High-Loan-to-Value Lending: Information on Loans Exceeding Home Value
(Letter Report, 08/13/98, GAO/GGD-98-169).
Pursuant to a congressional request, GAO provided information on
high-loan-to-value (HLTV) loans, focusing on the: (1) characteristics of
HLTV loans; (2) major organizations that provided HLTV lending; (3)
volume of HLTV lending in 1995, 1996, and 1997, and the expected volume
in 1998; and (4) benefits and risks of HLTV lending for borrowers,
lenders, investors, and regulated depository institutions.
GAO noted that: (1) HLTV loans are considered hybrid loans because they
have characteristics of both mortgage loans and unsecured consumer
loans; (2) like mortgages, HLTV loans are secured by a lien on a house;
(3) however, the lien itself may have less financial value than the
amount of the loan in the event of a borrower's defaulting since the
value of the HLTV loan and the original mortgage may exceed the value of
the house; (4) thus, as with unsecured consumer loans, HLTV lenders rely
more heavily on borrowers' creditworthiness; (5) according to industry
officials, most borrowers use HLTV loans primarily to consolidate credit
card debt or make home improvements; (6) while comprehensive industry
data were not available, data provided by a lender responsible for about
one-third of HLTV lending showed that, in 1997, HLTV loans averaged
about $30,000; (7) the data also showed that the average contract
interest rate was between 13 and 14 percent, with an average loan term
of 25 years; (8) according to industry officials and GAO's review of the
limited available industry data, from 1995 to 1997, HLTV loans were made
or managed primarily by 10 institutions; (9) according to public- and
private-sector officials, regulated depository institutions were not
heavily involved in originating HLTV loans; (10) the involvement of
these institutions would be important to any assessment of the potential
exposure of federal deposit insurance funds to defaults of HLTV loans;
(11) data indicate that HLTV lending has grown since its introduction in
1995 but remains small relative to other consumer lending; (12) data on
the volume of HLTV loans were limited to those loans that were
subsequently packaged into loan pools used to back securities sold to
investors; (13) the volume of securitized HLTV lending has more than
doubled from year to year from 1995 through 1997; (14) industry and
securities firm representatives expected HLTV lending to increase again
in 1998 to $12 billion or higher; (15) public- and private-sector
officials pointed to several benefits and risks associated with HLTV
lending to the borrower, lender, and investors; (16) while lenders and
investors have benefited from the high rate of return on these loans, it
is uncertain how these loans would perform during any future economic
downturn; (17) if defaults were to increase, the rate of return would
decrease; and (18) officials representing the two largest
government-sponsored enterprises did not believe that HLTV lending posed
greater risks to their portfolios than the existing credit card lending
that the HLTV lending generally refinances.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: GGD-98-169
TITLE: High-Loan-to-Value Lending: Information on Loans Exceeding
Home Value
DATE: 08/13/98
SUBJECT: Loan defaults
Mortgage loans
Homeowners loans
Credit
Lending institutions
Financial management
Debt
Losses
Personal loans
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO report. Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved. Major **
** divisions and subdivisions of the text, such as Chapters, **
** Sections, and Appendixes, are identified by double and **
** single lines. The numbers on the right end of these lines **
** indicate the position of each of the subsections in the **
** document outline. These numbers do NOT correspond with the **
** page numbers of the printed product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
Cover
================================================================ COVER
Report to the Chairman, Subcommittee on Financial Institutions and
Regulatory Relief, Committee on Banking, Housing, and Urban Affairs,
U.S. Senate
August 1998
HIGH-LOAN-TO-VALUE LENDING -
INFORMATION ON LOANS EXCEEDING
HOME VALUE
GAO/GGD-98-169
High-Loan-to-Value Lending
(233546)
Abbreviations
=============================================================== ABBREV
HLTV - High-loan-to-value
FannieMae - Federal National Mortgage Association
FreddieMac - Federal Home Loan Mortgage Corporation
APR - Annual percentage rate
HILA - Home Improvement Lenders Association
Letter
=============================================================== LETTER
B-279255
August 13, 1998
The Honorable Lauch Faircloth
Chairman, Subcommittee on Financial
Institutions and Regulatory Relief
Committee on Banking, Housing, and Urban Affairs
United States Senate
Dear Mr. Chairman:
This report responds to your request that we provide information on
high-loan-to-value (HLTV) loans. Since 1995, a segment of the
financial services industry has offered loans that are tied to the
value of a borrower's house but that, in combination with preexisting
first mortgages, exceed this value. The loans, referred to as HLTV
loans, provide for a combined loan-to-value ratio that could reach
125 percent or even more of a home's value.
As agreed with your office, this report provides information
regarding (1) the characteristics of HLTV loans; (2) the major
organizations that provided HLTV lending; (3) the volume of HLTV
lending in 1995, 1996, and 1997, and the expected volume in 1998; and
(4) the benefits and risks of HLTV lending for borrowers, lenders,\1
investors, and regulated depository institutions.
To compile this information, we interviewed officials representing
federal regulatory agencies, lending institutions, industry
associations, a rating agency, investment banks, and a consumer
advocacy group. We also reviewed publicly available information,
including published reports, prospectuses associated with the selling
of HLTV-related securities, and academic studies. Comprehensive
industry data on HLTV loans were not maintained by any single entity.
Therefore, in addition to interviews with officials, we relied on the
limited industry data that were available at various institutions.
Although we did not independently verify these data, we corroborated
evidence with other independent sources whenever possible. To ensure
that its contents were factually accurate, we provided draft copies
of this report to public sector and private sector officials. We
incorporated their technical comments where appropriate. We
conducted our work between January and June 1998 in accordance with
generally accepted government auditing standards. For more detailed
information concerning our scope and methodology, see appendix I.
