Chapter 5


Expanding Markets

After landing a full-time, $26,000-a-year job, Ray Evans of Normal, Illinois, was ready to move out of his parents’ home. Over three years, he had accumulated enough money for a down payment—with a little help from his mother and father.

The first lender Mr. Evans approached turned him down for a conventional mortgage to finance a modest $80,000 ranch house. His down payment, the lender said, was insufficient to consider him for anything but a government-insured mortgage.

Then Mr. Evans found a small local lender that uses automated underwriting. Within three weeks of loan application, the 27 year old was pocketing the keys to his first home.

Able to weigh Mr. Evans’ proven ability to manage his credit obligations, including a car loan and credit card, against other factors, Loan Prospector classified his application as an acceptable risk. “The loan officer told me my application was going off to an automated loan system, but I felt very comfortable with that,” Mr. Evans recalled.

Automated underwriting is opening the doors of homeownership to new groups of borrowers. By increasing the accuracy with which mortgage applications are underwritten, systems such as Freddie Mac’s Loan Prospector enable lenders to approve more first-time homebuyers like Mr. Evans. In addition, Loan Prospector can lower interest rates for thousands of borrowers with high-cost mortgages by bringing them into the conventional market.

Approving More Borrowers

Throughout this report, we present real-life examples of borrowers who have been helped by Loan Prospector. In the past, such borrowers might have been denied credit or been directed to a more expensive type of financing. Now, with the help of Loan Prospector, lenders have the confidence they need to offer low-cost conventional mortgage financing to borrowers who may not fit the traditional profile.

Estimating the number of potential borrowers who could be helped by the increased precision of Loan Prospector is admittedly difficult. However, based on preliminary evidence, Freddie Mac believes the impact will be significant. With its sharper tools to identify creditworthy borrowers who may have been overlooked by traditional underwriting, Loan Prospector can increase the number of homeowners by as many as 250,000 families. 1

Lower Aggregate Risk Boosts Lending

Automated underwriting also will identify as high-risk some applicants who might have been approved under traditional underwriting. A high-risk classification does not always result in a loan denial, however. 2 Some applicants have extenuating circumstances that make them acceptable credit risks. Nevertheless, as automated underwriting becomes more widespread, the number of high-risk applicants who receive loans will decline.

Reducing the number of high-risk borrowers initially will lower the average risk that the mortgage industry must absorb. As average risk declines, however, competitive pressures will spur the market to expand at the margin, bringing in borrowers whose current profiles put them just beyond existing standards.

The net effect on the size of the market inevitably will be positive. Substituting borrowers just at the margin for borrowers with significantly higher rates of default will enable the industry to serve a greater number of families without increasing overall risk.

This increase will be in addition to the 250,000 borrowers overlooked by traditional underwriting but found to qualify thanks to automated underwriting’s greater precision.

Lowering Interest Rates Across Markets

Not only will automated underwriting expand mortgage credit to a new generation of homeowners, it also will allow many existing borrowers to take advantage of lower mortgage rates. Early evidence from a pilot program shows that automated underwriting will classify many borrowers now relegated to the higher-cost subprime market as acceptable risks by conventional market standards. The shift of these borrowers into the lower-cost conventional market will benefit minority and low-income families the most.

Improving the Subprime Market. The subprime mortgage market accounts for about 5 percent of total mortgage originations. Typically, borrowers turn to this market when credit problems or other circumstances make them ineligible for conventional or FHA financing. Borrowers with subprime mortgages usually pay significantly higher interest rates than other borrowers.

To address lender demand for an automated underwriting service capable of evaluating loans in any mortgage market, Freddie Mac joined forces with S&P, a rating agency with significant experience evaluating subprime loans. Based on its independent assessment, S&P concluded that Loan Prospector “evidences a high degree of predictiveness coupled with consistency of scoring” for subprime (and jumbo) loans. 3

Buoyed by these results, we initiated a pilot program in October 1995 to make automated underwriting available to the subprime mortgage market. Although the mortgages originated under the pilot will not be sold to Freddie Mac, other investors stand ready to purchase these loans.

Bringing automated underwriting to the subprime market promises its borrowers increased efficiencies and lower processing costs. Interest rates on subprime mortgages also should fall as investors become more confident that the default risk on these mortgages has been reliably analyzed. 4

Identifying More Conventional Borrowers. In the course of the pilot program, Freddie Mac made a significant discovery. Not only did Loan Prospector improve the processing of subprime mortgages, it also identified applicants in the subprime market who would have qualified for lower-cost conventional financing.

Preliminary Freddie Mac estimates suggest that between 10 and 35 percent of borrowers who obtained mortgages in the subprime market could have qualified for a conventional loan. 5 A recent poll of the 50 most active subprime lenders supports this conclusion. The survey found that up to 50 percent of subprime mortgages could qualify as investment-grade mortgages, although some of these loans would fail to meet certain secondary market criteria. 6

Breaking out of the subprime market promises to confer significant benefits to families who believed they had no other choice. Minority and low-income borrowers are most likely to be affected. African-American and Hispanic borrowers are represented in the subprime market at nearly triple their rates in the conventional market, while low-income borrowers are represented at about one-and-one-half times their rates. 7

Subprime borrowers who would have qualified for conventional loans pay mortgage rates on the order of one to two-and-one-half percentage points higher in the subprime market, based on Freddie Mac estimates. 8 A one-and-one-half percentage point savings, for example, would mean paying 8.5 percent instead of 10.0 percent for a mortgage.

Moving to the conventional market means that qualifying subprime borrowers could save roughly $50 to $130 a month on a $75,000 30-year, fixed-rate mortgage—significant amounts for hardworking families. In the aggregate, these families could save up to $100 million in interest costs each year by obtaining conventional loan financing.


Footnotes:

1.This preliminary estimate was derived from a sample of about 2,300 conventional loan applications obtained from several industry sources that were denied during 1992 using traditional underwriting. We do not know whether some of these applicants were eventually approved by other lenders, perhaps in the subprime market. However, Freddie Mac’s analysis suggests that as many as 25 percent of the loan denials would have been accepted under automated underwriting. HMDA data report that approximately 1 million applicants for conventional home purchase loans were denied in 1995.
2. In the case of Loan Prospector, for example, all mortgage applications classified as caution are referred to human underwriters and are accompanied by feedback to help lenders work with families to shore up the weaknesses within their mortgage applications.
3. “Automated Systems Aid Mortgage Underwriting,” Standard & Poor’s CreditWeek, October 23, 1995. To reach this conclusion, S&P tested 50,000 subprime and jumbo (larger than Freddie Mac’s loan limit, which is $207,000 for one-unit homes in 1996) mortgages.
4. See “Technology Shrinking Spreads on B and C,” Origination News, August 1996.
5. To assess the potential size of this group, Freddie Mac used Loan Prospector technology to evaluate a sample of 15,000 subprime mortgages originated by four financial institutions.
6. “Half of Subprime Loans Categorized as ‘A’ Quality,” Inside B&C Lending, June 10, 1996.
7. Computed using HMDA data from 21 of the largest subprime lenders in 1994.
8. This estimate is based on a comparison of interest rates on approximately 1,000 subprime mortgages to the interest rates on mortgages



© 2008 Freddie Mac close window