Fuzzy logic in credit rating

Abstract

Credit assessments are meant to help a bank decide whether borrowers will be able to meet their obligations according to the loan agreement. In the very basic sense a credit rating is an opinion on the creditworthiness of an issuer indicating the probability that the issuer will not default on the payment obligation of the issue. Fuzzy logic systems in credit scoring can be said to be an extension of Expert systems. Expert systems are software solutions which aim to recreate human problem solving abilities in a specific area of application. In other words, expert systems attempt to solve complex, poorly structured problems by making conclusions on the basis of intelligent systems. The essential components of an expert system are the knowledge base and the inference engine. The Deutsche Bundesbank uses a fuzzy logic system as a module in its credit assessment procedure. The Bundesbank s credit assessment procedure for corporate borrowers first uses industry-specific discriminant analysis to process figures from annual financial statements and qualitative characteristics of the borrower s accounting practices. In this study an attempt is made to develop a model on the similar lines of the Deutsche Bank model. To develop the credit rating model, the first step is to decide the input parameters. To get good credit rating both the quantitative financial data and also the qualitative data is required as described earlier in the Deutsche Bank s Model. Due to lack of available data only quantitative data has been taken into consideration in this model. Even this simple model gave accuracy results which are higher compared to the multi discriminate analysis. This shows that fuzzy logic models have an inherent advantage because of their ability to take care of the fuzziness or ambiguity in the system. Unlike the statistical models, fuzzy logic models are easier to interpret and also they carry the experts knowledge with them. Hence sharing of knowledge becomes easier. Since fuzzy logic models don t follow any distribution, accurate ratings can be obtained even with smaller set of data as compared to statistical methods.