Oil Shock, Policy Uncertainty and Stock Return : An Analysis Based on the Tndp Model with Dominant Sectors

Based on an extension of the time-varying network dependence panel (TNDP) model proposed by Hauzenberger and Pfarrhofer (2021) with the identification of dominant units suggested by Pesaran and Yang (2020), this paper re-examines the impacts of oil shock and policy uncertainty on stock returns in a network framework with heterogeneity both over time and across sections. Using a sample of 38 sectors in China, we provide rich explanations to three critical questions in understanding the ambiguous influence of oil shocks on stock returns. First, after controlling the moderation of EPU, oil shocks directly stimulate stock returns of upstream sectors but depress stock returns of downstream sectors. Furthermore, oil shocks also trigger more substantial complementary spillovers among sector stock returns. Second, EPU serves as a cushion counteracting oil price volatility, with a cross effect more active and more effective on the supply side and in response to direct oil shocks. Third, the supply-side production network highly concentrated in stable dominant sectors is governed mainly by direct shocks, while the demand-side production network with evolving dominant sectors is heavily subject to indirect spillovers. Last, detailed risk propagation paths in two notable risk events—the 2008 financial crisis and the 2018 trade war—are illustrated as examples