POSSIBILITIES TO APPLY CLASSICAL BANKRUPTCY PREDICTION MODELS IN THE CONSTRUCTION SECTOR IN LITHUANIA

Abstract

ruptcy prediction models were chosen: 3 linear discriminant analytical models (Altman, Springate, Taffler) and 2 logistic regression models (Chesser, Zavgren). From the Altman’s models, the Altman’s model for companies whose shares are not quoted in the stock exchange markets, Altman’s Z”-Score Model for the service companies and Altman’s Z”-Score Model for emerging countries were investigated. From Taffler models, Taffler (1973) and Taffler & Tisshaw (1977) models were analysed. Having carried out the research, it is possible to come to the conclusion that the most accurate bankruptcy prediction models with the highest bankruptcy probability are the following: the logistic regression adapted Chesser and Zavgren models; the accuracy of the linear discriminant Springate models is also high. The research proved that the Taffler and Altman’s Z’’ Score Model for emerging countries models are least accurate. The results of the research might be useful for both the executive managers of companies in the construction sector and investors who analyse the problems of the operation continuity.