Performance evaluation of bankruptcy prediction models: An orientation-free super-efficiency DEA-based framework

Abstract

Prediction of corporate failure is one of the major activities in auditing firms' risks and uncertainties. The design of reliable models to predict bankruptcy is crucial for many decision making processes. Although a large number of models have been designed to predict bankruptcy, the relative performance evaluation of competing prediction models remains an exercise that is unidimensional in nature, which often leads to reporting conflicting results. In this research, we overcome this methodological issue by proposing an orientation-free super-efficiency data envelopment analysis model as a multi-criteria assessment framework. Furthermore, we perform an exhaustive comparative analysis of the most popular bankruptcy modelling frameworks for UK data including our own models. In addition, we address two important research questions; namely, do some modelling frameworks perform better than others by design? and to what extent the choice and/or the design of explanatory variables and their nature affect the performance of modelling frameworks?, and report on our findings.

Multi-criteria ranking of corporate distress prediction models: empirical evaluation and methodological contributions

“…Our survey of the existing studies concerned with the comparison of competing statistical prediction models, supports Bauer and Agarwal (2014) and Mousavi et al (2015) arguments in addressing three main drawbacks in the related literature. Firstly, most of the existing studies failed to have a comprehensive comparison between all types of statistical prediction models, i.e.…”

“…Obviously, the performance of models is not only dependent on the sample selection, modelling techniques and feature selection procedures but also reliant on the evaluation process and the chosen performance criteria. In practice, several studies have compared the performance of competing models taking into account different modelling frameworkse.g., Bauer and Agarwal (2014), Mousavi et al (2015) and Wu et al (2010); alternative sampling techniquese.g., Neves and Vieira (2006), and Zhou (2013), and various featurese.g., Tinoco and Wilson (2013), Trujillo-Ponce et al (2014). Furthermore, several criteria, including, discriminatory power, calibration accuracy, information content and correctness of categorical prediction have been used for the performance evaluation of alternative models.…”

An out-of-sample evaluation framework for DEA with application in bankruptcy prediction

“…DEA has witnessed a widespread use in many application areas-see Liu et al (2013) for a recent survey, and Mousavi et al (2015) and Ouenniche (2011, 2012a, b) for a recent application area-along with many methodological contributions-see, for example, Banker et al (1984), Andersen and Petersen (1993), Tone (2001Tone ( , 2002 and Seiford and Zhu (2003). Despite the growing use of DEA, to the best of our knowledge, no published work formally addressed out-of-sample evaluation in DEA.…”

Financial Risk Measurement and Prediction Modelling for Sustainable Development of Business Entities Using Regression Analysis

“…Mousavi et al [71] conclude that the choice and design of independent variables and their nature affect the overall performance of the model. It is obvious that there are significant differences among variables used in various models and that for different countries with different type of economy should be developed a unique model with appropriate variables.…”