Ranking of bankruptcy prediction models under multiple criteria
Document Type
Article
Publication Date
10-21-2016
Abstract
Prediction of corporate failure is one of the major activities in auditing firms' risks and uncertainties. In practice, the design of reliable models to predict bankruptcy is crucial in many decision-making processes. In this chapter we address two research questions related to the design of bankruptcy prediction models: namely, do some modelling frameworks perform better than others by design? and to what extent do the choice and/or the design of explanatory variables and their nature affect the performance of modelling frameworks? Elements of answers to these questions are devised through an exhaustive comparative analysis of the most popular bankruptcy-modelling frameworks, including our own models. Our comparative analysis is performed using a multicriteria performance evaluation framework based on data envelopment analysis, namely, an orientation-free super-efficiency slacks-based measure model. The proposed performance evaluation framework delivers a multicriteria ranking of bankruptcy prediction models, which overcomes the methodological issues related to the commonly used monocriterion framework. The performance of bankruptcy prediction models is assessed under a set of commonly used criteria and is tested on a sample that consists of all UK firms listed on the London Stock Exchange during an 18 year period.
Publication Title
Advances in DEA Theory and Applications: With Examples in Forecasting Models
First Page Number
357
Last Page Number
380
DOI
10.1002/9781118946688.ch24
Recommended Citation
Ouenniche, Jamal; Mousavi, Mohammad M.; Xu, Bing; and Tone, Kaoru, "Ranking of bankruptcy prediction models under multiple criteria" (2016). Kean Publications. 1708.
https://digitalcommons.kean.edu/keanpublications/1708