基于Logistic与Fisher的上市公司财务困境判别模型比较研究  

Comparative Research on Financial Distress Prediction Models in Listed Companies Based on Logistic Model and Fisher Model

作　　者：马若微[1] 张微[1] 

机构地区：[1]北京工商大学经济学院,北京100048

出　　处：《北京工商大学学报（社会科学版）》2014年第2期88-95,共8页JOURNAL OF BEIJING TECHNOLOGY AND BUSINESS UNIVERSITY:SOCIAL SCIENCES

基　　金：教育部人文社会科学研究资助项目(10YSC79088);北京市哲学社会科学规划项目(13JGB036);北京市青年拔尖人才培育计划(CIT&TCD201404037)

摘　　要：以我国股权分置改革后A股制造业公司为研究对象,采用大样本法选取样本,选用财务困境发生前2年内8个季度的财务数据进行分析,对财务困境公司与正常公司的财务指标分别进行Fisher判别分析、线性logistic回归和非线性logistic回归分析。研究结果表明,我国上市公司财务困境发生前2年的季度数据对公司财务困境是否发生具有预测效力;总资产净利润率、营运资金比率和成本费用利润率三个指标对财务困境的预测能力较强;非线性logistic模型的误判率最低,线性logistic模型其次,Fisher模型的误判率最高。This paper chooses China＇ s A-share manufacturing companies after the reform of non-tradable shares as the re- search object, with the use of large sample method to select the samples. The 8 quarterly data in the first two years before finan- cial distress are selected to make a regression analysis of the ST companies and normal companies respectively based on Fisher model, linear Logistic model and non-linear Logistic model. The results show that the quarterly data in the first two years before financial distress has an effect of prediction to the occurrence of financial distress; the net profit margin, working capital ratio and cost margins have better prediction to the financial distress ; and the error rate of non-linear Logistic model is the lowest, the rate of the linear Logistic model is in the second place, and the error rate of the Fisher model is the highest.

关 键 词：上市公司 财务困境 Fisher模型 LOGISTIC模型 

分 类 号：F832.5[经济管理—金融学] F275