Securitization and Federal Regulation of Mortgages for Safety and Soundness
Congressional research reportOct 31, 2008
Ask Donna
What actually matters in this document.
Text
ȱ
£ȱȱȱȱȱ
ȱȱ¢ȱȱȱ
ȱǯȱ¢ȱ
ȱȱȱȱ
ȱřŗǰȱŘŖŖŞȱ
ȱȱȱ
ŝȬśŝŖŖȱ
ǯǯȱ
ŘŘŝŘŘȱ
ȱȱȱ
Prepared for Members and Committees of Congress
ȱ
£ȱȱȱȱȱȱȱ¢ȱȱȱ
¢ȱ
Rising defaults in the subprime mortgage market have drawn attention to the regulatory
framework for mortgage lending. Traditionally, banks subject to federal regulation have extended
loans to potential home buyers and kept the loans in their own portfolios. Alternatively, mortgage
lenders can sell their loans to the secondary market, where the loans are transformed into
mortgage backed securities (MBS), in a process called securitization. Securitization allows banks
and non-banks to offer mortgages without retaining a long-term interest in the loans themselves.
Non-bank lenders are often outside the safety and soundness regulation of federal bank
examiners, although they are still subject to the consumer protection mandates of the Truth in
Lending Act (TILA). The Federal Reserve issued new rules pursuant to TILA for all mortgage
loans in July 2008. Federal Reserve implements TILA through Regulation Z.
Guidances issued by the financial regulatory members of the Federal Financial Institutions
Examinations Council (FFIEC) help maintain prudent lending standards for covered institutions.
Securitization of loans originated by non-banks, however, allows some mortgage lending to
escape these guidances. The underwriting standards of Fannie Mae and Freddie Mac, regulated
for safety and soundness by the Federal Housing Finance Agency (FHFA), could still influence
underwriting standards of non-bank lenders. Caps on the loans they could purchase, and other
factors, however, had limited the influence of these government-sponsored-enterprises’ (GSE)
underwriting standards. These caps were substantially raised (up to $625,000 in some high-cost
areas) by H.R. 3221 / P.L. 110-289.
There is some evidence that the underwriting standards of non-banks that chose to securitize
loans outside of Fannie Mae and Freddie Mac became weaker as the housing boom progressed.
Indicators of excessive debt appear to have weakened more than indicators of borrower payment
history. Potential reforms of securitization and the non-bank lending channel are now under
consideration. This report will be updated as conditions change.
ȱȱȱ
£ȱȱȱȱȱȱȱ¢ȱȱȱ
ȱ
D
isruptions in the mortgage market have drawn attention to the potential for lenders to
engage in imprudent underwriting. During the housing boom of 2002-2005, many
borrowers may have become overextended because their loans, in hindsight, were not
sustainable. Although federal bank examiners are primarily concerned with the financial stability
of the system, one byproduct of their regulations may be to limit the chances that borrowers will
be offered imprudent loans by regulated institutions. Securitization, which transforms pools of
loans into marketable securities, may have contributed to looser underwriting standards, because
it creates a non-bank source of mortgage funds, which is outside federal bank examination.
A non-bank mortgage originator can open a line of credit to fund its lending, rather than accepting
deposits or raising its own capital. The originator then draws down the line of credit to make
loans. The originator sells the mortgages to the secondary market and uses the proceeds to pay
back the line of credit and extends more loans. Once in the secondary market, the mortgages can
be packaged together and held in passive trusts. The trusts can then distribute the mortgage
payments by a pre-arranged formula to securities, called mortgage-backed securities (MBS),
which are purchased by investors. This securitization of mortgages increased the supply of funds
available for mortgage lending1, but has also reduced regulation; nothing in the non-bank
mortgage originators’ activities triggers safety and soundness regulation by traditional federal
bank examiners. Although disclosure rules for consumer protection are federally regulated and
apply to all loans, the prudence of the lenders’ underwriting is disciplined only by the perceived
willingness of investors to purchase the loans. This report discusses the network of federal
mortgage regulators and the impact of securitization on prudent mortgage underwriting.
