CREDIT SCORING MODELS IN ESTIMATING THE CREDITWORTHINESS OF SMALL AND MEDIUM AND BIG ENTERPRISES

• Main Author: Robert Zenzerović
• Format: Article
• Language: English
• Published: Croatian Operational Research Society 2011-02-01
• Series: Croatian Operational Research Review
• Subjects: financial instability; credit scoring models; small and medium and big enterprises; logistic regression analysis
• Online Access: http://hrcak.srce.hr/index.php?show=clanak&id_clanak_jezik=142215

This paper is focused on estimating the credit scoring models for companies operating in the Republic of Croatia. According to level of economic and legal development, especially in the area of bankruptcy regulation as well as business ethics in the Republic of Croatia, the models derived can be applied in wider region particularly in South-eastern European countries that twenty years ago transferred from state directed to free market economy. The purpose of this paper is to emphasize the relevance and possibilities of particular financial ratios in estimating the creditworthiness of business entities what was realized by performing the research among 110 companies. Along most commonly used research methods of description, analysis and synthesis, induction, deduction and surveys, the mathematical and statistical logistic regression method took the central part in this research. The designed sample of 110 business entities represented the structure of firms operating in Republic of Croatia according to their activities as well as to their size. The sample was divided in two sub samples where the first one consist of small and medium enterprises (SME) and the second one consist of big business entities. In the next phase the logistic regression method was applied on the 50 independent variables – financial ratios calculated for each sample unit in order to find ones that best discriminate financially stable from unstable companies. As the result of logistic regression analysis, two credit scoring models were derived. First model include the liquidity, solvency and profitability ratios and is applicable for SME’s. With its classification accuracy of 97% the model has high predictive ability and can be used as an effective decision support tool. Second model is applicable for big companies and include only two independent variables – liquidity and solvency ratios. The classification accuracy of this model is 92,5% and, according to criteria of predictive ability, it can be estimated as high. Credit scoring models represent scientifically based derived decision support tool. Their application on micro level can prevent the establishment of business relation with financially instable companies what can potentially result in losses while on macro level they can signal the forthcoming problems in economy as a whole and give the impulse for acting in appropriate direction.