Energy supply shocks' nonlinearities on output and prices

We use a Bayesian Threshold Vector Autoregression model identified through sign\r\n and narrative restrictions to uncover non-linearities in the propagation of energy supply\r\n shocks. We find that the transmission of energy supply shocks on consumer prices\r\n is stronger in high-inflation regimes, supporting nonlinear models. The faster pass\r\nthorough of energy supply shocks to consumer prices (excl. energy) cushions the drop\r\n in output in the short term. Energy supply shocks have a stronger impact on output\r\n in the medium-term with manufacturing being more adversely affected than GDP.\r\n Large energy supply shocks shift the economy to another state but after two and\r\n half years the mean-reversion to lower inflation implies a more moderate transmission\r\n mechanism, highlighting the importance of state-dependent impulse responses. The\r\n energy supply shocks between July 2021 and June 2022 are massive amounting to 3.9\r\n standard deviations on average each month.