A new Federal Reserve Board analysis of millions of home loans made in 2006 shows a correlation between higher-priced loans that carry heftier interest rates and the rate of serious delinquencies.
As in past years, data collected under the Home Mortgage Disclosure Act revealed blacks and Hispanics were more likely to take out such loans than whites, although it remains a matter of debate whether they are targeted for such loans.
Lenders covered by the act reported receiving 27.5 million home loan applications in 2006, including 14 million refinance loan applications, 10.9 million purchase loan applications and 2.5 million home improvement loan requests. The number of reported applications and purchased loans was down about 6 percent from 2005, driven in large part by a 12 percent decline in refinance applications.
The 8,886 reporting lending institutions — including 3,900 commercial banks, 946 savings institutions, 2,036 credit unions and 2,004 mortgage companies — made 14 million mortgage loans, and reported information on 6.2 million loans they purchased from other institutions.
At 29 percent, the denial rate for all home loans in 2006 was up slightly from 27 percent in 2005, but varied greatly by race and ethnicity. For home-purchase loans, the gross mean denial rate was 31.6 percent for blacks, 25.4 percent for Hispanics and 17 percent for Asians, compared with 13.1 percent whites.
Blacks and Hispanics were also more likely to be stuck with higher-cost loans than whites. Under HMDA, higher-cost loans are defined as first-lien loans with annual percentage rates that exceed the interest rate on Treasury securities of similar maturities by 3 percent or more. The threshold for junior loans is 5 percent.
Blacks got higher-cost loans 53.7 percent of the time and Hispanics 46.6 percent of the time in 2006, compared with 17.7 percent of the time for whites. Asians were the least likely of any racial or ethnic group to take out higher-priced loans, at 16.8 percent of the time.
Part of the difference in both denial rates and the incidence of higher-cost loans between ethnic groups can be explained by factors such as property location, income relied on in underwriting, and loan amount, Federal Reserve Board analysts said.
After adjusting for such factors, and factoring in the specific lending institution used by the borrower, the differences between groups were less pronounced.
The adjusted denial rate on purchase loans was 21.5 percent for blacks, 17.5 percent for Hispanics and 14.8 percent for Asians, compared with 13.1 percent for whites. After adjusting for borrower- and lender-related factors, the incidence of higher-cost loans was 30.3 percent for blacks, 24 percent for Hispanics, and unchanged for Asians and whites.
The reason for the remaining, unexplained disparities between different racial and ethnic groups may lie with other criteria lenders use in making underwriting decisions, but which is not collected under HMDA — including credit scores, loan-to-value and debt-to-income ratios, and differences in the choice of loan product.
“Differences in pricing and underwriting outcomes may also reflect discriminatory treatment of minorities or other actions by lenders, including marketing practices,” the study noted. “Further research is needed to assess the extent to which credit- or cost-related factors account for the unexplained differences in loan pricing and denial rates.”
After the Federal Reserve released HMDA data for 2005, a number of groups released their own studies, many of which gave greater credence to the likelihood that lenders were discriminating against minorities.
The U.S. Department of Justice has said it uses HMDA data to investigate whether minorities are targeted for higher-priced loans.
With a rise in delinquencies and foreclosures disrupting mortgage lending and financial markets, the Federal Reserve’s analysis of 2006 HMDA data also looked at whether higher-cost loans are associated with increased delinquencies.
An examination of delinquency rates at the end of March conducted at the metropolitan statistical area county level showed that, in general, areas with elevated rates of higher-priced lending also had elevated rates of serious delinquencies of 90 days or more.
Exceptions were parts of Florida, California and the middle Atlantic region where higher-priced lending was rampant, but which had moderate levels of serious delinquencies. Since the comparison was made, some of those areas have seen sharp increases in delinquency rates.
There were also many counties in Michigan, Indiana, Ohio, Colorado, western Pennsylvania and the south Atlantic region where the rate of higher-priced lending was not unusually high, but where delinquency rates were elevated because of economic conditions.
All in all, the results of the analysis suggested that the incidence of higher-priced lending can be a predictor of loan performance. An increase in the incidence of higher-priced lending of 1 percentage point implies an increase in the rate of serious mortgage delinquency of .03 percent, the study found.
“Although the effect may seem small, it is, in fact, fairly large given the relatively low level of mortgage delinquency,” the study concluded. In a county with the median level of serious delinquency, 1.27 percent, a 10 percent increase in the incidence of higher-priced lending could push delinquency rates into the next “quintile,” or between 1.46 percent and 1.91 percent.”