Popular in Course
verified elite notetaker
Popular in Business
This 77 page Document was uploaded by an elite notetaker on Saturday December 19, 2015. The Document belongs to a course at a university taught by a professor in Fall. Since its upload, it has received 6 views.
Reviews for 2008-HMDA-Fed-Reserve-Analysis
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 12/19/15
The 2008 HMDA Data: The Mortgage Market during a Turbulent Year Robert B. Avery, Neil Bhutta, Kenneth P. Brevoort, Glenn B. Canner, and Christa N. Gibbs, of the Division of Research and Statistics, prepared this article. Cheryl R. Cooper and Christine Coyer provided research assistance. [Note: This article is forthcoming in the Federal Reserve Bulletin.] The Home Mortgage Disclosure Act of 1975 (HMDA) requires most mortgage lending institutions with offices in metropolitan areas to publicly disclose information about their home- lending activity. The information includes the disposition of applications for mortgage credit, the characteristics of the home mortgages that lenders originate or purchase during a calendar year, the location of the properties related to those loans, and personal demographic and other 1 information about the borrowers. The disclosures are intended not only to help the public determine whether institutions are adequately serving their communities’ housing finance needs, but also to facilitate enforcement of the nation’s fair lending laws and to inform investment in both the public and private sectors. 2 The Federal Reserve Board implements the provisions of HMDA through regulation. The Federal Financial Institutions Examination Council (FFIEC) is responsible for collecting the 3 HMDA data and facilitating public access to the information. Each September, the FFIEC releases summary tables pertaining to lending activity from the previous calendar year for each reporting lender and aggregations of home-lending activity for each metropolitan statistical area 1 A description of the items reported under HMDA is provided in appendix A. 2 HMDA is implemented by Regulation C (12 C.F.R. pt. 203) of the Federal Reserve Board. Information about th3 regulation is available at www.federalreserve.gov. The FFIEC (www.ffiec.gov) was established by federal law in 1979 as an interagency body to prescribe uniform examination procedures and to promote uniform supervision among the federal agencies responsible for the examination and supervision of financial institutions. The member agencies are the Board of Governors of the Federal Reserve System, the Federal Deposit Insurance Corporation, the National Credit Union Administration, the Office of the Comptroller of the Currency, the Office of Thrift Supervision, and representatives from state bank supervisory agencies. 2008 HMDA Data 4 (MSA) and for the nation as a whole. The FFIEC also makes available a consolidated data file containing virtually all the reported information for each lending institution. 5 The 2008 HMDA data consist of information reported by about 8,400 home lenders, including all of the nation’s largest mortgage originators. The loans reported are estimated to represent the majority of home lending nationwide. Thus, they likely provide a broadly representative picture of home lending in the United States. This article presents a number of findings from our initial review of the 2008 HMDA data. Three of those findings are noted here. First, the 2008 HMDA data reflect the ongoing difficulties in the housing and mortgage markets. Reported loan application and origination volumes fell sharply from 2007 to 2008 after already falling considerably from 2006 to 2007. The reduction in lending occurred among all groups of borrowers regardless of race, ethnicity, or income, although lending for some groups declined more sharply than for others. Second, the Federal Housing Administration’s (FHA) role in the mortgage market expanded considerably during 2008. The increasing use of FHA-insured loans in 2008 appears to be related to a number of factors, including difficulties faced by private mortgage insurers and their pullback from the marketplace. Third, atypical changes in the interest rate environment, related primarily to widening spreads between the yields on Treasury securities and the interest rates on prime mortgage loans, resulted in a large number of loans being reported as higher priced in 2008 that would not have been so reported a year earlier. As a result, the decline in the incidence of reported higher-priced lending between 2007 and 2008 actually understates the true extent of the decline in higher- priced lending. Because the FHA insured substantially more loans in the second half of the year, when the distortions in the reporting of higher-priced loans were the largest, the effect of these distortions on the reported incidence of FHA higher-priced lending was particularly significant. 4For the 2008 data, the FFIEC prepared and made available to the public more than 51,100 MSA-specific HMDA reports on behalf of reporting institutions. The FFIEC also makes available to the public reports about private mortgage insurance (PMI) activity. Details about thePMI data are provided in appendix B of this article. All the HMDA and PMI reports are available on the FFIEC’s reports website at www.ffiec.gov/reports.htm. 5The only reported items not included in the data made available to the public are the loan application number, the date of application, and the date on which action was taken on the application. Those items are withheld to help ensure that the individuals involved in the application cannot be identified. 2 2008 HMDA Data 2008: A TURBULENT YEAR The 2008 HMDA data reflect a sharp deterioration in economic conditions during the year. The housing market’s continued decline was reflected in the Federal Housing Finance Agency’s (FHFA) nationwide home price index, which posted a year-over-year decline of more than 8 percent by November 2008, compared with less than 3 percent in January. At the same time, mortgage-related losses continued to weigh on the confidence of investors and the health of financial institutions. A number of major financial institutions either failed, merged under distress, or received government assistance. The government-sponsored enterprises (GSEs) Fannie Mae and Freddie Mac were placed into conservatorship by the FHFA in September. 6 Difficulties in the housing and financial markets advanced into a broad-based economic recession. By December 2008, the unemployment rate had risen to 7.2 percent from 4.9 percent a year earlier, and the number of employed individuals fell by nearly 3 million during the year. 8 The deterioration in household income and wealth as well as fears about buying into a falling market may have weakened demand for housing and mortgages. On the supply side, strained lending institutions, facing the risks posed by falling home prices and a weakening economy, were apprehensive or unable to offer loans that did not have some form of government backing. Potential borrowers, especially those with blemished credit histories and those seeking “jumbo” mortgages, likely found it more difficult than in previous years to obtain a mortgage. Those with adequate credit histories but little money for a down payment also faced a more challenging situation since private mortgage insurance (PMI) companies, which suffered large losses in 2007 and 2008, tightened their standards and raised prices.10 Lenders also sharply curtailed the issuance of second-lien loans used heavily in 6 To maintain the GSEs’ ability to purchase home mortgages, the Treasury announced plans to establish a backstop lending facility for the GSEs, to purchase up to$100 billion of preferred stock in each of the two firms, and to initiate a program to purchase agency mortgage-backed securities. See Board of Governors of the Federal Reserve System (2009), Monetary Policy Report to the Congress (Washington: Board of Governors, February), www.fede7alreserve.gov/monetarypolicy/mpr_20090224_part1.htm. 8 The National Bureau of Economic Research declared the start of the recession as December 2007. Employment statistics from the Bureau of Labor Statistics; based on individuals 16 years or older. 