The development of CRITA-Score ? A hybrid credit rating model for predicting financial distress
Credit ratingBankruptcyDefaultFinancial distressMultiple Discriminant Analysis (MDA)Z-Score modelManagement of enterprisesFöretagsledningManagementBusiness and Economics
Large companies in today?s business environment are to a great extent depending on their credit score by rating institutes like Moody?s and Standard & Poor?s. There is however indications that these big institutes are not always rating companies accurately. This is why this thesis aims at developing a new credit rating model, CRITA-Score, with the objective of determining a company?s financial health more accurately. The model is based on a number of financial ratios and was developed using a statistical model called MDA (Multiple Discriminant Analysis). To support the development of CRITA, interviews with the four larges banks in Sweden where made to capture the matters that are of essence for a creditor in today?s business climate. These interviews were then used as groundwork for choosing which financial ratios to incorporate in the statistical modelling. The model was developed using financial data from U.S. companies and incorporates a unique time-weighting technique. CRITA?s accuracy was tested by applying the model to a number of default and non-default companies. The results indicated that CRITA performed well in some cases and not so well in others. The conclusions of the analysis were however that CRITA has future potential to serve has a credit-tool if it is further tested and developed.