Research on the tail risk spillover between shanghai and shenzhen stock markets based on MODWT and time-varying Clayton Copula

This paper studies the tail risk spillover between Shanghai and Shenzhen stock markets based on maximal overlap discrete wavelet transform (MODWT) and time-varying Clayton Copula from the viewpoints of time and frequency simultaneously. The return rate series is decomposed into three scales. At each level scale, the propagation direction of risk spillover is judged by Granger causality test and the tail strength is measured based on time-varying Clayton Copula in this paper. The research shows there is unidirectional risk spillover at d1 scale. There exists bidirectional risk spillover at d2, d3 and a3 scale. d1 and d2 scales bring the risk spillover with higher power and larger tail strength. d3 scale brings the risk spillover with lower power, larger fluctuation range and relatively smaller tail strength. While a3 scale brings risk spillover with lower power and relatively larger tail strength.