Multilayer interbank networks and systemic risk propagation: Evidence from China

This paper proposes a multilayer bank network structure based on three channels: interbank lending, cross-shareholding and holding common assets. By investigating the dynamics of systemic risk contagion across networks, we extend the DebtRank model to a three-layer network. The empirical results suggest that the multilayer structure of the banking system has a non-linear superimposed effect on systemic losses, with interbank lending and holdings common asset being the main sources of risk contagion. We also simulate systemic shocks such as individual bank shocks, industry shocks, and macroeconomic shocks with the aim of identifying systemically important banks and industries. The results found that China’s four largest state-owned commercial banks are the systemically important banks, and the banks’ loan balances in manufacturing, transportation and other industries and personal credit should be the focus of attention. This study can feedback more correlations in the banking system, emphasizing the importance of research on multilayer network contagion mechanisms, and the content and findings of the study can provide empirical corroboration for systemic risk regulation.