Financial frictions versus physical scarcity: state-dependent drivers of oil prices

This study investigates the state-dependent drivers of crude oil prices employing a Structural Vector Quantile Autoregression framework. This approach explicitly addresses the aggregation bias inherent in standard linear models, which obscures regime-dependent risk transmission by assuming constant elasticities. The empirical analysis reveals a distinct dual asymmetry in market architecture. In the extreme lower tail, crude oil functions primarily as a financialized asset dominated by liquidity shocks that propagate liquidity spirals, consistent with the Financial Accelerator hypothesis. Conversely, in the extreme upper tail, crude oil behaves as a physical commodity driven by geoeconomic fragmentation and binding spare capacity constraints, providing evidence for a highly convex short-run supply curve. Accordingly, a state-contingent policy architecture is proposed: targeted liquidity provision is essential to mitigate fire-sales during crashes, whereas strategic reserve releases are effective primarily during scarcity-driven spikes.