9 Industry sources indicate that the dollar amount of originations of subprime loans fell 88 percent from 2007 to 2008, to a level of $23 billion. Jumbo loans are loans that exceed the size limits set for loans that Fannie Mae and Freddie Mac are permitted to purchase (commonly referred to as conforming loans). Available data indicate that the dollar amount of originations of jumbo loans fell 72 percent from 2007 to 2008, to a level of $97 billion. See Inside Mortgage Finance (2009), The 2009 Mortgage Market Statistical Annual, Vol. 1: The Primary 10rket (Bethesda, Md.: Inside Mortgage Finance Publications). See Mortgage Insurance Companies of America (2009), 2009–2010 Fact Book & Member Directory (Washington: MICA), www.privatemi.com/news/factsheets/2009-2010.pdf. 3 2008 HMDA Data previous years to help finance home purchases. Partly in response to difficulties in the private market, the government raised the size limits on loans eligible to be purchased by Fannie Mae or Freddie Mac and insured by the FHA as well as the guarantee limit for loans backed by the Department of Veterans Affairs (VA) as part of the Economic Stimulus Act of 2008. MORTGAGE MARKET TRENDS FROM THE HMDA DATA For 2008, 8,388 institutions reported under HMDA: 3,942 commercial banks, 913 savings institutions (savings and loans and savings banks), 2,026 credit unions, and 1,507 mortgage 11 companies (table 1). The number of reporting institutions fell nearly 3 percent from 2007, primarily because of a relatively large decline in the number of independent mortgage companies—that is, mortgage companies that were neither subsidiaries of depository institutions nor affiliates of bank or savings institution holding companies that reported data. Fifteen institutions that reported data in 2007 ceased operations during 2008 or early 2009 and did not report data on their 2008 lending activity. Reporting lenders submitted information on 14.2 million applications for home loans of all types in 2008, down 34 percent from 2007 and almost 50 percent from 2006 (table 2). Lenders also reported information on 2.9 million loans that they had purchased from other institutions and on 276,000 requests for preapprovals of home-purchase loans that did not result in an application for a loan (preapproval data not shown in table). The top panel of figure 1, which shows the monthly counts of loans, indicates a 12 downward trend in home-purchase lending from 2006 to 2008. For instance, the 2006 peak month for home-purchase lending (in June) was more than 400,000 loans, compared with less 11Not all mortgage lenders have to provide HMDA data. Depositories must have had an office in a metropolitan area and had assets of more than $37,000,000 atthe end of 2007 to report data for 2008. For filing year 2008, 55.7 percent of the commercial banks in existence on December 31, 2008, filed HMDA data. However, the filers had 93.0 percent of the total mortgage dollars outstanding on commercial bank portfolios at that time. For savings institutions, 70.9 percent of existing institutions holding 94.1 percent of the mortgage dollars filed. For credit unions, only 25.4 percent of the institutions filed;however, these institutions held 92.5 percent of the mortgage dollars outstanding on credit union balance sheets. Independent mortgage banks needed to meet other criteria related to their dollar volume of mortgage lending, the share of mortgage lending of their total lending,and their lending in metropolitan areas to be eligible for reporting. There is no comprehensive list of independent mortgage lenders, so it is difficult to know the full scope of HMDA data coverage of such lenders. 12Lenders report the date on which action on an application is taken. For originations, the “action taken” date is the closing date or date of loan origination for the loan. This date is the one we use to compile data at the monthly level. To help ensure the anonymity of the data, these dates are not released in the publically available HMDA data. 4 2008 HMDA Data than 300,000 loans at the peak month (June) in 2008. The bottom panel of figure 1 indicates that refinance lending jumped at the beginning of 2008 to a level in February exceeding any month in 2006 or 2007. Refinance lending then fell sharply during the remainder of 2008. Figure 1 also shows that the annual percentage rate (APR) for a 30-year fixed-rate prime mortgage fell sharply at the end of 2007 to levels not seen in several years; it continued to fall in early 2008 and dipped below 6 percent in January 2008, which may have triggered the jump in refinance lending. 13 The Potential Effect of Nonreporters on Lending Volume in the 2008 HMDA Data As part of the HMDA data collection effort, the Federal Reserve Board tracks each financial institution that is expected to report (including all lenders that reported data for the previous 14 calendar year) and then contacts those that did not submit a report. In some cases, nonreporting is due to a cessation of business; in others, it is the result of a merger, acquisition, or consolidation. When a merger, acquisition, or consolidation occurs, all lending by the institutions covered by HMDA in that year is reported by the surviving entity; only when an institution goes out of business is the volume of reported loans possibly affected. The Federal Reserve’s respondent tracking report records what happened to each institution that failed to report. For institutions that ceased operations, the tracking report also records, to the extent possible, the month that operations were discontinued. The tracking report indicates that 15 institutions that reported HMDA data for 2007 ceased operations during 2008 or 15 at the beginning of 2009 and did not report lending activity for 2008. Of the 15 nonreporting institutions, 3 were banking institutions and 12 were independent mortgage companies. Although it is not possible to know how many loans these 15 institutions originated in 2008 before discontinuing operations, one can gauge their potential importance by measuring their lending activity in 2007. In the aggregate, these 15 nonreporting companies accounted for about 5 percent of all conventional first-lien loans for site-built properties in the 2007 HMDA 13 The APRs for prime loans are based on data from Freddie Mac’s Primary Mortgage Market Survey and reflect interest rates and points offered to consumers during the first three days of each week. For more details, see note 29. Loan counts in figure 1 are aggregated to the monthly level using the date of loan origination, as opposed to an earlier date when the interest rate for the loan waslocked. If the HMDA data were aggregated using the “lock” date, th14spike in refinancings would likely occur closer to the January dip in the APR. Sometimes contacting a nonreporting lender is impossible because the firm has ceased operations. 