非核心负债、尾部依赖与中国银行业系统性风险

Non-core Liabilities,Tail Dependence,and China's Banking Systemic Risk

非核心负债是银行业系统性风险累积的重要驱动因素,本文将刻画时间维度系统性风险特征的非核心负债指标与刻画机构关联性的尾部依赖技术相结合,得到中国银行业系统性风险度量指标。在捕捉冲击事件有效性方面,本文指标可以捕捉样本期间的4次重大冲击事件,包括2008年国际金融危机、2013年银行业“钱荒”、2015年股市异常波动和2018年中美贸易摩擦;在捕捉规模特征有效性方面,本文指标能够规避传统风险指标中出现的“小机构、大贡献”这种不符合现实的问题;本文提出系统性风险度量领域的安慰剂检验方法,发现本文指标能排除股票市场风险等噪音信息。

Non-core liabilities are the key proxies of financial risk build-up.Combining the indicators of non-core liabilities,which capture the systemic risk accumulation within the temporal dimension,with the tail-dependence model,which captures the interlinkage between institutions within the cross-cutting dimension,this paper obtains the Chinese banking systemic risk metrics,which are effective at capturing the impacts of shocks,reflecting scale characteristics,and excluding noise information.With regard to the effectiveness of shock detection,the indicators in this paper reflect the impacts of four important events that have already occurred:(i)the 2008 global financial crisis;(ii)China's 2013“liquidity crisis”;(iii)the stock market crash of 2015;and(iv)the Sino-US trade friction of 2018.With regard to the validity reflected in the scale characteristics,the indicators enable avoiding the inconvenience of the“small banking institutions and large contributions”of traditional risk indicators,thereby becoming more consistent with the regulation of the relevant systemic institutions.In terms of noise information elimination,the results of the placebo test show that indicators may exclude a large proportion of noise information,such as that from stock markets.