Dynamics and causality of oil price shocks on commodities: Quantile-on-quantile and causality-in-quantiles methods

Abstract We aimed to use rolling quantile regression, quantile-on-quantile, and causality-in-quantiles methods to document the dynamics and causal relationships between oil price shocks and commodities using a dataset from January 1992 to June 2019. The empirical results indicated that the coefficients of the rolling window quantile regression varied over time periods and were most significant for oil supply shock effects on commodities. Furthermore, the causality results showed that oil shocks can provide some predictability for commodities in some quantiles. The quantile-on-quantile analysis revealed that the effects of oil shocks on commodities varied across quantiles and were heterogeneous and asymmetrical. The impact of supply shock on commodities was mainly negative, whereas for aggregate demand shock, the effects were positive in bearish markets. For oil-specific demand shock, a symmetric dependence structure was detected.