Systemic importance of Chinese financial institutions based on the QC-ISAM-ARMA temporal network with coupling

In this paper, we propose a novel QC-ISAM-ARMA temporal network with coupling² to investigate the systemic importance of 30 listed Chinese financial institutions under different market cycles and conditions. This network can better capture the complex correlation including non-linearity, periodicity, and time variability among financial institutions and facilitate to find the optimal model by an adjustment parameter. The empirical study concludes that the newly constructed network demonstrates superior performance in identifying systemic importance. Furthermore, the systemic importance of Chinese financial institutions varies across market cycles and conditions. Banks consistently hold higher systemic importance, while insurance institutions show increased sensitivity to economic cycles and the systemic importance of securities firms increases significantly under stable market. Specially, further comparative study about banks means that the character of “too connected to fail" cannot be ignored and dynamic supervision is indeed necessary. This research offers new perspectives and constructive insights for analyzing the systemic importance of financial institutions. Additionally, the proposed new network model can be applied to assess interdependence in other domains beyond the financial sector.