This is the accessible text file for GAO report number GAO-11-656
entitled 'Mortgage Reform: Potential Impacts of Provisions in the Dodd-
Frank Act on Homebuyers and the Mortgage Market' which was released on
July 19, 2011.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as
part of a longer term project to improve GAO products' accessibility.
Every attempt has been made to maintain the structural and data
integrity of the original printed product. Accessibility features,
such as text descriptions of tables, consecutively numbered footnotes
placed at the end of the file, and the text of agency comment letters,
are provided but may not exactly duplicate the presentation or format
of the printed version. The portable document format (PDF) file is an
exact electronic replica of the printed version. We welcome your
feedback. Please E-mail your comments regarding the contents or
accessibility features of this document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
United States Government Accountability Office:
GAO:
Report to Congressional Committees:
July 2011:
Mortgage Reform:
Potential Impacts of Provisions in the Dodd-Frank Act on Homebuyers
and the Mortgage Market:
GAO-11-656:
GAO Highlights:
Highlights of GAO-11-656, a report to congressional committees.
Why GAO Did This Study:
The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-
Frank Act) is intended, among other things, to reform residential
mortgage lending and securitization practices that contributed to the
recent financial crisis. The act provides some liability protection
for lenders originating mortgages that meet nine specified criteria,
as applicable, associated with a borrower’s ability to repay
(“qualified mortgages”). The act also requires securitizers of
mortgages not meeting separate criteria associated with lower default
risk to retain at least 5 percent of the credit risk, though federal
rulemaking agencies may vary this amount. The act directed GAO to
assess the effect of mortgage-related provisions on the availability
and affordability of mortgage credit and to issue a report by July
2011, but federal agencies are still developing implementing
regulations. This report discusses the potential impact of the act’s
(1) qualified mortgage criteria, (2) credit risk retention
requirement, and (3) provisions concerning homeownership counseling
and regulation of high-cost loans.
To do this work, GAO analyzed a proprietary database of residential
mortgages, reviewed relevant housing and mortgage market research, and
interviewed key mortgage industry stakeholders.
GAO provided a draft of this report to eight agencies. In a letter,
the National Credit Union Administration said, as noted in the report,
that the act’s impact would depend on regulatory decisions that had
yet to be made. Six other agencies provided technical comments that
have been incorporated as appropriate.
What GAO Found:
GAO examined five of the nine qualified mortgage criteria specified in
the Dodd-Frank Act for which sufficient data were available and
generally found that, for each year from 2001 through 2010, most
mortgages would likely have met the individual criteria. The five
criteria address payment of loan principal, length of the mortgage
term, scheduled lump-sum payments, documentation of borrower
resources, and borrower debt burden. The extent to which mortgages met
the individual criteria varied by year of origination, reflecting
changes in the mortgage market over the 10-year period. However, the
impact of the full set of qualified mortgage criteria is uncertain,
partly because data limitations make analysis of the other four
criteria difficult and partly because federal agencies could establish
different criteria as they develop final regulations. Consumer and
industry groups indicated that the criteria specified in the act would
likely encourage sound underwriting but could also restrict the
availability of and raise the cost of mortgage credit for some
homebuyers. Provisions in the act and proposed regulations attempt to
address some of these issues, in part by providing exemptions for
certain loan products in certain locales, such as rural areas. The
public comment period for these proposed regulations ends on July 22,
2011.
Mortgage industry stakeholders GAO spoke with indicated that the
implications of a risk retention requirement would depend on a variety
of regulatory decisions and potential changes in the mortgage market.
Rulemaking agencies are accepting public comments on proposed risk
retention regulations through August 1, 2011. Key decisions that have
yet to be made concern the characteristics of mortgages that would be
exempt from risk retention, the forms of risk retention that would be
allowed, the percentage that securitizers would be required to hold,
and risk-sharing arrangements between lenders and securitizers. These
factors could affect the availability and cost of mortgage credit and
the viability of a private mortgage securitization market.
Additionally, risk retention could complement other securitization and
mortgage reforms, such as those that promote greater transparency and
enforcement of loan underwriting standards.
Other provisions in the Dodd-Frank Act concerning homeownership
counseling and regulation of high-cost loans could enhance consumer
protections and improve mortgage outcomes for some borrowers, but
their specific impacts are difficult to assess at this time. The act
authorized a new Office of Housing Counseling within the Department of
Housing and Urban Development, but the office is still in the planning
stage. Findings from the limited research on housing counseling for
mortgage borrowers are mixed, with some studies suggesting that some
types of counseling can improve mortgage outcomes and others finding
no effect. The act also expands the definition of “high-cost loans,”
which have disclosures and restrictions designed to protect consumers.
Although lenders have generally avoided making these loans, additional
information on mortgage costs would be needed to assess the extent to
which the new definition would affect mortgages that may be made in
the future.
View [hyperlink, http://www.gao.gov/products/GAO-11-656] or key
components. For more information, contact William B. Shear at (202)
512-8678 or shearw@gao.gov.
[End of section]
Contents:
Letter:
Background:
Although Most Recent Mortgages Would Likely Have Met Certain Qualified
Mortgage Criteria, the Criteria Could Limit Mortgage Options for Some
Borrowers:
Regulatory Decisions and Other Factors Will Influence the Effect of a
Risk Retention Requirement on the Mortgage Market:
While Housing Counseling and High-Cost Lending Provisions May Enhance
Borrower Protections, Specific Impacts Are Unknown:
Agency Comments and Our Evaluation:
Appendix I: Objectives, Scope, and Methodology:
Appendix II: Proportions of Mortgages Meeting Selected Qualified
Mortgage Criteria in Geographic Groupings Based on Demographic and
Housing Market Characteristics:
Appendix III: Comments from the National Credit Union Association:
Appendix IV: GAO Contact and Staff Acknowledgments:
Bibliography:
Tables:
Table 1: Nine QM Criteria Specified in the Dodd-Frank Act:
Table 2: Percentage of Mortgages Meeting Qualified Mortgage Repayment
of Principal Requirement, by Demographic and Housing Market Grouping,
2001-2010:
Table 3: Percentage of Mortgages Meeting Qualified Mortgage Criterion
for Loan Terms of 30 Years or Less by Demographic and Housing Market
Grouping, 2001-2010:
Table 4: Percentage of Mortgages Meeting Qualified Mortgage Criterion
Restricting Balloon Payments by Demographic and Housing Market
Grouping, 2001-2010:
Table 5: Percentage of Prime, Near-Prime, and Government-Insured
Mortgages Meeting a Hypothetical Qualified Mortgage Criterion for DTI
Ratio of 41 Percent or Less by Demographic and Housing Market
Grouping, 2003-2010:
Figures:
Figure 1: Basic Steps in the Securitization of Residential Mortgages:
Figure 2: Proportions of Mortgages Meeting Qualified Mortgage
Repayment of Principal Requirement, 2001-2010:
Figure 3: Proportions of Mortgages Meeting Qualified Mortgage
Requirement for Loan Terms of 30 Years or Less, 2001-2010:
Figure 4: Proportions of Mortgages Meeting Qualified Mortgage
Restriction on Balloon Payments, 2001-2010:
Figure 5: Proportions of Prime, Near-Prime, and Government-insured
Mortgages Meeting Illustrative Qualified Mortgage Criterion for a Debt
Service-to-Income Ratio of 41 Percent or Less, 2003-2010:
Figure 6: Proportions of Prime, Near-Prime, and Subprime Mortgages
Originated in 2006 and 2010 That Would Have Met Different Requirements
for LTV Ratio:
Figure 7: Proportions of Prime and Near-Prime Mortgages Originated in
2006 and 2010 That Would Have Met Different Requirements for DTI Ratio:
Figure 8: Illustrative Example of the Implications of Horizontal and
Vertical Risk Retention on Risk-Based Capital Charges:
Abbreviations:
APR: annual percentage rate:
ARM: adjustable-rate mortgage:
CFPB: Bureau of Consumer Financial Protection or Consumer Financial
Protection Bureau:
Dodd-Frank Act: Dodd-Frank Wall Street Reform and Consumer Protection
Act:
DTI: debt service-to-income:
FAS: financial accounting statement:
FDIC: Federal Deposit Insurance Corporation:
Federal Reserve Board: Board of Governors of the Federal Reserve
System:
FHA: Federal Housing Administration:
FHFA: Federal Housing Finance Agency:
HMDA: Home Mortgage Disclosure Act:
HOEPA: Home Ownership Equity Protection Act of 1994:
HUD: Department of Housing and Urban Development:
LTV: loan-to-value:
NCUA: National Credit Union Administration:
NFMC: National Foreclosure Mitigation Counseling:
OCC: Office of the Comptroller of the Currency:
OTS: Office of Thrift Supervision:
QM: qualified mortgage:
QRM: qualified residential mortgage:
REIT: real estate investment trust:
RMBS: residential mortgage-backed securities:
SEC: Securities and Exchange Commission:
SPE: special purpose entity:
TILA: Truth in Lending Act:
[End of section]
United States Government Accountability Office:
Washington, DC 20548:
July 19, 2011:
The Honorable Tim Johnson:
Chairman:
The Honorable Richard C. Shelby:
Ranking Member:
Committee on Banking, Housing, and Urban Affairs:
United States Senate:
The Honorable Spencer Bachus:
Chairman:
The Honorable Barney Frank:
Ranking Member:
Committee on Financial Services:
House of Representatives:
The number of homes in foreclosure and of homeowners in financial
distress remains at historically high levels. In the first quarter of
2011, more than 3.5 million home mortgages were 90 or more days
delinquent or in the foreclosure process, and estimates indicate that
more than one in five mortgage borrowers owe more on their mortgages
than their homes are worth. The continuing foreclosure crisis was
fueled in part by the proliferation of mortgage products in the early
to mid-2000s that have come to be associated with poorer loan
performance. These products include mortgages with interest rates that
increased sharply after a few years, did not require a down payment or
full documentation of income, or allowed borrowers to defer principal
and interest payments, increasing their indebtedness over time. Some
mortgage brokers and originators had financial incentives to steer
borrowers who qualified for potentially more sustainable options into
such mortgages. After home prices began to stagnate or fall in 2005,
defaults and foreclosures increased rapidly. Complicating matters
during this period were securitization practices, which included
bundling higher-risk mortgages into residential mortgage-backed
securities (RMBS) that, in turn, were sometimes repackaged into more
complex investment products.[Footnote 1] As demand for RMBS grew,
lenders and securitizers were increasingly compensated based on loan
volume rather than loan quality, contributing to a decline in
underwriting standards.
To help prevent a recurrence of such problems in the mortgage market,
Congress passed the Dodd-Frank Wall Street Reform and Consumer
Protection Act (Dodd-Frank Act) on July 21, 2010.[Footnote 2] A key
challenge in implementing the Dodd-Frank Act's provisions is balancing
the goal of protecting borrowers from unsustainable mortgage products
with the goal of maintaining broad access to mortgage credit. Among
other things, the Dodd-Frank Act establishes minimum standards for
mortgages, requiring that consumers have a "reasonable ability to
repay" at the time a mortgage is made when the loan terms, applicable
taxes, homeowner's insurance, and assessments are taken into account.
This consumer protection provision creates due diligence standards for
mortgage lenders. According to the Dodd-Frank Act, a lender is
presumed to have satisfied the ability-to-repay requirement and
receives some protection from liability when it originates a
"qualified mortgage" (QM).[Footnote 3] The Dodd-Frank Act specifies
nine criteria that a loan must meet to be a QM:
(1) regular periodic payments do not result in an increase in the
principal balance or result in a deferral of the repayment of
principal;
(2) the loan term does not exceed 30 years;
(3) except for balloon loans under specified circumstances, the
mortgage does not include balloon payments;[Footnote 4]
(4) borrower income and financial resources are verified and
documented;
(5) the loan complies with guidelines or regulations established by
the Board of Governors of the Federal Reserve System (Federal Reserve
Board) relating to ratios of total monthly debt to monthly income or
alternative measures of ability to pay regular expenses after paying
total monthly debt;
(6) a fixed-rate loan is underwritten based on a fully amortizing
payment schedule that takes into account applicable taxes, insurance,
and assessments;
(7) an adjustable-rate mortgage (ARM) is underwritten based on the
maximum rate permitted during the first 5 years and on a fully
amortizing payment schedule that takes into account applicable taxes,
insurance, and assessments;
(8) total points and fees payable in connection with loan do not
exceed 3 percent of the total loan amount;[Footnote 5] and:
(9) a reverse mortgage that meets QM standards as set by the Federal
Reserve Board.[Footnote 6]
The Dodd-Frank Act gave federal rulemaking agencies the flexibility to
change these criteria.
The Dodd-Frank Act also requires mortgage securitizers to retain a
financial exposure of no less than 5 percent of the credit risk of any
securitized residential mortgage that does not meet a separate set of
criteria (to be defined by regulators) that are associated with a
lower risk of default.[Footnote 7] Securitized mortgages that meet
these criteria are exempt from this risk retention requirement and are
referred to as "qualified residential mortgages" (QRM). The risk
retention provision is designed to provide an economic incentive for
securitizers of non-QRMs to ensure that lenders originate well-
underwritten mortgages that protect investors from losses. Although
the Dodd-Frank Act contains a uniform 5 percent requirement, it gives
federal regulators the flexibility to specify a risk retention
requirement for nonexempt mortgages that varies depending on the
underwriting standards used.
Given the serious problems that continue in the mortgage market and
congressional interest in protecting consumers and ensuring credit
availability, we were required to assess the potential impact of the
mortgage-related provisions of the Dodd-Frank Act and issue a report
by July 21, 2011. Because regulations governing implementation of
these provisions are still being developed, the criteria we assessed
could change based on rulemakers' review of comments from the public
on the proposed rules. The public comment periods for proposed QM and
QRM rules will end on July 22 and August 1, 2011, respectively. Partly
for this reason, assessing the potential impact of the Dodd-Frank Act
provisions is challenging at this time. This report (1) assesses the
proportions of mortgages originated from 2001 through 2010 that would
have met selected QM criteria specified in the Dodd-Frank Act and
describes the views of mortgage industry stakeholders on the potential
effects of the QM criteria on the mortgage market, (2) discusses
relevant information and the views of mortgage industry stakeholders
on the potential impact of a risk retention requirement on the
mortgage market and the advantages and disadvantages of a uniform risk
retention requirement, and (3) describes what research and the views
of mortgage industry stakeholders suggest about the potential impact
of provisions in the Dodd-Frank Act regarding homeownership counseling
and changes to the Home Ownership Equity Protection Act (HOEPA).
[Footnote 8] For practical reasons, we examined these different parts
of the Dodd-Frank Act separately. Although the purpose and scope of
the QM and QRM provisions are somewhat different, they could be
expected to work together by increasing lenders' and securitizers'
exposure to the risks that are associated with mortgages whose
features and terms put borrowers at higher risk of default and
foreclosure.
Because recovery from today's restricted credit conditions could
expand the volume and types of mortgage products in the marketplace,
we used historical data to illustrate the potential effects of
selected QM criteria under different market conditions and lending
environments. Specifically, we analyzed a proprietary database of
loans from CoreLogic, Inc., to examine the proportions of loans
originated from 2001 through 2010 that likely would have met selected
QM criteria specified in the Dodd-Frank Act. This database contains
information from major mortgage servicers and covers a broad cross-
section of the mortgage market. For example, CoreLogic estimates that
the database captures 60 to 65 percent of the mortgages purchased by
Freddie Mac and Fannie Mae (the enterprises), respectively,
approximately 50 percent of subprime mortgages, and about 90 percent
of mortgages with government-insurance or guarantees (such as
mortgages insured by the Federal Housing Administration (FHA)).
[Footnote 9] Nevertheless, because of limitations in the coverage and
completeness of the data, our analysis may not be fully representative
of the mortgage market as whole. We examined five of the nine QM
criteria specified in the Dodd-Frank Act for which sufficient data,
including data from the CoreLogic database, were available (see the
first five criteria previously listed).[Footnote 10] In general, for
each year from 2001 through 2010, we identified the proportion of
mortgage originations that would have met the individual criteria. We
were not able to calculate relevant proportions for certain years and
mortgage market segments due to data limitations. Primarily due to
data limitations, we were also not able to assess the remaining four
QM criteria (see the last four criteria listed previously).
We assessed the reliability of the CoreLogic data by interviewing
CoreLogic representatives about the methods the firm used to collect
and ensure the integrity of the information. We also reviewed
supporting documentation about the database. In addition, we conducted
reasonableness checks on the data to identify any missing, erroneous,
or outlying figures. We concluded that the data elements we used were
sufficiently reliable for our purposes. To obtain additional
information and views on the potential effects of the QM criteria
specified in the Dodd-Frank Act, we reviewed proposed rules for
implementing the Dodd-Frank Act's QM provisions. We also reviewed
relevant research literature and interviewed officials from
organizations representing mortgage lenders, mortgage brokers,
investors, securitizers, and consumer interests. Additionally, we
interviewed officials from the Federal Reserve Board, Federal Deposit
Insurance Corporation (FDIC), Federal Housing Finance Agency (FHFA),
Department of Housing and Urban Development (HUD), National Credit
Union Administration (NCUA), Office of the Comptroller of the Currency
(OCC), and Office of Thrift Supervision (OTS).
To assess the potential impact of the Dodd-Frank Act's risk retention
requirement on the mortgage market, we reviewed relevant statutory
provisions and the rules that have been proposed to implement those
provisions. We also reviewed available information on mortgage
securitization practices prior to the financial crisis and factors
that could affect the impact of the risk retention requirement,
including potential changes to the roles of the enterprises and FHA.
We interviewed key mortgage industry stakeholders--including those
representing mortgage lenders, securitizers, investors, and consumers--
to obtain their views on the potential impact of a risk retention
requirement including how regulatory decisions regarding the form and
coverage of the requirement could affect the availability and
affordability of mortgage credit. We used the CoreLogic data to
examine selected criteria--loan-to-value (LTV) ratio and debt service-
to-income (DTI) ratio--that regulators are considering as part of the
QRM rulemaking to describe the proportion of mortgages that may have
met different LTV and DTI thresholds in 2006 (a period of relatively
lax underwriting standards) and 2010 (a period of relatively stringent
underwriting standards).[Footnote 11] To assess the impact of the risk
retention requirement on lenders, we reviewed relevant accounting
standards and risk-based capital requirements that could interact with
risk retention. We also interviewed industry stakeholders about the
impact of a risk retention requirement on different types and sizes of
mortgage lenders. To assess the advantages and disadvantages of a
uniform 5 percent risk retention requirement, we interviewed industry
stakeholders about the development, implementation, and enforcement of
both a uniform and a nonuniform requirement. Finally, we interviewed
officials from the previously cited federal agencies and the
Securities and Exchange Commission (SEC).
To describe the potential effects of the housing counseling and HOEPA
provisions in the Dodd-Frank Act, we reviewed relevant statutory
provisions and industry research. We identified and reviewed empirical
research and published literature on the impact of prepurchase and
foreclosure mitigation counseling on mortgage outcomes. We also
interviewed HUD officials about their plans for creating the new
housing counseling office required by the Dodd-Frank Act. We compared
the new HOEPA requirements in the Dodd-Frank Act to previous statutory
requirements and examined available research on the number of loans
originated from 2004 through 2009 that were covered by HOEPA
requirements. We also interviewed a wide range of mortgage and
counseling industry stakeholders, including federal agencies, consumer
groups, lenders, and academic researchers about the Dodd-Frank Act's
counseling and HOEPA provisions.
We conducted this performance audit from August 2010 to July 2011 in
accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe
that the evidence obtained provides a reasonable basis for our
findings and conclusions based on our audit objectives. Appendix I
explains our objectives, scope, and methodology in greater detail.
Background:
Mortgage Markets and Securitization:
Residential mortgages fall into several loosely defined categories and
encompass a range of loan products:
* Prime mortgages are made to borrowers with strong credit histories
and provide the most attractive interest rates and loan terms.
* Near-prime mortgages (also called Alt-A mortgages) generally serve
borrowers whose credit histories are close to prime but who have one
or more higher-risk characteristics, such as limited documentation of
income or assets.
* Subprime mortgages are generally made to borrowers with blemished
credit and feature higher interest rates and fees than prime loans.
* Government-insured or -guaranteed mortgages primarily serve
borrowers who may have difficulty qualifying for prime loans and
feature interest rates similar to those for prime loans. These
mortgages require insurance or charge guarantee fees. FHA and the
Department of Veterans Affairs (VA) operate the two main federal
programs that insure or guarantee mortgages.
