| 摘要: |
| 摘要：基于2008－2025年中国34家上市金融机构股票收益率数据，运用修正MES与DMST方法构建动态风险传染网络，考察金融风险传染路径及时变特征。研究发现：（1）金融机构的尾部风险具有顺周期性，受风险事件冲击骤升；（2）银行、证券及保险行业内关联较强，多元金融多与证券跨行业关联；（3）风险驱动与承担主体由银行转向证券；（4）银行和证券长期是重要传染桥梁，部分中小规模但关联密切的机构成为关键传播节点。研究结论可为监管部门评估金融机构风险重要性、完善系统重要性机构名单及实施动态风险监管提供经验证据。 |
| 关键词: 关键词：金融风险；风险传染；MES方法；网络中心度 |
| DOI： |
| 分类号:F832.5 |
| 基金项目:基金项目：重庆市社会科学规划项目博士培育项目“网络视角下系统性金融风险传染及防控研究”（2020PY47）；辽宁省教育厅项目“乡村振兴战略下金融支持农村一二三产业融合发展研究”（LN2019Q02）；东北财经大学校级科研项目“相对贫困阶段数字普惠金融的减贫增收效应研究”（DUFE2020Y01）。 |
|
| Contagion Pathways and Time-Varying Dynamics of Risk in Financial Institutions——An Empirical Study Based on a Modified MES Approach |
|
LIANG Yi-juan1，LI Jie1，BA Hong-jing21,2
|
|
1.(1.School of Economics and Management,Southwest University;2.Academy of Financial Inclusion,Dongbei University of Finance and Economics)
|
| Abstract: |
| Abstract：Based on the stock return data of 34 listed financial institutions in China from 2008 to 2025,this paper constructs a dynamic risk contagion network by the modified MES method and the DMST method to examine the contagion pathways and time-varying dynamics of financial risk.The findings include: (1) The tail risk of financial institutions exhibits procyclicality and surges sharply under the impact of financial risk events；(2) There are strong intra-industry linkages within the banking,securities,and insurance sectors,while diversified financial institutions predominantly exhibit cross-industry linkages with securities firms；(3) The role of risk drivers and risk bearers has shifted from banks to securities firms；(4) Banks and securities firms have long served as the important bridges for risk contagion,with some small and medium-sized institutions that maintain close interconnections emerging as key transmission nodes.The conclusions provide empirical evidence for regulatory authorities to assess the risk significance of financial institutions,refine the list of systemically important institutions,and implement dynamic risk supervision. |
| Key words: Key words：Financial Risk；Risk Contagion；MES Method；Network Centrality |
