the exchange rate changes on stock returns as a single coefficient with it and for this reason it does not help us analyze the reinforcing or offsetting interactions between direct and indirect exchange rate exposure effects. In this paper, we suggest an orthogonalized GJR-GARCH-t version of augmented CAPM that simultaneously addresses the above issues. Our findings have important implications for hedging and investment decision making.