HUD recently released Assessing Problems of Default in Local Mortgage Markets, a report that examines concentrations of Federal Housing Administration (FHA) defaults in certain neighborhoods and among certain lenders. A 1997 study from the National Training Information Center (NTIC) had raised a concern about such concentrations. Using statistical analysis HUD's research draws different conclusions than NTIC did about the extent and implications of FHA defaults. Some of the differences in default rates across neighborhoods and lenders are traceable to individual borrowers' and loans' characteristics. Borrowers in neighborhoods, and among lenders, with high default rates are more frequently first-time homebuyers, Black, have higher loan-to-value ratios, have lower incomes, and have smaller value of assets after closing than do borrowers in neighborhoods, and among lenders, with lower default rates. Although low incomes are associated with higher default rates, many neighborhoods with low incomes or substantial minority representation have default rates below the metropolitan average. Simple statistical analysis identifies a set of high-default neighborhoods and a set of high-default lenders. However, both high-default neighborhoods and high-default lenders vary with the loan-origination year, indicating that some conditions that generate high default rates are temporary and less responsive to remedial actions. By serving less affluent borrowers, FHA extends homeownership to those who are less well served by the conventional market. The report, therefore, concludes that putting further restrictions on FHA borrowers will reduce default rates but will also work against expanding homeownership among low- and moderate-income families. The original analysis, presented in the main body of the report, did not have access to applicant credit histories. It was thought that these histories might account for nonrandom concentrations of defaults not explained by other borrower or loan characteristics. However, later analysis reported in an appendix replicates a number of the initial analyses with the addition of credit history data. As expected, the updated analysis shows that high-default neighborhoods and lenders have more borrowers with poor credit. Including credit data in the analysis reduces the differential default rate among neighborhoods and lenders but not as much as might be expected. Default concentrations remain even after controlling for differential credit histories; thus, the findings of the original study still hold, albeit on a smaller scale. Readers will also find a detailed discussion of the data and the methodology, as well as appendixes with default data by metropolitan statistical area. Order Assessing Problems of Default in Local Mortgage Markets from HUD USER for $5. Use the order form. The report can also be downloaded from the HUD USER website at www.huduser.gov.
|