--------------------
\1 We use the term lender in this report to cover all businesses that
either originate these loans themselves or acquire these loans from
correspondents or other financial institutions.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :1
HLTV loans are considered "hybrid" loans because they have
characteristics of both mortgage loans and unsecured consumer loans.
Like mortgages, HLTV loans are secured by a lien\2 on a house.
However, the lien itself may have less financial value than the
amount of the loan in the event of a borrower's defaulting, because
the value of the HLTV loan and the original mortgage (which would
have first claim on the value of the house in the event of a default)
may exceed the value of the house. Thus, as with unsecured consumer
loans, HLTV lenders rely more heavily on borrowers'
creditworthiness--that is, the borrowers' likelihood of making timely
and complete payments--in making loans. According to industry
officials, most borrowers use HLTV loans primarily to consolidate
credit card debt. In addition, some borrowers used HLTV loans to
make home improvements. While comprehensive industry data were not
available, data provided by a lender responsible for about one-third
of HLTV lending showed that, in 1997, HLTV loans averaged about
$30,000. The data also showed that the average contract interest
rate was between 13 and 14 percent, with an average loan term of 25
years.\3 The average combined indebtedness of the first mortgage and
the HLTV loan represented about 110 percent of the borrower's
property value, although in some cases the combined loans reached or
exceeded 125 percent of value.
According to industry officials and our review of the limited
available industry data, from 1995 to 1997, HLTV loans were made or
managed primarily by 10 institutions. Often, these 10 institutions
obtained loans originally made by correspondents--that is,
institutions that dealt directly with the borrower and then
transferred the loans to one of the 10 institutions. According to
public and private sector officials, regulated depository
institutions were not heavily involved in originating HLTV loans.
The involvement of these institutions would be important to any
assessment of the potential exposure of federal deposit insurance
funds to defaults of HLTV loans. One nondepository
institution--FirstPlus Financial Group, Incorporated (FirstPlus), of
Dallas, Texas--has had about one-third of the market since 1995.
Available data indicate that HLTV lending has grown since its
introduction in 1995 but that it remains small relative to other
consumer lending. Data on the volume of HLTV loans were limited to
those loans that were subsequently packaged into loan pools used to
back securities sold to investors. Lenders and securities firms
involved in this process of securitization said that about 95 percent
of HLTV loans were sold as securities to investors. The volume of
securitized HLTV lending has more than doubled from year to year from
1995, the year it was introduced, through 1997. According to
industry representatives, about $1 billion worth of these loans were
made in 1995; $3 billion in 1996; and $8 billion in 1997. Industry
and securities firm representatives expected HLTV lending to increase
again in 1998 to $12 billion or higher. Although this has been a
growing market, total HLTV lending was small compared with other
types of consumer lending. For example, even if no HLTV loans had
been repaid, the total amount of outstanding HLTV loans would only
have been about $12 billion at the end of 1997. In contrast, total
outstanding residential mortgage debt reached $3.8 trillion in 1997,
and total outstanding consumer debt (other than mortgages) reached
$1.3 trillion in the same year.
While only limited data were available for documenting the
performance of HLTV loans, public and private sector officials
pointed to several benefits and risks associated with HLTV lending to
the borrower, lender, and investors. For example, while some of
these officials said that HLTV lending provided borrowers with a
quick way to consolidate credit card debt and lessen their monthly
debt payments, others noted that HLTV loans could make it more
difficult for homeowners to sell or refinance their houses. Also,
while lenders and investors have benefited from the high rate of
return on these loans, it is uncertain how these loans would perform
during any future economic downturn. If defaults were to increase,
the rate of return would decrease. In addition, officials
representing the two largest government-sponsored enterprises in the
secondary market for residential mortgages\4 did not believe that
HLTV lending, at its current level, posed substantially greater risks
to their portfolios than did the existing credit card lending that
the HLTV lending generally refinances. The risks to the portfolios
exist because a high total debt burden could increase the risk of
default on the mortgages that comprise the portfolios of these
government-sponsored enterprises.
--------------------
\2 A lien is a legal claim in the event of default on a loan that
gives the lender claim to the value of a property used as security
for the loan.
\3 We use the term contract interest rate to refer to the interest
rate stated in the loan agreement. If origination fees or points
were charged on the loan, its effective interest rate would exceed
the contract interest rate.
\4 Government-sponsored enterprises are federally chartered,
privately owned corporations designed to provide a continuing source
of credit nationwide to specific economic sectors. The nation's two
largest government sponsored enterprises are the Federal National
Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage
Corporation (Freddie Mac).
BACKGROUND
------------------------------------------------------------ Letter :2
According to industry officials, HLTV lenders used credit scoring
models as the primary basis for identifying creditworthy borrowers
and thus for approving HLTV loans. Generally speaking, lenders used
credit scoring models to evaluate credit risk--the possibility that
borrowers would be (1) delinquent (i.e., make late loan payments) or
(2) in default (i.e., cease to make loan payments). In general,
credit scores have been developed by assessing various types of
information from a large pool of borrowers, including borrowers with
good payment histories and others with poor payment histories.
Statistical analyses identifying the characteristics of borrowers who
were most likely to make loan payments have been used to create a
weight or score associated with each of the characteristics. For
instance, borrowers who did not have a history of delinquent payments
receive a higher credit score than borrowers who had many delinquent
payments. Most widely used credit scoring systems have a range of
scores from 350 to 900. Borrowers with higher scores are usually
considered more creditworthy because they would be more likely to pay
the loan on time, and in full, than would borrowers with lower
scores.\5
According to an official familiar with these proprietary models, a
key variable associated with higher credit scores is home ownership.
