Fuzzy Inference System for Credit Scoring: Legal Clients in Banking Industry

Abstract

Credit scoring is one of the fundamental concepts in bank industry, used for analysing and evaluating all of the customers requesting for facilities. Because of its importance to banks’ profitability, especially in the developing countries, this study aims to propose a fuzzy inference system (FIS) model for credit scoring of legal clients. This research is applied in terms of purpose and survey in terms of method. In the first step, after reviewing the literature, the evaluation criteria for legal client’s appraisal were identified, and 29 out of them were selected. In the next step, these criteria were analysed by the research experts and using a Delphi method, and 12 more important criteria were selected and organized in 4 categories for FIS modelling. Then, a researcher made questionnaire was developed to constitute rules for Main FIS and its 4 sub-FISs. Performance measure (Sub-FIS1) has four inputs: return of asset, fixed assets to equity, average customer account and customer capital. Leverage measure (Sub-FIS2) has two inputs: debt ratio and equity ratio. Borrowing measure (Sub-FIS3) has four inputs: ratio of deferred amount to current assets, amount of received facilities, borrowing capacity and amount of requested facilities, and finally, credit risk measure (Sub-FIS4) has two inputs; type of guarantee and the credit risk of the previous period. The proposed system designed based on Gaussian membership function and implemented in MATLAB. Finally, model’s validity was tested by extreme condition test. Comparing the results of the proposed FIS and the bank validation system, shows the proposed model can be considered suitable for credit scoring.