The spillover and comovement of downside and upside tail risks among crude oil futures markets

Abstract

Combining efficient transfer entropy and DY spillover indices, this study first explores the contagion effects of downside and upside tail risks among INE, Brent, and WTI. Then we utilize wavelet coherence to capture the comovement characteristics of tail risks at multi-time scales. Furthermore, by mapping wavelet coherence information into comovement networks, the diversity and regularity of mode conversions of tail interdependency are investigated in depth. The results show that INE is reactively susceptible to the tail risk spillovers from Brent and WTI, and in terms of the magnitude of spillover to INE, Brent surpasses WTI prominently. The coherence between the tail risks of INE and Brent or WTI is much weaker than that of the Brent-WTI pair, especially in the short term or in the upside tail case. The evolution of downside tail risk comovement is more difficult to stabilize. A few comovement modes show a significantly higher probability of occurrence than others, and they also tend to link each other. Of special note are some comovement modes with low probability of occurrence yet behave as the key “bridges” in the mode transformation process.