›› 2017, Vol. 33 ›› Issue (7): 68-76.

Previous Articles     Next Articles

Research on the FEW of Listed Companies in Growth Period based on the CombinationKalman Filtering Algorithm and Logistic Regression Model

ZHU Zhaozhen, DONG Xiaohong, WANG Jian   

  1. Business Institue Anhui University of Finance & Economics, Bengbu 233000, China
  • Received:2016-08-11 Online:2017-07-10 Published:2017-07-10

成长期上市公司财务危机预测——基于Kalman滤波与Logistic回归的实证研究

朱兆珍, 董小红, 汪健   

  1. 安徽财经大学商学院,安徽 蚌埠 233000
  • 作者简介:朱兆珍 (1982),女,安徽淮南人,安徽财经大学商学院讲师,博士;董小红(1981),女,安徽庐江人,安徽财经大学商学院讲师,博士;汪健(1972),男,安徽财经大学商学院副教授,博士。
  • 基金资助:
    安徽省高校自然科学研究重点项目(KJ2017A788);国家自然科学基金资助项目(71602001);国家社会科学基金资助项目(16BGL010)

Abstract: This research on the financial early-warning of the companies in the growth period is conducted by combining Kalman filtering algorithm and Logistic regression model. The results are as follows:(1) Corporate governance factors have significant effects on the occurrence of financial crisis; (2) By comparison,it is showed that the specificity generated from the Kalman filtering model is superior to that from the Logistic regression model; the sensitivity of the listed companies in the growth period of Model I & II on time T-1 and Model II on time T-3 is higher than that by Kalman filtering algorithm,but lower of Model I on time T-3 and Model I & II on time T-5. The paper puts forward suggestions to avoid financial distress.

Key words: Growth Period, Financial Distress, Kalman Filtering Algorithm, Logistic Regression

摘要: 本文将Kalman滤波智能算法与Logistic回归传统模型相结合,对成长期上市公司财务危机进行预测。结果表明:公司治理因素对上市公司是否发生财务危机具有显著影响;比较而言,Kalman滤波算法算得的专一性优于Logistic回归模型;临近被ST的T-1期模型I、II以及T-3期模型II计算所得敏感性高于Kalman滤波算法敏感性结果,但T-3期模型I及T-5期模型I、II敏感性皆低于Kalman滤波算法得到的敏感性。文章最后提出了成长期上市公司避免陷入财务困境的政策建议。

关键词: 成长期, 财务危机, Kalman滤波, Logistic回归

CLC Number: