Abstract: Based on the overall data of inter-bank assets and inter-bank liabilities of China’s commercial banks， the inter-bank borrowing-lending correlation network is inferred indirectly by the maximum entropy method， and the distress propagation process under a single shock or common shocks is investigated based on the DebtRank algorithm. The systemic impact and vulnerability of banks and their influencing factors are also investigated. The empirical results show that under certain circumstances， the negative results brought by bank distress are more severe than those brought by bank default. China’s commercial banks are not located in the area of “high vulnerability/strong impact”， indicating that China’s banking system is relatively safe. In terms of influencing factors， the average return on assets has a significant negative effect on banks’ systemic impact， while the loan reserve ratio， the tier-one leverage capital ratio and the inter-bank lending volume have significant positive effect on banks’ systemic impact. The tier-one leverage capital ratio and asset size have significant negative effect on the systemic vulnerability and the inter-bank lending volume has significant positive effect on the systemic vulnerability. The results not only help banks to understand their own situations， but also provide a basis for financial regulation.

Key words: distress propagation; borrowing-lending correlation network; DebtRank; systemic impact; systemic vulnerability