Across all of these market segments, two types of loans are common:
fixed-rate mortgages, which have interest rates that do not change
over the life of the loans and ARMs, which have interest rates that
change periodically based on changes in a specified index.
A number of loan features became more common in the 2000s. While these
features potentially expanded access to mortgage credit, they were
often associated with higher default rates. These features included
the following:
* Low-and no-documentation loans. Originally intended for borrowers
who had difficulty documenting income, such as the self-employed,
these loans were made with little or no verification of a borrower's
income or assets.
* High LTV ratios. As homebuyers made smaller down payments, this
ratio increased.
* Prepayment penalties. Some loans contained built-in penalties for
repaying part or all of a loan in advance of the regular schedule.
Other mortgage types that became more prevalent during this period
included different types of ARMs. Short-term hybrid ARMs had a fixed
interest rate for an initial period (usually 2 or 3 years) but then
"reset" to an adjustable rate for the remaining term of the loan.
Interest-only or payment-option ARMs allowed borrowers to defer
repayment of principal and possibly part of the interest for the first
few years of the loan.[Footnote 12] Payment-option ARMs enabled
mortgages to negatively amortize, meaning that the loan balance could
increase over time.
The secondary mortgage market, where loans are securitized, plays an
important role in providing liquidity for mortgage lending.
Securitization has a number of benefits for lenders. Among other
things, it is typically less expensive than raising funds directly and
it transfers some or all of the credit and interest rate risk from the
lender to the investor.[Footnote 13] To securitize mortgage loans,
mortgage lenders or originators sell their loans to third parties--
either directly to securitizing institutions or loan aggregators that
serve as intermediaries between originators and securitizers--
generating funds that could be used to originate more loans (see
figure 1). Securitization involves a number of players. Securitizing
institutions include investment banks, retail banks, mortgage
companies, and real estate investment trusts (REIT).[Footnote 14] As a
part of the securitization process, securitizers create a separate
legal entity ("special purpose entity" or SPE) to bundle mortgages and
sell them as investment products called RMBS. The purpose of creating
the SPE is to help ensure that securitized assets are protected in the
event of a bankruptcy of the securitizing or originating institutions.
Other parties to a securitization transaction include, but are not
limited to, credit rating agencies that assess the creditworthiness of
the securities based on the likelihood of default and the expected
value of dollar losses in the event of a default, and deal
underwriters hired by the securitizers to market and sell the
securities to investors. Finally, servicers are hired to collect
mortgage payments from the borrowers and disburse interest and
principal payments to the investors.
Figure 1: Basic Steps in the Securitization of Residential Mortgages:
[Refer to PDF for image: illustration]
1) Lenders originate or extend mortgage credit to borrowers.
1A) Borrowers obtain credit and make interest and principal payments
to servicers; go to step 5B.
2) Securitizers purchase mortgages from lenders and organize and
initiate residential mortgage-backed securitization transactions.
Securitizers create separate legal entities called Special Purpose
Entities (SPE) to transfer risk and protect the securitized assets in
the event of the securitizers’ bankruptcy.
3A) Credit rating agencies assess the creditworthiness of the
securities.
3B) SPEs receive mortgages from lenders, bundle them, and issue them
as securities.
4) Underwriters are hired by securitizers to market and sell
securities.
5A) Investors purchase securities and receive interest and principal
payments.
5B) Servicers collect mortgage payments from borrowers and disburse
interest and principal payments to investors. Go to step 5A.
Source: GAO.
[End of figure]
RMBS can be structured in different ways, but one common structure
involves a prespecified distribution of cash payments to investors in
different slices, or tranches of the security. Tranching allows
investors with different appetites for risk to invest in a security
with the same underlying pool of loans. In such credit-tranched
structures, also known as "senior subordinate" structures, cash from
the underlying loans is generally paid to the topmost, least risky
tranche first until the prespecified thresholds are met. Cash then
flows to the lower tranches in what is known as a "waterfall."
Conversely, the bottom-most tranche typically absorbs the losses from
defaults until it is depleted, with any additional losses flowing up
toward senior securities.
The secondary mortgage market consists of (1) Ginnie Mae-guaranteed
RMBS, which are backed by cash flows from federally insured or
guaranteed mortgages; (2) enterprise RMBS, which are backed by
mortgages that meet the criteria for purchase by Fannie Mae and
Freddie Mac; and (3) private-label RMBS, which are backed by mortgages
that do not conform to enterprise purchase requirements because they
are too large (i.e., jumbo mortgages) or otherwise do not meet
enterprise underwriting criteria. Most subprime and near-prime
mortgages, and many prime jumbo mortgages, were securitized into
private-label RMBS. However, the private-label market, which accounted
for most of the RMBS issuances in 2005 and 2006, collapsed in 2008 and
has not recovered. As a result, almost all RMBS issuances in recent
years are backed by the full guarantees of the enterprises and Ginnie
Mae. RMBS represent the biggest single piece of the larger
securitization market, accounting for over one-third of all new asset-
backed issuances from 2005 through the third quarter of 2010.
The composition of the mortgage market has changed dramatically in
recent years. In the early to mid-2000s, the volume of subprime and
near-prime mortgage originations grew rapidly and peaked in 2006,
accounting for nearly 40 percent of mortgage originations that year.
[Footnote 15] These market segments contracted sharply in mid-2007,
partly in response to increasing defaults and foreclosures, including
mortgages defaulting within a few months of origination, and a lack of
investor demand.[Footnote 16] The market segments comprising mortgages
backed by the enterprises and FHA had the opposite experience: a
decline in market share in the early to mid-2000s, followed by rapid
growth beginning in 2007 and 2008, respectively.[Footnote 17] For
example, the enterprises' share of the mortgage market decreased from
about one-half in 2003 to about one-third in 2006. By 2009 and 2010,
enterprise-backed mortgages had increased to more than 65 percent of
the market. Similarly, FHA-insured mortgages grew from about 2 percent
of the market in 2006 to about 20 percent in 2009 and 2010. Congress
and the administration are currently considering options to scale back
the role of the enterprises and FHA in the mortgage market and
increase the role of private capital. In addition to these potential
changes, a recovery from constrained credit conditions in the mortgage
market could expand the volume of mortgages extended to borrowers and
therefore subject to the Dodd-Frank Act's provisions.
Federal Mortgage Lending Laws:
The Dodd-Frank Act enacts numerous provisions intended to reform the
mortgage lending industry with an eye toward consumer protection. Many
of these provisions are contained in Title XIV of the act, which
amends provisions of the Truth in Lending Act (TILA) to reform and
provide accountability for consumer mortgage practices.[Footnote 18]
TILA, enacted in 1968, and HOEPA, which amended TILA in 1994, are
among the primary federal laws governing mortgage lending. TILA was
designed to provide consumers with accurate information about the cost
of credit. Among other things, TILA requires lenders to disclose
information about the terms of loans--including the amount financed,
the finance charge, and the annual percentage rate (APR)--that can
help borrowers understand the overall costs of their loans.[Footnote
19]
Congress enacted HOEPA in response to concerns about predatory
lending. HOEPA regulates and restricts the terms and characteristics
of certain kinds of "high-cost" mortgages--that is, those that exceed
certain thresholds in their APRs or fees (often referred to as "rate
and fee triggers"). The Dodd-Frank Act expands the definition of high-
cost loans to include mortgages for purchasing a home; reduces the APR
and points and fees triggers; and requires mandatory preloan
counseling for borrowers of high-cost mortgages, among other things.
The Federal Reserve Board implements TILA and HOEPA, but this
responsibility will transfer to the Bureau of Consumer Financial
Protection (also known as the Consumer Financial Protection Bureau or
CFPB) on July 21, 2011.
Minimum Lending Standards and Qualified Mortgage Provisions:
The Dodd-Frank Act reforms mortgage lending by amending TILA to
prohibit lenders from making mortgage loans without regard to
consumers' ability to repay them. As previously noted, lenders can
comply with the ability-to-repay standard by originating a QM. Lenders
are not prohibited from originating non-QMs, however. The Dodd-Frank
Act specifies nine QM criteria, but gives the Federal Reserve Board
the authority to add to, subtract from, or modify the criteria as it
develops implementing regulations (see table 1).[Footnote 20]
Table 1: Nine QM Criteria Specified in the Dodd-Frank Act:
1. Regular periodic payments do not result in an increase in the
principal balance or result in a deferral of the repayment of
principal (i.e., the mortgage cannot have a negative amortization
feature or interest-only period).
2. The loan term does not exceed 30 years. Rulemakers may extend loan
terms beyond 30 years for certain locales, such as high-cost areas.
3. Except for balloon loans under specified circumstances, the
mortgage does not include balloon payments.[A]
4. Borrower income and financial resources are verified and documented.
5. The loans comply with guidelines or regulations established by the
Federal Reserve Board relating to ratios of total monthly debt to
monthly income or alternative measures of ability to pay regular
expenses after paying monthly debt.
6. A fixed-rate loan is underwritten based on a fully amortizing
payment schedule that takes into account applicable taxes, insurance,
and assessments.
7. An ARM is underwritten based on the maximum rate permitted during
the first 5 years and on a fully amortizing payment schedule that
takes into account applicable taxes, insurance, and assessments.
8. Total points and fees payable in connection with loan do not exceed
3 percent of the total loan amount.
9. A reverse mortgage meets QM standards as set by the Federal Reserve
Board.[B]
Source: Dodd-Frank Act.
[A] According to Dodd-Frank Act provisions and proposed QM rules
issued in April 2011, some balloon mortgages can be considered to meet
the QM criteria, such as balloon mortgages made by creditors that
operate in predominantly rural or underserved areas.
[B] In proposed regulations, the Federal Reserve Board indicated that
QM requirements were generally not relevant to reverse mortgages
because the Dodd-Frank Act does not subject reverse mortgages to the
ability-to-repay requirement (see 76 Fed. Reg. 27390, 27407 (May 11,
2011)). As a result, the Federal Reserve Board has not proposed QM
standards for reverse mortgages at this time.
[End of table]
Risk Retention Requirement:
The Dodd-Frank Act requires securitizers of RMBS to retain no less
than 5 percent of the credit risk of any residential mortgage they
securitize that does not meet specified criteria.[Footnote 21] The
purpose of the requirement is to help align the interests of
participants in the securitization process and encourage sound loan
underwriting. The Dodd-Frank Act exempts government-insured or -
guaranteed mortgages from the risk retention requirement (excluding
mortgages backed by the enterprises, which are in government
conservatorship), and as noted previously, loans that meet the QRM
criteria.[Footnote 22] However, the Dodd-Frank Act requires that the
risk retention requirement be applied to any RMBS that contains one or
more non-QRMs, even if the vast majority of the security's mortgages
are QRMs.
Federal banking and other agencies are required by the Dodd-Frank Act
to jointly prescribe regulations for the risk retention requirement.
[Footnote 23] In crafting the risk retention regulations, the Dodd-
Frank Act requires rulemakers to specify, among other things,
* criteria for QRMs, taking into consideration underwriting and
product features that historical loan performance data indicate result
in a lower risk of default thereby ensuring high-quality loan
underwriting. The Dodd-Frank Act specifies that the QRM definition
cannot be broader than the QM definition described previously (i.e.,
the QRM criteria can be more restrictive than the QM criteria but not
less restrictive);
* permissible forms of risk retention and the minimum duration for
meeting the requirement;
* ways of allocating risk between securitizers and originators; and:
* the possibility of permitting a lower risk retention requirement
(less than 5 percent) for any non-QRM that meets underwriting
standards that the agencies develop in regulations.
Rulemakers issued proposed rules for the risk retention provisions in
March 2011.[Footnote 24] The proposed criteria for the QRM include,
but are not limited to, the following:
* the LTV ratio must be at least 80 percent for mortgages obtained for
a home purchase;[Footnote 25]
* the DTI ratio must be 36 percent or less;[Footnote 26]
* the loan term must not exceed 30 years;
* the loan cannot include negative amortization or payment deferral
features;
* points and fees cannot exceed 3 percent of the total loan amount;
* the borrower can neither be 30 or more days past due on any debt
obligation nor have been 60 or more days past due on any debt
obligation within the preceding 24 months; and:
* the originator must incorporate into the mortgage documents certain
requirements regarding policies and procedures for servicing the
mortgage, including procedures to promptly initiate activities to
mitigate the risk of default for delinquent loans.[Footnote 27]
Although this report focuses on risk retention for RMBS, the Dodd-
Frank Act's risk retention requirement also applies to securities
backed by other asset classes, such as credit cards and automobile
loans. In response to a mandate in the Dodd-Frank Act, the Federal
Reserve Board issued a report in October 2010 that, among other
things, describes historical issuance activity, securitization
structures, and incentive alignment mechanisms for nine categories of
asset-backed securities.[Footnote 28] The report noted that the
effects of a final set of risk retention requirements could not be
analyzed because implementing regulations were still being developed.
However, the report made a number of recommendations for rulemakers to
consider when crafting the risk retention requirement, including a
recommendation that the requirement be tailored to each major class of
securitized assets. Also in response to a mandate in the Dodd-Frank
Act, the Chairman of the Financial Stability Oversight Council issued
a report in January 2011 that examined the macroeconomic effects of a
risk retention requirement. While noting limitations in the
information available to assess the impacts of risk retention rules
not yet in place, the report offered several principles and
recommendations to inform the design of a risk retention framework
that facilitates economic growth by allowing market participants to
price credit risk more accurately and allocate capital more
efficiently.[Footnote 29]
Housing Counseling:
The Housing and Urban Development Act of 1968 authorized HUD to
provide housing counseling services.[Footnote 30] Specifically, it
authorized HUD to make grants to or contract with public or private
organizations to provide a broad range of housing counseling services
to homeowners and tenants to assist them in improving their housing
conditions and in meeting the responsibilities of homeownership or
tenancy.[Footnote 31]
The Dodd-Frank Act requires HUD to establish an Office of Housing
Counseling and gives the office a broad range of responsibilities
relating to homeownership and rental housing counseling, including
grant administration, policy development, public outreach, and
research.[Footnote 32] The Dodd-Frank Act requires HUD to appoint a
Director of Housing Counseling to report directly to the Secretary of
HUD and to establish an advisory committee consisting of mortgage and
real estate industry stakeholders and consumer groups and HUD-
certified housing counseling agencies.
Two key types of homeownership counseling are foreclosure mitigation
counseling and prepurchase counseling. Foreclosure mitigation
counseling focuses on helping financially distressed homeowners avoid
foreclosure by working with lenders to cure mortgage delinquency.
Prepurchase counseling topics can include the process of qualifying
for a mortgage, selecting a mortgage product, and successfully
maintaining a home. While prepurchase counseling was common prior to
the financial crisis, foreclosure mitigation counseling has gained
increasing attention and popularity as a means to assist homeowners
who are struggling to stay in their homes.
Although Most Recent Mortgages Would Likely Have Met Certain Qualified
Mortgage Criteria, the Criteria Could Limit Mortgage Options for Some
Borrowers:
Our analysis of the QM criteria specified in the Dodd-Frank Act
generally indicated that, for each year from 2001 through 2010, most
mortgages would likely have met the individual criteria for which
relevant data were available.[Footnote 33] The extent to which
mortgages met individual criteria varied by mortgage category and
origination year, reflecting changes in the mortgage market over the
10-year period. Consumer and industry groups that we spoke with noted
that the QM criteria would likely provide several benefits to
qualified borrowers, and housing research indicates that many of the
QM criteria are associated with a borrower's ability to repay a
mortgage. However, some consumer and industry groups stated that some
of the QM criteria could increase the cost and restrict the
availability of mortgages for some borrower groups, including lower-
income and minority borrowers.
Data on a Cross Section of Mortgages Suggest That Most Mortgages Would
Have Met Selected Qualified Mortgage Criteria Specified in the Dodd-
Frank Act:
To illustrate the potential significance of the QM criteria under
different lending environments and market conditions, we applied
selected criteria to CoreLogic data on mortgages originated from 2001
through 2010. We applied each criterion separately, calculating the
proportion of mortgages in each annual loan origination cohort that
likely would have met it. We were unable to determine the proportion
of mortgages that would have met all of the criteria we examined due
to the number of records in the database that had missing or
unreliable values for one or more of the criteria. For example, the
database contained no information on DTI ratio for subprime mortgages
and did not have reliable information on documentation of borrower
income and assets.[Footnote 34] As a result, we determined that
applying the criteria simultaneously would not have produced reliable
results. Because the CoreLogic data group mortgages into two broad
categories--the first containing prime, near-prime, and government-
insured or -guaranteed loans and the second containing subprime loans--
we examined these categories separately when possible.[Footnote 35]
The data did not contain information needed to examine all of the QM
criteria specified in the Dodd-Frank Act. As a result, our analysis
focused on the five criteria for which CoreLogic or other data were
available. Our analysis includes other limitations and assumptions, as
discussed in the rest of this section and in appendix I. Additionally,
appendix II contains additional breakdowns of our analysis by
geographic groupings based on racial, ethnic, income, and house price
patterns.
The five QM criteria in the Dodd-Frank Act that we were able to assess
were:
* regular periodic payments do not result in an increase in the
principal balance or deferred repayment of principal (e.g., due to
negative amortization features);
* the loan term does not exceed 30 years;[Footnote 36]
* except for balloon loans under specified circumstances, the loan
does not include balloon payments;[Footnote 37]
* borrower income and financial resources are verified and documented;
and:
* the loan complies with guidelines or regulations established by the
Federal Reserve Board relating to ratios of total monthly debt service
to monthly income. The Federal Reserve Board's proposed rules for QMs
do not provide a specific DTI ratio.[Footnote 38] Therefore, for
illustrative purposes, we used the 41-percent ratio that serves as a
guideline in underwriting FHA-insured mortgages.
Negative Amortization Features:
The significance of selected QM requirements varied by origination
year. Regarding the criterion for repayment of principal, our analysis
focused on mortgages with negative amortization features, which would
have been prohibited under the Dodd-Frank Act because they allowed
payments that did not cover the loan principal, resulting in
increasing loan amounts. Due to limitations in the CoreLogic data, our
analysis does not account for interest-only mortgages, which would
also have been prohibited because they deferred repayment of
principal.[Footnote 39] Negative amortization features can be
problematic because borrowers may experience payment shock when their
payments increase to include an amount that will fully amortize the
outstanding balance over the remaining loan term. As shown in figure
2, most mortgages originated from 2001 through 2010 would have met the
QM requirement related to repayment of principal. Among prime, near-
prime, and government-insured mortgages, the proportion of new
originations without a negative amortization feature declined from 99
percent in 2001 to 91 percent in 2005, then increased to essentially
100 percent in 2009 and 2010. This trend reflects the growth in near-
prime mortgages (many of which were payment-option ARMs that could
negatively amortize) early in the decade and their disappearance after
2007. In the subprime market, almost 100 percent of mortgage
originations from 2001 through 2007 did not have negative amortization
features. Because so few subprime mortgages were originated after
2007, we did not calculate corresponding percentages for 2008 through
2010 for this criterion or other QM criteria.
Figure 2: Proportions of Mortgages Meeting Qualified Mortgage
Repayment of Principal Requirement, 2001-2010:
[Refer to PDF for image: multiple line graph]
Year: 2001;
Subprime: 99.7311%;
Prime, near-prime, and government-insured: 98.9243%.
Year: 2002;
Subprime: 99.7569%;
Prime, near-prime, and government-insured: 98.6724%.
Year: 2003;
Subprime: 99.934%;
Prime, near-prime, and government-insured: 98.7117%.
Year: 2004;
Subprime: 99.999%;
Prime, near-prime, and government-insured: 94.7421%.
Year: 2005;
Subprime: 99.9977%;
Prime, near-prime, and government-insured: 90.8882%.
Year: 2006;
Subprime: 99.9883%;
Prime, near-prime, and government-insured: 92.5288%.
Year: 2007;
Subprime: 99.8206%;
Prime, near-prime, and government-insured: 95.9949%.
Year: 2008;
Prime, near-prime, and government-insured: 99.3448%.
Year: 2009;
Prime, near-prime, and government-insured: 100%.
Year: 2010;
Prime, near-prime, and government-insured: 99.9997%.
Source: GAO analysis of CoreLogic data.
Note: We do not report percentages for subprime mortgages after 2007
due to the low number of originations. The figure does not account for
interest-only mortgages, which would also be prohibited because they
defer repayment of principal. The CoreLogic database we used for this
analysis covers a broad cross-section of the mortgage market. However,
because of limitations in the coverage and completeness of the data,
our analysis may not be fully representative of the mortgage market
segments shown.
