A general partial risk-return relation is derived based on return decomposition to allowing for the effect of time-varying skewness and kurtosis on the risk-return trade-off. Empirically estimated for 12 international financial markets, the proposed risk-return trade-off is significantly positive even after controlling for time-varying higher moments. Moreover, the stochastic dominance test reveals that modeling time-varying skewness significantly lowers the level of the risk-return trade-off. More importantly, the empirical evidence shows that the risk-return trade-off is countercyclical in the U.S. markets, consistent with the theoretical habit-formation model of Campbell and Cochrane (1999), whereas the risk-return trade-offs in European and emerging markets appear to be procyclical over a 12-month horizon, but countercyclical for a shorter horizon of 3 months. Finally, common macroeconomic variables can significantly explain risk-return trade-off dynamics.

We utilize a fundamentals-based component volatility model to forecast the short-run volatility of exchange rate changes using monetary fundamentals quoted at different frequencies. Specifically, we allow the component volatility model to distinguish short-run exchange rate fluctuations from long-run movements that are directly linked to monetary fundamentals. Relative to more traditional time series volatility models, we find significant improvements in the ability to forecast the daily volatility of exchange rate changes by incorporating the monthly monetary fundamentals’ volatilities as predictors into the component volatility model. In the utility-based comparisons, we find that an investor is willing to pay a positive annual management fee of 5.72% on average to switch from the benchmark model to the fundamentals-based models. Of these models, the model with the symmetric and homogeneous Taylor rule and interest rate smoothing obtains the highest positive annual management fee.