Predicting Corporate Bankruptcy: A Cross-Sectoral Empirical Study - The Case of Greece

Abstract

Purpose: This article explores the prediction of bankruptcy of Greek companies, in particular of the manufacturing industry, wholesale, retail and service sectors. Design/methodology/approach: The Probit model was developed so as to try to highlight the differences in the predictive capacity of the model across the sectors but also to investigate any differences in the behavior of the financial indicators used in the model. Moreover, for the selection of these indicators, the technique of factor analysis was applied. Findings: The results showed significant explanatory capacity of the model in the four key sectors of the Greek economy up to four years before failure and bankruptcy, as well as a clear differentiation in the sector classification of companies. Research limitations/implications: This work can be used by managers, banks as well as by practitioners to identify the causes of firm’s failure. Originality/value: The limited investigation, to date, of the effects of sectoral features and the absence of sectoral samples of bankrupt companies with a higher degree of homogeneity in predicting bankruptcy may often lead prediction models to unreliable results. This paper has two main contributions to the relevant literature. At first, it serves as a work of distinguishing the differences between bankruptcy predictive power of the same financial indicators of enterprises belonging to different sectors. Secondly, the use of factor analysis in the selecting procedure of the appropriate variables provides better and more robust results in the field of bankruptcy prediction.