Financial risk propagation between Chinese and American stock markets based on multilayer networks

Stock networks, which are constructed from stock price time series, are useful tools for analyzing complex behaviors in stock markets. Following former researches, the epidemic model has been usually used to detect dynamic characteristics in a stock price complex systems. Recently, multilayer networks have been demonstrated well when working on heterogeneous nodes rather than integrated networks. In this paper, we proposed a two-layer SIR propagation model with an infective medium to analyze the spread of financial shocks. In consideration of strict financial regulation in the A shares, the model assumed that capital cannot flow directly between layers but through the Hong Kong stock market. By applying the model to constituent stocks included in three prominent indices, Standard & Poor 500, Shanghai and Shenzhen 300, and Hang Seng(medium), we established a two-layer Granger networks. Betweenness showed that the Hong Kong stock market had a promoting transition function of financial shocks between the US stock markets and the mainland China stock markets. In addition, with a big basic reproduction number, stock markets system appeared to be vulnerable during extreme financial shock such as the outbreak of COVID-19 epidemic and the meltdown of stock markets. Furthermore, sensitivity analysis and the spreading simulation indicated that the US stock markets were much more robust to financial shocks than the mainland China stock markets.