In lieu of an abstract, here is a brief excerpt of the content:

138 Over the last ten years, subprime mortgage lending has evolved from a small niche in home equity lending to a market valued at over $200 billion annually, or roughly 10 percent of the overall single-family residential mortgage market (Cutts and Van Order, 2005). The term subprime, which covers a wide-ranging set of mortgage products and practices, is also called nonprime . In simplest terms, it is mortgage lending where the cost of credit is higher than that offered by prime and FHA lending specialists. In most cases, the higher cost reflects the lower credit quality of approved applicants as measured by their credit scores. This chapter presents a stylized overview of the economic benefits that could be derived from the emergence of risk-based pricing in mortgage lending. It explores outcomes that are conditional on retail and wholesale practices in the marketplace as the mortgage industry matures, as well as policy and business implications of this shift. Shift to Risk-Based Subprime Lending Lenders evaluating loan applicants attempt to predict the default risks associated with a given loan. Loss of income, divorce, and severe illness are comparatively random events that could fall upon any class of borrowers to create a loan default. Other events are systematically related to the borrower and property characteristics and can be predicted based on the loan applicant’s past Exploring the Welfare Effects of Risk-Based Pricing in the Subprime Mortgage Market j. michael collins, eric s. belsky, and karl e. case 6 09 7409-5 ch06.qxd 7/7/2005 10:12 PM Page 138 effects of risk-based pricing 139 behavior. Lenders have developed more refined tools to predict systematic risks in the last decade, allowing risk-based pricing to achieve increasing degrees of granularity. A decade ago lenders manually examined payment ratios, loan-to-value ratios, employment histories, assessments of the value of the collateral, and credit histories of loan applicants to evaluate if a loan should be approved. Each was compared to relatively rigid standards established by decades of past industry practice. Stiglitz and Weiss (1981) argued that rationing by qualification standards in this way is the result of imperfect information about the uncertainty surrounding the systematic credit risks of a loan application. Since originators cannot observe the credit risk profiles of borrowers with certainty, they resort to rigid rationing systems. Chinloy and Macdonald (2004) build on this model of credit allocation, adding a secondary step of lenders sorting approved loans by loan-to-value (LTV) ratio and then pricing into two general categories based on collateral risk. Loans above 75, then 80 (and later higher) LTV ratios require the added cost of mortgage insurance, either at the loan closing in the form of an upfront premium or an increase in monthly debt service payments to cover premiums. These models ration credit to prime quality borrowers with a simple pricing structure. While these models prevailed, subprime borrowers were denied access to credit at any price. Chinloy and Macdonald (2004) suggest the advent of subprime lending has expanded credit allocation by a new dimension—credit quality. The lender now can accept most loan applications, but pricing becomes more complex, effectively expanding from two price levels into hundreds of risk-priced categories. True risk-based pricing implies each borrower’s unique observable systematic credit risk characteristics would be assessed and priced along a continuum of mortgage prices. In practice, most lenders continue to censor the riskiest corner of the credit pricing grid—sorting out the most risky applicants and rationing them out of the market. Lenders also use other mechanisms to govern the risk and revenue related to loans. For example, riskier loans have more restrictive terms than prime loans, including prepayment penalties, origination fees, and other features (Pennington-Cross, Yezer, and Nichols 2000). In practice, subprime loans are priced based on past loan payment behavior and credit scoring metrics. Temkin, Johnson, and Levy (2002) suggest the following categories of borrowers: — A, or prime, borrowers, generally have Fair Issac Company (FICO) scores above 660, have never missed a mortgage payment, and missed only one revolving debt payment in the last twenty-four months; — A-minus borrowers have scores above 620, and have missed no more than one mortgage payment or two credit-card payments; —B borrowers have missed several payments, one of which was at least sixty days late; 09 7409-5 ch06.qxd 7/7/2005 10:12 PM Page 139 [3.135.183.187] Project MUSE (2024-04-19...

Share