Global volatility connectedness and the determinants: evidence from multilayer networks

This paper proposes the multilayer network, including lagged and contemporaneous layers, to examine the volatility connectedness among global stock markets. We compare the topology structures of global lagged and contemporaneous spillovers based on the static-sample and dynamic-sample. Furthermore, we investigate the determinants of global volatility spillovers through the Shapley decomposition approach. Our results show that contemporaneous spillovers are stronger than lagged spillovers between global stock markets. Meanwhile, we observe that lagged and contemporaneous networks share similar risk propagation mechanisms during financial stress. Furthermore, we note that the European and US markets are more active in contemporaneous spillovers, and Singapore plays the transmitter role in global volatility connectedness. The global determinant analysis shows that the financial stress index and economic policy uncertainty are drivers of global lagged spillovers, while the term spread contributes to global contemporaneous spillovers. Finally, we find that cross-border capital inflows exacerbate vulnerabilities in developed markets, and exchange rate movements are the major cause of financial instability in developing markets.