15The list of lenders that ceased operations and did not report is as comprehensive as possible at this time. If additional information becomes available, the list will be updated on the Federal Reserve Board’s website. For a list of the institutions that ceased operations and did notreport, see appendix table A.1, which has been posted separately as an Excel file. 5 2008 HMDA Data 16 data (data not shown in tables). The tracking reports indicate that the 15 nonreporting institutions had exited the marketplace by the middle of 2008, so their effects on the completeness of the HMDA data are confined to the first half of the year. Government-Backed Lending Government-backed loans—those insured by the FHA and those backed by guarantees from the VA, the Farm Service Agency, or the Rural Housing Service—rose in 2008 relative to 2007. The rise in FHA-insured lending was particularly large. The number of reported FHA-insured loans was almost three times greater in 2008 than in 2007, and the FHA-insured share of home- purchase and refinance loans rose to more than 21 percent in 2008 from less than 6 percent in 17 2007 (table 3). Moreover, by December of 2008, the FHA’s share of home-purchase and refinance lending was about 30 percent (data not shown in tables). Lenders typically require borrowers to purchase mortgage insurance (through the FHA or PMI companies) or a credit guarantee (through the VA, for example) when the borrower provides a small down payment. 18 Such credit enhancements protect lenders against loss if the borrower defaults. The VA guarantees a percentage of the loan amount up to a certain limit (but with no cap on the loan size), while the FHA cannot insure mortgages that are larger than legislated limits. Historically, these limits have been set at levels that were sufficiently low that many homebuyers in areas with high home prices have not been able to use these programs. Under the Economic Stimulus Act of 2008, the limits were raised in high-cost areas. In a later section, “The Surge in FHA and VA Lending,” we will analyze more closely the contribution of increased limits to the increase in FHA and VA-backed lending. We will also examine whether difficulties facing PMI companies contributed to the shift to government-backed lending. 16 17Market shares reported in this article are based on the number of loans and not the dollar amounts. Loans are for owner-occupied, one- to four-family properties. Junior-lien loans and loans for manufactured homes are included because the HMDA data prior to 2004 do not separately identify these loans. The FHA share of home-purchase and refinance lending in 2008, excluding junior-lien and manufactured-home loans, was 22.5 percent. 18For more details about PMI, see appendix B, “Private Mortgage Insurance Data.” 6 2008 HMDA Data Loan Sales The HMDA data document the importance of the secondary market for home loans. Just over 73 percent of the first-lien home loans reported in 2008 were sold during the same year (table 4).19 Notably, the rise in government-backed lending between 2007 and 2008 described earlier has resulted in a sharp increase in the proportion of loans sold into pools guaranteed by the Government National Mortgage Association (Ginnie Mae). More prominent in the secondary market are the GSEs. For the most part, the purchases made by Fannie Mae and Freddie Mac consist of conventional loans originated to purchase homes or to refinance existing loans. Fannie Mae and Freddie Mac are restricted by law to purchase mortgages with origination balances below a specific amount, known as the conforming loan limit. As with the FHA loan limits mentioned earlier, the Economic Stimulus Act of 2008 increased the conforming loan limits. 20 In 2008, sales to Fannie Mae and Freddie Mac accounted for about 42 percent of the loans reported as sold, compared with about 28 percent in 2006. At least in part, this increase in market share reflects the reduction during this period in the higher-priced share of loans, which the GSEs typically do not purchase directly. Higher-priced loans were often sold through the private securitization process; indeed, loans sold through this process diminished considerably, from about 10 percent of sold loans in 2006 to less than 1 percent in 2008. Credit Unions A credit union is a cooperative financial institution formed by a group of people with a common bond, such as employees of a firm or members of a religious organization, university, or 21 governmental entity. Members of a credit union pool their funds to extend credit to their fellow members. In 2008, about 7,700 credit unions across the country served upward of 90 million members. The vast majority of credit unions are small measured by asset size, and many do little home lending. As such, only about 2,000 credit unions report under HMDA each year (table 1). Unlike other types of lenders, credit unions have not experienced a significant reduction 19Loans that are sold in a different calendar year than the year of origination are recorded as being held in the lender’s portfolio in the HMDA data. 20For more on the conforming loan limit, see www.fhfa.gov/Default.aspx?Page=185. 21The notion of a common bond has been expanded some in recent years, for example, to include individuals from broad geographic areas. 7 2008 HMDA Data in home-lending activity over the past couple of years (table 5). As a consequence, their share of one- to four-family site-built HMDA loans has risen, particularly for junior liens (28.3 percent in 2008). Their high market share of junior liens can be explained, in part, by the collapse of the piggyback market, discussed later in the section “Piggyback Lending.” Piggyback junior-lien home-purchase loans are issued as part of a purchase package. Less than 5 percent of credit union junior liens have been for home purchases, so they were not particularly affected by this collapse. The credit union data afford a unique opportunity to benchmark the HMDA data. Unlike other depositories, all credit unions are required to report their aggregate first- and junior-lien mortgage originations each year as part of their regulatory filings. These data allow a determination of the HMDA-filer coverage relative to all credit union mortgage lending. These data show that for 2008, about 88 percent of all credit union mortgage originations were made by lenders who reported under HMDA. For first liens, the numbers reported in regulatory filings by these lenders corresponded relatively closely to the number reported in HMDA (95 percent of first-lien loan originations are reported in HMDA, data derived from table 5). However, for closed-end junior liens, only about 50 percent appear to be reported, which suggests that many of these loans are new junior liens not reportable under HMDA rules. Lending for Manufactured Homes Since 2004, the HMDA data have distinguished between loans secured by site-built properties and those related to manufactured homes. 22 Manufactured-home lending differs from lending for site-built properties along a number of dimensions, including typical loan amounts, borrower incomes, and the share of such loans that are higher priced. The reported number of manufactured-home loans fell by about the same proportion as for site-built homes from 2007 through 2008 (table 6). However, when measured from 2005 (a year when mortgage markets were quite robust), the decline in loan activity was much steeper for site-built homes than for manufactured homes. Over this longer period, the number of loans to buy site-built homes fell 48 percent, and the number to buy manufactured homes fell 25 percent. 22 For more information about the reporting details, seeRobert B. Avery, Glenn B. Canner, and Robert E. Cook (2005), “New Information Reported under HMDA and Its Application in Fair Lending Enforcement,”Federal Reserve Bulletin, vol. 91 (Summer), pp. 344–94. 