Uncover the response of the U.S grain commodity market on El Niño–Southern Oscillation

Considering the sensibility of grain on the weather, this paper combines the GARCH-MIDAS model with the STL decomposition method to examine how the changes of El Niño–Southern Oscillation (ENSO) can help to forecast the futures markets price volatilities of three major U.S. grains, namely, corn, wheat, and soybean. We first explore whether the forecasts of grain can be improved by including the ENSO information. Empirical results show that ENSO plays a significantly important role in both in-sample estimation and out-of-sample prediction. To further accurately investigate the impact of ENSO on grain futures price volatility, we thus decompose the ENSO into the trend, seasonality, and remaining components with the STL decomposition. By constructing the univariate models, the empirical results show that the model based on seasonal components outperforms others. For multivariate models, the models involving all components show the best performance in statistics and economics. Thus, this study may provide a novel perspective for improving the ability to deal with the effects of climate change on grain commodities.