The identification of systemically important financial institutions from the aspect of “Too Connected to Fail”

This paper investigates the identification of Systemically Important Financial Institutions (SIFIs) from a “Too Connected to Fail” perspective by applying the K-shell algorithm to a dynamic tail-risk contagion network. The network is constructed for China’s financial system using the FARM-Selection method and CoVaR measure across the COVID-19 pandemic. We find that the pandemic elevated systemic risk and increased the number of SIFIs, which were securities companies, fintech firms, and city commercial banks during the crisis. This reflects the critical role of market-making, digital payment channels, and regional credit intermediation under conditions of heightened volatility and policy support. In the post-pandemic period, SIFIs are investment institutions, insurance companies, trust firms, and fintech firms, indicating a structural reorientation of the financial system toward long-term capital allocation, risk transfer, and embedded digital finance. This shift underscores the tail-risk contagion networks adapting from crisis-driven liquidity reliance to longer-term strategic intermediation. Furthermore, SIFIs show a negative relationship with market values, suggesting that highly tail-risk connected institutions are not necessarily the largest in market capitalization. Robustness is confirmed using the PageRank algorithm. The paper provides a network-based monitoring framework and policy insights for mitigating systemic risk in China, highlighting regulating connectivity-driven systemic importance.