Quantile-Based Spillover Network Analysis of Financial Institutions in Chinese Mainland and Hong Kong

Strong connection among financial institutions provides more possible channels for information spillovers under different market conditions. Therefore, this paper attempts to comprehensively capture the spillover effects among financial institutions in Chinese mainland and Hong Kong by constructing the spillover networks based on the Quantile Vector AutoRegression (QVAR) model. By taking the stock prices as the representatives of financial institutions, we find that the spillover effects under extremely positive and negative conditions are larger than those under normal conditions. The magnitudes of spillovers are dynamic and show a dramatically upward trend when the market encounters extreme shocks, such as the Sino-US trade friction and the COVID-19 pandemic. The directional spillovers of the financial sector and its subsectors from Chinese mainland to Hong Kong are greater than those from Hong Kong to Chinese mainland, and banking contributes more to the spillovers. In addition, most institutions in Chinese mainland are spillover transmitters, and whether the institutions are net receivers or not varies over shock sizes. These results are essential for risk management in financial institutions in the opening financial markets.