摘要
鉴于传统的以财务因素分析为主的企业危机预警模型研究的局限性,本文采用因子分析和逐步判别分析相结合的方法,在综合考虑财务和非财务因素的基础上,分别建立了基于原始财务指标(模型I)、行业相对财务指标(模型II)、行业相对财务指标和非财务指标(模型III)、行业相对财务指标和考虑宏观滞后影响(模型IV)的非财务指标Logit回归模型,并运用国内的相关数据进行了实证。研究及实证结果表明,消除行业影响后模型的分类准确度有所提高,而同时考虑财务与非财务因素及经济环境因素的滞后预警作用,将大大改善总体分类准确度和拟合度,且本文提出的8参数指标体系模型(模型IV)比国内外同类研究提出的不同指标体系构建的Logit模型具有更好的预测精度与模型拟合度。
Previous domestic studies always use financial ratios in the quantitative research of credit analysis. However, traditional financial analysis has some limitations. What＇s more, the credit enterprises are not in a closed system, they would be inevitably influenced and confined by macroeconomic and market environments. According to the current domestic situation, and based on previous domestic studies with only financial ratios, this paper extends the Logit regression model by integrating financial and non - financial factors, considering lag of macroeconemie factor and eliminating inter- industry effect for the prediction of financially distressed firms. Empirical results indicate that by introducing non - financial ratios （with lag of macroeconomie factors and eliminating inter- industry effect ） into traditional models based on only financial ratios, the total claasified accuracy and predictive power would be significantly improved.
出处
《金融研究》
CSSCI
北大核心
2006年第2期78-87,共10页
Journal of Financial Research
基金
教育部新世纪优秀人才支持计划(教技司[2005]2号)资助。
关键词
企业危机预警模型
财务指标
非财务指标Logit回归模型
the financial distress warning model
financial ratios
non- financial ratios
Logit regression model
