Impact of dynamic variables in bankruptcy prediction models based on logistic regression

The study focuses on logistic regression, a method applied with the majority of scorecards used in bank lending processes. The author is trying to answer the question whether considering the trends of financial indicators can improve ranking accuracy for bankruptcy prediction models. The research used the sample of 1527 enterprises in the construction industry. The findings of the study suggest dynamic variables have improved the accuracy of bankruptcy prediction particularly improving the ranking accuracy of operational enterprises. Thus, banks are advised to analyse the trends of financial indicators to make lending decisions. There is an important difference here between lending to start-ups or to enterprises with a long operational history. Obviously, you cannot study earlier values of indicators with a start-up company; for this reason, the spread of the practice may further deepen (restrict) the options of start-ups to have access to resources. Therefore, the method should rather be employed for companies with a long history. The findings of the study also indicate that large enterprises are less threatened in terms of bankruptcy