- 1、本文档共10页,可阅读全部内容。
- 2、原创力文档(book118)网站文档一经付费(服务费),不意味着购买了该文档的版权,仅供个人/单位学习、研究之用,不得用于商业用途,未经授权,严禁复制、发行、汇编、翻译或者网络传播等,侵权必究。
- 3、本站所有内容均由合作方或网友上传,本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺!文档内容仅供研究参考,付费前请自行鉴别。如您付费,意味着您自己接受本站规则且自行承担风险,本站不退款、不进行额外附加服务;查看《如何避免下载的几个坑》。如果您已付费下载过本站文档,您可以点击 这里二次下载。
- 4、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等,请点击“版权申诉”(推荐),也可以打举报电话:400-050-0827(电话支持时间:9:00-18:30)。
查看更多
毕博上海银行咨询Credit Risk Mgmt Sys Analytics Default modeling
Default Modeling and Credit Grading
Prepared properly, a credit score or risk grade for a borrower will provide a sufficient statistic for evaluating the likelihood of default over a chosen time interval such as a year. This paper discusses the steps involved in developing a default model that allows for such credit grading. We distinguish 3 essential steps, namely
Specifying a logically sound model
Estimating the model’s parameters
Testing forecasting performance.
We describe each of these steps below. We present the steps in their usual order, though the actual process may involve more than one pass through the sequence. Initial empirical results often motivate changes to a model.
Whenever possible, we believe that credit ratings should derive from an empirical study of the factors affecting default risk. We expect that both numerical, financial measures and categorical, subjective ones will help explain default. In either case, we’ll make every effort to assess a variable’s influence through statistical estimation and testing. This empirical approach contrasts with schemes that consolidate a series of judgmental evaluations using judgmental weights. We strongly believe in statistical estimation, since otherwise one has no firm basis for claiming that a credit score has more predictive power than a random-number generator.
Observe that if ratings derive from a statistical default model, one has more flexibility and granularity than offered by a small set of discrete risk grades. For some purposes such as migration analysis or limit setting, one might want to classify together borrowers with similar default rates. However, the continuous indicator provided by the empirical model stands out as the more informative risk grade for a borrower.
We focus here on borrower ratings. Starting with a borrower rating, one may characterize facility risk by appending descriptions of collateral, covenants, guarantees, amortization schedules, and other features of the l
您可能关注的文档
最近下载
- 010-数学形态学分析.ppt
- 2023年芜湖市镜湖区市场监督管理局招考工作人员笔试参考题库(共500题)答案详解版.docx VIP
- 励志班会:985博导桂海潮案例,读书可以改变命运主题班会.pptx
- Haier海尔洗衣机EG10014BD809LGU1使用说明书手册参数图解图示pdf电子版下载.pdf VIP
- 《中职高考英语总复习与同步练》(总复习分册)教案 第11课 语法知识——专题9 非谓语动词.docx VIP
- 供应室泛水应急演练.pptx VIP
- 化学品管理中的供应链安全和可追溯性.pptx
- 小学四年级上册心理健康教育教案.doc
- 《中职高考英语总复习与同步练》(总复习分册)教案 第8课 语法知识——专题7 形容词和副词(2).docx VIP
- 【方书】中医土单验方一百首(高清版).pdf
文档评论(0)