APPLICATION OF EXPERT SYSTEMS IN EVALUATION OF CREDIT CARD BORROWERS

An appropriate approach to determine and measure the characteristics of “good” borrowers has always been the subject of inquiry by various lenders both in domestic and international financial markets. Cost of bad debts incurring from non‐payers and slow payers is a major source of loss which would affect profits and consequently the value of the lending firm. A carefully designed and economically feasible method to select the right borrowers is consistent with the goal of a modern lending institution which is to minimize the risk of bad debts and increase the wealth of its shareholders. The issue of selecting good borrowers is more serious in the case of potential credit card holders where the number of applicants is far greater than the number of commercial, real estate, and other institutional borrowers. Lending institutions cannot afford spending more than a certain limited amount of time to scrutinize the application of each applicant. Competition and increasing cost of information are other reasons that a lender should approve or reject submitted applications in a reasonably short period of time with minimum decision errors.