Exploring dynamic extreme dependence of oil and agricultural markets

Abstract

This study examines the dynamic extreme dependence between oil markets and a set of non-exploited agricultural assets (Barley, coffee, maize, rice, sorghum, sugar, wheat) prices using 1990–2023 monthly prices data. Our time-varying copula models allow us to investigate nonlinear and asymmetric dependence between pairwise assets. We find evidence of positive exceedance correlation at small and medium quantiles/thresholds and negative exceedance correlation for quantiles higher than 0.6 for five out of seven food assets. Findings also indicate a tent-shaped quantile dependence for all food assets and demonstrate a time variation in the association path between oil and food asset returns. Important policy implications are derived from empirical findings.