管理评论

›› 2019, Vol. 31 ›› Issue (8): 84-96.

• 经济与金融管理 • 上一篇    下一篇

基于Copula尾部风险控制的行业贷款配置模型

张舒明, 周颖   

  1. 大连理工大学经济管理学院, 大连 116024
  • 收稿日期:2018-02-23 出版日期:2019-08-28 发布日期:2019-09-11
  • 作者简介:张舒明,大连理工大学经济管理学院硕士研究生;周颖,大连理工大学经济管理学院副教授,硕士生导师,博士。
  • 基金资助:

    国家自然科学基金重点项目(71731003；71431002);国家自然科学面上基金项目(71873103；71471027)；国家社科基金一般项目(16BTJ017)；国家自然科学基金青年科学基金项目(71601041)；爱德力智能科技(厦门)有限公司智能风险管控模型与算法项目(2019-01)。

Industry Loan Allocation Model Copula-based Tail Risk Control

Zhang Shuming, Zhou Ying   

  1. School of Economics and Management, Dalian University of Technology, Dalian 116024
  • Received:2018-02-23 Online:2019-08-28 Published:2019-09-11

PDF (PC)

337

摘要：

银行危机的本质是资产配置失误，行业资产配置又是银行资产配置的顶级层次。避免银行巨额亏损乃至银行危机的核心是控制资产配置的极端风险。本研究建立了基于尾部风险控制的行业贷款配置模型。本研究的创新和特色：一是通过建立不同行业的尾部相关系数与尾部风险价值的函数关系，建立了基于极端风险控制的行业贷款配置模型；改进了经典均值方差模型由于使用Pearson线性相关系数导致的无法度量尾部极端风险的弊端。二是建立了基于t分布的VaR约束，提高了VaR约束与“尖峰厚尾”特征的契合度，增强了VaR约束控制尾部极端风险的能力；解决了VaR约束的正态性假设与“尖峰厚尾”特征不一致的弊端。

关键词: 资产配置, 行业组合, 极端风险, 尾部风险, Copula函数

Abstract:

The root cause of banking crisis lies in misallocation of assets, and the industry asset allocation is the top level of bank asset allocation. The key to preventing banks from huge losses or even crisis is to control the extreme risk of asset allocation. This study establishes an industry loan allocation model based on tail risk control. The innovation and characteristics of our paper have two aspects. Firstly, we establish an industry loan allocation model based on tail risk control by using the functional relationship between the tail correlation coefficient and the tail risk value of different industries and this model solves the drawbacks of the classical mean variance model that cannot measure the extreme risk of the tail due to the use of the Pearson linear correlation coefficient. Secondly, we establish the VaR constraint based on t-distribution, which improves the fit to the distribution feature of fat-tailed enhances the ability of VaR constraint to control the extreme risk of tail, and solves the inconsistency between the normal distribution hypothesis in VaR and the actual distribution feature of fat-tailed.

Key words: asset allocation, industry portfolio, extreme risk, tail risk, Copula