[End of figure]
Loan Term:
As shown in figure 3, the large majority of mortgages originated from
2001 through 2010 would have met the QM criterion for a loan term of
30 years of less. A term of greater than 30 years increases the
borrower's total mortgage costs because more interest accrues than it
would in a shorter period. Among prime, near-prime, and government-
insured mortgages, essentially 100 percent met the criterion from 2001
through 2004. For this category of mortgages, the proportion declined
to 96 percent in 2007 and rose back to about 100 percent by 2009. For
subprime mortgages, the proportion that met the criterion was nearly
100 percent from 2001 through 2004, but declined to 85 percent in
2006. The trend in the middle of the decade toward mortgages with
longer loan terms suggests efforts by lenders to qualify borrowers for
mortgages that offered lower monthly payments during a period of
strong appreciation in house prices.
Figure 3: Proportions of Mortgages Meeting Qualified Mortgage
Requirement for Loan Terms of 30 Years or Less, 2001-2010:
[Refer to PDF for image: multiple line graph]
Year: 2001;
Subprime: 99.5957%;
Prime, near-prime, and government-insured: 99.7788%.
Year: 2002;
Subprime: 99.7661%;
Prime, near-prime, and government-insured: 99.7823%.
Year: 2003;
Subprime: 99.8275%;
Prime, near-prime, and government-insured: 99.7811%.
Year: 2004;
Subprime: 99.7776%;
Prime, near-prime, and government-insured: 99.5142%.
Year: 2005;
Subprime: 95.7375%;
Prime, near-prime, and government-insured: 98.7947%.
Year: 2006;
Subprime: 85.0127%;
Prime, near-prime, and government-insured: 96.6845%.
Year: 2007;
Subprime: 86.3418%;
Prime, near-prime, and government-insured: 96.0168%.
Year: 2008;
Prime, near-prime, and government-insured: 98.811%.
Year: 2009;
Prime, near-prime, and government-insured: 99.808%.
Year: 2010;
Prime, near-prime, and government-insured: 99.8164%.
Source: GAO analysis of CoreLogic data.
Note: We do not report the proportion of subprime mortgages after 2007
due to the low number of subprime originations. The CoreLogic database
we used for this analysis covers a broad cross-section of the mortgage
market. However, because of limitations in the coverage and
completeness of the data, our analysis may not be fully representative
of the mortgage market segments shown.
[End of figure]
Balloon Payments:
A high proportion of the mortgages originated over the 10-year period
we examined would have met the QM criterion restricting balloon
payments, although the percentages were somewhat different for prime,
near-prime, and government-insured mortgages compared with subprime
mortgages (see figure 4). A balloon mortgage does not fully amortize
over the term of the loan, leaving a balance due at maturity. The
final payment is called a balloon payment because it is generally much
larger than the other payments. Mortgages with balloon payments have
been associated with repayment problems, likely due to the payment
shock that occurs when the loan balance becomes due, or difficulty in
refinancing at the end of the loan term, especially if the home value
depreciated. Among prime, near-prime, and government-insured
mortgages, almost 100 percent of the originations each year did not
have balloon payments. For subprime mortgages, the proportions
increased from 96 percent in 2001 to 99 percent in 2003 and 2004, and
decreased to about 90 percent in 2007.
Figure 4: Proportions of Mortgages Meeting Qualified Mortgage
Restriction on Balloon Payments, 2001-2010:
[Refer to PDF for image: multiple line graph]
Year: 2001;
Subprime: 95.9666%;
Prime, near-prime, and government-insured: 99.4479%.
Year: 2002;
Subprime: 98.3489%;
Prime, near-prime, and government-insured: 99.2473%.
Year: 2003;
Subprime: 99.2907%;
Prime, near-prime, and government-insured: 99.0558%.
Year: 2004;
Subprime: 99.3153%;
Prime, near-prime, and government-insured: 99.3745%.
Year: 2005;
Subprime: 97.3576%;
Prime, near-prime, and government-insured: 99.7026%.
Year: 2006;
Subprime: 90.9082%;
Prime, near-prime, and government-insured: 99.7437%.
Year: 2007;
Subprime: 89.6104%;
Prime, near-prime, and government-insured: 99.8956%.
Year: 2008;
Prime, near-prime, and government-insured: 99.9988%.
Year: 2009;
Prime, near-prime, and government-insured: 100%.
Year: 2010;
Prime, near-prime, and government-insured: 100%.
Source: GAO analysis of CoreLogic data.
Note: We do not report the proportion of subprime mortgages after 2007
due to the low number of subprime originations. The CoreLogic database
we used for this analysis covers a broad cross-section of the mortgage
market. However, because of limitations in the coverage and
completeness of the data, our analysis may not be fully representative
of the mortgage market segments shown.
[End of figure]
Full Documentation:
A majority of the mortgages originated from 2001 through 2010 would
likely have met the QM criterion for full documentation of borrower
income and other financial resources, although low-or no-documentation
loans became common in certain market segments in the middle of the
decade. Low-or no-documentation of income or assets allows borrowers
to provide less detailed financial information than is traditionally
required. This feature was originally intended for borrowers who might
have difficulty documenting income, such as the self-employed, but
eventually became more widespread in the mid-2000s. As we previously
reported, mortgage originators or borrowers may have used the limited
documentation feature in some cases to overstate the financial
resources of borrowers and qualify them for larger, potentially
unaffordable loans.[Footnote 40]
The CoreLogic data on documentation level were not sufficiently
reliable for our purposes, but information from other sources provides
some insights on documentation practices during the 10-year period we
examined. FHFA analysis of mortgages purchased by the enterprises from
2001 through 2010 indicates that the proportion of mortgages
originated each year that were not "Alt-A," and therefore most likely
to have met the full documentation criterion, ranged from a low of
about 80 percent in 2006 (when enterprise-purchased mortgages
accounted for about one-third of the market) to a high of 100 percent
in 2010 (when the enterprises represented about two-thirds of the
market).[Footnote 41] According to FHA policy, all FHA-insured
mortgages, except the generally modest proportion that are streamlined
refinances (expedited refinancing from one FHA-insured loan into
another), are fully documented. As previously noted, FHA-insured
mortgages accounted for about 20 percent of new originations in 2009
and 2010 but for a substantially smaller share in prior years. As we
have previously reported, smaller proportions of subprime and near-
prime mortgages--which together grew to about 40 percent of mortgage
originations in 2006 but mostly disappeared after 2007--had full
documentation. Specifically, from 2001 to 2007, the proportion of
subprime mortgages with full documentation ranged from about 60
percent (in 2006) to 80 percent (in 2001), while the corresponding
proportion for near-prime mortgages ranged from about 20 percent (in
2006 and 2007) to just over 40 percent (in 2002).
Debt Service-to-Income Ratio:
Using an illustrative standard of 41 percent or less for the QM
criterion for DTI ratio, we found that more than half of the mortgages
originated from 2003 through 2010 for which reported DTI ratios were
available would likely have met the criterion; however, a sizable
proportion--from 25 to 42 percent--would not have.[Footnote 42] We did
not calculate corresponding percentages for 2001 and 2002 because our
CoreLogic data sample lacked DTI information for the large majority of
the mortgages originated in those years. The DTI ratio is a key
measure of a borrower's debt burden and is therefore a factor used in
assessing a borrower's ability to repay a loan. The proportion of
prime, near-prime, and government-insured mortgages that would have
met the illustrative 41-percent criterion decreased from about 75
percent in 2003 to 58 percent in 2008 and then increased to about 65
percent in 2009 and 2010 (see figure 5). Although the CoreLogic
database we used did not have DTI information for subprime mortgages,
we have previously reported that subprime mortgages originated from
2001 through 2005 had average reported DTI ratios of less than 41
percent and that those originated in 2006 and 2007 had average
reported DTI ratios of 41.1 and 41.5 percent, respectively.[Footnote
43] As a result, a substantial proportion of subprime mortgages would
not have met a 41-percent criterion.
Figure 5: Proportions of Prime, Near-Prime, and Government-insured
Mortgages Meeting Illustrative Qualified Mortgage Criterion for a Debt
Service-to-Income Ratio of 41 Percent or Less, 2003-2010:
[Refer to PDF for image: stacked vertical bar graph]
Year: 2003;
Equal to or less than 41%: 75%;
Greater than 41%: 25%.
Year: 2004;
Equal to or less than 41%: 72%;
Greater than 41%: 28%.
Year: 2005;
Equal to or less than 41%: 65%;
Greater than 41%: 35%.
Year: 2006;
Equal to or less than 41%: 62%;
Greater than 41%: 38%.
Year: 2007;
Equal to or less than 41%: 58%;
Greater than 41%: 42%.
Year: 2008;
Equal to or less than 41%: 58%;
Greater than 41%: 42%.
Year: 2009;
Equal to or less than 41%: 65%;
Greater than 41%: 35%.
Year: 2010;
Equal to or less than 41%: 65%;
Greater than 41%: 35%.
Source: GAO analysis of CoreLogic data.
Notes: Percentages reflect only those mortgages for which data were
available. About half of the mortgages in our CoreLogic data sample
did not have information on the DTI ratio for 2003 through 2010. We
concluded that those mortgages were likely not systematically
different from mortgages with DTI information based on a comparison of
the distribution of borrower credit scores associated with both groups
of mortgages, which showed little difference. Additionally, the
percentages shown are based on reported DTI ratios, which may
understate debt obligations or overstate income in some cases. As a
result, the proportions of mortgages we show as meeting the criterion
are likely somewhat higher than they would have been if all of the DTI
ratios had been calculated in a uniform and accurate manner. The Dodd-
Frank Act does not specify a maximum DTI ratio for QMs and authorizes
rulemakers to establish one. The CoreLogic database we used for this
analysis covers a broad cross-section of the mortgage market. However,
because of limitations in the coverage and completeness of the data,
our analysis may not be fully representative of the mortgage market
segments shown.
[End of figure]
Other Qualified Mortgage Criteria:
Our analysis suggests that for each year from 2001 through 2010, most
borrowers obtained mortgages with characteristics consistent with the
individual QM criteria we were able to examine. However, we were not
able to evaluate other QM criteria because of data limitations, and
rulemaking agencies have not yet established the final QM criteria.
The four criteria we were unable to examine were as follows:
* underwriting for fixed-rate mortgages is based on a fully amortizing
payment schedule that takes into account applicable taxes, insurance,
and assessments;
* underwriting for ARMs must be based on the maximum interest rate
allowed during the first 5 years and must take into account applicable
taxes, insurance, and assessments;
* total points and fees cannot exceed 3 percent of the total loan
amount; and:
* reverse mortgages must meet standards established by the Federal
Reserve Board.
The first two criteria address the practice of some lenders that
qualified borrowers for mortgages without assessing their ability to
pay taxes and insurance or make monthly payments that reflected
scheduled increases in interest rates.[Footnote 44] Requiring that
underwriting account for applicable taxes and insurance could help
ensure that borrowers can meet their responsibilities for paying these
costs in addition to their mortgage payment.[Footnote 45] Similarly,
requiring underwriting to be based on the maximum interest rate
allowed during the first 5 years could help ensure that borrowers have
the ability to pay scheduled increases in mortgage payments. Limiting
points and fees may protect borrowers against excessive up-front
charges that have been associated with predatory lending practices. We
were not able to examine the criterion concerning reverse mortgages
because the Federal Reserve Board did not propose standards for them.
In proposed regulations, the Federal Reserve Board indicated that QM
requirements were generally not relevant to reverse mortgages because
the Dodd-Frank Act does not subject reverse mortgages to the ability-
to-repay requirement.
The Dodd-Frank Act gives the Federal Reserve Board the authority to
add to, subtract from, or modify the QM criteria as implementing
regulations are developed. Additionally, the Department of
Agriculture's Rural Housing Service, FHA, and VA are required to
develop separate QM criteria for their loan programs through
regulations. The Federal Reserve Board issued proposed QM rules in
April 2011, and is accepting public comments through July 22, 2011.
The Dodd-Frank Act requires that the rules be finalized by no later
than January 2013.
Consumer and Industry Groups Cited Consumer Protection Benefits of the
Qualified Mortgage Criteria but also Raised Some Concerns:
Representatives from some consumer groups and the mortgage industry we
spoke with stated that they were generally supportive of certain QM
criteria in the Dodd-Frank Act because the criteria were associated
with a borrower's ability to repay a mortgage. Representatives from
one of the mortgage industry associations stated that the criteria
were consistent with their responsible lending policy, which was based
upon a consumer's ability to repay. Several consumer group
representatives stated that providing mortgages based upon a
borrower's ability to repay would ultimately benefit consumers by
providing them with sustainable products, such as 30-year fixed-rate
mortgages, that were easy to understand. In addition, some indicated
that QMs would protect eligible consumers from risky loan features,
such as abrupt interest rate increases that could cause payment shock.
Also, one consumer group indicated that the QM criteria could increase
the availability of affordable and sustainable mortgage credit by
encouraging lender competition in offering less risky mortgage
products and helping to increase investor confidence in private-label
RMBS.
Consistent with these views, research indicates that certain QM
criteria specified in the Dodd-Frank Act are associated with a
borrower's ability to meet their mortgage obligations. For example, we
and others have previously reported that no-and low-documentation
mortgages are associated with higher probabilities of default and
foreclosure, likely because borrowers' financial resources were
sometimes overstated, allowing for larger, potentially unaffordable
loans.[Footnote 46] Additionally, some research indicates that balloon
payments are associated with poorer loan performance, as some lenders
may use them to induce borrowers into mortgages with attractive
monthly payments without disclosing their long-term consequences. A
2007 study estimated that, controlling for other factors, subprime
refinance mortgages with balloon payments were 50 percent more likely
to experience a foreclosure than other loans.[Footnote 47] We
previously reported that mortgages with payment options that allowed
for negative amortization (by adding deferred interest payments to the
loan balance) could lead to payment shock when the interest-only or
payment option period expired.[Footnote 48] Homeowners who could not
afford the higher payments were more likely to enter foreclosure.
However, several of the mortgage industry representatives told us that
the QM criterion limiting total points and fees to 3 percent of the
total loan amount could increase the cost and decrease the
availability of mortgages for certain borrower groups, including
otherwise qualified low-income and minority borrowers. According to
these representatives, because certain costs for originating a
mortgage are fixed (i.e., do not vary with the size of the loan),
points and fees on smaller loans can easily exceed 3 percent of the
total (e.g., loans of $150,000 or less, according to one lender).
These representatives stated that a possible consequence of the cap
could be that lenders would increase interest rates on smaller loans
or be deterred from making them altogether. They indicated that this
outcome could disproportionately affect populations that tend to take
out smaller mortgages such as lower-income, first-time, rural, and
minority borrowers.
Several mortgage industry representatives also raised concerns about
the QM criteria that restrict mortgages with balloon payments and
create stricter underwriting standards for ARMs. According to these
representatives, both product types--which typically have lower
initial interest rates or monthly payments than comparable fixed-rate
mortgages--can be used responsibly under certain circumstances to make
mortgages more affordable in the short run. These representatives said
that these criteria could constrain mortgage options or delay
homeownership for borrowers that traditionally used such products,
including some rural and lower-income borrowers.[Footnote 49]
Concerns were also raised about DTI ratio requirements. While the Dodd-
Frank Act does not provide a maximum DTI ratio, it states that the QM
criteria must comply with any guidelines or regulations established by
the Federal Reserve Board relating to ratios of total monthly debt
service to monthly income (or alternative measures).[Footnote 50]
Several mortgage industry representatives stated that QM criteria that
include specific DTI ratios could restrict the availability of QMs for
retirees or those with irregular income streams. Industry
representatives that we met with also indicated that some retirees
might have small incomes but substantial assets to draw upon to meet
their mortgage obligations and that individuals with irregular
incomes, such as seasonal workers, could have trouble meeting income
documentation requirements. As a result, some creditworthy borrowers
might be prevented from obtaining QMs.
Representatives from several construction and mortgage industry
associations stated that the QM criteria could restrict new home
construction and mortgage refinancing. They said that the QM criteria
could make qualifying for a mortgage more difficult for some
borrowers, reducing demand for newly constructed homes. In addition,
officials from two mortgage industry associations stated that the QM
criteria could make it more difficult or expensive for some existing
homeowners to refinance their mortgages. In particular, the QM
criteria could affect homeowners who did not qualify for a QM or who
could not take advantage of "streamlined refinance" programs--which
allow qualified borrowers to refinance with their existing lenders
with less than full documentation and with reduced fees--because of
Dodd-Frank Act requirements for full documentation of borrower income
and assets.[Footnote 51]
Finally, a number of mortgage industry representatives expressed
concerns about the extent to which QMs would protect lenders from
legal claims by borrowers that the originating lenders had not
complied with the Dodd-Frank Act's ability-to-repay standard. Although
the Dodd-Frank Act provides some measure of protection from liability
for lenders of QMs, industry representatives we spoke with told us
that it was unclear whether that protection was intended to be a legal
"safe harbor" from liability--an interpretation they favored--or a
"rebuttable presumption of compliance" with the ability-to-repay
standard. A rebuttable presumption would allow borrowers to overcome
the presumption of compliance by providing evidence that the lender
did not, in fact, make a reasonable and good faith determination of
the borrower's ability to repay the loan. Consumer group
representatives told us that they favored this interpretation.
In April 2011, the Federal Reserve Board issued proposed regulations
concerning criteria for complying with the ability-to-repay standard,
including by originating a QM, and that addressed some of the concerns
related to the DTI ratio, the cap on points and fees, and balloon
loans. The proposed rules include two alternative definitions of a QM.
To help decide on a final definition, the Federal Reserve Board is
soliciting public comments on these two alternatives and invites
proposals for other definitions. The first alternative in the proposed
rule operates as a legal safe harbor and includes all of the QM
criteria described in the Dodd-Frank Act, with the exception of the
DTI ratio. The rules note that due to the discretion inherent in
making DTI ratio calculations, a requirement to consider the DTI ratio
would not provide certainty that a loan is a QM. The second
alternative provides a rebuttable presumption of compliance and
includes the QM criteria identified under the first alternative, as
well as other underwriting criteria--including consideration of
borrower DTI ratio or residual income, employment status, simultaneous
loans, current debt obligations, and credit history--drawn from the
Dodd-Frank Act's ability-to-repay standard. The proposed rules also
describe adjustments to and exclusions from the 3 percent cap on
points and fees (including for smaller loans) and the restrictions on
balloon payments for rural and underserved areas.
Regulatory Decisions and Other Factors Will Influence the Effect of a
Risk Retention Requirement on the Mortgage Market:
The risk retention requirement is intended to help align the interests
of key participants in the securitization market--securitizers,
lenders, and investors--and encourage sound loan underwriting. The
requirement mandates that securitizers of RMBS have an economic stake
in the securities they issue and therefore an incentive to ensure that
lenders originate well-underwritten mortgages that protect investors
from losses. Many industry stakeholders and consumer groups noted that
the implications of such a requirement would depend on a variety of
regulatory decisions and potential changes in the mortgage market.
These include decisions on the characteristics of QRMs that would be
exempt from the risk retention requirement, the forms of risk
retention that would be allowed, the percentage that securitizers
would be required to hold, and risk-sharing arrangements between
securitizers and lenders. These factors could affect the availability
and cost of mortgage credit and the future viability of the private-
label RMBS market. Some market participants and the rulemaking
agencies noted that risk retention may complement other securitization
and mortgage reforms, such as those that promote greater transparency
and enforcement of loan underwriting standards. Interactions between a
risk retention requirement and future changes to the federal
government's role in housing finance could also affect the cost of
mortgage credit and the private-label RMBS market.
Regulatory Decisions on Mortgage Characteristics Will Affect the Scope
and Implications of the Risk Retention Requirement:
Mortgage market participants and consumer groups that we interviewed
indicated that the effect of the risk retention requirement would
depend in large part on certain regulatory decisions. Rulemaking
agencies are accepting public comments on proposed risk retention
regulations through August 1, 2011. Restrictive criteria would limit
QRMs to mortgages with high credit quality, while less restrictive
criteria would expand QRMs to include mortgages with a wider range of
credit quality. Regulators' decisions about the criteria will
determine the proportion of securitized mortgages that are exempt from
a risk retention requirement and could affect the availability and
cost of mortgage credit for non-QRM borrowers. The Dodd-Frank Act
requires the rulemaking agencies to establish the definition by
considering mortgage underwriting and product features that historical
loan performance data indicate result in a lower risk of default. The
rulemaking agencies are considering a range of features, including LTV
and DTI ratios. Lower LTV ratios (indicative of larger borrower down
payments) and lower DTI ratios (indicative of smaller borrower debt
burdens) would represent more restrictive criteria.
