系统性风险传染的波动性研究——基于金融网络动态关联的视角

在市场流动性不足的前提下,建立资金融通的金融网络结构可以降低个体机构的破产概率,但转移的风险会通过网络节点之间的关联度和正反馈效应实现交叉传染,从而增加整个市场奔溃的概率。文章通过DCC-MGARCH模型和无向有权型网络阐释了包括银行、证券、保险和多元金融机构在内的金融市场系统性风险的时变机制。实证结果表明金融机构的动态关联能够较好解释系统性风险的波动性,且我国的金融市场符合无标度网络的风险传染特征。其中,银行部门的市场中介地位不断强化,银行与非银行金融机构的联系日益紧密,新型金融机构的发展潜力巨大。因此在系统性风险的监管中应强调关联度指标的重要性和金融机构的网络属性,构建具有风险包容性的金融体系。

Study on the Volatility of Systemic Risk Contagion: Based on the Dynamic Connectedness of Financial Network

Under the premise that the lack of liquidity in the financial market, to establish the financial network structure can reduce the bankruptcy probability of individual institutions, but the transferred risk can form cross infection by connectedness and positive feedback effect among the institutions, thus increasing the degree of systemic risk. The basis of the theoretical analysis is to take the connectedness degree between financial institutions as the node connection that linked in the financial network, and then analyze the mechanism of accumulated systemic risk in the nonlinear cross infection structure. The evaluation of systemic risk from the network angle combines both macro and micro analysis methods, getting riding of the defects that linear relation and individual effect of systemic risk contagion in traditional research. Through this way, the formation principle of systemic risk is well confirmed with the resonance effect of risk in the changing connected network. The relationship between financial institutions' risk correlation and systemic risk is clearly explained, and we can find that the cause of financial institutions failure is internal, not external. To prove the theory, the DCC-MGARCH model and weighted network structure without direction are chose as empirical models to testify the time-varying systemic risk in the financial market. In variable selection, the model expands the limitation of the institution types shown in the previous literature to consist of banks, securities, insurance and other financial institutions together to explain the systemic risk contagion. The empirical results showed that systemic risk in China's financial market remain high in the sample period, and the volatility of systemic risk has increased since January 2015. Banks are increasingly linked to non-bank financial institutions, and the innovative financial institutions have great potential for development. In general, the risk contagion in China's financial market meets the structural features of scale-free network, and the unbalanced network structure helps to control the transmission and diffusion of financial risk. So the systemic risk supervision should target on the importance of features in network structure, such as connectedness, network pressure and risk tolerance. At the same time, the transmission channels of systemic risk through key nodes in financial supervision should be taken seriously too. In the daily activity, the key points of supervision are about the risk tolerance and contagion in specific financial institutions which depend on the business cycle and circumstance. When the systemic risk become smaller, we should pay attention to the urban commercial banks with effective guidance and supervision of business innovation risk, and reducing its key role of risk contagion in the network. When the systemic risk become larger, financial regulation should emphasize the diversification of risk contagion characteristic and spillover effect about the state-owned banks and innovative financial companies. Only with the time-varying subjects and measures of supervision, can the regulatory efficiency be improved by controlling the core channels of risk transmission and mitigating risk volatility of the financial market in the same direction. If the conclusions talked above could be considered in the financial market reform schedule, a more tolerant financial system will show up with less systemic risk inside.