Predicting bankruptcy using recursive partitioning and a realistically proportioned data set

✍ Scribed by Thomas E. McKee; Marilyn Greenstein

• Publisher: John Wiley and Sons
• Year: 2000
• Tongue: English
• Weight: 167 KB
• Volume: 19
• Category: Article
• ISSN: 0277-6693
• DOI: 10.1002/(sici)1099-131x(200004)19:3<219::aid-for752>3.0.co;2-j

✦ Synopsis

Auditors must assess their clients' ability to function as a going concern for at least the year following the ®nancial statement date. The audit profession has been severely criticized for failure to `blow the whistle' in numerous highly visible bankruptcies that occurred shortly after unmodi®ed audit opinions were issued. Financial distress indicators examined in this study are one mechanism for making such assessments. This study measures and compares the predictive accuracy of an easily implemented two-variable bankruptcy model originally developed using recursive partitioning on an equally proportioned data set of 202 ®rms. In this study, we test the predictive accuracy of this model, as well as previously developed logit and neural network models, using a realistically proportioned set of 14,212 ®rms' ®nancial data covering the period 1981±1990. The previously developed recursive partitioning model had an overall accuracy for all ®rms ranging from 95 to 97% which outperformed both the logit model at 93 to 94% and the neural network model at 86 to 91%. The recursive partitioning model predicted the bankrupt ®rms with 33±58% accuracy. A sensitivity analysis of recursive partitioning cutting points indicated that a newly speci®ed model could achieve an all ®rm and a bankrupt ®rm predictive accuracy of approximately 85%. Auditors will be interested in the Type I and Type II error tradeos revealed in a detailed sensitivity table for this easily implemented model.