文章基本信息

标题：Predicting Bankruptcy with Robust Logistic Regression
本地全文：下载
作者：Richard P. Hauser ; David Booth
期刊名称：Journal of Data Science
印刷版ISSN：1680-743X
电子版ISSN：1683-8602
出版年度：2011
卷号：9
期号：4
页码：565-584
出版社：Tingmao Publish Company
摘要：Using financial ratio data from 2006 and 2007, this study usesa three-fold cross validation scheme to compare the classication and pre-diction of bankrupt firms by robust logistic regression with the Bianco andYohai (BY) estimator versus maximum likelihood (ML) logistic regression.With both the 2006 and 2007 data, BY robust logistic regression improvesboth the classication of bankrupt rms in the training set and the predictionof bankrupt rms in the testing set. In an out of sample test, the BY ro-bust logistic regression correctly predicts bankruptcy for Lehman Brothers;however, the ML logistic regression never predicts bankruptcy for LehmanBrothers with either the 2006 or 2007 data. Our analysis indicates thatif the BY robust logistic regression signicantly changes the estimated re-gression coecients from ML logistic regression, then the BY robust logisticregression method can significantly improve the classification and predictionof bankrupt firms. At worst, the BY robust logistic regression makes nochanges in the estimated regression coecients and has the same classifica-tion and prediction results as ML logistic regression. This is strong evidencethat BY robust logistic regression should be used as a robustness check onML logistic regression, and if a dierence exists, then BY robust logisticregression should be used as the primary classifier.
关键词：Bankruptcy prediction; robust logistic regression.