ȱ¢ȱȱȱȱȱȱȱȱȱ
¢ȱ ȱ
The United States has a complex financial regulatory structure that varies both by institutional
setting and banking function. Federally chartered national banks, for example, are subject to
safety and soundness examination by the Office of the Comptroller of the Currency (OCC).
Savings associations chartered at both the state and federal level are subject to safety and
soundness examination by the Office of Thrift Supervision (OTS). Bank holding companies are
subject to safety and soundness regulation by the Federal Reserve (FRB). Table 1 provides a list
of banking institutions and their safety and soundness regulators. The federal banking regulators
with examination powers cooperate through the Federal Financial Institutions Examinations
Council (FFIEC), which includes the OCC, OTS, FRB, the National Credit Union Administration
(NCUA), and the Federal Deposit Insurance Corporation (FDIC).
The agencies of the FFIEC, including the Federal Reserve, issue safety and soundness guidances
for their covered institutions, but these guidances do not have the force of regulation on lenders
who are not subject to the standards of one of the agencies. For consumer protection, in contrast,
the FRB issues binding regulations on all lenders through Regulation Z of the Truth in Lending
Act (TILA).2 Non-bank mortgage originators using the securitization channel are subject to
federal consumer protection regulation, but may escape federal safety and soundness regulation.
The FFIEC guidances on real estate lending, subprime lending, and alternative mortgages apply,
therefore, to only a portion of the mortgage market.
1
2
The decline of housing markets has coincided with a significant decline in securitization.
TILA is found at 15 USC 1601 et seq, Regulation Z is 12 CFR Part 226.
ȱȱȱ
ŗȱ
ȱ
£ȱȱȱȱȱȱȱ¢ȱȱȱ
Table 1. Regulators of Banking Institutions
National Banks
State Member Banks
Insured Federal Savings
Associations
Insured State Savings Associations
FDIC-insured State Nonmember
Banks
Non-FDIC-insured State Banks
Federal Credit Unions
State Credit Unions
Bank Holding Companies
Charter and
License
Safety/Soundness
Examination
Consumer
Protection
OCC
State
OCC
FRB & State
FRB & OCC
FRB & State
OTS
OTS
FRB & OTS
State
OTS & State
FRB, OTS, & State
State
FDIC & State
FRB, FDIC, & State
State
NCUA
State
FRB
State
NCUA
State
FRB
FRB, FTC, & State
FRB & NCUA
FRB, FTC, & State
FRB & FTC
Source: Table compiled by CRS.
Abbreviations:
FDIC—Federal Deposit Insurance Corporation
FRB—Federal Reserve Board
FTC—Federal Trade Commission
NCUA—National Credit Union Administration
OCC—Office of the Comptroller of the Currency
OTS—Office of Thrift Supervision
During the period 1997-2006, the share of mortgages securitized grew significantly, increasing
from 49.2% in 1997 to 67.7% in 2006. In dollar terms, the value of mortgages securitized grew
from $423 billion in 1997 to $2 trillion in 2006. The growth of this securitization channel may
have facilitated more lending by institutions not subject to federal bank examiners, although some
of the increase in securitization share came from regulated banks also selling to the secondary
market. Private securitization has since collapsed—the volume of non-agency MBS was 93%
lower for the first eight months of 2008 compared with the same period in 2007.3
ȱ ȱȱȱ ȱȱȬȱ ȱ
The absence of federal regulation of non-banks using securitization does not necessarily mean
that there is no federal influence on non-bank underwriting. Many mortgages are securitized by
government sponsored enterprises (GSEs), especially Fannie Mae and Freddie Mac. Lenders
planning on selling their mortgages to the GSEs would have to conform to the underwriting
standards of those institutions, which are subject to safety and soundness oversight by the Federal
Housing Finance Agency (FHFA), formerly known as the Office of Federal Housing Enterprise
Oversight (OFHEO). Standards enforced by FHFA could indirectly influence the willingness of
the GSEs’ lending partners to extend credit for more risky mortgages.
