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

中国管理科学 ›› 2009, Vol. 17 ›› Issue (1): 36-41.

• 论文 • 上一篇    下一篇

度量银行操作风险的POT幂律模型及其应用

司马则茜1,2, 蔡晨1, 李建平1   

  1. 1. 中国科学院科技政策与管理科学研究所,北京 100190;
    2. 中国科学院研究生院,北京 100180
  • 收稿日期:2008-01-27 修回日期:2008-12-15 出版日期:2009-02-28 发布日期:2009-02-28
  • 作者简介:司马则茜(1973- ),女(汉族),河北任丘人,中国科学院研究学院,中国科学院科技政策与管理科学研究所博士研究生,研究方向:项目与风险管理.
  • 基金资助:

    国家自然科学基金项目(70701033, 70531040)

Using the POT Power Law Model to Evaluate Banking Operational Risk

SIMA Ze-qian1,2, CAI Chen1, LI Jian-ping1   

  1. 1. Institute of Policy and Management, Chinese Academy of Sciences, Beijing 100190, China;
    2. Graduate School of Chinese Academy of Sciences, Beijing 100080, China
  • Received:2008-01-27 Revised:2008-12-15 Online:2009-02-28 Published:2009-02-28

PDF (PC)

1459

被引次数

85

摘要： 本文在极值理论的POT模型基础上,建立了基于分维的POT幂律模型,给出了POT模型的阈值选择的理论解释,还给出了满足尾部分布适合幂律的条件.分析表明,此模型较已有方法能更方便地估计尾部参数,对小样本情形研究厚尾问题提供了新的思路.此外,为计算操作风险损失和的在险价值(VaR)问题,本文引入保险理论里的随机和模型,并给出了计算操作风险在险值的简化公式.

关键词: 分维, 幂律, POT模型, 随机和, VaR

Abstract: In this paper, a new POTPL(POT-Power Law) model on the basis of fractal dimension theory by using POT model of extreme value theory is presented.It gives not only a theoretical explanation for the choice of the threshold of POT model, but also satisfying power law condition of estimating tail disfribution.The analysis result shows that this model can be more convenient to estimate the tail than former methods.And it provides a new idea for solving fat-tail problem in small-sample circumstances, In addition, the random sum model in insurance is introduced in order to solve calculating Sum VaR of the banking operational risk losses.Then a simplified formula of VaR is given on the basis of operational loss data in our banks.

Key words: fractal dimension, power law, PO T model, random sum, VaR

中图分类号: