Title: Highly frequent oil price shocks
Authors: Fabrizio Venditti - ECB (Germany) [presenting]
Giovanni Veronese - Banca d Italia (Italy)
Fabrizio Venditti - Queen Mary University of London (United Kingdom)
Abstract: A structural vector autoregression is constructed which uses information from financial markets to decompose daily changes in the price of crude oil into three structural drivers, namely a risk on/off shock, a global demand shock and an oil supply shock. We propose a novel identification strategy that rests on the use of instrumental variables blended with sign and narrative restrictions. Additional identification assumptions ensure that the response of macro variables to our daily shocks is consistent with that obtained in monthly models commonly used in the macro literature. We find that, while demand shocks were mostly responsible for the increase in the price of oil before the crisis, and for its collapse during the crisis, oil supply shocks had a decisive role in driving down the price of crude oil in the 2014-2016 slump, as well as in its recovery in 2017-2018.