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 ›› 2004, Vol. ›› Issue (6): 29-33.

• 论文 • 上一篇    下一篇

基于损失程度变化的CreditRisk+的鞍点逼近

蔡风景1,2, 杨益党3, 李元1   

  1. 1. 广州大学理学院 广东 广州 510405;
    2. 温州师范学院数学与信息科学学院 浙江 温州 325027;
    3. 昌吉学院 新疆 昌吉 831100
  • 收稿日期:2004-04-15 修回日期:2004-11-15 出版日期:2004-12-28 发布日期:2012-03-07
  • 基金资助:
    国家自然科学基金资助项目(10271033)

Saddlepoint approximation of Creditrisk+ Basing on Severity Variation

CAI Feng-jing1,2, YANG Yi-dang3, LI Yuan1   

  1. 1. School of Sciences, Guangzhou University, Guangzhou 510405, China;
    2. School of Mathematics & Information Science, Wenzhou Normal College, Wenzhou, Zhejiang 325027, China;
    3. Changji university, Xinjiang Chanji 831100, China
  • Received:2004-04-15 Revised:2004-11-15 Online:2004-12-28 Published:2012-03-07

摘要: 传统的CreditRisk+模型在度量信用风险过程,假定违约损失是给定不变的,而近年的实证研究表明在实际的金融市场中违约损失是变化的。针对传统模型这一假设的不合理性,本文对模型作了发展。新模型充分考虑了违约时损失程度的变化,用β分布来刻画这种变化,并利用鞍点逼近给出了信用风险的度量,改进了传统递推算法的不足。最后进行数值模拟以说明方法的有效性。

关键词: CreditRisk, 鞍点逼近, 损失程度

Abstract: The traditional Creditrisk+ assumes severity is given and invariable,but recent researchs prove that severity is variable in financial market.Due to this unreasonable assumption in the model,Credit Risk+ is improved in this paper.Severity variation which distribution is a beta distribution is taken into account and saddlepoint approximation,in stead of traditional recurrence relation,is used for credit risk measurement in the new model.Finally,simulations show that the method is effective.

Key words: Credit Risk+, saddlepoint approximation, severity

中图分类号: