Microeconomic scoring model of Czech firms' bankruptcy

by Jiří VALECKÝ , Eva SLIVKOVÁ

Download

JEL classification

  • Single Equation Models; Single Variables: Discrete Regression and Qualitative Choice Models; Discrete Regressors; Proportions
  • Bankruptcy; Liquidation

Keywords

Classification table, default probability, fractional polynomial, logistic regression, logit

Abstract

The paper is devoted to the proposing a scoring model of firm´s bankruptcy on the basis of logistic regression which could be used in the purpose of classifying the good and bad firms in the Czech Republic. Firstly, the general foundation of logistic regression is described and then the general procedure of model-building strategy including the statistical verification and in-sample validation are explained. Then, we have built the empirical scoring model using results of financial analysis carried out over 400 Czech firms in 2008 and we select such financial ratios with the most substantial prediction capability, i.e. return on assets, current and long-term indebtedness and cash ratio. Our proposed model is fully statistical verified and we also perform the in-sample validation via classification table and ROC curve. Finally, we compare our model with others studies made in this field.