The aim of this research is to analyze the accuracy of selected bankruptcy prediction models on the example of Lithuanian companies. The research involves financial statements of 23 companies that have gone bankrupt over the period of 2013-2019. We used three different groups of models. The first two are considered as classic models which were developed using discriminant analysis (Altman, modified Altman, Springate, Taffler and Tishaw, and Grover models) and logistic regression (Ohlson, Zmijewski, and Grigaravičius models). The third group is based on artificial intelligence (we used a decision tree model, which is the most innovative and the least explored model of all used). The analysis evidenced that the logistic regression models, such as Zmijewski and Ohlson, demonstrated the best results in the group of classic prediction models, i.e., high probability of bankruptcy even earlier than one year prior to actual bankruptcy in the case of most companies. However, the decision tree must be considered as the most accurate model as it predicted bankruptcy of all analyzed companies one year before actual bankruptcy; this could be interpreted as 100% accuracy. Too late bankruptcy process causes many negative consequences for company’s employees, partners, and the state. Though problems with financial resources such as growing accounts payable and the shortfall of working capital which contribute to insolvency can be seen in the financial statements, in addition to the analysis of financial indicators, it is particularly important to use the above-mentioned bankruptcy prediction models, which help to detect financial problems in time and make the right decisions concerning future activities.
CITATION STYLE
Stankevičienė, J., & Prazdeckaitė, G. (2021). Analysis of the accuracy of bankruptcy prediction models: the case of Lithuanian companies. Science and Studies of Accounting and Finance: Problems and Perspectives, 15(1), 44–53. https://doi.org/10.15544/ssaf.2021.05
Mendeley helps you to discover research relevant for your work.