宏观审慎监管需要识别出系统重要性机构(SIFIs),目前已有相关研究,但还没有就如何有效识别达成共识.首先阐述MES 法、SRISK 法和△ CoVaR 法三个主流的市场识别法原理与计算,在统一的DCC-GARCH 模型相关性分析框架中,对它们做了理论比较.然后,以中国上市银行为样本,收集市场交易数据,用DCC-GARCH 模型拟合相关结构,分别进行了中国SIFIs 识别,根据各指标的理论特征分析了各系统重要性排序结果的有效性及其经济意义.
Macro-prudential supervision firstly calls for there cognition of Systemically Important Fi-nancial Institutions(SIFIs). Nowadays there is no consensus about the effectiveness of several existing recognition methods of SIFIs. This paper introduces three most popular recognition methods. We pro-pose a theoretical comparison in a common framework and then collect market transaction data of Chi-nese listed banks to do the recognition of SIFIs by the DCC-GARCH model. According to the theoreti-cal and empirical analysis of the characteristics of the three recognition methods, we get the economic meanings of SIFIs rankings by each recognition methods.