Extreme dependence, connectedness, and causality between US sector stocks and oil shocks

Abstract

We scrutinize the possible impacts of disentangled oil price shocks on several equity index returns across multiple quantiles and throughout a sample period spanning from January 2014 to May 2023. The findings of the quantile coherency test demonstrate that the effect varies across sectors and quantiles, demonstrating that different industries react differently to crude oil shocks and that the impact is higher in the long run. The demand shocks are more powerful than the supply and risk shocks. Results show evidence of asymmetric impacts for all but the oil & gas index. The safe-haven property appears mostly against demand and supply shocks and is only visible in the short term. The pandemic outbreak has significantly shifted the direction and magnitude of total and net return connectedness over time. Across quantiles, the oil supply and risk shocks are the net receiver and transmitter, respectively, whereas the demand shocks appear to be the net receiver at extreme quantiles but act as the transmitter when markets are normal. Most sectors are net transmitters at extreme quantiles, while almost all are highly exposed to risk transmission at the median quantile. Regardless of the role they play in risk transmission, demand shocks tend to be the primary risk transmitters for the oil and gas index, while risk shocks appear to decisively spill risk to both equities and supply shocks across quantiles. Three oil shocks improve the prediction of equities during only major events, and the impact is stronger at median quantiles.