Decomposed Oil Shocks and State-Dependent Spillovers Across Commodity Markets: A QVAR-Based Quantile Connectedness Approach

Abstract This study investigates how structurally distinct oil price shocks interact with agricultural, precious, and industrial commodity markets across different market states and investment horizons. We combine the demand-supply-risk decomposition of oil shocks with a QVAR-based quantile connectedness framework and wavelet quantile correlation. The results reveal a pronounced U-shaped connectedness profile across the conditional distribution. System-wide connectedness is lowest around the median and rises sharply in both tails, and rolling estimates show that this distributional asymmetry persists through time. Directional measures further show that the oil shock components are not uniformly net transmitters. Supply and risk shocks are predominantly net receivers in the tails, while demand shocks switch roles across states. Copper is the most persistent net transmitter among the commodity markets. The wavelet evidence distinguishes the shocks along the horizon dimension. Demand shocks generate the broadest positive and persistent dependence, risk shocks are associated with strong negative long-horizon dependence in industrial metals, and supply shocks display weaker and more heterogeneous dependence. The findings show that oil-commodity interdependence depends jointly on shock origin, market state, and investment horizon.