Abstract:
We use variants of the Heterogeneous Autoregressive Realized Volatility (HAR-RV) model to examine the out-of-sample predictive value of climate-risk factors for the realized volatility of gold price returns as well as the realized volatility of for other metal price returns (Copper, Palladium, Platinum, Silver). We estimate the HAR-RV models using not only ordinary least squares, but also we use three different popular shrinkage estimators. Our main finding is that climate-risk factors improve the accuracy of out-of-sample forecasts prices at a monthly and, in some cases, also at a weekly forecast horizon.