To illustrate the potential impact of more and less restrictive QRM
criteria on the mortgage market, we used the CoreLogic data to
calculate the proportion of mortgages meeting certain LTV thresholds
and DTI ratio thresholds.[Footnote 52] We compared the percentage of
mortgages (prime, near-prime, and subprime combined) originated in
2006 and 2010 with LTV ratios of 80 percent or less (more restrictive
threshold) to the percentage of mortgages with LTV ratios of 90
percent or less (less restrictive threshold) (see figure 6). We made
similar comparisons for mortgages (prime and near-prime only) with
reported DTI ratios of 36 percent or less (more restrictive) or 41
percent or less (less restrictive).[Footnote 53] Our analysis showed
that about 82 percent of the mortgages originated in 2010 had LTV
ratios of 80 percent or less and that about 91 percent had LTV ratios
of 90 percent or less. The corresponding percentages of mortgages made
in 2006, just prior to the housing crisis, with restrictive and less
restrictive LTV ratios were about 78 percent and about 89 percent
respectively.
Figure 6: Proportions of Prime, Near-Prime, and Subprime Mortgages
Originated in 2006 and 2010 That Would Have Met Different Requirements
for LTV Ratio:
[Refer to PDF for image: vertical bar graph]
Year: 2006;
LTV ratio less than or equal to 80 percent: 78%;
LTV ratio less than or equal to 90 percent: 89%.
Year: 2010;
LTV ratio less than or equal to 80 percent: 82%;
LTV ratio less than or equal to 90 percent: 91%.
Source: GAO analysis of CoreLogic data.
Note: For this analysis, we excluded government-insured mortgages,
which typically have low down payments (high LTVs), because the Dodd-
Frank Act exempts them from the risk retention requirement. For the
mortgages that are included, we used the CoreLogic variable for LTV
ratio, which does not take any subordinate liens into account. We did
not use the variable for combined LTV ratio, which does take
subordinate liens into account, because it was not reliable. As a
result, the percentages we report are likely somewhat higher than they
would have been if we had been able to use combined LTV ratios. The
CoreLogic database we used for this analysis covers a broad cross-
section of the mortgage market. However, because of limitations in the
coverage and completeness of the data, our analysis may not be fully
representative of the market segments shown.
[End of figure]
Further, about 56 percent of the prime and near-prime mortgages
originated in 2010 with reported DTI ratios met the more restrictive
DTI threshold of 36 percent or less (see figure 7).[Footnote 54] About
70 percent of the prime and near-prime mortgages met the less
restrictive DTI threshold of 41 percent or less. The corresponding
percentages for mortgages originated in 2006 were about 44 percent and
about 62 percent respectively. An FHFA analysis of mortgages
originated in 2009 that were purchased by the enterprises illustrates
the impact of simultaneously applying multiple criteria in the
proposed QRM rules (including DTI and LTV thresholds, a requirement
that mortgage payments pay down principal, and a requirement for full
documentation).[Footnote 55] FHFA estimated that only about 31 percent
of these mortgages would have met the proposed QRM criteria they
examined. FHFA estimated that this percentage was even lower for
mortgages originated in previous years.
Figure 7: Proportions of Prime and Near-Prime Mortgages Originated in
2006 and 2010 That Would Have Met Different Requirements for DTI Ratio:
[Refer to PDF for image: vertical bar graph]
Year: 2006;
DTI ratio less than or equal to 36 percent: 44%;
DTI ratio less than or equal to 41 percent: 62%.
Year: 2010;
DTI ratio less than or equal to 36 percent: 56%;
DTI ratio less than or equal to 41 percent: 70%.
Source: GAO analysis of CoreLogic data.
Note: For this analysis, we excluded government-insured mortgages.
Percentages reflect only those mortgages for which data were
available. About 56 percent of the prime and near-prime mortgages from
2006 in our CoreLogic data sample and 43 percent of the 2010 mortgages
lacked information on the DTI ratio. We concluded that these mortgages
were likely not systematically different from mortgages with DTI
information based on a comparison of the distribution of borrower
credit scores associated with both groups of mortgages, which showed
little difference. Additionally, the percentages shown are based on
reported DTI ratios, which may understate debt obligations or
overstate income in some cases. As a result, the proportions of
mortgages we show as meeting the different DTI criteria are likely
somewhat higher than they would have been if all of the DTI ratios had
been calculated in a uniform and accurate manner. The CoreLogic
database we used for this analysis covers a broad cross-section of the
mortgage market. However, because of limitations in the coverage and
completeness of the data, our analysis may not be fully representative
of the market segments shown.
[End of figure]
FHFA's findings suggest that relatively restrictive QRM criteria could
ultimately subject a large proportion of mortgages that are not
insured or guaranteed by the government to a risk retention
requirement if the mortgages are securitized. Some mortgage industry
and consumer representatives we spoke with expressed concern that this
approach would subject some mortgages with relatively low default
risks to risk retention and make mortgage credit less affordable for
many borrowers, because the increased securitization costs would be
passed on to borrowers in the form of higher mortgage interest rates
and fees. Several also indicated that a risk retention requirement
that applied to a broad segment of the market could make
securitization a less attractive method for financing mortgages.
Because lenders may rely on securitization (as opposed to bank
deposits, for example) to provide funds for mortgage lending, actions
that make securitization more costly could hamper recovery of the
private-label RMBS market, according to some market participants.
Additionally, a range of industry stakeholders, including lenders and
consumer groups, told us that many creditworthy borrowers--
particularly low-and moderate-income households--are not able to make
a down payment of 20 percent and would therefore not qualify for QRMs
under the proposed rules.
However, federal regulators and other industry stakeholders favored
relatively restrictive QRM criteria and indicated that interest rates
for non-QRMs would likely be only modestly higher than those for QRMs.
For example, the Chairman of the FDIC has stated that it is
appropriate for the QRM definition to be narrowly drawn because QRMs
are intended to be the exception and not the rule. She said she
anticipates that QRMs will account for a small part of the mortgage
market and that mortgages securitized with risk retention or held in
lender's portfolios will provide more flexible options for borrowers
who cannot meet the QRM criteria. Further, rulemaking agencies
indicated that more restrictive QRM criteria could help ensure that a
sufficient volume of non-QRMs subject to risk retention would be
available for an active, liquid securitization market for such
mortgages. Federal regulators have also stressed that historical loan
performance data show that the mortgage characteristics in the
proposed QRM definition significantly influence the risk of mortgage
default. For example, FHFA analyzed mortgages originated from 1997
through 2009 and purchased by the enterprises. They estimated that of
the mortgages that would have met all of the other proposed QRM
criteria, those with LTV ratios of 80 percent or less had 90-day
delinquency rates that were 2.0 to 3.9 times lower than those with LTV
ratios greater than 80 percent and less than 90 percent.[Footnote 56]
Finally, FDIC officials have stated that risk retention should not
result in substantially higher interest rates for non-QRM borrowers--
less than half a percentage point, according to their estimates--and
comes with the benefit of safer and sounder lending practices.
[Footnote 57] They stressed that a 5 percent risk retention
requirement would increase costs to borrowers only to the extent that
it exceeded what investors would demand in the absence of the
requirement.
Finally, the Dodd-Frank Act exempts government-insured or -guaranteed
mortgages from the risk retention requirement but does not apply this
exemption to mortgages backed by the enterprises.[Footnote 58]
However, rulemaking agencies have proposed that the full guaranty
provided by the enterprises would satisfy the requirement while these
institutions are in conservatorship. Some market participants told us
that in the short term, this provision would limit the impact of a
risk retention requirement on the availability and cost of mortgage
credit because most mortgages, including many that would be non-QRMs,
are currently securitized by the enterprises. However, others have
argued that the proposed rules would help preserve the enterprises'
dominant market position by not subjecting them to the costs
associated with retaining 5 percent of the securities they issue. In
contrast, FHFA has indicated that requiring the enterprises to hold 5
percent of their securities would have little impact on the
enterprises' costs (because they already bear 100 percent of the
credit risk) and would be inconsistent with federal efforts to reduce
the mortgage assets held for investments by each enterprise.
Different Forms of Risk Retention Could Have Different Financial
Implications:
Several industry stakeholders we spoke with stated that different
forms of risk retention could have different implications for
securitizers' incentives and costs that in turn could affect mortgage
borrowers differently. The proposed risk retention rules provide
securitizers with a number of options for meeting the 5 percent risk
retention requirement, in recognition of the different securitization
structures and practices that exist for different classes of assets.
Federal Reserve Board officials said that this flexibility was
designed to reduce the proposed rules' potential to negatively affect
the availability and costs of credit. However, it is possible that
investors in RMBS will demand particular forms of risk retention or
amounts greater than 5 percent. Stakeholders we spoke with primarily
discussed two options that illustrate the differences in the potential
financial impacts of a risk retention requirement on securitizers:
retention of a tranche or multiple tranches of a securitization that
are the first to absorb losses ("horizontal" risk retention)--and
retention of a pro rata portion of each tranche of a security
("vertical" risk retention).[Footnote 59]
A number of securitization market participants and regulatory
officials indicated that retaining a horizontal slice of a
securitization would potentially provide greater incentives for
quality loan underwriting and would carry substantially higher capital
costs than vertical risk retention. Certain regulated securitizing
institutions, including banks and bank holding companies, are subject
to risk-based regulatory capital requirements, meaning they must hold
a minimum level of capital ("capital charges") to cover their risk
exposures, including assets held on their balance sheets.[Footnote 60]
Horizontal risk retention would require securitizers to retain an
economic interest in the part of the security that absorbs losses
first and carries a higher risk weight under regulatory capital
requirements.[Footnote 61] One credit rating agency with which we
spoke saw this approach as an advantage because the securitizers'
exposure to first losses would create incentives to ensure that the
mortgages backing the security were well underwritten. However,
because the high risk weighting would require securitizers to hold a
substantial amount of capital against the horizontal slice, these
capital costs would be expected to be passed on to borrowers in the
form of higher interest rates. Figure 8 illustrates the amount of
capital a securitizer may have to hold--applying risk-based capital
charges to each portion of a security--for a 5 percent horizontal
slice of a hypothetical $750 million RMBS, compared with the
corresponding amount of capital for a 5 percent vertical slice. In
this example, the securitizer would have to hold $37.5 million in
regulatory capital for horizontal risk retention (5 percent of $750
million--or $37.5 million--times the 100 percent capital charge for a
first-loss equity tranche). With vertical risk retention, the
securitizer would hold $2.6 million (the sum of the capital charges
for 5 percent of each tranche of the security).[Footnote 62] Because
of anticipated changes in capital requirements and calculations for
securitization exposures, capital charges for future RMBS may differ
from this illustrative example.[Footnote 63]
Figure 8: Illustrative Example of the Implications of Horizontal and
Vertical Risk Retention on Risk-Based Capital Charges:
[Refer to PDF for image: illustration]
Horizontal risk retention:
Total collateral: $750 million;
Credit rating: AAA: $637.5 million;
Credit rating: A: $37.5 million;
Credit rating: BBB: $37.5 million;
Equity: $37.5 million.
Percentage capital charge: 100%;
Total dollar capital charge: $37.5 million.
Total capital charge: $37.5 million.
Vertical risk retention:
Total collateral: $750 million;
Credit rating: AAA: $637.5 million (retained part of each tranche);
Percentage capital charge: 1.60%;
Total dollar capital charge: $510,000.
Credit rating: A: $37.5 million (retained part of each tranche)
Percentage capital charge: 4.00%;
Total dollar capital charge: $75,000.
Credit rating: BBB: $37.5 million (retained part of each tranche)
Percentage capital charge: 8.00%;
Total dollar capital charge: $150,000.
Equity: $37.5 million;
Percentage capital charge: 100%;
Total dollar capital charge: $1,875,000.
Total capital charge: $2.6 million.
Source: GAO.
Note: We used the regulatory risk weights that federal banking
regulators use to calculate the capital charges for horizontal and
vertical risk retention. Because this example is meant to be
illustrative, we did not apply all regulatory capital or accounting
standards that could influence the capital impacts of vertical and
horizontal risk retention.
[End of figure]
Additionally, market participants indicated that interactions between
horizontal risk retention and recent changes to accounting standards
for securitizations could increase the cost of securitizing mortgages.
Securitization typically involves the transfer of assets to an SPE
that removes assets from the securitizers' balance sheets and ensures
that investors still receive payments in the event of the bankruptcy
or failure of the securitizers. In 2009, the Financial Accounting
Standards Board issued financial accounting statement (FAS) 166, which
addresses whether securitizations and other transfers of financial
assets are treated as sales or financings, and FAS 167, which requires
securitizers or lenders with a controlling financial interest in an
SPE to "consolidate" the securitized assets on their balance sheets.
[Footnote 64] Although the need for accounting consolidation would
depend on the specific characteristics of each securitization
transaction, added on-balance sheet exposure from any consolidated
assets would generally result in higher regulatory capital
requirements for securitizers than if the assets were off-balance
sheet. A number of securitization market stakeholders indicated that
securitization could be economically unattractive in cases in which
accounting consolidation was triggered.
Vertical risk retention potentially exposes the securitizer to less
credit risk than horizontal risk retention because it involves
retaining a portion of every tranche of a security--some of which have
a relatively low risk of loss--rather than just the tranche in which
credit losses are concentrated. Vertical risk retention is also
considered less likely to result in accounting consolidation because
it potentially represents less of a financial interest in an SPE.
While the securitizer could therefore have less financial incentive to
securitize higher-quality mortgages, some market participants
indicated that vertical risk retention would help to align the
securitizer's interest with those of investors in each tranche.
Investor representatives, in particular, noted that when the
securitizer held only the bottom-most (first-loss) tranche and was
also the mortgage servicer, it could have an incentive to service the
mortgages in ways that favored just its tranche rather than all
tranche holders. For example, because lower tranches absorb initial
losses, they generally benefit from actions that delay the realization
of losses from mortgage defaults, which may include extending
repayment periods and postponing foreclosure. In contrast, senior
tranches generally benefit from actions that pay down mortgage
principal as quickly as possible, which may include expeditiously
foreclosing on a delinquent borrower and selling the foreclosed
property. Investor representatives indicated that having securitizers
hold a vertical slice of a security would help to ensure that
mortgages were serviced equitably for all tranche holders.
The Implications of a Risk Retention Requirement Will Depend on Other
Key Regulatory Decisions:
Implications of a Nonuniform Requirement:
The Dodd-Frank Act specifies a risk retention requirement for non-QRMs
of at least 5 percent but authorizes the rulemaking agencies to create
a different requirement--for example, greater than 0 and less than 5
percent--for non-QRMs that meet underwriting standards the agencies
prescribe. In March 2011, the rulemaking agencies proposed a uniform 5
percent level of risk retention for securitized non-QRMs. In the
proposed rule, the agencies indicated that they considered 5 percent
to be a minimum level of risk retention and suggested that levels
below 5 percent might not provide sufficient incentive for sound
mortgage underwriting in all circumstances.
A range of mortgage and securitization industry groups told us that a
nonuniform requirement would have both advantages and disadvantages,
some of which have implications for the cost and availability of
mortgage credit and risks to the mortgage market. On the one hand,
stakeholders noted that a nonuniform requirement could, in principle,
be more economically efficient than a uniform requirement because it
could allow the risk retention amount to be scaled to the risk level
of the mortgages being securitized. Further, some noted that a 5
percent requirement could be excessive for non-QRMs with relatively
low default risk, unnecessarily raising the cost of those mortgages
and tying up capital that could be used to securitize additional
mortgages. As previously discussed, some mortgage industry and
consumer group representatives indicated that if the final QRM
criteria were highly restrictive, non-QRMs could include some lower-
risk mortgages. On the other hand, industry stakeholders also stated
that a nonuniform requirement could potentially be difficult to
develop and enforce. To develop such a requirement, rulemaking
agencies would have to divide non-QRMs into different categories based
on risk level and develop an appropriate risk retention percentage for
each category. Several industry analysts indicated that it would be
challenging to calibrate a risk retention requirement that finely.
Additionally, assessing compliance with a requirement that had
multiple risk categories and retention levels could be more difficult
than assessing compliance with a uniform requirement.
Drawing general conclusions about whether a uniform or a nonuniform
risk retention requirement would be preferable is difficult, for two
reasons. First, historical and marketwide information about the
amount, form, and impact of risk retention in the secondary mortgage
market is limited. Industry stakeholders told us that risk retention
practices for private-label RMBS varied in terms of the slice (if any)
of the security that securitizers retained and how long and for what
purpose they retained it. More specifically, they described a range of
risk retention practices in the years leading up to the financial
crisis, including retaining 1 to 3 percent horizontal slices of near-
prime and subprime RMBS and nothing of prime jumbo RMBS. They also
indicated that when risk retention did occur, some securitizers held
the retained piece as part of an investment strategy, while others
sold it soon after issuance of the security. Analysis by FDIC of a
limited sample of prime and near-prime RMBS deals from 2001 through
2007 suggests that risk retention levels in the private-label market
varied considerably, ranging from less than 1 percent to over 8
percent for the deals they examined. Mortgage and securitization
industry participants also indicated that a lack of systematic
marketwide data on these practices had prevented analysis of how
different practices affected the incentives of market participants and
the quality of mortgage underwriting. Without this information, it is
difficult to determine whether a particular level of retained risk
would be optimal for all non-QRM mortgages or whether varying the
level depending on the credit quality of the mortgages would better
achieve the goals of risk retention. Given this uncertainty, the
requirement may need to be adjusted once regulators have assessed how
the private-label RMBS market has reacted to it. For example, if a
uniform 5 percent requirement was perceived as too high for some non-
QRMs and limited the availability of mortgage credit for certain
borrowers, regulators might want to consider a lower risk retention
requirement for those mortgages. Alternatively, if the regulation
established a requirement that was lower than that dictated by
investors in the market, some increase might be warranted.
Second, rulemaking agencies have not made final decisions about the
QRM criteria or other aspects of the risk retention requirement, and
these decisions could influence whether a uniform standard would be
more appropriate. For example, a nonuniform requirement could be more
appropriate if the final QRM definition were restrictive (i.e.,
limited to mortgages of very high credit quality), because non-QRMs
would potentially include mortgages with a wide range of credit risks.
Conversely, a uniform requirement could be more appropriate if the QRM
definition were less restrictive, because non-QRMs would potentially
encompass a narrower range of credit risks.
Implications of Risk-Sharing for Lenders:
The Dodd-Frank Act places the responsibility for retaining risk on
securitizers but authorizes rulemaking agencies to require that
lenders share the risk retention obligations.[Footnote 65] The
proposed rules do not require lenders to retain risk but would permit
a securitizer to allocate a portion of its risk retention requirement
to any lender that contributed at least 20 percent of the underlying
assets in the pool. Additionally, the proportion of risk retained by
each lender could not exceed the percentage of the securitized assets
it originated, and the lender would have to hold its allocated share
in the same manner (e.g., vertical or horizontal) as the securitizer.
The impact of the risk retention requirement on lenders will depend,
in part, on how the risk retention requirement is shared.[Footnote 66]
If lenders are required to share risk (either directly by regulation
or indirectly though an allocation from a securitizer), they would
have to hold capital against this risk exposure.[Footnote 67] Several
mortgage industry representatives indicated that smaller lenders, such
as independent mortgage companies and small community banks, could
lack sufficient capital resources to share risk retention obligations
or hold non-QRMs that were not securitized on their balance sheets.
[Footnote 68] A few of the mortgage and securitization market
participants we spoke with said that, in contrast, large lenders had
the financial capacity to share risk retention obligations with
securitizers or hold non-QRMs on their balance sheets, giving these
lenders an advantage over smaller lenders that could ultimately reduce
competition in mortgage lending. While acknowledging some of these
concerns, the rulemaking agencies have estimated that the proposed
requirement would not have a significant impact on a substantial
number of small banking institutions, at least under current market
conditions. They cited data indicating that small lenders generally
did not securitize mortgages themselves, did not contribute 20 percent
or more of the mortgages to private-label securitizations, and
primarily securitized mortgages through the enterprises. A number of
market participants noted that even if lenders were not required to
share risk in the manner prescribed by the Dodd-Frank Act,
securitizers could be expected to take steps to transfer the cost of
risk exposure by paying lenders less for the mortgages they sold or
requiring additional collateral to ensure the underwriting quality of
the mortgages. However, others noted that lenders would pass this cost
on to borrowers and that the cost would likely be marginal.
