Predicting bankruptcy among U.S. companies : a study based on Altman’s Z-score and Almamy’s J-UK model

This thesis aims to compare the famous two bankruptcy prediction models (the Altman Z-score and the U.K. based J-UK model) and investigates which model is a better predictor of bankruptcy when applied for US companies. Expectations were that Altman’s Z-score would have a lower predictive ability due to its focus on manufacturing firms and lower significance compared to when it was first revealed in 1968. It was also expected that the J-UK model would have a higher predictive power due to the additional cash flow variable. The models were tested based on a multiple discriminant analysis (MDA) which revealed that the discriminating ability of the J-UK model was significantly higher. However, both models had a high classification ability which is like due to my sample size and availability of data which was not the case when the J-UK model was initially applied in the UK.