Homeowners tend to have better payment histories than borrowers with
otherwise similar characteristics. Industry officials told us that
homeowners acted as if they had a vested interest or "psychological
drive" to make prompt loan payments in an effort to keep the homes in
which they lived.
In general, in making HLTV loans, lenders sought to assemble the HLTV
loans into pools of loans, which then were used to back securities
sold to investors. According to securities firm officials, about 95
percent of HLTV loans were combined into pools and securitized. In
general, securitization involves a lender's packaging financial
assets--typically loans--into pools, the creation of securities based
on the cash flows from the pools, and the sale of these securities to
investors. Securitization is commonly done for a wide range of
financial assets, including mortgages, car loans, and other forms of
consumer debt.
The process of HLTV securitization has three major steps. In the
first step, the borrowers take out loans from the lender, and the
lender organizes the pool and sells the HLTV loans to the pool. In
addition, the lender hires a credit rating agency and a securities
underwriter. The pool is established as a legal entity that is
independent of the lender. Second, the credit rating agency rates
the securities that will be issued by the pool and the securities
underwriter sets initial prices for the securities, buys the
securities from the pool, and immediately sells the securities to
investors. Third, the lender services the pool by collecting
payments from HLTV borrowers and disbursing payments to investors.
This securitization process is depicted in figure 1.
Figure 1 : Steps in a HLTV
Securitization
(See figure in printed
edition.)
Note: In this diagram, the
lender both originates the
loans and creates the pool.
(See figure in printed
edition.)
Source: GAO
(See figure in printed
edition.)
Since the pool is a separate legal entity, its credit quality is
considered to be separate from that of the lender that formed the
pool. The credit quality of the pool depends on the credit quality
of the loans in the pool. In addition, the credit quality of the
assets in the pool may also be enhanced, either internally or
externally. One form of internal credit enhancement occurs if the
value of loans in the pool exceeds the value of the securities sold;
this is a form of over-collateralization. Common forms of external
credit enhancements include insurance policies that pay in the event
of loan defaults.
--------------------
\5 Even though some borrowers with higher scores could become
delinquent or default and some borrowers with lower scores could pay
the loan as required, statistics showed that borrowers with higher
scores would be more likely to pay the loan on time, and in full,
than would borrowers with lower scores.
HLTV LOANS HAVE UNIQUE
CHARACTERISTICS
------------------------------------------------------------ Letter :3
HLTV loans have often been referred to as "hybrid" loans because they
have characteristics of both mortgage loans and unsecured consumer
loans. According to lenders, borrowers have used HLTV loans
primarily to consolidate credit card debt or to make home
improvements. Although industrywide data were not available,
FirstPlus officials had compiled an "average HLTV loan profile" that
identified the average amount of an approved HLTV loan, the interest
rate charged, the duration of the loan, and the percent of property
value the loan represented. Further, FirstPlus officials identified
the "average HLTV borrower profile," which was a creditworthy
borrower (based on credit scores) who had a stable income. Officials
representing federal regulators and a consumer advocacy group were
not aware of any borrowers' complaints about the disclosure of the
terms and conditions of HLTV loans.
HLTV LOANS ARE HYBRID LOANS
---------------------------------------------------------- Letter :3.1
HLTV loans have characteristics of both mortgage loans and unsecured
consumer loans. On the one hand, HLTV loans share characteristics
with mortgage loans in that both types of loans are secured with a
lien on the property in the event of borrower default. On the other
hand, HLTV loans are like unsecured consumer loans because, in both
cases, the lending decisions are based primarily on the
creditworthiness of the borrower. Creditworthiness is the primary
basis for the lending decision because, in the event of a default,
there might be little or no value in the house left to pay off the
HLTV loan, once the first mortgage loan is covered by the sale of the
house.
AVERAGE HLTV LOAN PROFILE
---------------------------------------------------------- Letter :3.2
Although industrywide data were not available, FirstPlus--which
accounted for about one-third of the HLTV loan volume in
1997--provided us with information on its average HLTV loan profile.
According to FirstPlus officials, the average HLTV loan they made in
1997 was for about $30,000 with a 25-year term. Also, in the same
year, FirstPlus charged an average contract interest rate of 13 to 14
percent on its HLTV loans, but any origination fees or "points
charged" would raise the effective interest rate of the loan. A
FirstPlus official told us that FirstPlus charged an average of five
"points."\6 Thus, a 25-year loan with a 13.5 percent contract rate
and 5 points would have an effective interest rate of 14.3 percent.
In addition, the average combined debt of a first mortgage and a
FirstPlus HLTV loan with a second lien generally represented about
110 percent of a borrower's property value, although some reached as
high as 125 percent of value.
--------------------
\6 OTS officials told us that some lenders have charged as many as 12
points.
AVERAGE HLTV BORROWER
PROFILE
---------------------------------------------------------- Letter :3.3
Although there were no comprehensive industry data, FirstPlus
provided us with information on its borrowers. Generally speaking,
FirstPlus made HLTV loans to borrowers with a stable average annual
income of $60,000. These borrowers had, on average, at least 5 years
of job tenure and were in their late thirties. Also, while the
borrower had an average of about $20,000 in outstanding nonmortgage
debt, as previously mentioned, the HLTV loans were on average about
$30,000. Virtually all of these borrowers occupied the house they
were using as collateral. The borrower owed an average of about
$110,000 on a first mortgage, backed by property worth about
$130,000. However, a key FirstPlus official told us that approval of
a HLTV loan was not strongly tied to the amount of equity the
borrower had in the house; rather, it was primarily tied to the
borrowers' creditworthiness, as measured by a credit score.\7 In
addition, FirstPlus officials told us that they review the borrower's
credit histories, debt-to-income ratios, and disposable-income
levels.
