Speaker: Hilde Bjørnland, BI Norwegian Business School

Link to personal website: www.bjornland.no

Title: The Anatomy of Asymmetric Oil Supply Shocks

Abstract

We study how oil supply shocks propagate across the conditional distribution of macroeconomic outcomes. To do so we develop a structural Quantile Vector Autoregression (QVAR) that embeds external instruments and sign restrictions within a Bayesian factor model estimated at multiple quantiles, yielding quantile-specific impulse responses and historical decompositions. We apply the model to monthly U.S. and global data from 1974 to 2019, identifying oil supply news shocks with two independently constructed instruments. The results reveal sharp distributional asymmetry: normalizing all shocks to a 10% oil price increase, U.S. industrial production expands at the 10^thpercentile but initially contracts at the 90^thpercentile. Because the normalization equates the oil price impact, these differences reflect distinct propagation mechanisms rather than differences in shock magnitude. An identification anatomy shows that the two instruments recover closely related oil supply shocks, with the tightest agreement at the median and somewhat wider dispersion at the tails. Historical decompositions pin the distributional heterogeneity to specific episodes: the pre-crisis oil price run-up of 2007–08, seen through the left tail, appears driven by favorable demand conditions, while the right tail attributes the subsequent output decline to a supply constraint.

Host: Niels Haldrup

Organisers: Stefán Guðmundsson and Chen Huang