The Research On Extreme Risk Spillover In Global Crude Oil Market

In recent years,the financialization degree of crude oil market is deepening,and the behavior of crude oil price is becoming more complex.With the frequent occurrence of extreme risk events and the sharp fluctuation of crude oil prices,the need for extreme risk management and risk warning in crude oil market is becoming more and more urgent.The existing literature focuses on the mean spillover and volatility spillover of crude oil market,while few studies on the extreme risk spillover of crude oil market.This paper studies extreme risk spillovers between global crude oil markets.On the one hand,it supplements the research blank in the field of extreme risk spillovers within the crude oil market.On the other hand,it provides theoretical support for government departments,financial institutions,oil-related enterprises and investors to formulate policies and manage risks.In this paper,seven crude oil benchmark prices,including WTI,Brent,OPEC,Tapis,Dubai,and Minas,are taken as the research object.The sample interval is from December 31,2002 to November 18,2022.This paper studies the static and dynamic extreme risk spillovers in the global crude oil market.First,the CAVia R method was used to measure extreme risk levels in seven crude oil markets,namely 1%Va R and 5%Va R.Secondly,the spillover index model proposed by Diebold and Yilmaz(2012,2014)and the Granger causality in risk test model proposed by Hong et al.(2009)are respectively used to study the extreme risk spillover level of the global crude oil market from the perspectives of generalized prediction error variance decomposition and lead lag relationship.Then the time variability of the extreme risk spillover effect is studied through the method of rolling window.Finally,the DCC-GARCH model is used to measure the dynamic correlation coefficient between VIX,EPU and total extreme risk spillovers,which provides a possible explanation for the change of extreme risk spillover in crude oil market.The results show that:(1)Brent,OPEC and WTI are the price benchmarks for global crude oil,but Brent and OPEC are getting stronger while WTI is getting weaker.(2)The spillover effect of extreme risks between global crude oil markets is time-varying.The impact of extreme risk events will deepen the integration of the crude oil market and have a long-term impact.(3)In general,when market panic increases and economic policy uncertainty decreases,the contribution of individual crude oil market extreme risk to the extreme risk of the crude oil system increases,and the lead lag relationship of the extreme risk of the pair crude oil market in the system decreases.(4)Both the spillover index model and the Granger causality in risk test model can measure the extreme risk spillover effect,but their internal mechanisms and influencing mechanisms are different.The main innovations of this paper are as follows:(1)This paper studies the spillover effect between global crude oil markets from the perspective of extreme risk,which complements the research gap in related fields.(2)In this paper,CAVia R-spillover index method and CAVia R-Granger causality in risk method are used to study the spillover effect of extreme risks from different perspectives and form a contrast,which enriches the application of extreme risk spillover research methods.(3)Going further than previous studies,this paper attempts to analyze the extreme risk spillover mechanism between global crude oil markets using VIX and EPU.