Public and private sector officials--including representatives of
FirstPlus--told us that credit scores for HLTV borrowers were, on
average, about 680 and that this score was considered to be "high."
According to the leading provider of credit scoring models, a score
of 700 predicts the odds of 30 to 1 that the borrower will be current
on the loan payments.\8 An official with a government-sponsored
enterprise active in the secondary mortgage market told us that 85
percent of all mortgage borrowers have credit scores of 620 or
higher.
The Survey of Consumer Finances, a triennial survey of family
finances sponsored by the Federal Reserve with the cooperation of the
Department of the Treasury, provides a basis to compare the profile
of the HLTV borrower with the broader U.S. population. In 1995, the
most recent Survey at the time of our review, average family income
was estimated to be $44,300 for all families and $54,600 for
homeowners. The Survey also found that 64.7 percent of families were
homeowners, with the median value of the house being $90,000.
Further, 41.1 percent of the homeowners in the Survey had mortgages
or home equity loans, the median total loan amount being $51,000.
Credit cards were held by 47.8 percent of the homeowners in the
Survey, who owed a median amount of about $1,500.
A comprehensive comparison of the HLTV and Survey profiles was not
possible because some Survey data on families are reported as
medians, while the HLTV data are based on averages. Further,
available data we obtained on HLTV borrowers were from the FirstPlus
borrower profile; other lenders' borrower profiles may be different.
With these cautions in mind, however, some general similarities and
differences between the HLTV and Survey profiles can be delineated.
The HLTV borrower had a slightly higher average income: the
FirstPlus borrower had an average annual income of about $60,000,
while the average homeowner in the Survey had an income of $51,000.
The average value of housing in the HLTV profile ($130,000) is higher
than the median value of housing in the Survey ($90,000), although
this comparison of an average and a median is inherently limited.
Since housing values do not have an upper limit, the mean or average
value would increase more than the median with the inclusion of
high-valued houses. HLTV borrowers appear to have had higher
mortgage balances even before taking on the HLTV loans: the
FirstPlus profile reported an average first mortgage of $110,000,
while the median mortgage and home equity balance for homeowners in
the Survey was $51,000.
--------------------
\7 HLTV lending differed from subprime lending, which is lending to
less creditworthy borrowers. HLTV lending primarily depended on the
borrowers' creditworthiness, not on the value of equity in their
collateral. Subprime lending, on the other hand, heavily depended on
collateral equity and, to a lesser extent, borrower creditworthiness.
\8 The precise definition of being current on the loan payments that
this provider used was that the borrower would not be over 60 days
delinquent in payments.
HLTV LOAN DISCLOSURE OF
REQUIREMENTS TO BORROWERS
---------------------------------------------------------- Letter :3.4
Federal law and regulations set forth requirements concerning the
information on a loan that must be disclosed to a borrower. In
addition, state law may require disclosure to borrowers.
Specifically, to promote the informed use of consumer credit, the
Truth in Lending Act\9 requires that creditors disclose credit terms
and the cost of credit as an annual percentage rate (APR). For loans
secured by a consumer's home, additional disclosures are required,
and the act permits consumers to cancel certain transactions that
involve a lien on their principal dwelling. The Home Ownership and
Equity Protection Act of 1994\10 added new disclosure requirements
for home equity loans that have credit terms in excess of a certain
amount. To be covered by these new disclosure requirements, the loan
must meet either of the following tests: (1) the APR must be more
than 10 percentage points over the yield on Treasury securities with
a maturity comparable to the loan, or (2) total points and fees
payable by the borrower must exceed the greater of either 8 percent
of the total loan amount or $435.\11 A special early disclosure is
required at least 3 days before the loan closing. This disclosure
generally includes (1) a warning that the house may be lost in the
event of a default, (2) the APR, and (3) the amount of payments. The
Home Ownership and Equity Protection Act also limits the use of
certain contract terms and other practices in loan transactions
covered by the act. These limitations address, among other things,
rebates, prepayment penalties, advance payments, and negative
amortization. Moreover, the Real Estate Settlement Procedures Act\12
requires use of a standard form for the statement of settlement
costs, as prescribed by the Secretary of Housing and Urban
Development, in all transactions involving federally related mortgage
loans.\13
Officials representing banking and thrift regulators, the Federal
Trade Commission, and a consumer advocacy group told us that they
were not aware of any complaints regarding disclosure of HLTV loan
terms to the borrower. We reviewed a standard, blank package of
settlement papers provided by FirstPlus. While we did not determine
compliance with applicable disclosure requirements, our review found
that the papers contained information that appeared to address these
requirements. Also, our review of these settlement papers indicated
that some of the HLTV loans may be covered by the Home Ownership and
Equity Protection Act.
--------------------
\9 Truth in Lending Act, 15 U.S.C. 1601 et seq. The act is
implemented by the Federal Reserve Board's Regulation Z, 12 CFR part
226.
\10 The Home Ownership and Equity Protection Act of 1994, Subtitle B
of the Riegle Community Development and Regulatory Improvement Act of
1994, P.L. 103-325, 108 Stat. 2106 (1994).
\11 The $435 figure applies to 1998. The Federal Reserve Board
adjusts the amount annually based on changes to the consumer price
index.
\12 12 U.S.C. 2601 et seq. The act is designed to, among other
things, provide for more effective advance disclosure to home buyers
and sellers of settlement costs.
\13 Federally related mortgage loans include first or subordinate
mortgages on residential homes made by depository institutions or by
certain creditors who make more than $1,000,000 in residential loans
per year, or which are intended to be sold to Fannie Mae or Freddie
Mac.