8 2008 HMDA Data The mean loan amount used to purchase manufactured homes in 2008 was $75,000, which was much smaller than the mean loan amount of $217,000 for site-built homes. Similarly, the mean income of borrowers purchasing manufactured homes in 2008 was $48,400, which was much smaller than the mean income of $93,300 for purchasers of site-built homes for the same period. Non-Owner-Occupant Lending One factor contributing to the strong performance of housing markets over the first half of this decade was the growth in sales of homes to investors or individuals purchasing second or vacation homes, which are collectively referred to here as non-owner-occupied units. 23 From 1996 through 2005, the share of non-owner-occupant lending used to purchase one- to four- family site-built homes rose each year, increasing from 6.4 percent to 17.3 percent over the period (table 7). This share has since fallen to 13.5 percent in 2008. Currently, loans for non-owner occupants are not eligible for the FHA or VA programs. However, the GSEs can purchase non-owner-occupied loans that otherwise meet their requirements, but they typically demand interest rates that are about 3/8 of a percentage point higher than the interest rates on loans for similar owner-occupied properties. Perhaps reflecting less of an appetite for such loans on the part of private lenders, the GSE market share of both home-purchase and refinance non-owner-occupied lending grew about 10 percentage points from 2007 to 2008 (33.8 percent to 43.1 percent for home-purchase lending and 28.4 percent to 39.2 percent for refinance lending). Nevertheless, non-owner-occupied lending remained a comparatively small part of overall GSE lending in 2008 (17.9 percent of home-purchase lending and 11.3 percent of refinance lending; data not shown in tables). Piggyback Lending In recent years, piggyback loans emerged as an important segment of the conventional mortgage market, particularly regarding loans to purchase homes. In piggyback lending, borrowers simultaneously receive a f rst-lien mortgage and a junior-lien (piggyback) loan. The piggyback 23An investment property is a non-owner-occupied dwelling that is intended to be rented or resold for a profit. Some non-owner-occupied units—vacation homes and second homes—are for the primary use of the owners and thus would not be considered investment properties. The HMDA data do not, however, distinguish between these two types of non-owner-occupied dwellings. 9 2008 HMDA Data loan finances the portion of the purchase price not being financed by the first mortgage and sometimes any cash payment that might have been made; the junior-lien loan may amount to as much as 20 percent of the purchase price. In many cases, borrowers used piggyback loans to avoid the need to obtain PMI. Sometimes, piggyback loans were used to keep the size of the first-lien loan within the Fannie Mae and Freddie Mac conforming loan limits so the borrower could take advantage of the lower interest rates available on conforming loans. The HMDA data help document the extent of piggyback lending over time. However, because not all lenders submit HMDA data, some of the junior-lien loans that are reported may not have the corresponding first-lien loan reported, and some of the first-lien loans that are reported may not have the associated junior-lien loan reported. Also, some piggyback loans may be open-end loans which do not need to be reported under HMDA. The HMDA data for 2005 and 2006 show that lenders extended about 1.3 million junior- lien loans to help individuals purchase homes (for both owner-occupant and non-owner-occupant purposes) in each of these years (data not shown in tables). The number of reported junior-lien loans contracted sharply in 2007 to about 600,000 such loans. This contraction continued as the number of junior-lien loans declined by 84 percent from the 2007 level to only about 98,000 loans in 2008. A loan-matching process can be undertaken to determine which reported junior-lien loans in the HMDA data appear to be associated with the appropriate reported first-lien loans. 25 Our matching algorithm indicates that in 2008, 2.7 percent of the nearly 1.6 million first-lien conventional loans to purchase one- to four-family site-built owner-occupied homes involved a piggyback loan reported by the same lender, a proportion that was down 77 percent from 2007 (table 8). 24 One advantage of piggyback loans over those backed by PMI insurance was that PMI payments made by the borrower did not qualify as deductible interest under Internal Revenue Service (IRS)guidelines, whereas interest payments on many piggyback loans did. The Congress allowed the deductibility of PMI premiums of some borrowers starting in 2007, which reduced the relative attractiveness of piggybacks. 25For the analysis here, a junior-lien loan was identified as a piggyback loan to a reported first-lien loan if both loans (1) were conventional loans involving property in the same census tract; (2) were originated by the same lender with approximately the same dates of loan application and closing; and (3) had the same owner-occupancy status and identical borrower income, race or ethnicity, and sex. 10 2008 HMDA Data THE DISPOSITION OF APPLICATIONS BY LOAN CHARACTERISTICS IN 2008 Thus far, our analysis of the 2008 HMDA data has focused primarily on how the mortgage market has evolved over the past few years. In this section, we examine the information provided by HMDA about what home lending looked like in 2008. Tables 9 and 10 categorize every loan application reported in 2008 into 25 distinct product categories characterized by loan and property type, purpose of the loan, and lien and owner-occupancy status. Each product category contains information on the number of total and preapproval applications, application denials, originated loans, loans with prices above the reporting thresholds established by HMDA reporting rules for identifying higher-priced loans, loans covered by the Home Ownership and Equity Protection Act (HOEPA), and the mean and 26 median APR spreads for loans reported as higher priced . The 2008 HMDA data include information on 14 million loan applications, about 12 million of which were acted upon by the lender (table 9). The vast majority of these applications were for first-lien loans on one- to four-family site-built homes. Among these applications, about two-thirds of home-purchase applications and four-fifths of refinance applications were for conventional loans. These shares of applications for conventional loans are considerably lower than were observed in earlier years (data not shown in tables). Patterns in the denial rates are consistent with what has been observed in earlier years. Denial rates on applications for home-purchase loans are generally lower than those observed for either refinance or home-improvement loans. Denial rates on applications backed by manufactured housing are generally higher than those backed by site-built homes. Furthermore, requests for a first-lien, conventional, home-purchase loan backed by a manufactured home is the only one of the 25 product categories for which the majority of applications are denied. In addition to the application data provided under HMDA, about 734,000 requests for preapprovals that were acted on by the lender were reported under HMDA (table 10). Almost one-quarter of these requests for preapproval were denied by the lender. Of the applications 26 The type of information provided in tables 9 and 10 is identical to that provided in analyses of earlier years of HMDA data. Comparisons of the numbers in these two tables with earlier years can be made by consulting the following articles: Robert B. Avery, Kenneth P.Brevoort, and Glenn B. Canner (2008), “The 2007 HMDA Data,” Federal Reserve Bulletin, vol. 94, pp. A107–A146; Robert B. Avery, Kenneth P. Brevoort, and Glenn B. Canner (2007), “The 2006 HMDA Data,”Federal Reserve Bulletin, vol. 93, pp. A73–A109; Robert B. Avery, Kenneth P. Brevoort, and Glenn B. Canner (2006), “Higher-Priced Home Lending and the 2005 HMDA Data,” Federal Reserve Bulletin, vol. 92, pp. A123–A166; and Avery, Canner, and Cook, “New Information Reported under HMDA.” 11 2008 HMDA Data acted on by the lender and preceded by requests for preapproval, more than 88 percent were approved (data derived from table 10). The HMDA data also indicate which loans were covered by HOEPA. Under HOEPA, certain types of mortgage loans that have rates or fees above specified levels require additional 27 disclosures to consumers and are subject to various restrictions on loan terms. For 2008, 2,281 lenders reported extending about 7,750 loans covered by HOEPA (data regarding lenders not shown in tables). In comparison, lenders reported on about 11,500 loans covered by HOEPA in 2007. In the aggregate, HOEPA-related lending made up less than 0.2 percent of all the originations of home-secured refinance mortgages and home-improvement loans reported for 2008 (data derived from table 9). 28 Relative to previous years, a smaller proportion of loans were reported as higher priced in 2008, and a larger proportion of reported higher-priced loans had an APR less than 1 percentage point above the reporting threshold. Furthermore, a substantial fraction of loans in 2008 were reported as higher priced because of atypical changes in the interest rate environment, rather than because the loans represented relatively high credit risk. We discuss this issue in detail in the next section and formulate an adjusted measure of higher-priced loans that is more consistent over time. THE 2008 HMDA DATA ON LOAN PRICING When analyzing the loan pricing information in the HMDA data, one may be tempted to assume that changes in the incidence of reported higher-priced lending reflect changes in subprime lending activity. This interpretation, however, ignores a number of factors that can alter the incidence of reported higher-priced lending without any corresponding changes in subprime lending activity. In 2008, we identify two such factors that were related to the overall interest rate environment and may have led to variation over time in whether a loan was reported as higher priced in HMDA. Understanding how these changes in the interest rate environment 27The requirement to report HOEPA loans in HMDA relates to whether the loan is subject to the original protections of HOEPA, as determined by the coverage test in the Federal Reserve Board’s Regulation Z, 12 C.F.R. pt. 226.32(a). The required reporting is not triggered by the more recently adopted protections for “higher-priced mortgage loans” under Regulation Z, notwithstanding that those protections were adopted under authority given to the Board by HOEPA. See 73 Fed. Reg. 44522 (July 30, 2008). The more recent HOEPA regulations do not take effect until October 1, 2009. 28HOEPA does not apply to home-purchase loans. 12 2008 HMDA Data affected the reported incidence of higher-priced lending is important when attempting to draw inferences about how lending to high-risk borrowers has changed. In the following sections, we discuss how changes in the interest rate situation during 2007 and 2008 may have affected the reported incidence of higher-priced lending. We then present the methodology we use to adjust for changes in the interest rate environment in a manner that provides a clearer picture of how home lending to high-credit-risk borrowers has changed. We then discuss what the 2008 HMDA data indicate about lending to high-risk borrowers. How the Interest Rate Situation Affected the Reporting of Higher-Priced Loans The reporting rules governing HMDA require lenders to use the yield on a Treasury security with a comparable term to maturity in determining whether a loan was required to be reported as higher priced under HMDA. Because most mortgages prepay well before the stated term of the loan, lenders typically use relatively shorter-term interest rates when setting the price of mortgage loans. For example, lenders often price 30-year fixed-rate mortgages based on the yields on securities with maturities of fewer than 10 years, and they typically set interest rates on adjustable-rate mortgages (ARMs) on the basis of securities with much shorter terms. Thus, a change in the relationship between shorter-term and longer-term yields can affect the reported incidence of higher-priced lending. For example, if short-term interest rates fall relative to long- term rates, then the number and proportion of loans reported as higher priced will fall even if other factors, such as lenders’ underwriting practices or borrowers’ characteristics, are unchanged. For ARMs, this effect is further exacerbated by the manner in which APRs are calculated. The interest rates on most ARM loans, after the initial interest rate reset date, are typically set based on interest rates for one-year securities. As a result, the APRs for ARMs— which take into account the expected interest rates on a loan, assuming that the loan does not prepay and that the index rates used to establish interest rates after the reset do not change—will be particularly sensitive to changes in one-year interest rates. Consequently, higher-priced lending reported for ARMs will fall when one-year interest rates decline relative to other rates even if the relationship between long-term and intermediate-term rates is constant. The relationship between shorter- and longer-term interest rates can be seen in the yield curve for Treasury securities, which displays how the yields on these securities vary with the 13 2008 HMDA Data term to maturity. Through the first seven months of 2007, the yield curve was relatively flat and then began to steepen, so that the differences between the yield on a 30-year Treasury security and the yields on the five-year and one-year Treasury securities increased (figure 2). Overall, this steepening continued in 2008; while spreads did narrow during the spring and at the very end of 2008, they remained consistently above the spreads observed in 2007. As discussed earlier, this change would be expected to decrease the incidence of reported higher-priced lending, particularly for ARMs, even in the absence of any changes in high-risk lending activity. In addition to the steepening yield curve, a second change in the interest rate environment affected the likelihood that a loan was reported as higher priced in HMDA in 2008. As a result of the “flight to quality” and liquidity concerns caused by the financial crisis late in 2008, the spreads between the yields on Treasury securities and other securities and loans, including 30-year fixed-rate loans, widened considerably. At the beginning of 2008, the HMDA reporting threshold was 7.66 percent, and the APR on a 30-year fixed-rate prime loan, based on the rates reported by Freddie Mac’s Primary Mortgage Market Survey (PMMS), was 6.12 percent (figure 3).29 This difference resulted in a gap between the HMDA reporting threshold and the APR on a prime 30-year fixed-rate loan of 1.54 percentage points. By the end of 2008, this gap had narrowed to approximately 0.77 percentage point, as the falling yields on Treasury securities pulled the HMDA reporting threshold closer to the prime mortgage rate. As a result, an increasing share of near-prime loans would have been reported as higher priced toward the end of 2008 over what had been reported earlier in the year. Widening spreads between the interest rates on Treasury securities and the rates on prime mortgage loans would be expected to increase the overall incidence of higher-priced lending, even if the credit- risk profile of borrowers remained unchanged. These two changes in the interest rate environment in 2008, therefore, worked in opposite directions. The expected net effect of these two competing forces can be discerned from figure 3. The top line in that figure shows the HMDA reporting threshold in effect from 2006 through 2008. The middle three lines show the APRs calculated from the interest rates reported in Freddie Mac’s PMMS for the three 30-year loan products reported in that survey: a fixed-rate 29 The weekly Freddie Mac Primary Mortgage Market Survey reports the average contract rates and points for all loans and the margin for adjustable-rate loans forloans offered to prime borrowers (those with the lowest credit risk). The survey currently reports information for two fixed-rate mortgage products (30 year and 15 year) and two ARM products (one-year adjustable rate and a five-year adjustable rate). For more information, see www.freddiemac.com/dlink/html/PMMS/display/PMMSOutputYr.jsp. 14 2008 HMDA Data loan, a 5-year ARM, and a 1-year ARM. As expected, the steepening of the yield curve had a much larger effect on the APRs associated with ARMs than on fixed-rate loans, though rates on all three products were generally lower in 2008 than they had been in earlier years. The change during 2008 in the spreads between the APRs on these prime loans and the HMDA reporting threshold (shown by the bottom three lines in figure 3) suggests that the net effect of these changes depended upon whether the loan had either a fixed or adjustable rate. For ARMs, the spreads appeared to have widened substantially in 2008, suggesting that the incidence of reported higher-priced lending for these loans should have decreased in 2008 even without changes in borrower characteristics. For fixed-rate loans, spreads appear to have narrowed relative to earlier years. Consequently, the incidence of reported higher-priced lending for fixed- rate loans should have increased. The difference in the net effects of the changes in the interest rate environment between fixed- and adjustable-rate loans complicates an analysis of the HMDA data, because one cannot determine whether a loan in the HMDA data is a fixed- or adjustable-rate loan. Using industry data, however, it is possible to estimate the monthly volume of both loan types. 30 These data show that at the beginning of 2007, ARMs accounted for about 17.8 percent of the market, falling to a range of between 5 and 6 percent at the beginning of 2008 (table 11). During 2008, ARM activity continued to fall (particularly in the latter portion of the year) to less than 2 percent. Given the small share of ARMs in the marketplace in 2008, any distortions in the incidence of reported higher-priced lending caused by changes in the interest rate environment can be attributed to fixed-rate lending. Adjusting for Changes in the Interest Rate Environment, 2006-2008 The changes in the interest rate environment discussed in the previous section can result in loans of a given level of credit risk being reported as higher priced in the HMDA data at some points in time but not others. This variation makes drawing inferences about changes in high-credit- risk lending based upon changes in the incidence of reported higher-priced lending much more complicated. To better isolate the credit-risk component of pricing so that we have a definition of a “higher-priced loan” that is more constant over time and, therefore, more fully reflective of 30 Source: Lender Processing Services, Inc. (LPS). LPS claims coverage of about 70 percent of the mortgage market, including all loans of 9 of the top 10 mortgage servicers (see www.lpsvcs.com). 15 2008 HMDA Data high-risk lending activity, we constructed an adjusted measure. We defined the credit-risk component of a loan as the difference between the APR on that loan and the APR available to the lowest-risk prime borrowers at that time. This credit risk 31 component is assumed to be constant over time. In other words, we assume that a nonprime borrower who received a loan with an APR that was 0.25 percentage point above the APR available to prime borrowers at that time would receive, if the nonprime borrower’s characteristics remained constant, a loan that was 0.25 percentage point above the available rate for prime borrowers at all other times, regardless of any changes in the interest rate environment. We then examined the share of loans over time with credit-risk components above specific thresholds. This approach should provide a more accurate depiction of how the extent of high- risk lending has changed that is relatively free of the distortions introduced in the incidence of reported higher-priced lending by changes in the interest rate environment. In estimating the credit-risk component of loans in the HMDA data, we used, as the measure of the rate available to prime borrowers, the APR derived from the information reported 32 in the Freddie Mac PMMS for a 30-year fixed-rate loan. As an approximation of the APR on loans in HMDA, we added the reported spread (for higher-priced loans) to the appropriate HMDA reporting threshold for a 30-year loan. We refer to the resulting estimate of the credit- 33 risk component as the PMMS spread. PMMS spreads can only be calculated for loans with reported spreads in HMDA. Loans with PMMS spreads below 0.95 percentage point would not have been reported as higher priced at any time between 2006 and 2008. We are therefore unable to identify these loans in the data. Loans with PMMS spreads between 0.95 and 1.75 percentage points would have been reported as higher priced at some points during the three years but not at others, so we can only identify 31 The credit-risk component that we are defining here may include other risk components besides credit risk (for example, prepayment risk). 32By using the APR for the 30-year fixed-rate mortgage, we are implicitly treating all loans in the HMDA data as though they were 30-year fixed rate loans. Because of the small market share for ARMs and the prevalence of 30-year loans, we do not expect this methodological approach to have a substantive effect on our analysis. However, note that the share of loans that were ARMS in 2006 and early 2007 was much higher than in 2008. As such, one should exercise caution when comparing incidences of adjusted higher-priced lending across these periods.33 Under new rules adopted by the Federal Reserve Board in 2008, the spread between a loan’s APR and the APR of comparable prime PMMS loan will be used to determine whether a loan is reported as higher priced in HMDA. The new rules take effect for all loans with application dates on or after October 1, 2009, and for loans regardless of application date if originated in 2010. APRs of first–lien loans with a PMMS-APR spread of 1.50 percentage points or more must be reported. For second-lien loans, the reporting threshold is a PMMS-APR spread of 3.50 percentage points. 