Building a Bankruptcy Prediction Model for Chemical Industry Companies

This paper researches the models of predicting the bankruptcy of Russian chemical industry companies. The problem of bankruptcy is crucial nowadays, because it helps to estimate risks of investments and highlight the weaknesses of a company. This work will analyze the existing models of bankruptcy prediction such as Altman Z-score, Taffler Z-score, Zaytseva model and Lis model. These models will be tested for our dataset to estimate their accuracy, to find the key indicators and to build new model. Besides the indicators from the previous researches, new indicators will be provided. The dataset will be collected from open source and will contain financial reports of Russian joint-stock and limited liability companies that manufacture chemicals. Two ways of building a new model will be tested - Multiple discriminant analysis and Logistic regression. As a result of this study we will gain a new model that will have higher predictive power than the previous ones and will help to predict bankruptcy for the companies in the selected industry.