主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2014, Vol. 22 ›› Issue (7): 10-17.

• 论文 • 上一篇    下一篇

我国金融机构的系统性金融风险评估——基于极端分位数回归技术的风险度量

陈守东1,2, 王妍2   

  1. 1. 吉林大学数量经济研究中心, 吉林 长春 130012;
    2. 吉林大学商学院, 吉林 长春 130012
  • 收稿日期:2012-12-07 修回日期:2014-02-19 出版日期:2014-07-20 发布日期:2014-07-24
  • 作者简介:陈守东(1955-),男(汉族),天津蓟县人,吉林大学商学院,教授,博士生导师,研究方向:金融计量分析。
  • 基金资助:

    教育部人文社科重点研究基地重大项目(2009JJD790015);国家社科基金项目(12BJY158)

Measuring Systemic Financial Risk of China’s Financial Institution——Applying Extremal Quantile Regression Technology and CoVaR Model

CHEN Shou-dong1,2, WANG Yan2   

  1. 1. Center for Quantitative Economics, Jilin University, Changchun 130012, China;
    2. School of Business, Jilin University, Changchun 130012, China
  • Received:2012-12-07 Revised:2014-02-19 Online:2014-07-20 Published:2014-07-24

摘要: 本文将极值理论引入到系统性金融风险度量中,通过极端分位数回归技术估计我国33家上市金融机构对金融系统整体的风险贡献,并识别出我国系统重要性金融机构。研究结果表明,我国金融机构的市场价值总资产收益率呈现明显的非正态分布特征,使用极端分位数回归技术可以更准确的度量尾部的风险联动性;银行类金融机构的系统性风险贡献水平最高且波动变化最大,系统性风险贡献排名前十的金融机构基本为银行类机构;证券类、保险类、信托类金融机构的风险贡献水平相对较低;通过与其他研究的对比发现,考虑到极端情形下的尾部风险联动性时,股份制商业银行对金融系统的风险贡献上升。本文的研究为系统重要性金融机构的宏观审慎监管提供了实证依据。

关键词: 系统性金融风险, CoVaR, 极端分位数回归, 风险度量

Abstract: Based on the extreme theory, a new approach is presented for measuring systemic financial risk. Using extremal quantile regression, 33 listed financial institutions'contributions to the systemic risk of financial system are estimated and systemically important institutions in China' s financial system are recagnjzed. It is found that, the distributions of the growth rate of market valued total assets and the co-movement of the tail risk can be accurately estimated using the extremal quantile regression. The level and variance of the systemic risk contribution in bank sector are the highest. The top ten systemically important financial institutions are almost in bank sector. The levels of systemic contribution in security sector, insurance sector and trust sector are relatively low. Moreover, systemic contributions of joint-stock commercial banks are higher in our study. An empirical tool is provided in this paper for further macro-prudential regulation of systemically important institutions.

Key words: systemic financial risk, CoVaR, extremal quantile regression, risk measure

中图分类号: