The Federal Financial Institutions Examination Council (FFIEC) has announced the availability of data on mortgage lending transactions at 7,400 U.S. financial institutions covered by the Home Mortgage Disclosure Act (HMDA). Covered institutions include banks, savings associations, credit unions, and mortgage companies. The HMDA data made available today cover 2012 lending activity — applications, originations, purchases and sales of loans, denials, and other actions related to applications.
The data being released today include disclosure statements for each financial institution, aggregate data for each metropolitan statistical area (MSA), nationwide summary statistics regarding lending patterns, and Loan/Application Registers (LARs) for each financial institution (LARs are modified to protect borrower privacy). The FFIEC prepares and distributes this information on behalf of its member agencies.
The HMDA data show the disposition of loan applications and include information on loan amount; loan type (such as conventional, Federal Housing Administration, or Veterans Administration); purpose (home purchase, home improvement, or refinancing); property type (1- to 4-family, multifamily, or manufactured housing); property location (MSA, state, county, and census tract); applicant characteristics (race, ethnicity, sex, and income); and pricing-related data. The data also show whether a loan is subject to the Home Ownership and Equity Protection Act and whether a loan is secured by a first or subordinate lien, or is unsecured. The data as released by the FFIEC include census tract characteristics (minority composition and income).
The 2012 HMDA data are the first to use the census tract delineations and population and housing characteristic data from the 2010 Census and from the American Community Survey (ACS). For the past decade such information was drawn from the 2000 decennial census. Because the annual ACS includes a relatively small sample of households, the FFIEC has determined that the combined 2006-2010 ACS data will serve as the basis for some of the population and housing characteristics used in preparing the disclosure statements covering the HMDA data for 2012 through 2016. The boundaries of many census tracts have been revised in the process of transitioning to the 2010 Census. Users of the HMDA data should be aware that boundary changes and updates to the population and housing characteristics of census tracts complicate intertemporal analysis of the annual HMDA data.
For 2012, the number of reporting institutions of 7,400 fell three percent from the number in 2011, continuing a downward trend since 2006 when HMDA coverage included just over 8,900 lenders. The decline reflects mergers, acquisitions, and the failure of some institutions. The 2012 data include information on 15.3 million home loan applications (of which nearly 9.8 million resulted in loan originations) and 3.2 million loan purchases, for a total of nearly 18.5 million actions. The data also include information on about 477,000 requests for preapprovals related to a home purchase loan. The total number of originated loans of all types and purposes reported increased by about 2.7 million, or 38 percent, from 2011, in part because of a 54 percent increase in the number of refinancings. Home purchase lending also increased, but by a more modest 13 percent.
Since the emergence of difficulties in the mortgage market several years ago, homebuyers have been heavily reliant on government-backed mortgages, particularly those insured by the Federal Housing Administration (FHA) or guaranteed by the Veterans Administration (VA) to finance their purchases. For home purchase lending, the FHA’s share of first-lien loans increased from about five percent in 2006 to a high of 37 percent in 2009. In 2012, the FHA share was 27 percent, down some from 31 percent in 2011. First-lien lending for home purchases backed by VA guarantees also has increased in recent years, although VA-backed lending represents a smaller share of the market than FHA-backed lending. The VA market share of first-lien home purchase loans increased from about two percent in 2006 to nearly eight percent in 2011. Although the number of reported VA home purchase loans increased about 11 percent from 2011 to 2012, the VA market share of home purchase lending remained unchanged from 2011, at about eight percent.
Although FHA- and VA-backed loans play a role in the refinance market, the vast majority of such lending is conventional. In 2012, conventional loans accounted for 85 percent of all refinancings while FHA- and VA-backed loans accounted for about nine percent and 6sixpercent, respectively. Although the shares of FHA- and VA-backed loans in the refinance market are much smaller than are conventional loans, they experienced more growth from 2011 to 2012. Among refinancings, conventional loans increased about 51 percent from 2011 to 2012 while those backed by FHA insurance or VA guarantees increased 78 percent and 90 percent, respectively.
The 2012 HMDA data also include information on loan pricing. Lenders report pricing information for loans classified as “higher-priced.” Higher-priced loans are defined as loans with annual percentage rates (APRs) that are 1.5 percentage points for first-lien loans and 3.5 percentage points for subordinate lien loans above the average prime offer rates (APORs), estimated using data reported by Freddie Mac in its Primary Mortgage Market Survey.
The data on the incidence of higher-priced lending shows that a small minority of first-lien loans in 2012 have APRs that exceeded the loan price reporting thresholds. The principal exception was for conventional first-lien loans used to purchase manufactured homes; for such loans 82 percent exceeded the reporting threshold in 2012. For conventional first-lien loans used to purchase site-built properties, about 3.2 percent of the reported loans exceeded the reporting threshold (down from 3.9 percent in 2011). The incidence of higher-priced lending for FHA-insured loans on the purchase of site-built properties (4.2 percent in 2012) is somewhat higher than for conventional loans. The incidence of higher-priced lending for home purchase loans backed by VA guarantees is notably smaller than for either conventional or FHA-insured loans; only about 0.1 percent of VA-guaranteed loans were higher priced in 2012.
Regarding the disposition of applications for conventional homepurchase loans in 2012, black and Hispanic white applicants experienced higher denial rates than non-Hispanic white applicants. The denial rate for Asian applicants was virtually the same as the corresponding denial rate for non-Hispanic white applicants. These relationships are similar to those found in earlier years.
The HMDA data can facilitate the fair lending examination and enforcement process and promote market transparency. When examiners evaluate an institution's fair lending risk, they analyze HMDA data in conjunction with other information and risk factors, in accordance with the Interagency Fair Lending Examination Procedures. Risk factors for pricing discrimination include, but are not limited to, the presence of broad pricing discretion and consumer complaints.
The HMDA data alone cannot be used to determine whether a lender is complying with fair lending laws. They do not include many potential determinants of creditworthiness and loan pricing, such as the borrower's credit history, debt-to-income ratio, and the loan-to-value ratio. Therefore, when examiners conduct fair lending examinations, including ones involving loan pricing, they analyze additional information before reaching a determination regarding institutions’ compliance with fair lending laws.