Regime-dependent wheat price volatilities

We estimate the volatility dynamics of wheat prices across the United States, Ukraine, and China between 2013 and 2023. To capture differing volatility regimes caused by various factors including political unrest, we consider a mixture of two trivariate GARCH processes. We find distinctive features between the mixtures: component 1 is characterized by higher volatility and greater spillover effects whereas component 2 is characterized by lower and more persistent volatility. By modelling volatilities as mixtures, we are able to consider what factors drive component membership. We find the probability of the higher volatility regime increased by 10% under Trump’s presidency and by 40% during the Russia-Ukraine war.