Nonlinear tail dependence between energy and agricultural commodities

This paper examines the tail dependence structure between energy commodities (Brent oil, natural gas and gasoline) and agricultural commodities (wheat, soybean, corn, cotton, sugar, rice, oat, coffee and cocoa) from 01.06.2017 to 09.06.2023, spanning periods before, during and after Covid-19 pandemic. We employ the tail-restricted integrated regression function (IRF), a novel approach for analyzing nonlinear tail dependence, as it offers further insights into tail events by considering a continuum of quantiles, rather than focusing on a single quantile. The results reveal significant and persistent lower and upper tail dependence across all commodity pairs throughout each period, indicating asymmetric risk transmissions from energy commodities to agricultural commodities. Additionally, the findings are corroborated using cross-quantilogram analysis and nonparametric tests for Granger causality in distribution. •We study the nonlinear tail dependence from energy commodities to agriculture commodities.•We consider 3 main energy commodities as conditioning variables and 9 main agricultural commodities as dependent variables.•We conduct our analysis for the pre, during, and post Covid-19 periods from June 01, 2017, to September 09, 2023.•There is a significant and persistent both upper and lower tail dependence for all commodity pairs for each period.•There is an idiosyncratic relation for each commodity pair during shocks, rather than a certain dependence pattern.