16 2008 HMDA Data these loans at some points in time. Only those loans with a PMMS spread of more than 1.75 percentage points have been consistently identified in the HMDA data as higher priced. Therefore, we focused on loans with a PMMS spread greater than 1.75 percentage points in examining how high-risk lending has changed over time, as this measure should be free of the distortions introduced by changes in the interest rate environment and should more accurately reflect changes in high-risk lending activity over time. We refer to loans with a PMMS spread in excess of 1.75 percentage points as adjusted higher-priced loans. Incidence of Higher-Priced Lending As in earlier years, most loans reported in 2008 were not higher priced as defined under HMDA reporting rules. Among all the HMDA-reported loans secured by one- to four-family properties, 11.6 percent were higher priced in 2008, down significantly from the historic high point of 28.7 percent in 2006 and from 18.3 percent in 2007 (data for 2008 shown in table 3; data for 2006 and 2007 are not shown in tables). The incidence of higher-priced lending fell from the 2007 levels for all conventional loan product categories, with the exception of those related to manufactured homes. Looking exclusively at changes in the annual rates of higher-priced lending can obscure the information about how the mortgage market is developing over time. To better illustrate how changes in higher-priced lending have played out in recent years, we examined monthly patterns in higher-priced lending activity. The top line in the upper panel of figure 4 shows the incidence of reported higher-priced, home-purchase lending. The monthly data show that the overall annual decline in the incidence of higher-priced lending between 2007 and 2008 obscures a substantial rebound in the incidence of reported higher-priced lending in the second half of 2008. A similar rebound in the incidence of reported higher-priced lending is observed for the refinance loans (shown in the bottom panel of figure 4). This rebound in the incidence of reported higher-priced lending appears to reflect changes in the interest rate environment and not changes in actual high-risk lending activity. Using our methodology to correct for distortions caused by changes in the interest rate environment, we see that the share of adjusted higher-priced loans (shown in figure 4 as “PMMS + 1.75”) continued to decline in 2008 and remained at historically low levels, even when the incidence of reported higher-priced lending in HMDA began to increase. There does appear to 17 2008 HMDA Data have been something of a rebound in the share of adjusted higher-priced home-purchase loans at the very end of 2008, though, even after this increase, the incidence of adjusted higher-priced lending remained below the levels observed throughout 2007. The pattern for refinance lending appears somewhat different than that for home- purchase lending. The incidence of adjusted higher-priced refinance lending fell at the beginning of 2008 and then remained relatively flat throughout the rest of the year. The timing of this decline, and the fact that a similar decline was not observed for home-purchase lending, suggests that this may be the result of a changing mix of borrowers caused by the refinancing boom in early 2008. This refinancing boom, which coincided with a sharp decline in the prime mortgage rate, may have encouraged a large number of high-credit-quality borrowers to refinance their prime mortgages in order to take advantage of relatively low mortgage rates. A tendency of high-credit-quality borrowers to refinance when rates are low and to refrain when rates are high may explain why the incidence of adjusted higher-priced refinancing lending exhibits more variation than home-purchase lending. A comparison of the incidence of adjusted higher-priced lending and volume of refinancing suggests that increases (decreases) in refinancing activity often occur at the same time as decreases (increases) in the incidence of adjusted higher-priced lending (figures 3 and 4). Figure 4 also shows the share of home-purchase and refinance lending that was composed of loans with PMMS spreads of more than 2.75 percentage points (shown in the figure as “PMMS + 2.75”) and more than 3.75 percentage points (“PMMS + 3.75”). Most of the adjusted higher-priced loans had PMMS spreads in excess of 2.75 percentage points for most of 2006. In 2007, this circumstance changed dramatically as the shares of both home-purchase and refinance lending accounted for by these loans fell precipitously. While starting 2008 from much lower levels than previous years, the share of loans made up of these loans that were very higher priced continued to fall in 2008, though the decline seems to have slowed somewhat. Nevertheless, loans with PMMS spreads in excess of 2.75 percentage points now account for a negligible share of home-purchase lending and for a very small share of refinance lending. This suggests that, as in 2007, the decline in the incidence of adjusted higher-priced lending has been greater for the highest-risk borrowers. 18 2008 HMDA Data Higher-Priced Lending by Lender Type Higher-priced lending activity can also differ by type of lender. Three types of lender are considered here: depository institutions, subsidiaries or affiliates of depository institutions, and independent mortgage companies. In 2006, independent mortgage companies originated almost one-half of all higher-priced loans and accounted for about 31.7 percent of all first-lien loans (table 12). For that year, depository institutions accounted for a smaller share of higher-priced lending (26.8 percent of adjusted higher-priced lending) than independent mortgage companies. Since 2006, the share of higher-priced loans originated by independent mortgage companies has fallen dramatically. Independent mortgage companies accounted for 18.5 percent of reported higher-priced loans in HMDA in 2008, down from 45.7 percent of such loans in 2006. When using the adjusted higher-priced loan definition, the decline has been even steeper (particularly between 2007 and 2008), with the share of higher-priced loans extended to independent mortgage companies falling to 12 percent. The share of adjusted higher-priced loans originated by depository institutions has increased substantially from 26.8 percent in 2006 to 61.6 percent in 2008, though the incidence of adjusted higher-priced lending has also fallen for depository institutions over this period from 14.7 percent to 5.6 percent. These numbers suggest that the increased share of adjusted higher- priced lending of depository institutions reflects the sharp decline in high-risk lending by independent mortgage companies and not an increased focus on high-risk lending by depository institutions. Some of the increased share for the depository institutions may reflect acquisitions of previously independent mortgage companies. THE SURGE IN FHA AND VA LENDING Figure 5 illustrates the changing structure of the mortgage market between 2006 and 2008. It groups first-lien owner-occupied site-built mortgages for home purchase and refinance into six distinct categories: (1) loans sold to an affiliate or held in the portfolio of the originating lender (“Portfolio”), (2) loans sold into the private securitization market or to unaffiliated institutions (“Private”), (3) loans sold to Fannie Mae or Freddie Mac (GSEs), (4) loans insured by the FHA, (5) loans backed by the VA, and (6) loans insured by the Farm Service Agency or Rural Housing Service. The data show that approximately 40 percent of loans in early 2006 were sold into the 19 2008 HMDA Data 34 private securitization market or to an unaffiliated institution. By the end of 2008, nearly one- half of home-purchase loans and one-quarter of refinance loans were backed by either the FHA or the VA, and fewer than 15 percent of originations were sold to unaffiliated institutions or into the private securitization market (however, recall table 4, which indicates that almost no loans were sold into the private securitization market in 2008). The two GSEs increased their market share in 2007, but then relinquished much of these gains during 2008. While the decline of the subprime-based private securitization market was well under way by 2007, FHA and VA lending did not surge until 2008. At least two events in early 2008 may help explain the timing of this surge. First, as part of the Economic Stimulus Act passed in February, the Congress authorized an increase in the loan-size limits applicable for the FHA and VA programs and GSE purchases. Second, beginning in the early part of 2008, PMI companies started limiting their issuance of PMI insurance and raising prices because of rising claims and binding capital restrictions in certain states. As a consequence, Fannie Mae and Freddie Mac substantially reduced their purchases of loans with loan-to-value ratios (LTV) above 80 percent, which by statute require PMI (or other credit enhancement). Both GSEs also raised their credit guarantee fees for such loans at this time as well. We examine the effects of these events in the following two sections. The Effect of Higher Loan-Size Limits New standards released on March 6, 2008, raised the GSE and FHA loan-size limits up to $729,750 in certain areas designated by the Department of Housing and Urban Development as “high cost”. 