Bankruptcy Prediction Models: How to Choose the Most Relevant Variables?

B ankruptcy prediction has given rise to an extensive body of literature since the late 1960's, the aim of which is to assess the determinants of financial failure and to design prediction rules. Beginning with Altman (1968) and continuing through Agarwal and Taffler (2008), nearly all factors that may strongly influence the accuracy of a model have been analyzed through empirical research. Ample supporting evidence may be found in the literature, which provides insights into the issues regularly emerging in this field (Balcaen and Ooghe, 2006). However, a comprehensive examination of the main research topics shows that they have not received the same degree of attention. We have analyzed 190 papers on bankruptcy prediction models written in the last forty.