Interbank Dynamic Network Driven by EG-MOPSO Strategy: A Portrait-Style Simulation Approach

Financial risks exert a significant influence on the interbank market, contributing to systemic risk within the banking system. Simulating risk contagion is essential for the prevention of systemic risks. However, existing simulation methods encounter several challenges: the dynamic nature of interbank risk contagion, the effectiveness of interbank lending relationships, and biases inherent in risk contagion theories concerning loss measurement. In response to these challenges, this article establishes an autonomous lending mechanism for banks to endogenously form a dynamic interbank network and introduces the lending relationship generation method evolutionary game method based on multiobjective particle swarm optimization (EG-MOPSO), grounded in real scenarios. Furthermore, to accurately reflect risk contagion losses in practical contexts, this study presents a contagion method that integrates Eisenberg–Noe (EN) and DebtRank, characterizing the contagion process from both bankruptcy risk and actual default risk perspectives. Finally, utilizing the real indicator SRISK, model comparisons, and ablation experiments are conducted to validate the practical effectiveness of various innovative solutions, alongside simulation experiments designed to explore the impact of interest rates on the interbank lending market, thereby providing valuable risk prevention insights for both individual banks and government entities.