3
Calculated from monthly MBS issuance available from the Securities Industry and Financial Markets Association
(SIFMA), available at http://www.sifma.org/research/pdf/Mortgage_Related_Issuance.pdf.
ȱȱȱ
Řȱ
ȱ
£ȱȱȱȱȱȱȱ¢ȱȱȱ
At least three factors limited the influence of the GSEs’ underwriting standards as the housing
boom progressed. First, there is a cap on the size of the loan that the GSEs are allowed to
purchase, called the conforming loan limit. Loans larger than the cap, ranging from $417,000 up
to $615,000 in some high-cost areas, are called jumbo loans and can only be securitized outside
the GSEs. Securities from issuers other than the GSEs are called non-agency MBS. Because the
housing markets in some high-cost areas, such as California, were particularly active during the
boom, and mortgages in these areas are more likely to be above the cap, non-agency MBS grew
faster than the overall market.
Second, the GSEs did not enter the risky subprime market directly, instead, they purchased the
more senior (and therefore less risky) securities of non-agency subprime MBS. A lender planning
to sell to a non-agency MBS issuer would be unlikely to alter underwriting standards for GSE
purchases of senior securities. One reason the GSEs purchased non-agency subprime MBS was
that the Department of Housing and Urban Development’s (HUD) housing goals were rising.
HUD’s housing goals mandate that the GSEs purchase a minimum share of their mortgages for
lower-income borrowers and in underserved areas. The GSEs received pro-rated credit toward
their housing goals for their share in non-agency MBS. In this way, the GSEs provided additional
funds to subprime markets without a corresponding extension of their underwriting standards.
Third, there was a relatively high proportion of refinances during the boom. The decline in
interest rates caused a drop in the share of mortgages that were goals-qualifying for GSE
purchase. Higher-income home owners disproportionately took advantage of the opportunity to
refinance. This meant that relatively large mortgages, which are generally not goals-qualifying,
grew as a share of the GSE-eligible market. As a reference, the share of GSE-eligible mortgages
that were refinances in the first quarter of 1995 were 26%, but the share of mortgages that were
refinances in the first quarter of 2003 were 80%.
Ȭ¢ȱȱȱȱ ȱȱ
It is difficult to assess the underwriting standards of non-agency MBS because the information is
proprietary. There is some evidence, however, that underwriting standards loosened as the
housing boom progressed. The results of one study of the boom, by UBS, are presented in Table
2.4 The table shows an increase in the average risk of loans underwritten in 2005. For example,
interest-only mortgages (I/Os) rose from 0.0% of subprime loans in 2000 to 26.5% of subprime
loans in 2005, before falling back to 16.3%. An interest-only requires a reset to a higher payment
even if interest rates do not change. Other risk indicators, such as debt-to-income ratio (DTI) and
combined-loan-to-value (CLTV), also increased during 2001-2005. Interestingly, the primary
indicator of borrower payment history, the FICO5 score, improved during 2000-2005 from 590 to
627, although it fell back to 624 in 2006. This suggests that the use of nontraditional products
such as I/Os and debt-burdens may have played as important a role as the payment histories of the
borrowers. On the other hand, improving economic conditions and rising house prices could
generally improve FICO scores and increase the size of loans relative to incomes even if
underwriting criteria had not loosened.
4
“The U.S. Subprime Market: An Industry in Turmoil,” Thomas Zimmerman, UBS presentation,
http://www.prmia.org/Chapter_Pages/Data/Files/1471_2576_Zimmerman%20Presentation_presentation.pdf.
5
The term FICO comes from scores developed by the Fair Isaacs credit reporting firm.
ȱȱȱ
řȱ
£ȱȱȱȱȱȱȱ¢ȱȱȱ
ȱ
Table 2. Selected Risk Indicators in Non-Agency Subprime MBS During the
Housing Boom
2000
2001
2002
2003
2004
2005
2006
I/O %
0.0
0.0
0.7
3.7
15.3
26.5
16.3
FICO
590
598
612
621
623
627
624
CLTV
78.1
79.6
80.5
82.0
83.9
85.7
86.0
Full Doc
73.8
72.9
67.5
64.9
62.2
58.3
56.8
DTI
38.6
39.1
39.4
39.7
40.3
41.0
41.8
Source:
Abbreviations:
UBS.