WHILE HUNDREDS OF INSTITUTIONS
ORIGINATED HLTV LOANS, 10
INSTITUTIONS COLLECTIVELY
SERVED MOST OF THE HLTV MARKET
------------------------------------------------------------ Letter :4
According to industry officials, hundreds of institutions originated
HLTV loans. There is no comprehensive, publicly available
information regarding the identity of these institutions or their
HLTV activity, and they are not tracked by regulatory agencies or
industry associations. However, many of the institutions acted as
correspondents, generally transferring the loans to other
institutions that in turn securitized the loans.
According to interviews with numerous industry officials and our
review of limited available industry data, 10 institutions have
collectively led the HLTV market since 1995. The 10 institutions
either received HLTV loans originated by correspondent institutions
or originated HLTV loans themselves. One institution in
particular--FirstPlus--has made about one-third of all HLTV loans
since 1995, either directly or through correspondent relationships
with other lenders. The following are the 10 institutions that led
the market, in order of the volume of HLTV loans they provided in
1997.
-- FirstPlus,
-- Cityscape,
-- Empire Funding,
-- Master Financial,
-- PSB Lending Corporation,\14
-- Life Financial,\15
-- First Keystone,
-- The Money Store,\16
-- Mego Mortgage, and
-- Preferred Mortgage.
According to public and private sector officials, regulated
depository institutions were not heavily involved in originating HLTV
loans from 1995 to 1997. The degree of involvement of these
institutions would be important to assess any potential exposure of
federal deposit insurance funds to defaults on HLTV loans. For
instance, of the 10 institutions that led the market, only one--Life
Financial--was a regulated depository institution. Other
institutions were affiliated with depository institutions.
--------------------
\14 In June 1998, PSB Lending Corporation was in the process of being
purchased by Bay View Capital Corporation. However, the acquisition
was put on temporary hold pending review of possible federal
regulatory guidelines affecting HLTV lending.
\15 As of June 1998, FirstPlus had applied to acquire Life Financial,
a thrift holding company. According to OTS officials, this
application is subject to regulatory review.
\16 The Money Store was purchased by First Union Corporation in June
1998.
HLTV LENDING IS GROWING BUT
STILL REPRESENTS ONLY A SMALL
PERCENTAGE OF THE TOTAL
MORTGAGE AND UNSECURED LENDING
MARKETS
------------------------------------------------------------ Letter :5
Data on the volume of HLTV loans are limited to those loans that were
subsequently sold to investors as securities. Lenders and securities
firms involved in this process of securitization agreed that about 95
percent of HLTV loans were sold as securities. According to private
sector officials, the total securitized volume of HLTV lending has
more than doubled from year to year from 1995, the year it was
introduced, through 1997. The officials also expected total
securitized HLTV lending to increase in 1998, with estimates of total
volume of $12 billion or higher. (See figure 2.)
Figure 2: Available data show
total Industry Volume of HLTV
Lending Has Continued to Grow
(See figure in printed
edition.)
Note: 1998 estimates include $12 billion and higher estimates.
Source: GAO review of reports from an investment bank and the Home
Improvement Lenders Association (HILA), interviews with private
sector officials, and various published articles.
While available data show that HLTV lending has continued to grow,
the total outstanding debt represents a small percentage of the total
debt outstanding for all consumer-oriented lending. Even if no loans
were repaid, total debt outstanding for HLTV loans would be almost
$12 billion in 1997. In contrast, total residential mortgage debt
outstanding reached about $3.8 trillion in 1997, and total unsecured
consumer debt outstanding reached about $1.3 trillion in 1997,
according to Federal Reserve statistics.
ESTIMATED SECURITIZED VOLUME
OF MAJOR ORGANIZATIONS
---------------------------------------------------------- Letter :5.1
Industry officials we interviewed agreed that the 10 institutions
shown in table 1 have provided almost all HLTV loans since their
introduction in 1995. As shown, FirstPlus has provided a significant
part of the HLTV lending--offering about one-third of HLTV loans
since 1995.
Table 1
HLTV Loan Volumes by the 10 Leading
Lenders, 1995-1997
(Dollars in billions)
Lender 1995 1996 1997
---------------------------------- ---------- ---------- ----------
FirstPlus\a $0.30 $1.10 $3.10
Cityscape \ 0 0 0.80
Empire Funding 0 0 0.90
Master Financial 0 0 0.70
PSB Lending Corporation n/a n/a 0.70
Life Financial\b n/a n/a 0.40
First Keystone 0.50 0.80 0.40
Money Store 0 0 0.40
Mego Mortgage 0 0.03 0.30
Preferred Mortgage\b n/a n/a 0.20
======================================================================
Estimated Total\c $1.0 $3.0 $8.0
billion billion billion
----------------------------------------------------------------------
Note: n/a = not available.
\a According to a key company official, in 1997, FirstPlus provided
$3.10 billion in HLTV loans in 42 states; about 27 percent of these
loans were made to borrowers in California.
\b Volume data for 1995 and 1996 for these institutions were not
readily available because they were not members of HILA as of May
1998, when the Executive Director sent questionnaires to industry
officials.
\c Totals represent estimates made by industry officials and
estimates reported in various published articles. The data in the
columns may not add up to the estimated totals because of rounding or
because not all data were available.
Source: Compiled by GAO from industry responses to a HILA
questionnaire and other industry data.
BENEFITS AND RISKS ASSOCIATED
WITH HLTV LENDING
------------------------------------------------------------ Letter :6
There were limited data on the losses resulting from HLTV lending.
However, based on the characteristics of the loans and available
performance data on HLTV loans, public and private sector officials
identified many inherent trade-offs between the benefits and the
risks associated with HLTV lending that face borrowers, lenders, and
investors. These officials told us that regulated depository
institutions have not made many HLTV loans, but that the practice
appears to be growing.
