企业财务风险预测模型实证比较分析  

An Empirical Comparison of the Models for Predicting the Corporation's Financial Risk

作　　者：刘杰 袁韵锋 黄媛梅 

机构地区：[1]重庆市综合经济研究院经济咨询研究中心,重庆401147 [2]重庆市大渡口区行政中心,重庆400084

出　　处：《统计与信息论坛》2007年第5期97-100,共4页Journal of Statistics and Information

摘　　要：企业财务风险一直是风险管理理论和实务界关心的热点话题。运用判别分析和计量经济方法对重庆市某商业银行的461个样本企业2002-2005年的违约特征进行实证检验和预测。结果发现最重要的决定变量是资产负责率、酸性试验比率、资产净利率等7个财务比率以及企业所处的产业部门,考虑了异方差性的probit模型有更好的预测能力。This paper estimates and predicts the default characteristics for a sample of 461 corporations in Chongqing between 2002 - 2005 with discriminate analysis and econometric methods. The most important determinants are found to be seven financial ratios including debt ratio, acid test ratio, net profit on asset etc, and dummy variables representing industry sectors. By comparison to the accuracy of prediction, we find that the probit model taking account of the heteroskedasticity has the strongest explanatory power. This research will be crucial to construction of more accurate internal credit risk evaluation system for commercial banks or other financial institutions and to improving the efficiency of credit risk evaluation, and reducing credit risk exposure of financial institutes.

关 键 词：财务风险 判别分析 PROBIT模型 LOGIT模型 异方差 

分 类 号：F064.1[经济管理—政治经济学]