嵌入网络舆情指数的中国金融机构系统性风险传染效应研究

首先基于文本挖掘技术构建反映投资者情绪的网络舆情指数，然后将所构建的网络舆情指数嵌入到系统性风险传染效应度量模型，得到修正的单指标非对称CoVaR模型，并运用线性分位数LASSO算法与局部多项式估计方法进行参数估计，以此为基础构建金融有向网络，进而对中国金融机构系统性风险传染效应进行实证分析。实证研究表明：(1)以单指标非对称CoVaR为代表的金融机构风险指标与网络舆情的协同变化趋势明显；(2)证券类和银行类金融机构对外部风险非常敏感，极易受到其他金融机构的影响，也极易影响其他金融机构；(3)非银行类机构在风险积累阶段占据重要位置，银行在风险爆发时刻占据重要位置；(4)相对于非银行类金融机构，银行类机构具有较强的传染能力。

Research on Systemic Risk Contagion Effect of Chinese Financial Institutions Considering Network Public Opinion Index

The global financial crisis of 2007-2009 has aroused considerable interest in systemic risk, and has triggered rethinking of systemic financial risks by international organizations, financial regulators and scholars in various countries. The crisis connected socio-economic entities such as banks, real estate, insurance companies, hedge funds, and consumers into a community of economic interests, sharing risk factors. Financial crisis refers that the values of most financial assets drop together or one institution’s failure could propagate to other institutions, thus, financial crisis is systemic and results in the collapse of whole financial system. In fact, during the financial crisis, losses spread across financial institutions and the financial system is threatened. At the same time, the contagion of financial risks is becoming more and more frequently, and extreme risk events, including “cash crunch” and “curcuit breakers”, have caused the spread of network public opinion, and made systemic financial risks spread rapidly in the capital market.