LIMITED DATA AVAILABLE TO
IDENTIFY LOSSES ON HLTV
LOANS
---------------------------------------------------------- Letter :6.1
Public and private sector officials told us that data on losses from
HLTV loans experienced by institutions making loans were limited for
a variety of reasons, such as that the loans (1) were a recent
lending practice adopted in 1995 (and losses are usually experienced
over longer periods of time); (2) were offered during a period of
strong economic growth; and (3) were not specifically tracked by
regulators because HLTV loans were generally made by unregulated,
nondepository institutions.
A recent academic study\17 concluded that there is "no credible
economic argument" to suggest that HLTV lending would increase
consumer default risk or destabilize the economy. The study noted,
however, that several HLTV lenders had substantial losses in 1997.
According to the study, some lenders experienced problems due to
unrealistic assumptions on how rapidly the loans would be repaid,
inadequate underwriting standards, and poor management.
In addition, one securities firm reported on possible HLTV loan
losses by comparing them to losses expected in mortgage and credit
card lending. The firm noted that possible HLTV losses might be
higher than losses experienced in the mortgage lending market.
First, the delinquency and default rates on HLTV loans might be
higher than those rates on mortgages. Second, in the event of a
default and subsequent sale of the house securing the loan, the
proceeds of the sale may not be sufficient to pay off both the first
mortgage and the HLTV loan. Because first mortgages have first claim
on the house sale proceeds, there may be little or nothing left to
pay off the HLTV loan. On the other hand, the securities firm also
noted that possible HLTV charge-off rates (i.e., the percentage of
total loan value that would be written off because no further
payments were expected) would compare favorably with losses on
"higher quality credit card pools." For instance, the HLTV charge-off
rates were estimated to peak at a 4.7 percent annualized rate for
recently originated pools. By comparison, the charge-off rate on
Moody's aggregate credit card index was reported as 6.65 percent in
December 1997.
--------------------
\17 Charles W. Calomiris and Joseph R. Mason, "High Loan-to-Value
Mortgage Lending: Problem or Cure?" July 10, 1998, unpublished
working paper, American Enterprise Institute.
BORROWERS WITH HLTV LOANS
EXPERIENCED BENEFITS AND
RISKS
---------------------------------------------------------- Letter :6.2
Lenders and independent analysts agreed that most borrowers decreased
their monthly debt burden payments by using HLTV loans to consolidate
their credit card debts, for two reasons. First, because borrowers
rescheduled their debts over a longer period of time, the monthly
debt burden would decrease even if they paid the same interest rates
they were paying on the credit cards. Second, to the extent that the
interest rate was lower than the credit card rates, the monthly debt
service would be further decreased.\18 In 1997, the average contract
HLTV loan charged 13 to 14 percent interest, while the average
interest rate on credit cards was 16 percent. However, any
advantages of a HLTV loan relative to credit card debt would depend
on the terms and conditions of the credit card debt agreement, and
these vary across the industry. In addition, as previously
mentioned, the effective interest rate on an HLTV loan would vary
depending on the number of points charged on the loan. Finally, some
industry officials said that HLTV loans might allow borrowers some
tax deductions for interest expenses that would not be available for
credit card interest expenses.\19
Public and private sector officials pointed out that HLTV lending
involves three primary risks on the part of the borrower. First,
officials questioned whether borrowers would have the ability to make
HLTV loan payments in the future if they encountered economic
difficulties, such as job losses or unemployment. While borrowers
might have benefited from restructuring debt by using HLTV loans
during good economic conditions, officials pointed out that there
were no data on how well the borrowers could make loan payments
during any future economic downturn. Default rates on all loans
generally increase during economic downturns.
Second, some analysts expressed concern that HLTV borrowers could
increase their credit card debt after taking out an HLTV loan because
they could retain their credit cards and/or be approved for new
credit cards. Given lower monthly debt service, once credit card
debt was restructured as an HLTV loan, it is possible that borrowers'
credit scores could rise. For instance, Brittain Associates, a
marketing research firm, reported in 1998 that some HLTV borrowers
had already begun to build up sizable credit card debt soon after
getting the previous debt paid off with a home equity loan.
Finally, some officials believed that HLTV loans might decrease
borrowers' financial flexibility by limiting their ability to sell
their houses or to refinance first mortgages. The combined value of
the HLTV loan and first mortgage often exceeds the value of the
house. Upon sale, the borrower would usually need to pay off both
loans. If sufficient funds were not available, either from the sale
or from other sources, the borrower might not be able to sell. In
addition, officials at a government-sponsored enterprise noted that
the presence of second loans secured by the value of houses generally
made it more complicated to refinance first mortgages.
In addition, housing prices can be particularly sensitive to changes
in economic conditions. In some cases, this sensitivity can affect
loan payments by borrowers. For example, when the price of oil fell
during the mid 1980s, residential housing values in Houston, Texas,
fell 23 percent. Similarly, when the California economy went into a
recession in the early 1990s, average housing prices declined by 21.1
percent in Southern California and 9.6 percent in Northern
California. During these times, defaults on mortgages increased
because borrowers lost jobs and, in some cases, were unwilling to pay
on mortgages whose balances far exceeded the current market values of
their houses.
--------------------
\18 Conversely, the credit card monthly debt service would decrease
over time if no new debt was added to the credit cards and at least
minimum payments were being made regularly. The monthly debt service
on the HLTV loan would remain unchanged for the full term of the
mortgage if payments were made regularly.
\19 The Internal Revenue Code provides that interest paid can only be
deducted to the extent that the amount of the home equity loan plus
the first mortgage does not exceed the market value of the house.
