Volatility dynamics of agricultural futures markets under uncertainties

Abstract

The objective of this study is to examine the effect of various uncertainty measures on the realized volatility of agricultural futures markets. In doing so, we use a range of uncertainty indicators in our analysis to investigate whether news-based uncertainty measures (e.g., geopolitical risk and economic policy uncertainty) have better predictive contents than the market-based uncertainty measures (e.g., crude oil volatility index, the US equity market VIX and exchange rate VIX). This comparison is important given that employing both measures has some specific benefits. Methodologically, we consider the application of the LASSO (least absolute shrinkage and selection operator) method as well as the heterogenous autoregressive (HAR) process. The in-sample estimates indicate that among the various news-based and market-based risk measures the latter provide better forecasts for the realized volatility of agricultural futures markets. The out-of-sample forecasts also confirm the same with the LASSO method outperforming the HAR process.