【摘要】 本文以我国上市公司为研究对象 ,选取了 70家处于财务困境的公司和 70家财务正常的公司为样本 ,首先应用剖面分析和单变量判定分析 ,研究财务困境出现前 5年内各年这二类公司 2 1个财务指标的差异 ,最后选定 6个为预测指标 ,应用Fisher线性判定分析、多元线性回归分析和Logistic回归分析三种方法 ,分别建立三种预测财务困境的模型。研究结果表明 :(1 )在财务困境发生前 2年或 1年 ,有 1 6个财务指标的信息时效性较强 ,其中净资产报酬率的判别成功率较高 ;(2 )三种模型均能在财务困境发生前做出相对准确的预测 ,在财务困境发生前 4年的误判率在 2 8%以内 ;(3)相对同一信息集而言 ,Logistic预测模型的误判率最低 ,财务困境发生前 1年的误判率仅为 6 .47%更多还原

【Abstract】 This paper studies how to establish models for predicting financial distress in China’s listed companies.We firstly discuss three key issues,namely,definition of financial distress, prediction variables and models. Then both univariate and multivariate models,including Linear Probability Model,Discriminant Analysis and Logistic Regression are used to conduct an empirical study of 70 companies with and 70 without financial distress as samples in order to select variables and models. In is noted that the 6 of 21 financial ratios are significantly related with financial distress,and the model by Logistic Regression is more precise in predicting fumancial distress with lowest error.更多还原