LENDERS OF HLTV LOANS
EXPERIENCED BENEFITS AND
SOME RISKS
---------------------------------------------------------- Letter :6.3
HLTV loans offered several benefits to HLTV lenders. For example,
when the loans were securitized, the lender could continue to service
the loans, which would generate service fee income for the lender. A
securities industry study noted that several securitization offerings
during 1997 provided for service fees of 0.75 to 1.0 percent of the
loan amount. On a $30,000 loan, this fee would equal between $225
and $300.
More generally, officials representing securities firms told us that
HLTV loans could be profitable for lenders because these loans could
be sold at a premium over the face value of the loan to the pools,
which subsequently would be securitized. As noted earlier, the
lenders organize the pools of assets in the securitization process,
but the pools are established as separate legal entities. The
lenders sell the HLTV loans to the pool. Under current accounting
rules, any such profits must be immediately recognized on those sales
where certain equity interests are maintained by the seller as part
of the deal and recorded as an asset. The estimated fair value of
the asset is added to sales proceeds when calculating profit.
Officials representing securities firms also noted several ways in
which the lenders would retain some risk. First of all, some HLTV
loans were not securitized and were held on the lenders' books. The
officials we talked with estimated that 5 percent of the HLTV loans
were not securitized, but there is no way to reliably know how many
loans were not securitized or to identify their value. In addition,
even when the HLTV loans were securitized (about 95 percent), the
officials said the lenders often retained a stake in the retained
assets pools. The lender might hold a class of securities that was
designed to absorb credit risk to protect investors, for example. By
taking on the credit risk with these securities, the lender provides
an internal credit enhancement to the other investors in the pool.
In addition, by absorbing these credit risks, the lender may protect
the pool and its reputation as a securitizer from the consequences of
securities that do not perform. In cases where the seller retains an
equity interest as part of the securitization, the retained asset
carries risk for the lender because the value of the asset is based
on assumptions about the future performance of the loans sold and
assumptions about future cash flows. If these assumptions do not
hold and the value of the retained asset decreases, previously
recognized accrued income will not be realized in cash.
INVESTORS OF HLTV LOANS
EXPERIENCED A TRADE-OFF
BETWEEN BENEFITS AND RISKS
---------------------------------------------------------- Letter :6.4
Several classes of securities were created from each pool of HLTV
loans and sold to investors. Different securities classes offered
different mixes of expected returns, credit risk, and interest rate
risk (the risk that cash flows will vary as interest rates change in
the future) to investors. In general, classes with higher risks were
designed to pay higher expected returns. Although it is difficult to
identify investors of HLTV lending pools because of a lack of data,
officials from a securities firm told us that investors were mainly
large insurance companies.
To protect investors, securitization redistributes credit risks so
that one class of securities is designed to absorb all or much of the
credit risk in return for a higher expected return. Such structuring
is a common internal credit enhancement. Another common internal
credit enhancement is over-collateralization, where the difference
between the interest paid on the loans and interest paid to investors
(the "excess spread") absorbs losses. In addition, to improve the
liquidity or marketability of certain securities, lenders can also
provide external credit enhancements to protect the owners of these
securities from losses from loan defaults. These external credit
enhancements can include insurance policies that protect a pool's
securities against default losses on the loans in the pool. Private
sector officials told us that external credit enhancements were a
popular technique in the securitization of HLTV loans in 1995 and
1996 but that during 1997 and 1998 internal credit enhancements were
more prevalent.
Different classes in the securitization would be attractive to
investors taking different approaches to managing interest rate risk
because each class of securities would receive different parts of the
pool's interest and principal payments over different time periods.
For example, investors can manage interest rate risk by buying
securities with specific payoff periods.
Nevertheless, there is always some intrinsic risk in any investment,
and this is increased in the case of securities only recently
introduced in the market. Investment bank and government-sponsored
enterprise officials noted several risks associated with the
securities based on the HLTV loans. While these securities were
designed to mitigate risks to the investors, there were risks
associated with the underlying loans. For instance, the rate at
which defaults are expected to occur on HLTV loans is a crucial
factor in putting together the pools and the resulting securities.
The expected average default rate on HLTV loans, with an average term
of maturity of 25 years, is based on the creditworthiness of the
borrower as predicted by credit scoring models. According to the
leading provider of credit scoring models, the models are designed to
optimize their predictive power over a 2-year period. Although the
models will rank borrowers over longer periods, the separation
between good and bad borrowers will degrade over a longer time
horizon (for instance, 10 years).
Similarly, the likelihood that HLTV loans will be paid before they
are due has only been estimated based on the prepayment rate of other
forms of consumer debt. If the actual rate of prepayment differs
from the predicted rate, the cash flow from the pool of loans will
not match the predicted cash flow that would yield the expected
payments to investors. This, in turn, would change the realized rate
of return on pools of mortgages and on those securities sensitive to
interest rate risk created by prepayments.
REGULATED DEPOSITORY
INSTITUTIONS WERE NOT
HEAVILY INVOLVED IN
ORIGINATING HLTV LOANS
---------------------------------------------------------- Letter :6.5
It does not appear that regulated depository institutions have
experienced many benefits or risks because these institutions did not
originate many HLTV loans or invest in the securitized pools. Their
minimal benefits, however, might have included service fee income
from any HLTV loans that banks originated and sold.
Officials at Fannie Mae and Freddie Mac told us that HLTV lending, at
current levels, did not pose a greater risk to their portfolios of
first mortgages than did the credit card debt HLTV loans generally
refinanced. The risks to the portfolios exist because a high total
debt burden on consumers could increase the risk of default on the
mortgages that comprise the portfolios of these government-sponsored
enterprises. HILA's executive director said, however, that HLTV
loans might reduce the possibility of defaults on first mortgages
because HLTV loans allowed borrowers to lower their monthly payments
on debt. In any event, Fannie Mae and Freddie Mac officials noted
that they, as first mortgage holders, would maintain their equity
position and be the first to obtain the proceeds from the sale of a
house in the event of default.