Percent of loans that are interest-only
FICO—Average borrower credit score under Fair-Isaacs
CLTV—Average loan-to-value ratio (combined with any 2nd)
Full Doc—Percent of loans with full documentation
DTI—Debt-to-income ratio
I/O%—
ȱȱ ȱ ȱ
¡ȱȱȱ¢ȱ ȱ
Testimony by the financial regulatory agencies suggests that loans subject to their guidance have
fared much better than non-agency MBS originated by non-banks.6 On the one hand, the
guidances of the FFIEC provided for more prudent underwriting standards and closer scrutiny of
subprime loans even before the housing markets cooled off. On the other hand, the guidances are
administered by bank examiners within an existing institutional framework and it is unclear how
non-bank lenders would be incorporated. Could they be subject to examination by existing
agencies and personnel, or would there need to be significant changes to agency structure or
staffing?
ȱȱ
The market has already improved underwriting standards and punished the riskiest lenders with
bankruptcy and the investors in the riskiest securities with significant losses. Underwriting
standards for non-agency MBS were essentially set by the willingness of investors to accept the
estimated risk of the mortgage pools. Because investors rarely had detailed knowledge of the loan
pools, they often relied on ratings agencies to evaluate the risk. While house prices were rising, a
troubled borrower could sell the house rather than default, which held down expected default
rates. Because housing markets have slowed down and loan defaults have been rising, markets
have been re-evaluating the risks in non-agency MBS. As a result, MBS ratings have been falling,
and funding for the riskiest mortgages has already all but dried up. A disadvantage of taking no
action is that while the market has already raised current underwriting standards there is no
assurance that a future boom and bust cycle will not be repeated.
6
Statement of Sheila C. Bair, Chairman, Federal Deposit Insurance Corporation on Recent Events in the Credit and
Mortgage Markets and Possible Implications for U.S. Consumers and the Global Economy before the Financial
Services Committee, September 5, 2007.
ȱȱȱ
Śȱ
ȱ
£ȱȱȱȱȱȱȱ¢ȱȱȱ
ȱ¢ȱ
Mortgage originators, including brokers and lenders, could be made liable for defaults if
underwriting standards are unsuitable for the borrower’s circumstances. One advantage of this
approach is that originators have direct contact with borrowers and have a great deal of
information about each borrower’s circumstances, relative to MBS investors or financial
regulators. Originator liability could ensure that mortgage brokers and lenders retain a stake in the
long-term performance of their loans. A disadvantage of this approach is that suitability is
difficult to define, subject to significant uncertainty and litigation risk, and determined only after
events occur that trigger defaults.
ȱ¢ȱ
The secondary purchasers of mortgage loans, assignees, could be held liable for unfair, deceptive,
or unsuitable mortgage originations. The advantage of this approach is that it would encourage
secondary market participants to be more vigilant in monitoring the practices of mortgage brokers
and lenders. This approach also gives aggrieved borrowers potential redress if a thinly capitalized
mortgage originator goes bankrupt before the borrower can seek compensation. A disadvantage of
this approach is that if liability is unclear, investors will not be able to quantify and price it, and
the market may shut down.
ȱȱ ȱ ȱȱ ȱ
National underwriting guidelines could be set in statute or an agency could be authorized to
establish national underwriting guidelines by regulation. Official standards for debt-to-income
ratios, FICO scores, and other risk indicators could be announced. Banks, non-banks, and
borrowers could all be made aware of a single set of prudential limits on loan terms. On the other
hand, mortgage markets would become less flexible and borrowers with nontraditional sources of
income or other characteristics would have difficulty qualifying for loans. One such proposal,
H.R. 3915, passed the House of Representatives in the 110th Congress but has not as yet been
considered by the Senate.
ȱȱ ȱ
Edward V. Murphy
Specialist in Financial Economics
tmurphy@crs.loc.gov, 7-6201
ȱȱȱ
śȱ
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.