›› 2019, Vol. 31 ›› Issue (11): 3-19.

• 经济与金融管理 •    下一篇

基于Wilcoxon检验的小型工业企业债信评级模型及实证

李刚1, 张亚京2, 王基凡2, 迟国泰2   

  1. 1 东北大学秦皇岛分校经济学院, 秦皇岛 066004;
    2 大连理工大学经济管理学院, 大连116024
  • 收稿日期:2017-03-27 出版日期:2019-11-28 发布日期:2019-11-30
  • 通讯作者: 迟国泰,大连理工大学经济管理学院教授,博士生导师,博士。
  • 作者简介:李刚,东北大学秦皇岛分校经济学院副教授,硕士生导师,博士;张亚京,大连理工大学经济管理学院,博士;王基凡,大连理工大学经济管理学院硕士研究生;
  • 基金资助:

    国家自然科学基金重点项目(71731003);国家自然科学基金面上项目(71471027,71971051);国家自然科学基金青年科学基金项目(71601041);河北省自然科学基金面上项目(G2019501105);2019年河北省高等学校青年拔尖人才计划项目(BJ2019213)。

Research on the Small Industrial Business Credit Rating Model and Empirical Analysis Based on Wilcoxon Test

Li Gang1, Zhang Yajing2, Wang Jifan2, Chi Guotai2   

  1. 1 School of Economics, Northeastern University at Qinhuangdao, Qinhuangdao 066004;
    2 School of Economics and Management, Dalian University of Technology, Dalian 116024
  • Received:2017-03-27 Online:2019-11-28 Published:2019-11-30

摘要:

债信评级的本质是衡量一笔贷款违约的概率大小,违约概率越大,债信等级越低。由于小型工业企业财务信息不健全等自身特点的限制,国内外金融机构对此缺乏相应研究,导致现有研究中小型工业企业债信评级模型极不完善。而本文通过Wilcoxon检验构建了小型工业企业债信评级模型。本文的特色与创新:一是通过Wilcoxon检验进行指标筛选建立评级指标体系,保证了指标体系可以有效判别企业的违约状态;二是通过游程检验方法检验指标体系的合理性,解决了对债信评级模型的检验问题;三是通过指标对违约状态判别能力越大权重越大的思路确定指标权重,弥补了现有研究中指标赋权过程权重与违约状态甄别能力不匹配的不足。

关键词: 债信评级, 小企业评级, Wilcoxon检验, 游程检验, 违约状态判别

Abstract:

The essence of credit rating is to measure the default probability of a loan, specifically, credit rating decreases with the increase of the default probability. Small industrial enterprises with less information available are not fully covered by financial analysts at home and abroad and that's why the credit rating model proposed in the existing researches for small industrial enterprises is defective. In this paper we use Wilcoxon test to establish a credit rating model for small industrial enterprises. There are three aspects of innovations and features in this paper. Firstly, the Wilcoxon method is applied to select features to establish a credit rating index system, which ensures that the index system can distinctly discriminate the default status of an enterprise; secondly, the run test is applied to evaluate the performance of the index system, which can do the validation of the credit rating model. Thirdly, index weight is determined on the basis of the regular pattern that the stronger ability an index has to discriminate the default status, the greater the weight is. The index system established this way successfully solves the problem in existing researches that the weight of an indicator does not match its ability of distinguishing the default state.

Key words: credit rating, small business credit rating, wilcoxon test, run test, default discrimination