Realized skewness of oil price returns and the short-term predictability for exchange rate
研究发现油价收益的实现偏度能有效预测九国汇率，优于随机游走模型，其预测力与通胀、行为偏差和风险传导有关。
When predicting exchange rate returns, existing models find it difficult to defeat naive random walk models. We propose the realized skewness of oil returns to predict exchange rates for nine countries. The results show that it is statistically and economically more effective than the random walk model. Moreover, realized skewness can offer complementary predictive information about what is previously known about oil prices and oil volatility. It is noteworthy that the superior predictability of realized skewness is linked to inflation rates, behavioral bias, and risk transmission channels. Furthermore, the predictive power of realized skewness for exchange rates stems from the jump and drift components.