Time-varying correlation between agricultural commodity and energy price dynamics with Bayesian multivariate DCC-GARCH models

By Yegnanew A. Shiferaw. Physica A: Statistical Mechanics and its Applications, 2019

51 citations recorded by the source corpus.

Abstract This article investigates the dependence structure between the agricultural commodity prices (white maize, yellow maize, wheat, sunflower and soya) and energy prices (oil, natural gas and coal) dynamics of South Africa based on the Bayesian multivariate GARCH (MGARCH) model with skewness and heavy tails. A computationally intensive Markov chain Monte Carlo (MCMC) algorithm was adopted and implemented for both parameter estimation and model comparison. Based on the information criteria, the Bayesian DCC-MGARCH model with the error skewed-mvt distribution assumption performed better than other competitive methods. Moreover, the correlation between the agricultural commodity and energy price returns is dynamic (time-varying) in South Africa, indicating that the prices of agricultural commodities and energy prices exhibit strong co-movement. The findings have significant implications in the domain of agricultural commodity policy and financial sector.

Subfields: Economics and Econometrics, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/3a1678c6629cd1f9cc9098c0577877d86588e1b8