35 FHA loan limits were also raised above their 2007 levels in many other areas to new levels. Prior to these changes, the GSEs could not purchase single-family home loans above $417,000 in most states, while the FHA could not insure single-family home loans above 34 Classifying loans by their ultimate disposition is complicated by HMDA reporting rules. A loan is classified as sold if the sale takes place within the HMDA reporting year. In other words, a loan originated in the December must be sold within the same month to be classified as sold. Since lenders often hold loans for several months before selling them, there is an “underreporting” in loan sales in HMDA for loans originated toward the end of the year. Analysis of the HMDA dataindicates that most loans are sold within three months if they were sold. To adjust for the underreporting in October-December, we used an imputation formula based on the allocation of loans originated in September (and the following January for 2006 and 2007 data) to allocate conventional loans among the first three groups shown in figure 5. Data in all of the tables presented in this section are based on this imputation. 35More than one-half of the 2008 loans in the high-cost areas were in California. One-third of the loans were in the mid-Atlantic states of New York, New Jersey, Maryland, and Virginia. 20 2008 HMDA Data 36 $271,050 in most areas of the country. VA loans do not have a size limit, but they do have a guarantee limit that is tied to GSE 37 loan limits. The VA guarantees the smaller of 25 percent of the loan amount or 25 percent of the applicable GSE loan limit. As such, increases to the GSE loan limit raise the maximum VA guarantee amount. To understand the potential effect of the higher limits, we divided loans originated in 2008 into four categories based on the size of the loan and the location of the property securing the loan: (1) loans smaller than the applicable 2007 FHA limit; (2) loans larger than the applicable 2007 FHA loan limit, but less than $417,000 and the applicable 2008 FHA limit; (3) loans larger than $417,000 but under the 2008 high-cost area limit common to the FHA, VA, and GSEs; and (4) all other loans. Changes in the loan-size limits directly affected the options available to borrowers for loans in categories 2 and 3 but did not affect those in categories 1 and 4. Table 13 displays the share of loans in these four categories by month, loan purpose, and loan product type (FHA, VA, GSE, and other). 38 Among FHA loans, there is a noticeable rise in the share of “newly FHA-eligible” loans (categories 2 and 3) in the first half of the year when the limits were increased for both home-purchase and refinance loans. For 2008 overall, the share of FHA-insured home-purchase loans in categories 2 and 3 was 9.7 percent, compared with 2.4 percent in 2007. 39 This increase implies that the limit changes lifted FHA home-purchase lending by 7.4 percent in 2008, assuming that the share of FHA lending in each of these categories would have remained at its 2007 level in the absence of limit changes (derived from table). This same assumption would imply that FHA refinance lending was 8.9 percent higher 36 The GSE loan limits were higher in Alaska and Hawaii; the maximum loan size for the FHA program was as l37 as $201,160 in some low-cost areas. VA loans larger than the GSE limits, however, cannot be sold into Ginnie Mae security pools. 38 The other category includes portfolio loans, private loans, and loans insured by the Farm Service Agency or the Rural Housing Service (a very small part of the category). Loan growth during 2008 (particularly for the first halfof the year), shown in table 13, is likely understated because of the omission of data from the 15 lenders who failed to report HMDA data, as discussed earlier. In December 2007, these lenders accounted for 3.4 percent of home-purchase loans and 6.0 percent of refinance loans in HMDA; however, these loans were not proportionately distributed among the four loan types examined here. For the same period, these lenders represented less than 2 percent of FHA loans and 0.01 percent of VA loans. Their market share of GSE loans was 3.1 percent for home-purchase loans and 5.7 percent for refinance loans; for other loans, t39ir share was 4.1 percent for home-purchase loans and 6.7 percent for refinance loans. FHA-insured loans in the 2007 HMDA data for amounts that exceed the single-family loan limit can be attributed to recording errors in the data or to loans for two-, three-, or four-family structures, which have higher loan limits and are not identified separately in the HMDA data. 21 2008 HMDA Data because of the limit changes. In contrast to the patterns for FHA lending, the proportion of VA loans in the four categories changed little over the course of the year, suggesting that the limit increases had little effect on VA lending. GSE lending showed only a modest boost from the limit increases (category 3). Under the same assumption used above, we estimate that GSE home purchase lending would have been 1.9 percent lower and GSE refinance lending only 0.8 percent lower in 2008 had the GSE limits not been changed. In sum, the effect of the limit increases on FHA, VA, and GSE lending appears to have been modest because the vast majority of the growth in both FHA and VA lending was in the categories in which there was no change in the eligibility standards. Pullback by PMI Companies and its Implication for FHA and VA lending With losses mounting in 2007 and 2008, PMI companies started raising prices and limiting coverage in some areas in the spring of 2008. These changes likely reduced the ability of the GSEs to purchase higher-LTV loans (loans with LTVs above 80 percent) because of the statutory requirement that such loans carry PMI (or a comparable credit enhancement) in order to be eligible for GSE purchase. The GSEs also raised their own underwriting fees for relatively high LTV loans in March 2008 and further in June. 40 Both the FHA and VA loan programs offer a form of credit insurance and, consequently, compete with the PMI companies. The two government programs likely increased their market share, at least to some extent, because the PMI and GSE price increases pushed the price of 41 conventional higher-LTV loans above that for the FHA and VA programs for some borrowers. 40PMI annual premiums for loans with LTVs above 80 percent range from 0.50 percentage point to greater than 1.00 percentage point. On March 1, 2008, Fannie Mae and Freddie Mac raised their one-time delivery fees for 30-year loa
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'