Media evidence and previous research have established that geopolitical risk is an important driver of crude oil price volatility. In this paper, we assess whether the importance of geopolitical uncertainty is also ”translated” into valuable predictive information for oil price volatility forecasts. To do so, we construct a ”beauty contest” where we assess the incremental predictive content of geopolitical risk against several other highly important uncertainty indicators, for forecasting horizon up to 22-days ahead. Initially, we use a HAR model which is augmented by each of the uncertainty indicators. Subsequently, we develop a Dynamic Model Averaging (DMA) methodology, where we assess whether the combination of all uncertainty indices (DMA-all), vis-a-vis a DMA model without the geopolitical uncertainty index, exhibits superior predictive performance. Our findings show that geopolitical uncertainty offers superior predictive information when combined with other uncertainty indicators. More importantly, we show that the inclusion of geopolitical uncertainty in a DMA framework generates superior trading profits and risk management measures’ predictions, in comparison with benchmark models, especially in longer-run horizons. Several implications are drawn from these results.

中文翻译：

---

原油波动预测导航：评估地缘政治风险的影响

媒体证据和先前的研究已经确定，地缘政治风险是原油价格波动的重要驱动因素。在本文中，我们评估了地缘政治不确定性的重要性是否也被“转化为”有价值的预测信息，用于预测油价波动。为此，我们构建了一个“选美比赛”，根据其他几个非常重要的不确定性指标评估地缘政治风险的增量预测内容，以预测未来 22 天的视野。最初，我们使用一个 HAR 模型，该模型由每个不确定性指标进行增强。随后，我们开发了一种动态模型平均 （DMA） 方法，在该方法中，我们评估所有不确定性指数 （DMA-all） 的组合相对于没有地缘政治不确定性指数的 DMA 模型是否表现出卓越的预测性能。我们的研究结果表明，当地缘政治不确定性与其他不确定性指标相结合时，它提供了卓越的预测信息。更重要的是，我们表明，与基准模型相比，将地缘政治不确定性纳入 DMA 框架会产生卓越的交易利润和风险管理措施的预测，尤其是在长期范围内。从这些结果中可以得出几个含义。