Moreover, because the data available at the time of our review show
that most institutions making HLTV loans were not regulated
depository institutions and did not have federally insured deposits,
a failure of these institutions was unlikely to impose any direct
costs on the government. This condition could change, however, in
the event that (1) the size of the HLTV market were to increase and
(2) regulated depository institutions were to become substantially
involved in making HLTV loans and carrying them on their books.
Federal banking and thrift regulatory officials have recognized this
possibility, and some have begun efforts to identify and monitor HLTV
lending done by regulated financial institutions. In addition,
officials from the Office of Thrift Supervision told us that they are
clarifying their regulatory guidelines as they affect HLTV lending.
---------------------------------------------------------- Letter :6.6
As agreed with you, unless you publicly release its contents earlier,
we plan no further distribution of this report until 30 days from its
issue date. At that time, we will provide copies to the Ranking
Minority Member of your Subcommittee, the Chairman and Ranking
Minority Member of other congressional committees with jurisdiction
over financial issues, the Chairman of the Board of Governors of the
Federal Reserve System, the Comptroller of the Currency, the Director
of the Office of Thrift Supervision, the Chairman of the Federal
Deposit Insurance Corporation, the Chief Executive Officer of the
Federal National Mortgage Association, the Chief Executive Officer of
the Federal Home Loan Mortgage Corporation, the chief executive
officers of various HLTV lenders, and other interested parties. We
will also make copies available to others upon request.
This report was prepared under the direction of James McDermott,
Assistant Director, Financial Institutions and Markets Issues. Other
major contributors include Becky Kennedy, Evaluator-in-Charge;
Mitchell Rachlis, Senior Economist; and Edwin Lane, Evaluator. If
you have any questions about this report, please call me on (202)
512-8678.
Sincerely yours,
Susan S. Westin
Associate Director, Financial Institutions
and Markets Issues
OBJECTIVES, SCOPE, AND METHODOLOGY
=========================================================== Appendix I
This report was prepared in response to a request from the Chairman,
Senate Subcommittee on Financial Institutions and Regulatory Relief,
Committee on Banking, Housing, and Urban Affairs. Our objectives
were to provide information regarding (1) the characteristics of HLTV
loans; (2) the major organizations that provided HLTV lending; (3)
the volume of HLTV lending in 1995, 1996, and 1997, and the expected
volume in 1998; and (4) the benefits and risks of HLTV lending for
borrowers, lenders, investors, and regulated depository institutions.
To determine the characteristics of HLTV loans, we gathered
background information on various aspects of individual loans by
interviewing officials representing FirstPlus (the institution that
provided about one-third of total HLTV lending from 1995 through
1997), industry associations, a rating agency, and investment banks,
as well as by reviewing publicly available information, including
published articles that reported such characteristics. To identify
the average loan profile and the average borrower profile of HLTV
loans, we obtained data on HLTV loans made by FirstPlus. We
interviewed company officials and reviewed their literature as well
as other published reports. Although we did not independently verify
this--or any--industry data, we corroborated evidence with other
independent sources whenever possible.
To identify the major organizations that provided HLTV lending, we
interviewed numerous officials representing the private sector. We
selected officials to talk to, in part, on the basis of industry
recommendations of knowledgeable people. We also conducted a
literature search and reviewed selected articles that reported on
HLTV lenders and their activities. Moreover, we collected and
reviewed numerous mail solicitations sent to GAO staff from many
different lenders advertising HLTV loans. We compiled a list of the
top 10 HLTV lenders by corroborating information we collected. The
Executive Director of the Home Improvement Lenders Association (HILA)
confirmed that we had identified the top 10 HLTV lenders.
We obtained the industry volume of HLTV lending in 1995, 1996, and
1997, as well as the expected volume in 1998, by interviewing various
officials representing HLTV lending institutions, industry
associations, and investment banks and by reviewing published
information, such as company prospectuses and relevant articles
obtained through our literature searches. We obtained volume
amounts--for the same years--for individual HLTV lenders from the
Executive Director of HILA. For our review, the Executive Director
surveyed individual lenders that were members of HILA and provided us
with the survey responses. In addition, we contacted officials
representing another selected lender to obtain additional volume
data. We did not independently verify this industry data.
To identify the benefits and risks of HLTV lending for borrowers,
lenders, investors, and regulated depository institutions, we
interviewed public sector officials representing federal banking and
thrift regulatory agencies and the Federal Trade Commission, as well
as private sector representatives from First Plus, industry
associations, a rating agency, investment banks, and a consumer
advocacy group. We also reviewed numerous published journal
articles, academic and industry studies, and congressional
testimonies that reported benefits and risks associated with HLTV
lending and investing. In addition, we reviewed published literature
to generally denote the securitization process as well as the common
accounting treatment of profits. Finally, we interviewed public and
private sector officials and reviewed selected federal and state
regulations and laws to gain an understanding of lender protection
laws relevant to HLTV lending.
We conducted our work between January and June 1998 in accordance
with generally accepted government auditing standards. To ensure
that its contents were factually accurate, we provided a draft of
this report to officials at federal banking and thrift regulatory
agencies, the Federal Trade Commission, Fannie Mae, and Freddie Mac.
We also provided a draft of the relevant sections of this report to
the officials representing FirstPlus for their review. In addition,
to ensure that we accurately reported volume data, we contacted
officials representing the other nine institutions that collectively
served most of the HLTV market. We incorporated their technical
comments where appropriate.
*** End of document. ***