A Risk Retention Requirement Will Complement or Interact With Other
Efforts to Reform Housing Finance:
Risk retention is a part of legislative and regulatory efforts to
reduce undisclosed risks to the overall credit market by addressing
the weaknesses in the securitization process that contributed to the
housing crisis. The risk retention requirement complements other parts
of the Dodd-Frank Act that are intended to improve the securitization
markets, as well as existing mechanisms to encourage lenders to sell
high-quality loans. For example, the Dodd-Frank Act requires issuers
of asset-backed securities, including RMBS, to conduct reviews of the
assets underlying the securities and disclose the nature of the
reviews.[Footnote 69]
The Dodd-Frank Act also requires credit rating agencies and
securitizers of residential mortgages (and other assets) to publicly
disclose information about representations and warranties--the
assertions lenders make about a loan's underwriting standards and
contractual obligations to refund the value of the loan if the
assertions later prove to be untrue. Representations and warranties
existed prior to the mortgage crisis, and in some cases lenders failed
to repurchase loans that violated these terms because they could not
afford to and subsequently went out of business. Several market
participants and researchers told us that these mechanisms would be
more effective if they were better monitored and enforced, possibly by
using third parties to verify loan information or requiring that
originators demonstrate the financial ability to honor warranties and
repurchase requests. The Dodd-Frank Act requires credit rating
agencies to disclose the representations, warranties, and enforcements
available to investors when the agency issues a credit
rating.[Footnote 70] Securitizers are required to disclose fulfilled
and unfulfilled repurchase requests so that investors can identify
lenders' records related to such requests. In addition, a trade
association representing the securitization industry has an ongoing
effort to make representations and warranties for RMBS more
standardized and transparent.
The risk retention requirement will also interact with efforts to
reduce the federal government's role in mortgage finance, which could
have implications for mortgage borrowers and the private-label RMBS
market. Mortgages backed by the enterprises and FHA currently dominate
the mortgage market, and the private-label RMBS market is largely
dormant. However, the administration and Congress are considering
options that would diminish the federal role and help transition to a
more privatized market by winding down the enterprises and reducing
the size of FHA.[Footnote 71] Several mortgage market participants
indicated that in the long run, if the enterprises were eliminated or
their activities scaled back, more non-QRMs would be subject to risk
retention, potentially raising the cost of these mortgages for
borrowers. Potential changes in FHA's role also could influence how a
risk retention requirement would affect mortgage borrowers. In
addition to non-QRMs, FHA-insured mortgages, which are exempt from
risk retention, are a potential alternative for borrowers who may not
qualify for mortgages that do not meet the QRM criteria. FHA borrowers
often make down payments that are low (generally less than 5 percent)
compared with the 20 percent that has been proposed as the QRM down
payment requirement for home purchase mortgages.[Footnote 72] However,
in recent years, FHA has tightened its underwriting standards and
raised insurance premiums as it tries to reduce its market share and
strengthen its financial condition. For example, beginning in 2010 FHA
began requiring borrowers with lower credit scores to make larger down
payments. Additionally, FHA is considering further steps, such as
increasing down payment requirements more broadly. As a result, FHA
may not provide mortgage alternatives for as many non-QRM borrowers as
it would have in the past.
While Housing Counseling and High-Cost Lending Provisions May Enhance
Borrower Protections, Specific Impacts Are Unknown:
HUD has initiated plans to establish an Office of Housing Counseling,
as required under the Dodd-Frank Act.[Footnote 73] HUD already
performs a number of activities that are consistent with the new
office's authorized functions and plans to move these functions into
the new office. Industry and consumer groups we spoke with identified
opportunities for the counseling office to enhance HUD's role in
housing counseling, but the financial resources for the office are
uncertain. Findings from the limited research available on housing
counseling are mixed, with some studies suggesting that some types of
counseling can improve mortgage outcomes and others finding no effect.
The Dodd-Frank Act supports another consumer protection by changing
the definition of high-cost loans under HOEPA. This change could
prevent some high-cost lending, although whether the definition would
affect mortgages currently available to consumers is unclear.
HUD Is Creating a New Office for Its Housing Counseling Activities,
but Funding Is Uncertain:
To enhance consumer protections for homebuyers and tenants, the Dodd-
Frank Act requires HUD to establish an Office of Housing Counseling.
This office will perform a number of functions related to
homeownership and rental housing counseling, including establishing
housing counseling requirements, standards, and performance measures;
certifying individual housing counselors; conducting housing
counseling research; and performing public outreach. The office is
also mandated to continue HUD's role in providing financial assistance
to HUD-approved counseling agencies in order to encourage successful
counseling programs and ensure that counseling is available in
underserved areas.
Currently, HUD's housing counseling program operates out of the
Program Support Division within the Office of Single-Family Housing.
[Footnote 74] HUD supports housing counseling through the division in
two ways. First, it approves and monitors housing counseling agencies
that meet HUD criteria and makes information about these agencies
available to consumers on HUD's Web site. According to HUD officials,
as of May 2011, about 2,700 counseling agencies were HUD-approved.
Second, HUD annually awards competitive grants to approved agencies to
help them carry out their counseling efforts. HUD's housing counseling
program provides funding for the full spectrum of housing counseling,
including prepurchase counseling, foreclosure mitigation counseling,
rental housing counseling, reverse mortgage counseling for seniors,
and homeless assistance counseling. HUD-approved agencies report to
HUD on the type the number and type of service interactions (e.g.,
counseling sessions) they have with clients. Self-reported data on
homeownership counseling conducted by these agencies indicate that
service interactions for foreclosure mitigation counseling rose from
about 171,000 in 2006 to more than 1.4 million in 2010, while service
interactions for prepurchase counseling declined from about 372,000 to
about 245,000 over the same period.[Footnote 75]
Besides these two main functions, the Program Support Division and
other HUD staff perform other counseling-related activities, some of
which are similar to the functions the Dodd-Frank Act requires of the
new counseling office. For example, HUD has developed standards and
protocols for reverse mortgage counseling, certifies individual
reverse mortgage counselors, is conducting research on the impact of
homeownership counseling, and recently launched a public awareness
campaign on loan modification scams.
A working group within HUD is in the process of developing a plan for
the new counseling office. According to HUD officials, the primary
change needed to create the new office is the reassignment of the
approximately 190 staff who spend most of their time on housing
counseling activities but also have other responsibilities. HUD
expects the new office to consist of approximately 160 full-time staff
members. In order to move forward with the establishment of the office
and the appointment of a Director of Housing Counseling, HUD must
submit a plan to Congress for approval.
HUD officials told us that the new counseling office would have
advantages over their current organizational structure. They indicated
that having dedicated resources, staff, and leadership would raise the
profile of the housing counseling function and help the agency build a
more robust capacity in this area. One official noted that getting
sufficient information technology resources for housing counseling had
been difficult and said that a separate counseling office might be
able to compete more effectively with other parts of the agency for
these resources. HUD officials also indicated that the new office
would be organized to help the agency better anticipate and respond to
changing counseling needs and improve interaction with counseling
industry stakeholders. For example, the officials said that the new
office would be organized around functional areas such as policy,
training, and oversight, making it easier for industry stakeholders to
direct their questions or concerns to the appropriate HUD staff.
Additionally, HUD officials told us that the office would work with
the CFPB's Office of Financial Literacy in the future to coordinate
the housing counseling activities of both organizations.
Mortgage industry participants, consumer groups, and housing
researchers we spoke with were supportive of the new housing
counseling office and believed that it offered opportunities to
enhance HUD's role in the housing counseling industry. For example,
some of the consumer groups stated that the office could help
standardize counseling practices and publicize best practices, further
elevating and professionalizing the counseling industry. In addition,
representatives from several of the consumer groups and researchers we
met with stated that the office could help enhance coordination among
counseling agencies by providing opportunities for improved training,
networking, and communication. Furthermore, they said that the office
could potentially support improved data collection for research on the
impact of housing counseling.
Budget constraints could delay the establishment of the new counseling
office and reduce the scale of HUD's housing counseling activities.
Although the Dodd-Frank Act authorized $45 million per year for fiscal
years 2009 through 2012 for the operations of the new office, HUD had
not received any appropriations for this purpose as of May 2011. In
addition, appropriations for fiscal year 2011 eliminated HUD's housing
counseling assistance funds, which are primarily grant funds for
approved counseling agencies.[Footnote 76] According to a HUD
official, as a result of this funding reduction, HUD is revising its
proposal for the new counseling office and is unable to estimate when
it will submit the proposal to Congress. HUD officials said they would
begin the awards process for about $10 million in unspent fiscal year
2010 counseling assistance funds in May 2011 but expressed concern
that some counseling agencies would run out of funds soon and might
not receive additional HUD funding until well into fiscal year 2012.
Housing counseling groups we spoke with said that the cuts in HUD
funding, which they use to leverage private funds, ultimately could
result in fewer counseling services for prospective and existing
homeowners unless private funds make up the difference.[Footnote 77]
Findings from the Limited Research Available on Homeownership
Counseling Are Mixed:
Empirical research on outcomes for homeownership counseling is
limited, with some studies suggesting that foreclosure mitigation
counseling can be effective in improving mortgage outcomes (e.g.,
remaining current on mortgage payments versus defaulting or losing the
home to foreclosure).[Footnote 78] However, findings on prepurchase
counseling are less clear. Considered to be one element of financial
literacy, these types of homeownership counseling are based on the
idea that providing information and advice can help consumers make
better decisions about home purchases and maintenance and work more
successfully with lenders and mortgage servicers to obtain loan
modifications or refinancing.
Conducting research on homeownership counseling outcomes is
challenging for a variety of reasons, and limitations in the
methodologies used in existing studies make it hard to generalize the
results. According to housing counseling researchers we spoke with,
the primary barrier in the study of housing counseling is a lack of
data. Long-term data on counseling outcomes are limited because of the
difficulty of tracking counseling recipients after the counseling
ends. In addition, many counseling agencies are hesitant to request
sensitive personal information from clients. One researcher we spoke
with told us that the ability to track loan performance over time is
critical to an effective assessment of housing counseling programs.
For this reason, some counseling researchers have begun working with
lenders and mortgage servicers to access information on the payment
status (e.g., current or delinquent) of counseling recipients and the
long-term outcomes of their mortgages.
Another limitation of the current research is the lack of experimental
research design, which is considered the best approach for evaluating
differences in an intervention such as counseling and comparing it to
no intervention.[Footnote 79] Studies that employ experimental designs
are often difficult and costly to conduct. We did not identify any
published studies that evaluated homeownership counseling using an
experimental design. Further, researchers have not been able to
overcome another inherent limitation: the fact that consumers choose
counseling themselves, generally voluntarily, and those who choose
counseling may differ in unknown ways from those who do not.[Footnote
80] Both of these issues make researchers hesitant to draw firm
conclusions from the published literature. Finally, differences among
counseling programs--in terms of curriculum, intervention method
(e.g., one-on-one, telephone, classroom), level of intervention (e.g.,
intensity or amount of time spent counseling), and outcome measures--
also make it difficult to draw general conclusions about the impact of
housing counseling. A selected bibliography of research we reviewed on
outcomes for prepurchase and foreclosure mitigation counseling appears
at the end of this report.
The limited body of evidence available is not conclusive on the impact
of all types of housing counseling. However, recent research on
foreclosure mitigation counseling suggests that it can help struggling
mortgage borrowers avoid foreclosure and prevent them from lapsing
back into default, especially if counseling occurs early in the
foreclosure process. A 2010 evaluation of the National Foreclosure
Mitigation Counseling (NFMC) Program found that homeowners who
received counseling under the program were more likely to receive loan
modifications and remain current after counseling, compared with a
group of non-NFMC borrowers with similar observable characteristics.
[Footnote 81] Specifically, the authors estimated that borrowers who
received NFMC counseling were 1.7 times more likely to "cure" their
foreclosure (i.e., be removed from the foreclosure process by their
mortgage servicer) than borrowers who did not receive NFMC counseling.
The authors also estimated that loan modifications received by NFMC
clients in the first 2 years of the program resulted in monthly
mortgage payments of $267 less on average than what they would have
paid without the help of the program. Additionally, the study found
that for borrowers counseled in 2008, the relative odds of bringing
their mortgages current were an estimated 53 percent higher if they
received counseling prior to receiving a loan modification than if
they did not receive NFMC counseling. Other studies of foreclosure
prevention counseling have also found that the timing of the
counseling was critical and that the earlier in the foreclosure
process borrowers received counseling, the more likely they were to
have a positive outcome.[Footnote 82]
The findings on prepurchase counseling are less clear. For example, a
2001 study analyzed data on the performance of about 40,000 mortgages
made under a Freddie Mac program for low-to moderate-income
homebuyers, a large majority of whom received prepurchase counseling.
[Footnote 83] The authors compared the loan performance of program
participants who received different types of prepurchase counseling to
the loan performance of participants who did not. The study found that
borrowers who underwent individual and classroom counseling were 34
and 26 percent less likely, respectively, to become 90 days delinquent
on their mortgages than similar borrowers who did not undergo
counseling.[Footnote 84] However, subsequent studies have found either
no effect on loan performance or effects that were potentially
attributable to other factors. For example, a 2008 study of about
2,700 mortgage borrowers found that prepurchase counseling had no
effect on a borrower's propensity to default.[Footnote 85] A 2009
study examined a legislated pilot program in 10 Illinois ZIP codes
that mandated prepurchase counseling for mortgage applicants whose
credit scores were relatively low or who chose higher-risk mortgage
products such as interest-only loans. Although the authors found that
mortgage default rates for the counseled low-credit score borrowers
were lower than those for a comparison group, the authors attributed
this result primarily to lenders tightening their screening of
borrowers in response to stricter regulatory oversight.[Footnote 86]
Additional empirical research on the impact of housing counseling is
under way at HUD and Fannie Mae. HUD's Office of Policy Development
and Research issued a broad overview of the housing counseling
industry in 2008 and is currently conducting two studies on mortgage
outcomes related to foreclosure mitigation and prepurchase counseling
programs.[Footnote 87] The foreclosure mitigation study will follow
880 individuals and evaluate mortgage outcomes 12 months after
counseling ends. HUD officials said that they expected the study to be
published in 2012. The prepurchase counseling study will use an
experimental design and will track 1,500 to 2,000 individuals who
receive different types of counseling (one-on-one, group, Internet, or
telephone) or no counseling. HUD officials said they expected data
collection for this study to begin in 2012. In addition, Fannie Mae is
conducting both prepurchase and postpurchase counseling studies.
According to Fannie Mae officials, the prepurchase study will track
over a 2-year period the loan performance of borrowers who received
counseling prior to purchasing a home. The postpurchase study will
evaluate the impact of telephone counseling on existing homeowners who
receive loan modifications through the Department of the Treasury's
Home Affordable Modification Program.[Footnote 88]
Expanded HOEPA Provisions Could Protect Homebuyers by Providing
Additional Restrictions For High-Cost Lending:
As previously noted, HOEPA regulates and restricts the terms and
characteristics of mortgages that exceed specified APR and fee
triggers. For these "high-cost loans," HOEPA requires enhanced
preclosing disclosures to borrowers, restricts certain loan contract
terms, and imposes penalties on lenders for noncompliance. In
addition, HOEPA imposes liabilities on purchasers or securitizers
("assignees") of high-cost loans for violations of law committed by
the mortgage originators.[Footnote 89] Because of the associated
penalties and liabilities, lenders have generally avoided making high-
cost loans, and the secondary market for these loans has been
negligible. Data collected under the Home Mortgage Disclosure Act
(HMDA) indicate that in 2004 (the first year for which marketwide data
on high-cost loans are available), lenders reported making 23,000 high-
cost loans, which accounted for only 0.003 percent of all the
originations of home-secured refinance or home improvement loans
reported for that year.[Footnote 90] The number of reported high-cost
loans rose to about 36,000 in 2005 but fell every year thereafter. In
2009, the most current year for which HMDA data are available, these
loans numbered only 6,500, which, in aggregate, made up less than 0.1
percent of all the originations of home-secured refinancing and home
improvement loans reported for that year.
The Dodd-Frank Act expanded the definition of high-cost mortgages in
several ways. Prior to the Dodd-Frank Act, the definition of such
loans applied only to refinance loans and closed-end home equity loans
(e.g., home improvement loans) secured by the borrower's principal
dwelling.
However, the Dodd-Frank Act expanded the definition of high-cost
mortgages by:
* Applying the high-cost triggers to a wider range of loan types,
including mortgages for purchasing a home, open-end loans, and any
other home-secured loan other than a reverse mortgage.
* Lowering the APR trigger from 8 percentage points to 6.5 percentage
points over the average prime offer rate for first liens, and from 10
percentage points to 8.5 percentage points over the average prime
offer rate for subordinate liens.[Footnote 91]
* Lowering the points and fees trigger from 8 percent to 5 percent of
the total loan amount and banning the financing of points and fees.
[Footnote 92]
* Adding a third trigger for prepayment penalties extending beyond 36
months from mortgage closing or exceeding 2 percent of the outstanding
balance of the mortgage.
In addition, the Dodd-Frank Act prohibits prepayment penalties for
high-cost loans and requires that borrowers undergo counseling with a
HUD-approved counselor before taking out a high-cost loan.
Data limitations make assessing the potential impact of the new
definition difficult, but the views of industry stakeholders and prior
research provide some useful perspectives. Additional information
would be needed to assess the extent to which the new definition would
affect mortgages currently available to consumers. As we have
previously reported, marketwide data on APRs, points, and fees are not
readily available to researchers.[Footnote 93] As a result,
determining the proportion of mortgages made in recent years that
might have met the new high-cost triggers is difficult. Industry
stakeholders we spoke with indicated that the new definition of high-
cost loans would further expand disincentives for originating
mortgages with potentially predatory terms and conditions.
Additionally, they said that lenders would likely continue to avoid
offering high-cost loans because the strict penalties and liabilities
attached to these loans make them risky to originate and difficult to
securitize. In prior work, we examined research on the impact of state
and local anti-predatory lending laws--some of which are similar to
HOEPA--on subprime mortgage markets. This research provides some
evidence that anti-predatory lending laws can have the intended effect
of reducing loans with problematic features without substantially
affecting credit availability.[Footnote 94]
Implementing mortgage-related provisions in the Dodd-Frank Act will
involve tradeoffs between providing consumer protection and
maintaining credit availability. Additionally, potential interactions
with plans to scale back government involvement in the mortgage market
and expand the role of private capital add complexity to
implementation efforts. Limited data and research show that certain
provisions could provide benefits to homebuyers and the larger
mortgage market. However, the ultimate impact of the Dodd-Frank Act's
mortgage-related requirements is not yet known and will depend, in
part, on regulatory actions, decisions to fund housing counseling, and
mortgage market adjustments that have not yet occurred.
Agency Comments and Our Evaluation:
We provided a draft of this report to the Federal Reserve Board, FDIC,
FHFA, OCC, OTS, NCUA, HUD, and SEC for their review and comment. We
received written comments from the Chairman of the NCUA that are
reprinted in appendix III. We also received technical comments from
the Federal Reserve Board, FDIC, FHFA, HUD, OCC, and SEC, which we
incorporated as appropriate. OTS did not provide comments on the draft
report.
In its written comments, NCUA indicated, as we do, that the impact of
the Dodd-Frank Act would depend on regulatory decisions that had yet
to be made. NCUA also said that while our report found that most
mortgages would have met individual QM criteria, applying the criteria
simultaneously would narrow the population of loans that would qualify
as QMs. While this is a reasonable conclusion, as stated in our
report, we were unable to determine the proportion of mortgages
meeting all of the QM criteria we examined because of limitations in
the data (e.g., missing or unreliable values) available for our
analysis. We added language to the report to clarify the impact of
these limitations on our analysis.
With respect to rulemaking efforts, NCUA expressed concern about the
lack of a mechanism for non-QMs to receive QM status after some period
of performance given the potential difficulty some borrowers,
including those of modest means, may have in meeting the QM criteria.
NCUA suggested that creating such a mechanism could help achieve the
goal of protecting borrowers from unsustainable mortgage products
while maintaining broad access to mortgage credit.
We are sending copies of this report to the appropriate congressional
committees, the Chairman of FDIC, the Chairman of the Federal Reserve
Board, the Acting Director of FHFA, the Secretary of Housing and Urban
Development, the Chairman of NCUA, the Acting Comptroller of the
Currency, the Chairman of SEC, the Acting Director of OTS, the Bureau
of Consumer Financial Protection, and other interested parties. In
addition, the report is available at no charge on the GAO Web site at
[hyperlink, http://www.gao.gov].
