基于金融机构尾部风险网络的金融系统性风险形成演化机制及预警控制研究

The financial systemic risk is the risk of collapse of an entire financial system. It is imposed by interdependencies among financial institutions, where the failure of a single financial institution can cause a cascading failure, which could potentially bankrupt or bring down the entire system. The cascading failure will be reflected simultaneously as the financial statement data and market data tail-risk contagion of the financial institutions. To reveal the cross contagion relationships among financial institutions, we can construct the tail risk network, in which the nodes represent the financial institutions and the edges represent the tail risk contagion relationships. As a result, the tail risk network is co-evolved with financial systemic risk. Based on tail risk networks of financial institutions, this project studies the formation and evolution mechanisms of financial systemic risk and its early warning and controlling by applying the complex network theories and methods. Firstly, we analyze the tail risk correlation mechanisms of financial institutions and construct the correlation model. Secondly, the tail risk networks are established, and the topology structures and their dynamic evolution rules are empirically investigated. After that, we construct the network dynamic evolution models. Thirdly, we analyze the risk contagion processes on the tail risk networks, the co-evolution rules between network topology structures and financial systemic risk, and the temporal and temporal-spatial evolution rules of financial systemic risk. All of these studies will demonstrate the formation and evolution mechanisms of financial systemic risk. Fourthly, the financial systemic risk early warning model will be constructed. Lastly, we will investigate the financial systemic risk controlling strategies from the perspective of controllability of tail risk networks. In conclusion, from the aspects of tail risk correlations, tail risk network evolution and tail risk contagion, financial systemic risk formation and evolution, and financial systemic risk early warning and controlling, the project intends to establish a financial systemic risk management theory and method system under the framework of complex network theories and methods.

金融机构间相继出现经营困境或破产倒闭等极端事件，客观上表现为其公开财务或市场数据的尾部风险传染。金融机构尾部风险网络，揭示了大量机构间的尾部风险交叉传染模式，其与金融系统性风险是协同演化的。本项目基于金融机构尾部风险网络，运用复杂网络理论与方法，研究金融系统性风险的形成演化机制及其预警控制。具体内容包括：分析金融机构尾部风险关联机制并建立关联模型；构建金融机构尾部风险网络，实证分析网络拓扑结构特征及其动态演化规律，建立网络动态演化模型；研究尾部风险网络上的风险传染过程、网络结构与金融系统性风险的协同演化规律、金融系统性风险的时间及时空演化规律，进而揭示系统性风险的形成演化机制，并建立风险预警模型；从网络结构控制角度，研究系统性风险控制策略。本项目从尾部风险关联、网络结构演化及风险传染、系统性风险形成演化及其预警控制等方面，探索建立复杂网络理论方法框架下的金融系统性风险管理理论及方法体系。

本课题首先构建金融机构尾部风险关联度量指标CoCVaR并进行实证研究；分别利用条件在险价值、非线性双变量Copula的GARCH模型衡量金融机构之间的尾部风险依赖性，进而构建金融机构尾部风险网络，并分析了网络全局和局部拓扑结构特征及其动态演化规律。.其次，研究了金融机构间的信息溢出（收益、波动和尾部风险）关联，并依此分析机构的系统重要性及系统性风险传染规律。具体地，分别基于格兰杰因果关系检验、VAR-GARCH-BEKK模型、广义方差分解、LASSO和高维方差分解等模型方法，构建金融机构间收益、波动或尾部风险溢出网络，分析了网络拓扑结构特征及其动态演化规律，根据节点中心性衡量信息传染及承受视角的机构系统重要性，并挖掘了其基本面及市场层面的影响因素。.再次，基于商业银行的资产负债表，分别利用最大熵方法和最小密度法，估算银行间借贷关联网络，比较分析了两种网络的拓扑结构特征。研究了单个银行破产的随机冲击情形下，银行系统的风险传染过程和结果。此外，还从主权风险跨国溢出视角，研究了金融系统性风险的空间及时空溢出规律。.然后，基于金融机构尾部风险网络，研究了金融系统性风险对实体经济的溢出影响。具体地，利用TENETs模型研究了金融机构的网络连接性，通过网络整体拓扑结构及其动态演化分析金融体系的连通性水平以及尾部风险溢出情况，研究了尾部风险溢出网络的变化对经济产出的预测能力；利用金融机构动态收益溢出网络的连通性量化金融系统的互连性及潜在的风险传递渠道，研究了网络连接性对总体宏观经济指标的预测能力。.最后，从金融网络动态演化视角研究了金融系统性风险的控制策略。具体地，给出了银行同业资金借贷双方理性决策下的同业借贷网络的形成演化机制，构建了网络上的违约传染模型，通过理论分析和数值仿真研究发现，征收系统性风险税可以显著降低网络的系统性风险。 . 本研究聚焦于金融机构尾部风险网络下的金融系统性风险形成演化机制及预警控制问题，研究成果对于金融系统性风险的防范化解具有重要的现实意义。.