Commodity Price Persistency and Dynamics: A Smooth Transition Regression Approach

Abstract

AbstractIn this study, a nonlinear two-regime Smooth transition Regression (STR) approach is applied to examine the behavior of three commodity prices (energy, non-energy, and precious metals). In particular, the study seeks to reveal the main factors that determine the commodity price changes. The results are compared between logistic versus exponential transition functions. The data used is monthly data from 1995 to 2018. The result shows that the transition in energy commodity and precious metal sectors are more sensitive and reacts immediately to the threshold variable of its history lag path. Meanwhile, the transition in the non-energy commodity sector fluctuates the least. Besides, crude oil prices influence the prices of the commodities for all three sectors, and the impact is the highest on the energy commodity sector. Consumer price inflation is one of the determinants for commodity price inflation in non-energy commodity and precious metal sectors, where the effect is higher in the latter. In addition, gross domestic product and the United States central bank policy rate affect the commodity price inflation of the non-energy commodity and precious metal sectors, respectively. Furthermore, the dynamic effects in energy commodity and precious metal sectors are more persistent, which indicates a longer time for the inflation to smooth down. Conversely, the less persistency of the dynamic effects in the non-energy commodity sector shows that inflation takes a shorter time to smooth down. The findings might provide useful information to the policymaker in policy plan/decision. Early preventive action can be taken to reduce the negative shocks due to commodity price fluctuations.KeywordsCommodity priceNonlinearitySmooth transitionPrice persistencyThreshold variable