New paths of the time-varying Granger-causality test under extreme shocks: An analysis of geopolitical risk and crude oil markets

Motivated by the significant market disruptions caused by recent geopolitical crises and natural disasters, we develop an extreme time-varying Granger-causality test. This approach integrates the recursive evolution window algorithm into the Granger-causality test framework under a data-driven environment, which enables us to uncover dynamic causal differences between extreme and nonextreme financial conditions. Specifically, we apply this novel method to comprehensively investigate the dynamic impact of geopolitical risk on the crude oil market. The empirical analysis not only reveals that extreme geopolitical risk shocks drive the price of the crude oil market in a dynamic manner but also reveals the asymmetric and nonlinear characteristics of these impacts. The degrees and durations of such impacts vary significantly across different periods of geopolitical risk. To ensure the reliability of our findings, we conduct a series of robustness tests, and all the considerations verify the stability of the novel method. Thus, our work contributes to a better understanding of the real-time impact of geopolitical risk on the crude oil market, especially under extreme shock scenarios. The insights provided herein may offer valuable reference for decision-makers to formulate energy-related policies, devise business strategies and make investment decisions.