If you or your staffs have any questions about this report, please
contact me at (202) 512-8678 or shearw@gao.gov. Contact points for our
Offices of Congressional Relations and Public Affairs may be found on
the last page of this report. Key contributors to this report are
listed in appendix IV.
Signed by:
William B. Shear:
Director, Financial Markets and Community Investment:
[End of section]
Appendix I: Objectives, Scope, and Methodology:
Our objectives were to (1) assess the proportions of mortgages
originated from 2001 through 2010 that would have met selected
qualified mortgage (QM) criteria specified in the Dodd-Frank Wall
Street Reform and Consumer Protection (Dodd-Frank Act) and describes
the views of mortgage industry stakeholders on the potential effects
of the QM criteria on the mortgage market, (2) discuss relevant
information and the views of mortgage industry stakeholder on the
potential impact of a risk retention requirement on the mortgage
market and discuss the advantages and disadvantages of a uniform risk
retention requirement, and (3) describe what research and the views of
mortgage industry stakeholders suggest about the potential impact of
provisions in the Dodd-Frank Act regarding homeownership counseling
and changes to the Home Ownership Equity Protection Act (HOEPA).
Assessment of Qualified Mortgage Criteria:
To assess the proportions of recent loans that would likely have met
selected QM criteria, we reviewed relevant statutory provisions and
proposed rules to implement those provisions. We applied the QM
criteria to mortgages in a proprietary loan-level servicing database
from CoreLogic, Inc. This database contains information from major
mortgage servicers and covers a broad cross-section of the mortgage
market. For example, CoreLogic estimates that for the period we
examined, the database captures 60 to 65 percent of the mortgages
purchased by Fannie Mae and Freddie Mac (the enterprises),
approximately 50 percent of subprime mortgages, and about 90 percent
of mortgages with government insurance or guarantees. Nevertheless,
because of limitations in the coverage and completeness of the data,
our analysis may not be fully representative of the mortgage market as
whole. For our analysis, we used a random 10 percent sample of the
database that amounted to about 6.6 million mortgages for the 2001
through 2010 period. Our sample included purchase and refinance
mortgages and mortgages to owner-occupants and investors, and excluded
second-lien mortgages. We assessed the reliability of the CoreLogic
data by interviewing CoreLogic representatives about the methods the
firm used to collect and ensure the integrity of the information. We
also reviewed supporting documentation about the database. In
addition, we conducted reasonableness checks on the data to identify
any missing, erroneous, or outlying figures. We concluded that the
data elements we used in this objective and the following objective
were sufficiently reliable for our purposes.
We focused on mortgages originated from 2001 through 2010 to provide
insight into the potential effects of the Dodd-Frank Act's provisions
under different market conditions and lending environments. We applied
each QM criterion separately, calculating the proportion of mortgages
in each annual loan origination cohort that likely would have met the
criterion. We were unable to determine the proportion of mortgages
that would have met all of the criteria we examined due to the number
of records in the database that had missing or unreliable values for
one or more of the criteria. For example, the database contained no
information on DTI ratio for subprime mortgages and did not have
reliable information on documentation of borrower income and assets.
As a result, we determined that applying the criteria simultaneously
would not have produced reliable results. Because the CoreLogic data
group mortgages into two broad categories--one containing prime, near-
prime, and government-insured loans and another containing subprime
loans--we examined these categories separately when possible.
The data did not contain information needed to examine all of the QM
criteria specified in the Dodd-Frank Act. As a result, our analysis
focused on five of the nine QM criteria specified in the Dodd-Frank
Act for which sufficient data, including data from the CoreLogic
database, were available. These criteria were:
* regular periodic payments do not result in an increase in the
principal balance or result in a deferral of the repayment of
principal;
* the loan term does not exceed 30 years;
* except for balloon loans under specified circumstances, the mortgage
does not include balloon payments;
* borrower income and financial resources are verified and documented;
and:
* the loan complies with guidelines or regulations established by the
Board of Governors of the Federal Reserve System (Federal Reserve
Board) relating to ratios of total monthly debt to monthly income
(e.g., debt service-to income (DTI) ratio).
In general, for each year from 2001 through 2010, we identified the
proportion of mortgage originations that would have met the individual
criteria. We were not able to calculate relevant proportions for
certain years and mortgage market segments due to data limitations.
For example, because few subprime mortgages were originated after
2007, we only present data for 2001 through 2007 for that market
segment. Regarding the criterion for repayment of principal, our
analysis focused on mortgages with negative amortization features. Our
analysis does not account for interest-only mortgages because most
loan records in the CoreLogic data had missing values for the interest-
only indicator. The Dodd-Frank Act does not contain a specific
threshold for DTI ratio and leaves that decision to rulemakers. For
illustrative purposes, we used the 41 percent ratio that is used as a
guideline for underwriting mortgages insured by the Federal Housing
Administration (FHA). The CoreLogic database did not contain
information on DTI for any subprime mortgages, nor for many prime,
near-prime, and government-insured mortgages. Because DTI information
was missing for the large majority of mortgages originated in 2001 and
2002, we only present DTI data for 2003 through 2010. For the latter
period, about 53 percent of the mortgages in the CoreLogic data sample
did not have DTI information. We concluded that those mortgages were
likely not systematically different from mortgages with DTI
information based on a comparison of the distribution of borrower
credit scores associated with both groups of mortgages, which showed
little difference. Additionally, our analysis was based on reported
DTI ratios, which may understate debt obligations or overstate income
in some cases. As a result, the proportions of mortgages we show as
meeting the criterion are likely somewhat higher than they would have
been if all of the DTI ratios had been calculated in a uniform and
accurate manner. To examine the subprime market segment, we drew upon
information from a prior analysis we conducted of mortgage
characteristics using a separate CoreLogic database that captures a
large majority of subprime mortgages.[Footnote 95] We also examined a
fifth QM criterion in the Dodd-Frank Act--documentation of borrower
income and assets--using information from that prior analysis, data
from the Federal Housing Finance Agency (FHFA) on mortgages purchased
by the enterprises, and information on FHA policies concerning
borrower documentation. (Although the CoreLogic database contained
information on documentation level, we determined that it was not
sufficiently reliable for our purposes.) For certain QM criteria
(repayment of principal, loan term, balloon payment, and DTI ratio),
we performed a similar analysis by geographic groupings based on
racial, ethnic, income, and house price patterns. Appendix II contains
the results and methodology for this analysis. Data limitations
prevented us from assessing three of the remaining four QM criteria
contained in the Dodd-Frank Act--specifically, those relating to
interest rates used for underwriting adjustable-rate mortgages,
consideration of applicable taxes and insurance in underwriting, and
limitations on points and fees. We were not able to examine the fourth
criterion concerning reverse mortgages because the Federal Reserve
Board did not establish QM standards for them. In proposed
regulations, the Federal Reserve Board indicated that QM requirements
were not relevant to reverse mortgages because the Dodd-Frank Act does
not subject reverse mortgages to the ability-to-repay requirement.
To obtain additional information and views on the potential effects of
the QM criteria specified in the Dodd-Frank Act, we reviewed relevant
research literature and conducted interviews with nearly 40 individual
mortgage and securitization industry stakeholders. These stakeholders
included representatives from financial services companies (major
mortgage lenders and mortgage securitizers); groups representing
mortgage lenders, brokers, securitizers, and investors; groups
representing consumer interests; and academics. Additionally, we
interviewed officials from the Federal Reserve Board, Office of the
Comptroller of the Currency (OCC), Federal Deposit Insurance
Corporation (FDIC), Office of Thrift Supervision (OTS), National
Credit Union Administration (NCUA), Department of Housing and Urban
Development (HUD), and FHFA. We also reviewed testimonies and
published papers from these stakeholders that documented their views.
Assessment of Risk Retention Requirement:
To assess the potential impact of the Dodd-Frank Act's risk retention
requirement on the mortgage market, we reviewed relevant statutory
provisions and proposed rules to implement those provisions. We also
reviewed available information on mortgage securitization practices
prior to the financial crisis and factors that may affect the impact
of the risk retention requirement, including information from two
other studies on risk retention required by the Dodd-Frank Act from
the Federal Reserve Board and the Financial Stability Oversight
Council. To assess the implications of interactions between the risk
retention requirement and other mortgage market and securitization
reforms, we reviewed other provisions in the Dodd-Frank Act intended
to improve the securitization process and information on proposed
changes to the federal government's role in housing finance.
Because the risk retention regulations were being developed during the
course of our audit work, we interviewed the key private sector
mortgage and securitization industry stakeholders mentioned previously
as well as representatives from two credit rating agencies to obtain
views on the potential impact of a risk retention requirement. We
obtained their views on how regulatory decisions regarding the form
and coverage of the requirement may affect the availability and cost
of mortgage credit for borrowers and the viability of a private-label
residential mortgage-backed securities (RMBS) market. Additionally, we
interviewed officials from the previously cited federal agencies and
the Securities and Exchange Commission (SEC) and reviewed their
research, testimonies, and other public statements on risk retention.
We also reviewed comment letters, testimonies, and published papers
from these stakeholders that documented their views.
To illustrate the potential impact of regulatory decisions regarding
the coverage of the risk retention requirement, we used the CoreLogic
data to examine selected criteria (loan-to-value (LTV) ratio and DTI
ratio) being considered by regulators as part of the qualified
residential mortgage (QRM) rulemaking. We included only conventional
mortgages in the analysis because mortgages that are insured or
guaranteed by the federal government are exempt from the risk
retention requirement. We analyzed the data to describe the
proportions of mortgages that may have met more restrictive and less
restrictive versions of these criteria in 2006 (a period of relatively
lax underwriting standards) and 2010 (a period of relatively stringent
underwriting standards).[Footnote 96] For the LTV analysis, we used 80
percent as the more restrictive criterion (based on proposed QRM rules
for purchase mortgages) and 90 percent as the less restrictive
criterion. We used the CoreLogic variable for LTV ratio, which does
not take any subordinate liens into account. We did not use the
variable for combined LTV ratio, which does take subordinate liens
into account, because it was not reliable. As a result, the
percentages we report are likely somewhat higher than they would have
been if we had been able to use combined LTV ratios. For the DTI
analysis, we used 36 percent as the more restrictive criterion (based
on proposed QRM rules) and 41 percent as the less restrictive
criterion. We limited the analysis of DTI ratios to prime and near-
prime mortgages because the CoreLogic database did not contain DTI
ratios for subprime mortgages. As in the DTI analysis in the previous
section of the report, we used reported DTI ratios in the CoreLogic
database, which may understate debt obligations or overstate income in
some cases. As a result, the proportions of mortgages we show as
meeting the different DTI criteria are likely somewhat higher than
they would have been if all of the DTI ratios had been calculated in a
uniform and accurate manner. To provide additional perspective on the
potential coverage of the risk retention requirement, we reviewed an
analysis by FHFA, which examined the proportion of mortgages purchased
by the enterprises that would have met the proposed QRM criteria,
including those for LTV and DTI ratios.
To assess the financial impact of the risk retention requirement on
lenders and securitizers, we reviewed relevant accounting standards
and federal risk-based regulatory capital requirements that may
interact with risk retention. In particular, we reviewed financial
accounting statement (FAS) 166 (which addresses whether
securitizations and other transfers of financial assets are treated as
sales or financings), FAS 167 (which requires securitizers or lenders
with a controlling financial interest in an SPE to "consolidate" the
securitized assets on their balance sheets), and regulatory capital
standards based on the Basel accords. We also reviewed provisions in
the Dodd-Frank Act and the proposed risk retention rules that applied
to lenders specifically and interviewed industry stakeholders about
the potential impact of a risk retention requirement on different
types and sizes of mortgage lenders. To illustrate the potential
capital impacts of different forms of risk retention, we developed a
hypothetical securitization based on research and industry information
about the size and structure of RMBS. We used regulatory capital risk
weights used by federal banking regulators to calculate the capital
charges for horizontal and vertical risk retention to estimate the
total amount of regulatory capital that a securitizer would have to
hold for each option. Because this example is meant to be
illustrative, we did not apply all regulatory capital or accounting
standards that could influence the capital impacts of the risk
retention requirement, including the FAS 166 and 167 accounting
statements.
To assess the advantages and disadvantages of a uniform 5 percent risk
retention requirement, we reviewed available information on past risk
retention practices. We also interviewed industry stakeholders about
these practices and information that should be considered in comparing
the merits of a uniform and a nonuniform requirement. Additionally, we
interviewed federal rulemakers and mortgage industry stakeholders
(including representatives from financial services companies and
mortgage and securities analysts) about the development,
implementation, and enforcement of both a uniform and a nonuniform
requirement.
Assessment of Housing Counseling and HOEPA Provisions:
To describe the potential effects of consumer protection provisions in
the Dodd-Frank Act for housing counseling and high-cost HOEPA loans,
we reviewed relevant statutory provisions and regulations. We also
reviewed information from HUD regarding its current housing counseling
assistance program, including data on the counseling services provided
by HUD-approved counseling agencies from 2006 through 2010. We
identified and reviewed empirical research on the impact of
foreclosure mitigation and prepurchase housing counseling, HUD
reports, and relevant academic and industry literature about housing
counseling research and policy. We also interviewed officials from
HUD's Office of Single-Family Housing and Office of Policy Development
and Research, and officials from organizations currently conducting
housing counseling research, including Fannie Mae, the Federal Reserve
Bank of Philadelphia, and the Urban Institute. With respect to HOPEA,
we compared the Dodd-Frank Act's new requirements for high-cost loans
to previous statutory requirements and examined available research on
the number of loans originated from 2004 through 2009 that were
covered by HOEPA. We interviewed a wide range of mortgage and
counseling industry stakeholders, including consumer groups, lenders,
academic researchers, and housing counseling intermediaries
(organizations that channel HUD counseling funds to local, affiliated
counseling agencies) about the Dodd-Frank Act's counseling and HOEPA
provisions.
We conducted this performance audit from August 2010 to July 2011 in
accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audit to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objectives. We believe
that the evidence obtained provides a reasonable basis for our
findings and conclusions based on our audit objectives.
[End of section]
Appendix II" Proportions of Mortgages Meeting Selected Qualified
Mortgage Criteria in Geographic Groupings Based on Demographic and
Housing Market Characteristics:
In contemplating the potential impact of QM criteria, one
consideration is the extent to which mortgages made to different
borrower groups and within different housing markets would have met
selected QM criteria. Using the CoreLogic database described in more
detail in appendix I, we examined the percentages of mortgages
originated within various geographic groupings that would have met
selected QM criteria from 2001 through 2010 and compared them with the
corresponding percentages for all borrowers.[Footnote 97] We applied
each criterion separately. We looked at ZIP codes grouped by race,
ethnicity, and income level to examine the proportions of mortgages in
each grouping that likely would have met the four QM criteria for
which the CoreLogic database had relevant information (mortgage does
not have a negative amortization feature, mortgage term does not
exceed 30 years, mortgage does not include balloon payments, and
mortgage complies with regulations relating to DTI ratio).[Footnote
98] Using 2000 Census data, the most recent data available as of June
2011, we grouped ZIP codes associated with the mortgages in the
CoreLogic database into three categories: black or African-American
households made up 75 percent or more of the population, Hispanic or
Latino households made up 75 percent or more of the population, and
median incomes were less than 80 percent of the median income of the
associated metropolitan statistical area (low income).[Footnote 99] We
also grouped states into two categories: one containing states that
experienced rapid house price appreciation followed by rapid
depreciation (severe housing bubble states) during the 2000s, and all
other states. The severe housing bubble states were Arizona,
California, Florida, and Nevada. Except where noted below, for this
appendix we used a dataset that combined the mortgages in the prime,
near-prime, and government-insured category of the CoreLogic database
with the mortgages in the subprime category.
Our analysis of the QM criterion prohibiting negative amortization
features found that in ZIP codes with high proportions of black or
African-American households, the percentages of mortgage originations
that met the criterion were generally similar to the percentages for
all borrowers, with the exception of 2004 through 2006 when the
proportions were approximately 2 to 4 percentage points higher (see
table 2). In ZIP codes with high proportions of Hispanic or Latino
households, the percentages of mortgage originations that met the
criterion were similar to those for all borrowers, although they were
somewhat lower (about 3 percentage points) from 2005 through 2007. In
low-income ZIP codes, the proportions of mortgage originations that
met the criterion in all years were similar to the proportions for all
borrowers. In severe housing bubble states, the proportions of
mortgage originations from 2003 through 2007 that met the criterion
were about 2 to 7 percentage points lower than they were for all
borrowers. In all other states, the proportions of mortgage
originations that met the criterion were similar in all years to the
proportions for all borrowers, except from 2005 through 2007, when
they were from 2 to 3 percentage points higher.
Table 2: Percentage of Mortgages Meeting Qualified Mortgage Repayment
of Principal Requirement, by Demographic and Housing Market Grouping,
2001-2010:
Grouping: All borrowers;
Year: 2001: 99.0%;
Year: 2002: 99.8%;
Year: 2003: 99.8%;
Year: 2004: 96.0%;
Year: 2005: 93.4%;
Year: 2006: 94.2%;
Year: 2007: 96.4%;
Year: 2008: 99.4%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: ZIP codes with 75% or greater black or African-American
population;
Year: 2001: 99.7%;
Year: 2002: 99.5%;
Year: 2003: 99.3%;
Year: 2004: 98.1%;
Year: 2005: 97.1%;
Year: 2006: 96.7%;
Year: 2007: 96.7%;
Year: 2008: 99.1%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: ZIP codes with 75% or greater Hispanic or Latino population;
Year: 2001: 99.0%;
Year: 2002: 98.6%;
Year: 2003: 98.9%;
Year: 2004: 95.5%;
Year: 2005: 90.1%;
Year: 2006: 89.8%;
Year: 2007: 92.3%;
Year: 2008: 98.5%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: Low-income ZIP codes;
Year: 2001: 99.0%;
Year: 2002: 98.6%;
Year: 2003: 98.5%;
Year: 2004: 95.9%;
Year: 2005: 92.9%;
Year: 2006: 93.4%;
Year: 2007: 95.2%;
Year: 2008: 99.1%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: Severe housing bubble states;
Year: 2001: 97.5%;
Year: 2002: 97.2%;
Year: 2003: 97.4%;
Year: 2004: 92.3%;
Year: 2005: 86.5%;
Year: 2006: 87.3%;
Year: 2007: 91.7%;
Year: 2008: 98.4%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: All other states;
Year: 2001: 99.6%;
Year: 2002: 99.4%;
Year: 2003: 99.4%;
Year: 2004: 97.7%;
Year: 2005: 96.7%;
Year: 2006: 97.3%;
Year: 2007: 98.2%;
Year: 2008: 99.7%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Source: GAO analysis of Census 2000 and CoreLogic data.
[End of table]
Our analysis of the QM criterion for amortization terms of 30 years or
less found that in ZIP codes with high proportions of black or African-
American households, the percentages of mortgage originations that met
the criterion were generally similar to those for all borrowers,
exception of 2006 through 2008, when the proportions were
approximately 2 to 4 percentage points lower (see table 3). In ZIP
codes with high proportions of Hispanic or Latino households, the
percentages of mortgage originations that met the criterion were
similar to those for all borrowers, although they were somewhat lower
(about 2 to 3 percentage points) from 2006 through 2008. In low-income
ZIP codes, the proportions of mortgage originations that met the
criterion were similar to those for all borrowers, with the exception
of 2006 and 2007, when the proportion was nearly 2 percentage points
lower. In severe housing bubble states, the proportions of mortgage
originations that met the criterion were similar to the proportions
for all borrowers, with the exception of 2006 and 2007, when it was 3
to 4 percentage points lower. In all other states, the proportions of
mortgage originations that met the criterion in all years were similar
to the proportions for all borrowers.
Table 3: Percentage of Mortgages Meeting Qualified Mortgage Criterion
for Loan Terms of 30 Years or Less by Demographic and Housing Market
Grouping, 2001-2010:
Grouping: All borrowers;
Year: 2001: 99.8%;
Year: 2002: 99.8%;
Year: 2003: 99.8%;
Year: 2004: 99.6%;
Year: 2005: 98.0%;
Year: 2006: 94.1%;
Year: 2007: 95.1%;
Year: 2008: 98.8%;
Year: 2009: 99.8%;
Year: 2010: 99.8%.
Grouping: ZIP codes with 75% or greater black or African-American
population;
Year: 2001: 99.4%;
Year: 2002: 99.4%;
Year: 2003: 99.6%;
Year: 2004: 99.5%;
Year: 2005: 97.6%;
Year: 2006: 91.7%;
Year: 2007: 91.1%;
Year: 2008: 97.2%;
Year: 2009: 99.5%;
Year: 2010: 99.9%.
