Predicting bankruptcy based on the full population of Croatian companies

This paper analyses the bankruptcy prediction based on population of companies representative total business sector in Croatia. The representativity sample is achieved through propensity score matching full bankrupt and similar non-bankrupt companies. robust estimation carried out multiple discriminant analysis (MDA) logistic regression (logit). results indicate high classification accuracy both models, but more favourable performance logit estimation. Overall MDA model was 73.7%, while overall 76.3%. serve as a benchmark for