ABSTRACT Consumer credit is ubiquitous and lending poses credit risk--the risk of economic loss due to the failure of a borrower to repay according to the terms of his or her contract with the lender. And so, managing credit risk entails estimating the potential ability of borrowers to repay their debts. Researchers have sought to identify factors that contribute to consumer risk, by using quantitative models. However, the presence of data mining techniques to identify credit risk cannot be ignored. There is a paucity of research to demonstrate the use of data mining techniques in this context, and such studies could be instructive to practitioners and academicians. This study fills that void. Using a data mining tool, this study shows that consumers can be segmented by their characteristics such as education, income, years on the job, and payment habits. The study showed that the rich were highly educated and always paid in full. Delinquency experiences were more frequent in the lower income segments. Knowledge about the risk of delinquency can be useful for lenders to price for credit risk and therefore to expand the reach of credit to consumers without having to compromise on profitability. INTRODUCTION Scott (2007) reports that the average United States household has 8 credit cards, which are used to charge nearly $2 trillion in goods and services annually. Further the study reports that consumers use credit cards inappropriately and spend beyond their means thereby accumulating inessentials that they cannot reasonably afford. According to the Federal Reserve Board's survey of consumer finances done in 2004, 76% of U.S. families carried some form of debt. Credit use was prevalent among families of all types regardless of their age, race, ethnicity, housing status, net worth and work-force status of the household head. Further, the percentage of families holding credit cards issued by banks rose from 16% in 1970 to about 71% in 2004. Debt was carried by 90% of the families in the top income groups and 53% in the lower income groups. The amount of outstanding debt was over $900 billion on bank-type credit cards at the end of 2006 (Scott, 2007). Further, about 1.56 million households, or about 1.4% of all U.S. households, filed for bankruptcy. Job loss was reported as one of the reasons for filing for bankruptcy. Until the late 1970s state usury laws established limits on the interest rates credit card issuers could charge on outstanding balances, which limited issuers' ability to price for credit risk. But after that, legislation relaxed the restrictions on credit card interest rates, and it allowed national banks to charge market-determined rates throughout the country. This reduction in legal impediments, along with improvements in information technology, allowed for the development of risk-based pricing nationally and contributed to the growth of revolving consumer credit. (1) Therefore, estimating the risk of potential borrowers is very important. Risky customers are those who have been delinquent. Credit card companies need to identify low-risk and high-risk customers because card rates are set based on the degree of risk that consumers reflect. For instance, Citigroup claims that it added 5.99% to the prime rate (which was 6.28%) in the case of low-risk customers, and in the case of high-risk customers it added 9.99% to the prime rate (Peng, Kou, Shi, Wise & Xu, 2005; Hsing, Gibson, Lin & Wallace, 2003). Determining and managing credit card rates are a challenge for credit card companies, and they must do well in this challenge if they are to survive in this competitive sector (Park, 2004; Hogarth, Hilgert & Kolodinsky, 2004). There is immense competition among firms, as some firms try to offer very low teaser rates to attract potential customers. Offering low rates can lead to rising charge-offs, increase in delinquency rates and declining credit quality. …
Read full abstract
Jan 1, 2017
Neha Gupta + 1
Open Access