Grouping: ZIP codes with 75% or greater Hispanic or Latino population;
Year: 2001: 99.8%;
Year: 2002: 99.8%;
Year: 2003: 99.8%;
Year: 2004: 99.7%;
Year: 2005: 96.9%;
Year: 2006: 90.9%;
Year: 2007: 92.6%;
Year: 2008: 97.2%;
Year: 2009: 99.8%;
Year: 2010: 99.9%.
Grouping: Low-income ZIP codes;
Year: 2001: 99.7%;
Year: 2002: 99.7%;
Year: 2003: 99.7%;
Year: 2004: 99.6%;
Year: 2005: 97.4%;
Year: 2006: 92.6%;
Year: 2007: 93.6%;
Year: 2008: 98.4%;
Year: 2009: 99.8%;
Year: 2010: 99.8%.
Grouping: States that experienced severe housing bubbles;
Year: 2001: 99.8%;
Year: 2002: 99.8%;
Year: 2003: 99.8%;
Year: 2004: 99.4%;
Year: 2005: 96.5%;
Year: 2006: 90.2%;
Year: 2007: 92.3%;
Year: 2008: 98.3%;
Year: 2009: 99.9%;
Year: 2010: 99.9%.
Grouping: All other states;
Year: 2001: 99.8%;
Year: 2002: 99.8%;
Year: 2003: 99.8%;
Year: 2004: 99.6%;
Year: 2005: 98.6%;
Year: 2006: 95.8%;
Year: 2007: 96.1%;
Year: 2008: 99.0%;
Year: 2009: 99.8%;
Year: 2010: 99.8%.
Source: GAO analysis of Census 2000 and CoreLogic data.
[End of table]
Our analysis of the QM criterion restricting balloon payments found
that in ZIP codes with high proportions of black or African-American
households or Hispanic or Latino households, the percentages of
mortgage originations that met the criterion were generally similar to
those for all borrowers, with the exception of 2006, when the
proportions were nearly 2 percentage points lower for both ZIP code
groupings (see table 4). In low-income ZIP codes, severe housing
bubble states, and all other states, the proportions of mortgage
originations that met the criterion in all years were similar to the
proportions for all borrowers.
Table 4: Percentage of Mortgages Meeting Qualified Mortgage Criterion
Restricting Balloon Payments by Demographic and Housing Market
Grouping, 2001-2010:
Grouping: All borrowers;
Year: 2001: 99.2%;
Year: 2002: 99.2%;
Year: 2003: 99.1%;
Year: 2004: 99.4%;
Year: 2005: 99.1%;
Year: 2006: 97.7%;
Year: 2007: 98.9%;
Year: 2008: 100.0%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: ZIP codes with 75% or greater black or African-American
population;
Year: 2001: 97.8%;
Year: 2002: 98.9%;
Year: 2003: 99.4%;
Year: 2004: 99.3%;
Year: 2005: 98.4%;
Year: 2006: 96.0%;
Year: 2007: 97.5%;
Year: 2008: 100.0%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: ZIP codes with 75% or greater Hispanic or Latino population;
Year: 2001: 99.6%;
Year: 2002: 99.7%;
Year: 2003: 99.8%;
Year: 2004: 99.6%;
Year: 2005: 98.5%;
Year: 2006: 95.8%;
Year: 2007: 98.2%;
Year: 2008: 100.0%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: Low-income ZIP codes;
Year: 2001: 98.9%;
Year: 2002: 99.3%;
Year: 2003: 99.3%;
Year: 2004: 99.4%;
Year: 2005: 98.7%;
Year: 2006: 97.0%;
Year: 2007: 98.4%;
Year: 2008: 100.0%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: Severe housing bubble states bubbles;
Year: 2001: 99.5%;
Year: 2002: 99.4%;
Year: 2003: 99.3%;
Year: 2004: 99.6%;
Year: 2005: 98.9%;
Year: 2006: 96.7%;
Year: 2007: 98.4%;
Year: 2008: 100.0%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Grouping: All other states;
Year: 2001: 99.1%;
Year: 2002: 99.1%;
Year: 2003: 99.0%;
Year: 2004: 99.3%;
Year: 2005: 99.2%;
Year: 2006: 98.2%;
Year: 2007: 99.1%;
Year: 2008: 100.0%;
Year: 2009: 100.0%;
Year: 2010: 100.0%.
Source: GAO analysis of Census 2000 and CoreLogic data.
[End of table]
Due to limitations in the CoreLogic database, our examination of the
criterion for DTI ratio was restricted to mortgages in the prime, near-
prime, and government-insured category for 2003 through 2010. Using a
hypothetical standard of 41 percent or less for DTI ratio, we found
that in ZIP codes with high proportions of black or African-American
households, the percentages of mortgage originations that met the
criterion were generally lower than those for all borrowers, ranging
from 1 percentage point less in 2004 and 2006 to 10 percentage points
less in 2009 (see table 5). In ZIP codes with high proportions of
Hispanic or Latino households, the proportions of mortgage
originations that met the criterion were also generally lower than the
proportions for all borrowers, ranging from about 3 percentage points
less in 2006 and 2007 to 15 percentage points less in 2010. In low-
income ZIP codes, the proportions of mortgage originations that met
the criterion were generally lower than the proportions for all
borrowers, ranging from 2 percentage points less in 2004 to nearly 5
percentage points less in 2010. In severe housing bubble states, the
proportions of mortgage originations that met the criterion were
generally lower than the proportions for all borrowers, ranging from
less than 1 percentage point lower in 2004 and 2005 to 5 percentage
points lower in 2010. In all other states, the proportions of mortgage
originations that met the criterion in all years were similar to those
for all borrowers, except in 2010, when the proportion was 2
percentage points higher.
Table 5: Percentage of Prime, Near-Prime, and Government-Insured
Mortgages Meeting a Hypothetical Qualified Mortgage Criterion for DTI
Ratio of 41 Percent or Less by Demographic and Housing Market
Grouping, 2003-2010:
Grouping: All borrowers;
Year: 2003: 75.2%;
Year: 2004: 71.9%;
Year: 2005: 65.5%;
Year: 2006: 62.1%;
Year: 2007: 58.3%;
Year: 2008: 58.5%;
Year: 2009: 64.9%;
Year: 2010: 64.8%.
Grouping: ZIP codes with 75% or greater black or African-American
population;
Year: 2003: 70.6%;
Year: 2004: 70.6%;
Year: 2005: 63.2%;
Year: 2006: 60.8%;
Year: 2007: 56.8%;
Year: 2008: 51.9%;
Year: 2009: 55.1%;
Year: 2010: 57.2%.
Grouping: ZIP codes with 75% or greater Hispanic or Latino population;
Year: 2003: 69.6%;
Year: 2004: 65.4%;
Year: 2005: 61.2%;
Year: 2006: 59.2%;
Year: 2007: 55.5%;
Year: 2008: 49.7%;
Year: 2009: 50.5%;
Year: 2010: 49.8%.
Grouping: Low-income ZIP codes;
Year: 2003: 72.6%;
Year: 2004: 69.9%;
Year: 2005: 65.8%;
Year: 2006: 62.2%;
Year: 2007: 57.9%;
Year: 2008: 55.9%;
Year: 2009: 61.3%;
Year: 2010: 60.2%.
Grouping: Severe housing bubble states;
Year: 2003: 74.9%;
Year: 2004: 70.9%;
Year: 2005: 64.4%;
Year: 2006: 60.4%;
Year: 2007: 56.2%;
Year: 2008: 54.9%;
Year: 2009: 60.7%;
Year: 2010: 59.5%.
Grouping: All other states;
Year: 2003: 75.3%;
Year: 2004: 72.3%;
Year: 2005: 66.1%;
Year: 2006: 62.9%;
Year: 2007: 59.2%;
Year: 2008: 59.7%;
Year: 2009: 66.3%;
Year: 2010: 66.6%.
Source: GAO analysis of Census 2000 and CoreLogic data.
Note: This analysis only includes mortgages in the prime, near-prime,
and government-insured category of the CoreLogic database. Because the
CoreLogic database did not contain sufficient information on DTI for
prime, near-prime, and government-insured mortgages originated in 2001
and 2002, we only present data for 2003 through 2010. For this latter
period, about 53 percent of the mortgages did not have information on
the DTI ratio. We concluded that those mortgages were likely not
systematically different from mortgages with DTI information based on
a comparison of the average borrower credit scores associated with
both groups of mortgages, which showed little difference.
[End of table]
[End of section]
Appendix III: Comments from the National Credit Union Association:
National Credit Union Administration:
Office of the Chairman:
1775 Duke Street:
Alexandria, VA 22314-3428:
703-518-6300:
Via E-Mail:
July 11, 2011:
Mr. William B. Shear:
Director:
Financial Markets and Community Investment:
U.S. Government Accountability Office:
441 G Street, NW:
Washington, DC 20548:
Dear Mr. Shear:
We reviewed the draft report "Potential Impacts of Provisions in the
Dodd-Frank Act on Homebuyers and the Mortgage Market." The Dodd-Frank
Wall Street Reform and Consumer Protection Act (the Act) directed GAO
to assess the effect of mortgage-related provisions on the
availability and affordability of mortgage credit. As noted in the
report, federal agencies are still developing the regulations for some
of these provisions, with rulemaking agencies accepting public
comments on proposed risk retention regulations through August 1,
2011. Therefore, our comments on this draft report will be general,
since the impact of the Act largely depends upon other key regulatory
decisions which have not been made.
According to the Act, a lender is presumed to have satisfied the
ability-to-repay requirement when it originates a "qualified mortgage"
(QM). The report contains thorough analysis on five of the nine QM
criteria specified in the Act for which sufficient data were
available. The five criteria were (1) payment of loan principal, (2)
length of the mortgage term, (3) scheduled lump-sum payments, (4)
documentation of borrower resources, and (5) borrower debt burdens.
The report generally found that for each year from 2001 to 2010, most
mortgages would likely have met the individual criteria for these five
elements prescribed in the Act. We also note the trends provided in
the report show mortgage originators improved their underwriting
standards in 20092010 based on the adverse loss trends, which address
many of the criteria studied by GAO.
One of NCUA' s key concerns is the lack of a mechanism for non-QM
loans to receive QM status through seasoning. While the report found
that most loans originated for the time period studied met the
criteria when analyzed individually, the impact of simultaneously
applying multiple criteria creates a much narrower population for QM
loans. Moreover, borrower debt burdens (the fifth element) appear to
be the most difficult element for borrowers to achieve. This may be
especially true for those of modest means. Creating a mechanism for
lenders to originate non-QM loans that could become QM loans after a
specific period of time expires may help achieve the goal of
protecting borrowers from unsustainable mortgage products while
maintaining broad access to mortgage credit.
NCUA appreciates the detailed analysis performed by GAO. Thank comment
you for the opportunity to on this draft report.
Sincerely,
Signed by:
Debbie Matz:
Chairman:
[End of section]
Appendix IV: GAO Contact and Staff Acknowledgments:
GAO Contact:
William B. Shear, (202) 512-8678 or shearw@gao.gov:
Staff Acknowledgments:
In addition to the individual named above, Steve Westley (Assistant
Director), Meghana Acharya, Serena Agoro-Menyang, William Bates,
Stephen Brown, Emily Chalmers, Matthew McDonald, John McGrail, Timothy
Mooney, Lisa Moore, Alise Nacson, and Jim Vitarello made key
contributions to this report.
[End of section]
Bibliography:
Selected Bibliography of Research on Homeownership Counseling:
AARP Public Policy Institute. Homeownership Education and Counseling.
PPI Issues Brief. Washington, D.C.: 2003. Available at
http://www.aarp.org/money/budgeting-saving/info-
2003/homeownership_education_counseling.html:
Agarwal, Sumit, Gene Amromin, Itzhak Ben-David, Souphala
Chomsisengphet, and Douglas D. Evanoff. Do Financial Counseling
Mandates Improve Mortgage Choice and Performance? Evidence from a
Legislative Experiment. Federal Reserve Board of Chicago Working Paper
No. 2009-7, October 2009.
------. "Learning to Cope: Voluntary Financial Education and Loan
Performance During a Housing Crisis." American Economic Review, vol.
100, no. 2 (2010).
Carswell, Andrew. "Does Housing Counseling Change Consumer Financial
Behaviors? Evidence from Philadelphia." Journal of Family and Economic
Issues, vol. 30, no. 4 (2009).
Collins, J. Michael. "Exploring the Design of Financial Counseling for
Mortgage Borrowers in Default." Journal of Family and Economic Issues,
vol. 28, no. 2 (2007).
Collins, J. Michael., and Collin O'Rourke. Homeownership Education and
Counseling: Do We Know What Works? Policylab Consulting for Research
Institute for Housing America (Special Report) and Mortgage Bankers
Association. Washington, D.C.: April 2011.
Ding, Lei, Roberto G. Quercia, and Janneke Ratcliffe. "Post-Purchase
Counseling and Default Resolutions Among Low-and Moderate-Income
Borrowers." Journal of Real Estate Research, vol. 30, no. 3 (2008).
Hartarska, Valentina, and Claudio Gonzalez-Vega. "Credit Counseling
and Mortgage Termination by Low-Income Households." Journal of Real
Estate Finance and Economics, vol. 30, no. 3 (2005).
------. "Evidence on the Effect of Credit Counseling on Mortgage Loan
Default by Low-Income Households." Journal of Housing Economics, vol.
15, no. 1 (2006).
Herbert, Christopher E., Jennifer Turnham, and Christopher N. Rodger.
The State of the Housing Counseling Industry. Abt Associates for the
U.S. Department of Housing and Urban Development. Washington, D.C.:
September 2008.
Hirad, Abdighani, and Peter M. Zorn. A Little Knowledge is a Good
Thing: Empirical Evidence of the Effectiveness of Pre-Purchase
Homeownership Counseling. Joint Center for Housing Studies of Harvard
University, Low Income Homeownership Working Paper Series 01.4.
Cambridge, Mass.: August 2001.
Hornburg, Steven P. Strengthening the Case for Homeownership
Counseling: Moving Beyond "A Little Bit of Knowledge." Joint Center
for Housing Studies of Harvard University Report W04-12. Cambridge
Mass.: August 2004.
Jones, Katie. "Housing Counseling: Background and Federal Role."
Congressional Research Service Report to Congress R41351. Washington,
D.C.: July 2010.
Mallach, Alan. Homeownership Education and Counseling: Issues in
Research and Definition. Community Affairs Discussion Paper 00-01,
Federal Reserve Bank of Philadelphia, 2001.
Mayer, Neil S., Peter A. Tatian, Kenneth, Temkin, and Charles A.
Calhoun. National Foreclosure Mitigation Counseling Program
Evaluation: Preliminary Analysis of Program Efforts, September 2010
Update. Prepared for NeighborWorks America. Washington, D.C.: Urban
Institute, 2010.
McCarthy, G. and Roberto Quercia. Bridging the Gap between Supply and
Demand: The Evolution of the Homeownership Education and Counseling
Industry. Research Institute for Housing America Report No. 00-01.
Washington, D.C.: May 2000.
Quercia, Roberto, and Spencer M. Cowan. "The Impacts of Community-
based Foreclosure Prevention Programs." Housing Studies, vol. 23, no.
3 (2008).
Quercia, Roberto, and Jonathan S. Spader. "Does Homeownership
Counseling Affect the Prepayment and Default Behavior of Affordable
Mortgage Borrowers?" Journal of Policy Analysis and Management, vol.
27, no. 2 (2008).
Quercia, Roberto and Susan M. Wachter. "Homeownership Counseling
Performance: How Can It Be Measured?" Housing Policy Debate, vol. 7,
no. 1 (1996).
[End of section]
Footnotes:
[1] Securitization allows lenders to sell loans from their portfolios,
transferring credit risk to investors, and use the proceeds to make
more loans.
[2] Pub. L. 111-203.
[3] We use the term "lender" to refer to what the Dodd-Frank Act calls
a mortgage "originator" or "creditor." A lender can also meet the Dodd-
Frank Act's ability-to-repay requirement by originating a mortgage
that satisfies eight underwriting factors which emphasize
consideration of borrower characteristics such as employment and
current or expected income. We focus on the QM criteria, which
emphasize mortgage features, because data available to us primarily
contained information on mortgage characteristics.
[4] A balloon payment is a large lump-sum payment scheduled at the end
of a series of smaller periodic payments. Section 1412 of the Dodd-
Frank Act defines a balloon payment as a scheduled payment that is
more than twice as large as the average of earlier scheduled payments.
[5] A point is a loan charge, usually paid at loan closing, expressed
as a percentage of the loan amount (1 point is 1 percent of the loan
balance).
[6] A reverse mortgage is a loan that converts the borrower's home
equity into payments from a lender and typically does not require any
repayments as long as the borrower continues to live in the home.
[7] The Dodd-Frank Act defines a securitizer as an issuer of an asset-
backed security or a person who organizes and initiates an asset-
backed securities transaction by selling or transferring assets,
either directly or indirectly, including through an affiliate, to the
issuer.
[8] HOEPA, enacted in 1994, regulates and restricts the terms and
characteristics of certain kinds of high-cost mortgages.
[9] The enterprises purchase mortgages that meet specified
underwriting criteria from approved lenders. Most of the mortgages are
made to prime borrowers with strong credit histories. The enterprises
bundle the mortgages into securities and guarantee the timely payment
of principal and interest to investors in the securities. On September
6, 2008, the enterprises were placed under federal conservatorship
because of concern that their deteriorating financial condition and
potential default on $5.4 trillion in outstanding financial
obligations threatened the stability of financial markets.
[10] As presented in appendix II, we used data from the Census Bureau
and information on state-level house price trends to examine the
proportions of mortgages within different geographic groupings (based
on demographic and housing market characteristics) that likely would
have met four of these criteria.
[11] The LTV ratio is the loan amount divided by the value of the home
at mortgage origination. The DTI ratio represents the percentage of a
borrower's income that goes toward all recurring debt payments,
including mortgage payments.
[12] GAO, Alternative Mortgage Products: Impact on Defaults Remains
Unclear, but Disclosure of Risks to Borrowers Could Be Improved,
[hyperlink, http://www.gao.gov/products/GAO-06-1021] (Washington,
D.C.: Sept. 19, 2006).
[13] Interest rate risk is the risk that an increase in interest rates
will reduce the value of a fixed-rate loan.
[14] REITs are companies that own income-producing real estate and in
some cases engage in financing real estate. To qualify as a REIT, a
company must have most of its assets and income tied to real estate
investment and must distribute at least 90 percent of its taxable
income to shareholders annually in the form of dividends.
[15] All of the market share figures in this paragraph are calculated
based on data from Inside Mortgage Finance, are expressed in terms of
dollar volume (rather than number of loans), and exclude home equity
loans.
[16] For additional information about the characteristics and
performance of subprime and near-prime mortgages, see GAO, Nonprime
Mortgages: Analysis of Loan Performance, Factors Associated with
Defaults, and Data Sources, [hyperlink,
http://www.gao.gov/products/GAO-10-805] (Washington, D.C.: Aug. 24,
2010). Lenders were often required by contract to repurchase mortgages
for which the borrower failed to make a payment in the first 3 months
after origination. Some lenders ended up in bankruptcy due to their
inability to satisfy these repurchase requests.
[17] FHA insures lenders against losses from borrower defaults on
mortgages that meet FHA criteria. FHA historically has served
borrowers who would have difficulty obtaining prime mortgages but in
recent years has increasingly served borrowers with stronger credit
histories.
[18] TILA, as amended, is codified at 15 U.S.C. §§ 1601 - 1666j.
[19] APR is a measure of credit cost to the borrower that takes
account of the interest rate, points, and certain lender charges.
[20] The FHA, Department of Veterans Affairs (VA), and the Department
of Agriculture's Rural Housing Service, in consultation with the
Federal Reserve Board, are required to develop separate QM criteria
for their loan programs through regulations. Additionally, rulemaking
authority for TILA is scheduled to transfer to CFPB on July 21, 2011.
Accordingly, the rulemaking for the QM provisions will be finalized by
CFPB rather than by the Federal Reserve Board.
[21] Dodd-Frank Act, sec. 941(b) (codified at 15 U.S.C. sec. 78o-11).
[22] The act also does not apply the exemption for government-insured
or -guaranteed mortgages to mortgages backed by Federal Home Loan
Banks, which form a system of regional cooperatives that support
housing finance through advances and mortgage programs, among other
activities.
