The implementation of credit scoring as a risk management tool

Abstract

Consumer lending is a natural niche for community banks, offering opportunities to expand customer relationships and diversify risks. Both small and large community-based institutions are turning their attention to growing these Iines of business. Accompanied by the growth in consumer debt, however, is the reality of consumer bankruptcies and delinquencies. As a result, lending organizations are pressured to find the best methods to estimate risk as accurately as possible in the credit granting process. If risk is estimated poorly, the entire enterprise can fail either because of excessive default or lack of business. If you can estimate risk better than your competitor can, his business is in danger, since you are able to offer better terms to more people than he can. The challenge facing any lending organization then, is to try to manage the credit losses in this high volume consumer lending business in order to improve pricing margins. Turning a profit in the credit industry rests on many factors beyond the loan agreement. Long-term profitability can be affected by the credit grantor's ability to close deals quickly at the point-of-sale, to accurately determine applicant creditworthiness, and to assess various stages of risk throughout the course of the agreement. As elusive as these capabilities may seem, some credit grantors are already performing these activities economically with the aid of new technologies. The capacity to instantly access credit bureau report information and assess creditworthiness in seconds, rather than hours, is a reality reaping dividends for credit grantors representing a wide variety of industries. One of these technologies widely used in consumer lending today is credit scoring. Credit scoring is a tool designed to give credit management the ability to measure risk and to take the action best suited to each risk that is identified. Through a simple, single transaction, patrons of this burgeoning technology have gained instantaneous access to numerous consumer and credit information databases, as well as the ability to score applications consistently, without bias, and according to their own specific lending strategies. Additional technologies have also emerged which are extending these capabilities to mobile professionals, as well as allowing online users to instantly verify applicant information and identify fraud . The technique involves awarding points to answers given to questions listed on an application form. Questions cover such areas as: • Marital status and dependants; • Age; • Number of years at present address; • Nature of property occupied - whether it is owned or rented; • Occupation; • Length of time with employer; • Previous credit history. Each question will carry a maximum possible score that will be higher for important questions such as occupation and previous credit history, and lower for questions such as age and marital status. The score awarded to each answer will vary. For example, if the question asks for the number of years at present address, ten points might be awarded for each year, up to a maximum of, say, 200, in the same way certain occupations will be awarded higher marks than others. The score will determine whether an account may be opened. If the score is above a certain number, the account may be opened and vice versa, if the score is below a certain point, the applicant will not be accorded an account. The technique of credit scoring can be used for all routine applications where some exercise of judgment would formerly have been needed. It is commonly in use for the opening of current accounts, for car finance, issuing credit cards, personal loans and more recently for mortgage lending and small business applications. Good resources on the topic are however, scarce and there does not seem to be any published information in South Africa on credit scoring. The primary objective of the study was to try and prove that by using credit scoring the decision to grant or not to grant credit can be more scientifically based in order to reduce risk in lending, and for management to have better control over the granting of credit. In order to test the effectiveness of credit scoring as a risk management tool, Absa bank is used as a benchmark, being one of South Africa's largest financial institutions. Absa, realizing the need to improve on systems and procedures in the credit granting process in order to remain competitive, appointed specialist consultants, the Experian Group (UK) during 1993 , to perform a feasibility study. Questions were designed from the results of the study where certain inherent problems were identified within the credit management environment at Absa. These questions based on the problems identified by Experian, form the basis of the Empirical study. The position before and after the implementation of crepit scoring at Absa is compared using various research methodologies such as Pareto analysis, Statistical analysis and interviews. To obtain the general viewpoint of bankers countrywide on the topic 'Credit scoring as a risk management too l', a questionnaire was designed that was distributed amongst Absa personnel and various other registered banks in South Africa for completion. The results emanating from the Empirical study and the Literature study prove that although a scoring model is not perfect and does have its drawbacks, it is effective as a risk measurement tool. A large number of benefits can accrue to an organization . following the successful implementation of a well-designed credit scoring system . Using credit scoring, competitive advantages could well be attained through efficient service and improved turnaround times.