Abstract: Studies on systemic risk based on the traditional CCA method were undertaken within the BS framework. However, the results could be biased since the homoscedasticity assumption of BS model is inconsistent with reality. In this paper, the time-varying volatility of bank assets was characterized by using the asymmetric GARCH process, and the systemic risk was measured based on the GARCH option valuation model. Since existing studies on systemic risk mostly focus on the period of subprime crisis, the systemic risk of the banking industry from 2012 to 2016 was empirically analyzed with a sample of listed banks in China. The results show that systemic risk increases moderately at 2012 and at 2015 after the stock market crash, while it increases dramatically at 2016. The systemic risk indicator based on the traditional CCA method could not fully reflect the rise of systemic risk at 2015 and 2016. The systemic risk indicator of this paper is significantly better than that based on the traditional CCA method in predicting the dynamics of real economy.

Key words: systemic resk, time-varying volatility, contingent claims analysis (CCA) method, GARCH option valuation model

摘要： 基于传统CCA方法的系统性风险研究都是在BS框架下进行的，而BS模型的同方差假定与现实不符，可能引致偏误。利用非对称GARCH过程刻画了银行资产波动率的时变特性，并基于GARCH期权定价模型对系统性风险进行了度量研究。鉴于现有系统性风险研究大多聚焦于“次贷危机”期间，以中国上市银行为样本，对2012~2016年银行业的系统性风险状况进行了实证检验。结果显示：银行业系统性风险分别在2012和2015年“股灾”之后小幅上升，在2016年大幅上升；基于传统CCA方法的系统性风险指标无法对2015年和2016年系统性风险的上升做出充分响应；基于时变波动率CCA方法的系统性风险指标对宏观经济动态的预测能力显著优于基于传统CCA方法的指标。

关键词: 系统性风险, 时变波动率, 未定权益分析方法, GARCH期权定价模型