[23] The federal banking agencies (with the exception of NCUA), FHFA,
HUD, and SEC must jointly issue regulations related to the risk
retention provisions of the Dodd-Frank Act as they pertain to
residential mortgages. For the purposes of the Dodd-Frank Act, the
federal banking agencies include FDIC, the Federal Reserve Board, and
OCC.
[24] See 76 Fed. Reg. 24090 (Apr. 29, 2011). Under the Dodd-Frank Act,
the risk retention requirement also applies to other asset classes.
[25] The proposed rules also contain an LTV ratio cap of 70 percent
for cash-out refinance mortgages (i.e., refinancing at a higher amount
than the loan balance to convert home equity into money for personal
use) and 75 percent for rate and term refinance mortgages (i.e.,
refinancing to change the interest rate or length of the mortgage with
no cash out).
[26] The proposed criteria for the QRM also specify a maximum level of
28 percent for the percentage of a borrower's income that goes toward
mortgage and other housing-related payments such as private mortgage
insurance, property taxes, and homeowner association fees.
[27] The proposed rules also request public comments on a possible
alternative approach to QRMs. This approach would allow QRMs to have
higher LTV and DTI ratios than those described above and take into
account mortgage insurance or other third-party credit enhancements.
Non-QRMs would be subject to stricter (e.g., less flexible or higher)
risk retention requirements than described in the main approach.
[28] Dodd-Frank Act, sec. 941(c). Board of Governors of the Federal
Reserve System, Report to the Congress on Risk Retention (October
2010), available at [hyperlink,
http://www.federalreserve.gov/newsevents/press/other/20101019a.htm].
[29] Dodd-Frank Act, sec. 946. Financial Stability Oversight Council,
Macroeconomic Effects of Risk Retention Requirements (January 2011),
available at [hyperlink, http://www.treasury.gov/press-center/press-
releases/Pages/tg1027.aspx].
[30] 12 U.S.C. 1701x.
[31] Several other federal agencies also provide limited support or
funding for housing counseling, often for specific populations,
including the Departments of Defense (DOD) and the Treasury (Treasury)
and the VA. For example, Treasury provided funding for financial
education and counseling through the Financial Education and
Counseling Pilot Program, authorized pursuant to Section 1132 of the
Housing and Economic Recovery Act of 2008 (Pub. L. 110-289). Through
this program, Treasury awarded grants to nine eligible organizations,
including HUD-approved housing counseling agencies. Grant recipients
are required to identify successful methods of financial education and
counseling services that result in positive behavioral change for
financial empowerment and to establish program models for
organizations to deliver effective financial education and counseling
services to prospective homebuyers. Congress appropriated $2.0 million
for the program in fiscal year 2009 and $4.15 million in fiscal year
2010. P.L. 110-289 also directed DOD to set up a foreclosure
counseling program for servicemembers returning from active duty
abroad. Similarly, the VA employs loan counselors through its nine
Regional Loan Centers to help veterans who are facing foreclosure or
other financial problems. The VA's counselors assist veterans whether
or not their mortgages are guaranteed by the VA. The VA also relies on
HUD's housing counseling program for prepurchase housing counseling.
[32] The Dodd-Frank Act also requires the CFPB to establish an Office
of Financial Education to improve financial literacy through
activities that include financial counseling. The duties this office
is charged with are in some ways similar to those of the separate
Office of Financial Education and Financial Access within Treasury. We
have previously reported on the need for federal entities to
coordinate their roles and activities to avoid unnecessary overlap and
duplication. See GAO, Financial Literacy: The Federal Government's
Role in Empowering Americans to Make Sound Financial Choices, GAO-11-
504T (Washington, D.C.: Apr. 12, 2011).
[33] As previously noted, the CoreLogic database we used for this
analysis covers a broad cross-section of the mortgage market. However,
because of limitations in the coverage and completeness of the data,
our analysis may not be fully representative of the mortgage market as
whole.
[34] As discussed later in this section, we relied on other data
sources to examine the proportion of mortgages that would have met the
QM criterion for full documentation.
[35] In terms of dollar volume, mortgages in the first category
accounted for roughly 90 percent of mortgage originations from 2001
through 2003. This proportion declined to approximately 77 percent in
2005 and 2006, then rose from about 91 percent in 2007 to almost 100
percent in 2009 and 2010. For ease of presentation, we refer to
mortgages with government insurance or guarantees as government-
insured mortgages in the remainder of this report.
[36] As previously noted, rulemakers may extend loan terms beyond 30
years for certain locales, such as high-cost areas.
[37] As previously noted, the Dodd-Frank Act defines a balloon payment
as a scheduled payment that is more than twice as large as the average
of earlier scheduled payments. According to proposed QM rules issued
in April 2011, some balloon mortgages can be considered to meet the QM
criteria, such as balloon mortgages with terms of 5 or more years made
by creditors that operate in predominantly rural or underserved areas.
[38] The proposed rules describe two alternative sets of QM criteria:
one that does not include DTI ratio, and one that requires
consideration of DTI ratio.
[39] Although most of the mortgages in the CoreLogic dataset had
missing values for the interest-only indicator, the data suggest that
the interest-only feature was especially prominent among prime and
near-prime hybrid ARMs, a product type that became more common in the
mid-2000s. An FHFA analysis covering the period from 2006 through 2010
indicates that interest-only mortgages accounted for about 15 percent
of Fannie Mae's mortgage purchases in 2006 but that this percentage
declined to 0 to 1 percent in 2009 and 2010. The analysis showed that
for Freddie Mac, the percentage of interest-only mortgages peaked in
2007 at 22 percent before falling to 0 percent in 2009 and 2010. See
FHFA, Conservator's Report on the Enterprises' Financial Performance,
Fourth Quarter 2010, available at [hyperlink,
http://www.fhfa.gov/Default.aspx?Page=172].
[40] [hyperlink, http://www.gao.gov/products/GAO-10-805].
[41] Alt-A generally refers to a mortgage loan originated under a
lender's program offering reduced or alternative documentation than
that required for a full documentation mortgage loan but may also
include other alternative product features. Both enterprises classify
mortgages as Alt-A if the lenders delivering the mortgages classify
them as Alt-A based on documentation or other product features.
[42] According to OCC officials, mortgage originators may have
calculated DTI ratios differently depending on their definitions of
debt and income. The officials also indicated that in some cases,
mortgage originators used mortgage payments that did not fully
amortize the mortgage to determine a borrower's total recurring debt
payments and that borrower income was sometimes overstated for
mortgages without full documentation of income. As a result, the
proportions of mortgages we show as meeting the criterion are likely
somewhat higher than they would have been if all of the DTI ratios had
been calculated in a uniform and accurate manner.
[43] [hyperlink, http://www.gao.gov/products/GAO-09-848R].
[44] By one estimate, about three-quarters of subprime borrowers lack
escrow accounts, which are bank accounts set up by lenders into which
monthly payments from the borrower are deposited for property taxes.
[45] In March 2011, the Federal Reserve Board issued a proposal to
implement Sections 1461 and 1462 of the Dodd-Frank Act, which provide
certain escrow requirements for higher-priced loans.
[46] [hyperlink, http://www.gao.gov/products/GAO-10-805].
[47] Robert G. Quercia, M. Stegman, and W. Davis, "The Impact of
Predatory Loan Terms on Subprime Foreclosures: The Special Case of
Prepayment Penalties and Balloon Payments," Housing Policy Debate,
vol. 18, no. 2 (2007).
[48] [hyperlink, http://www.gao.gov/products/GAO-08-78R].
[49] ARMs with initial fixed-rate periods of less than 5 years became
a common product type during the mid-2000s. They accounted for over
one-third of the first-lien mortgages in our CoreLogic data sample
that were originated in 2005.
[50] As previously noted, the Federal Reserve Board's proposed QM
rules do not provide a specific DTI ratio. Rather, they describe two
alternative sets of QM criteria: one that does not include DTI ratio,
and one that requires consideration of DTI ratio.
[51] The Dodd-Frank Act allows streamlined refinancing without
verification of borrower income and assets in cases where the borrower
is refinancing from a "nonstandard mortgage" into a "standard
mortgage" with the same lender. According to the Federal Reserve
Board's proposed implementing rules, a nonstandard mortgage is (1) an
ARM with an introductory fixed rate for a period of years, (2) an
interest-only loan, or (3) a negative amortization loan. A standard
mortgage is one that does not have a negative amortization, interest-
only, or balloon payment feature and that limits the points and fees.
The consumer's monthly payment must be reduced through the refinancing
and the consumer must not have had more than one payment more than 30
days late on the existing nonstandard mortgage during the 24 months
preceding the application for the standard mortgage.
[52] For this analysis, we excluded government-insured mortgages,
which typically have low down payments (high LTVs), because the Dodd-
Frank Act exempts them from the risk retention requirement. We also
applied the 80 and 90 percent LTV thresholds to all mortgage types
(e.g., purchase mortgages and refinance mortgages).
[53] As in the previous section of this report, this analysis used
reported DTI ratios, which may understate debt obligations or
overstate income in some cases. As a result, the proportions of
mortgages we show as meeting the different criteria are likely
somewhat higher than they would have been if all of the DTI ratios had
been calculated in a uniform and accurate manner.
[54] The CoreLogic data did not contain information on DTI ratios for
subprime mortgages.
[55] FHFA's analysis used the separate LTV thresholds proposed for
purchase mortgages (80 percent) and different types of refinance
mortgages (75 percent for no-cash-out refinances and 70 percent for
cash-out refinances). In addition, FHFA used the proposed thresholds
for both the DTI ratio (36 percent) and the percentage of a borrower's
income that goes toward mortgage payments (28 percent). They also used
a credit score threshold as a proxy for proposed criteria concerning
borrower delinquency history. For a discussion of these and other
proposed QRM criteria FHFA applied, see FHFA, Mortgage Market Note 11-
02 (Apr. 11, 2011), available at [hyperlink,
http://www.fhfa.gov/Default.aspx?Page=77].
[56] Although there is no uniform definition of default across the
lending industry, 90-day delinquency rates (i.e., the percentage of
mortgages for which the borrower is least 90 days late on the
payments) are sometimes used as an indicator of mortgage default.
[57] Estimates of the impact of a risk retention requirement on
borrower interest rates depend on a number of assumptions, including
the form of risk retention, capital costs, and the liquidity of RMBS
backed by non-QRMs. FDIC officials told us that their estimate
pertains to both horizontal and vertical risk retention, assumes
capital costs are not affected by the accounting consolidation
scenario discussed in this section, and assumes a liquid market for
RMBS backed by non-QRMs.
[58] As previously noted, the Dodd-Frank Act also does not apply the
exemption for government-insured or -guaranteed mortgages to mortgages
backed by Federal Home Loan Banks.
[59] The other options are L-shaped (a hybrid of the horizontal and
vertical options), seller's interest (typically a shared interest with
all of the investors in a security backed by a pool of revolving
loans, such as credit cards), representative sample (a randomly
selected representative sample of assets that is equivalent, in all
material respects, to the securitized assets and is commonly used in
connection with securities backed by automobile loans), and an option
specifically designed for structures involving asset-backed commercial
paper. To help ensure that securitizers do not reduce or offset their
retained economic interest by monetizing "excess spread" (i.e., the
difference between the gross yield on a pool of securitized assets
minus the cost of financing those assets) generated over time,
rulemaking agencies also proposed requiring securitizers to place
these funds into a "premium capture cash reserve account." This
account would be in addition to the base risk retention requirement
and would be used to absorb the first losses.
[60] Federal banking and thrift regulators require banking
institutions to maintain a minimum amount of capital and generally
expect them to hold capital above these minimums, commensurate with
their risk exposure, to ensure they remain solvent in the event of
unexpected losses. These requirements were established under
international Basel Accord frameworks.
[61] Under regulatory capital requirements, all assets are assigned a
risk weight according to the credit risk of the obligor and the nature
of any qualifying collateral or guarantee, where relevant. These
requirements are broadly intended to assign higher risk weights to--
and require banks to hold more capital for--higher-risk assets.
[62] For example, with vertical risk retention, a securitizer would
have to hold capital against 5 percent of the $37.5 million equity
tranche (i.e., $37.5 million times 5 percent times a capital charge of
100 percent, which equals $1.875 million) instead of the entire $37.5
million equity tranche in the case of horizontal risk retention.
[63] For example, section 939A of the Dodd-Frank Act requires federal
agencies to remove references to credit ratings, changing how capital
against securitization exposures is calculated. Additionally, federal
banking regulators are currently contemplating changes to regulatory
capital rules, including changes related to the Basel III Accord.
[64] Consolidation is the process by which the financial statements of
a parent company are combined with those of its subsidiaries (in this
case the SPE), as if they were a single economic entity. A securitizer
would have a controlling financial interest in an SPE if it had (1)
the power to direct the activities of the SPE that most significantly
affected the SPE's economic performance, and (2) the obligation to
absorb the losses of, or the rights to receive benefits from, the SPE
that could potentially be significant to the SPE. FAS 166, Accounting
for Transfers of Financial Assets, an Amendment of FASB Statement No.
140, is codified within Accounting Standards Codification (ASC) Topic
860, and FAS 167, Amendments to FASB Interpretation No. 46(R) is
codified within ASC Topic 810.
[65] For example, assuming a 5 percent overall requirement, a
securitizer would only need to retain 3 percent if a lender retained
the remaining 2 percent.
[66] When the lender is also the securitizer, it would retain the full
amount of risk retention.
[67] In the case of banking institutions, risk-based regulatory
capital requirements would apply.
[68] According to Home Mortgage Disclosure Act data, independent
mortgage companies originated about 20 percent of all mortgages in
2009. Although precise figures are not available for small community
banks, community banks in general have originated about 15 to 20
percent of mortgages in recent years, according to a trade association
that represents these institutions.
[69] Dodd-Frank Act, sec. 945 (codified at 15 U.S.C. sec. 77g(d)). The
SEC issued regulations implementing this requirement in January 2011.
See 76 Fed. Reg. 4231 (Jan. 25, 2011).
[70] SEC issued implementing regulations for Section 943 in January
2011. See 76 Fed. Reg. 4489 (Jan. 26, 2011).
[71] Department of the Treasury and Department of Housing and Urban
Development, Reforming America's Housing Finance Market: A Report to
Congress (February 2011).
[72] FHA requires a minimum borrower contribution of 3.5 percent of
the sales price of the home.
[73] Dodd-Frank Act, sec. 1442 (codified at 42 U.S.C. sec. 3533(g)).
[74] The Program Support Division has staff in HUD headquarters and in
HUD's four homeownership centers located in Atlanta, Georgia; Denver,
Colorado; Philadelphia, Pennsylvania; and Santa Ana, California.
[75] Counseling agencies may have multiple service interactions with
the same client. Some of the HUD-approved counseling agencies that
report service interaction data do not receive HUD funds, and those
that do receive HUD grants also rely on other funding sources,
according to HUD officials. As a result, the service interaction data
do not represent just the counseling services provided with HUD funds.
[76] In fiscal year 2010, HUD was appropriated $88 million for housing
counseling assistance. The President's budget for fiscal year 2012
requests $88 million for HUD housing counseling assistance.
[77] The federal government funds homeownership counseling through a
number of programs and has provided targeted support for foreclosure
mitigation counseling in recent years. For example, Congress
appropriated $65 million in fiscal year 2011 to the National
Foreclosure Mitigation Counseling (NFMC) Program, which was designed
to rapidly expand the availability of foreclosure mitigation
counseling. NFMC is administered by NeighborWorks®, a government-
chartered, nonprofit corporation with a national network of affiliated
organizations. NeighborWorks® competitively distributes NFMC funds to
three types of authorized recipients: HUD-approved counseling
intermediaries (i.e., organizations that channel HUD counseling funds
to local, affiliated counseling agencies), state housing finance
agencies, and NeighborWorks® affiliates.
[78] Section 1013(d)(7) of the Dodd-Frank Act requires us to conduct a
study on financial literacy programs. As part of this work, we are
conducting a literature review of financial literacy programs.
[79] Experimental design involves random assignment of subjects to
treatment and control groups to isolate the impact of the treatment.
In the context of prepurchase homeownership counseling, one group of
prospective homebuyers would receive counseling (treatment group) and
the other would not (control group).
[80] Individuals who receive housing counseling, either on their own
or by enrolling in a research study, represent a "self-selected"
population. As noted, they may be systematically different than
individuals who do not seek counseling, and this potential bias makes
generalizing research results for the self-selected population
problematic.
[81] Neil S. Mayer, Peter A. Tatian, Kenneth, Temkin, and Charles A.
Calhoun, National Foreclosure Mitigation Counseling Program
Evaluation: Preliminary Analysis of Program Efforts, September 2010
Update, prepared for NeighborWorks America (Washington, D.C.: Urban
Institute, 2010). The study focused on the approximately 800,000
borrowers who received NFMC counseling from January 2008 through
December 2009 and a comparison sample of non-NFMC-counseled homeowners.
[82] Lei Ding, Roberto G. Quercia, and Janneke Ratcliffe, "Post-
purchase Counseling and Default Resolution among Low-and Moderate-
Income Borrowers," Journal of Real Estate Research, vol. 30, no. 3
(2008).
[83] Abdighani Hirad and Peter M. Zorn, A Little Knowledge is a Good
Thing: Empirical Evidence of the Effectiveness of Pre-Purchase
Homeownership Counseling, Joint Center for Housing Studies of Harvard
University, Low Income Homeownership Working Paper Series 01.4
(Cambridge, Mass.: August 2001).
[84] Hirad and Zorn, A Little Knowledge Is a Good Thing.
[85] Roberto Quercia and Jonathan S. Spader, "Does Homeownership
Counseling Affect the Prepayment and Default Behavior of Affordable
Mortgage Borrowers?" Journal of Policy Analysis and Management, vol.
27, no. 2 (2008).
[86] Sumit Agarwal, Gene Amromin, Itzhak Ben-David, Souphala
Chomsisengphet, and Douglas D. Evanoff, Do Financial Counseling
Mandates Improve Mortgage Choice and Performance? Evidence from a
Legislative Experiment, working paper 2009-07 (Federal Reserve Board
of Chicago, 2009).
[87] Christopher E. Herbert, Jennifer Turnham, and Christopher N.
Rodger, The State of the Housing Counseling Industry, Abt Associates
for the U.S. Department of Housing and Urban Development (Washington,
D.C.: September 2008).
[88] The purpose of the Home Affordable Modification Program is to
enable borrowers who meet eligibility requirements to avoid
foreclosure by modifying loans to a level that borrowers can afford
and sustain in the long-term.
[89] Assignee liability is intended to discourage secondary market
participants from purchasing loans that may have predatory features
and to provide an additional source of redress for victims of abusive
lenders.
[90] HMDA requires lending institutions to collect and publicly
disclose information about housing loans and applications for such
loans. HMDA data traditionally capture about 80 percent of the
mortgages funded each year and are one of the most comprehensive
sources of information on mortgage lending. Some high-cost loans are
extended by institutions not covered by HMDA, and some high-cost loans
made by HMDA-covered institutions are not required to be reported.
[91] The Dodd Frank Act also amended the APR trigger to be based upon
the average prime offer rate, to be published monthly by the Federal
Reserve Board, rather than the yield on Treasury securities having
comparable periods of maturity.
[92] The Dodd-Frank Act also expanded the definition of points and
fees to include all compensation paid by the consumer or creditor
directly or indirectly to the mortgage originator.
[93] [hyperlink, http://www.gao.gov/products/GAO-10-805].
[94] [hyperlink, http://www.gao.gov/products/GAO-09-741].
[95] [hyperlink, http://www.gao.gov/products/GAO-09-848R]. For that
report, we used CoreLogic's Asset-backed Securities database, which
contains information on securitized subprime and near-prime mortgages.
[96] The LTV ratio is the amount of the loan divided by the value of
the home at mortgage origination.
[97] We did not examine the reasons for differences among the various
groupings or the performance of the mortgages for each grouping as
part of our analysis. In a prior report, we examined statistical
associations between a number of loan and borrower characteristics--
including borrower race, ethnicity, and reported income--and the
probability of default. See GAO-10-805.
[98] For DTI ratio, we used the 41 percent figure that serves as a
guideline in underwriting FHA-insured mortgages.
[99] The groupings we examined are not mutually exclusive, but our
analysis did not allow us to assess the separate effects of borrower
income, race, and ethnicity.
[End of section]
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Phone:
The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s Web site,
[hyperlink, http://www.gao.gov/ordering.htm].
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional
information.
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]:
E-mail: fraudnet@gao.gov:
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548: