Homeownership: FHA's Role in Helping People Obtain Home Mortgages
(Chapter Report, 08/96, GAO/RCED-96-123).

Pursuant to a congressional request, GAO provided information on the
Federal Housing Administration's (FHA) role in helping people to obtain
home mortgages.

GAO found that: (1) the FHA and Department of Veterans Affairs (VA)
programs allow borrowers to make smaller downpayments and accumulate
higher total debt to income ratios than private mortgage insurers (PMI);
(2) FHA programs finance closing costs as a part of the mortgage, insure
loans up to $155,250, and provide full insurance coverage to lenders;
(3) FHA insured 15 percent of the single-family housing market in 1994;
(4) FHA insures low-income homebuyers with incomes no greater than 80
percent of the median income of the metropolitan statistical area; (5)
FHA insures more home purchase mortgages than PMI or VA; (6) two-thirds
of FHA approved loans would not have qualified for PMI; (7) the maximum
loan amount for a FHA single-family home mortgage is the lesser of 95
percent of the median house price or 75 percent of the Federal Home Loan
Mortgage Corporation's (Freddie Mac) loan limit; (8) the federal
government promotes affordable homeownership through several HUD and
other Federal programs; (9) these programs require federal funds and
assist homebuyers in combining their assistance with FHA mortgage
insurance; and (10) FHA programs promote homeownership among home buyers
that are typically underserved by other agencies and PMI.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-96-123
     TITLE:  Homeownership: FHA's Role in Helping People Obtain Home 
             Mortgages
      DATE:  08/96
   SUBJECT:  Mortgage programs
             Mortgage loans
             Government sponsored enterprises
             Mortgage-backed securities
             Federal aid for housing
             Housing programs
             Minorities
             Mortgage protection insurance
             Government guaranteed loans
IDENTIFIER:  FHA Single-Family Mortgage Insurance Program
             VA Home Loan Guaranty Program
             HUD Community Development Block Grant Program
             HUD Home Investment Partnership Program
             HUD Housing Opportunities for People Everywhere Program
             FHLB Affordable Housing Program
             FHLB Community Investment Program
             VA Lender Appraisal Processing Program
             FHA Section 203(b) Program
             Mutual Mortgage Insurance Fund
             RTC Affordable Housing Program
             Dept. of Commerce American Housing Survey
             
******************************************************************
** 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 Housing and Community
Opportunity, Committee on Banking and Financial Services, House of
Representatives

August 1996

HOMEOWNERSHIP - FHA'S ROLE IN
HELPING PEOPLE OBTAIN HOME
MORTGAGES

GAO/RCED-96-123

Homeownership

(385460)


Abbreviations
=============================================================== ABBREV

  AHP - Affordable Housing Program
  AHS - American Housing Survey
  CDBG - Community Development Block Grant
  CIP - Community Investment Program
  CRA - Community Reinvestment Act
  FHA - Federal Housing Administration
  FHFB - Federal Housing Finance Board
  FHLB - Federal Home Loan Bank
  GAO - General Accounting Office
  GSE - government-sponsored enterprise
  HFA - housing finance agency
  HMDA - Home Mortgage Disclosure Act
  HOME - Home Investment Partnership program
  HOPE - Housing Opportunities for People Everywhere
  HOPE3 - Housing Opportunities for People Everywhere
  HUD - Department of Housing and Urban Development
  LTV - loan-to-value
  MBA - Mortgage Bankers' Association
  MCC - mortgage credit certificate
  MICA - Mortgage Insurance Companies of America
  MRB - mortgage revenue bond
  MSA - Metropolitan Statistical Area
  NHSA - Neighborhood Housing Services of America
  NRC - Neighborhood Reinvestment Coalition
  NWO - NeighborWorks organization
  PHA - public housing authority
  PMI - private mortgage insurer
  RHS - Rural Housing Service
  RTC - Resolution Trust Corporation
  VA - Department of Veterans Affairs

Letter
=============================================================== LETTER


B-271628

August 13, 1996

The Honorable Rick Lazio
Chairman
Subcommittee on Housing
 and Community Opportunity
Committee on Banking and
 Financial Services
House of Representatives

Dear Mr.  Chairman: 

This report responds to your request that we review the role of the
Federal Housing Administration (FHA) in providing mortgage credit to
home buyers.  Specifically, the report discusses (1) the terms of the
mortgage insurance offered by FHA, private mortgage insurers, and the
U.S.  Department of Veterans' Affairs; (2) the characteristics of
borrowers of insured mortgages and the overlap between FHA-insured
mortgages and privately insured mortgages; and (3) other methods used
by the federal government to promote affordable homeownership. 

We are sending copies of this report to the Secretaries of Housing
and Urban Development, Agriculture, and Veterans Affairs; the
Chairman of the Federal Housing Finance Board; the Presidents of the
Federal National Mortgage Association and Federal Home Loan Mortgage
Corporation; and the President of the Mortgage Insurance Companies of
America.  We will make copies available to others on request. 

Please call me at (202) 512-7631 if you or your staff have any
questions.  Major contributors to this report are listed in appendix
VII. 

Sincerely yours,

Judy A.  England-Joseph
Director, Housing and Community
 Development Issues


EXECUTIVE SUMMARY
============================================================ Chapter 0


   PURPOSE
---------------------------------------------------------- Chapter 0:1

Many changes have occurred in the single-family housing finance
system since the Department of Housing and Urban Development's (HUD)
Federal Housing Administration (FHA) was established in 1934 to
insure housing loans made by private lenders.  These changes include
the advent of modern private mortgage insurance, the development of a
secondary mortgage market, and the emergence of a number of public-
and private-sector initiatives designed to expand affordable housing
opportunities for home buyers.  Given these developments, recent
debate and legislative proposals have centered on whether there is
still a need for FHA's single-family mortgage insurance program and,
if so, what changes need to be made to the program.  Critics of FHA
contend that other housing finance players, such as the private
mortgage insurers, are filling the need once filled exclusively by
FHA.  Supporters of FHA argue that its single-family program, which
has insured about 24 million home mortgages since its inception,
remains the only way for some families to become homeowners and
should be expanded. 

To obtain more information about the role of FHA's single-family
program in today's housing finance system, the Chairman of the
Subcommittee on Housing and Community Opportunity, House Committee on
Banking and Financial Services, asked GAO to address three questions: 
(1) How do the insurance terms available through FHA's primary
single-family mortgage insurance program compare with private
mortgage insurance and guaranties from the U.S.  Department of
Veterans' Affairs (VA)?  (2) How many of the home buyers in 1994 used
FHA insurance, what are the characteristics of these home buyers, and
how many of them might also qualify for private mortgage insurance? 
(3) What other federal activities promote affordable homeownership? 


   BACKGROUND
---------------------------------------------------------- Chapter 0:2

About 6 million borrowers took out mortgages in 1994 for purchasing
homes and for refinancing existing mortgages, according to
information collected through the Home Mortgage Disclosure Act.\1
About 32 percent of these mortgages were insured.  Lenders usually
require mortgage insurance when a home buyer has a down payment of
less than 20 percent of the value of the home.  In these cases, the
loan-to-value ratio of the mortgage--the mortgage amount as a
percentage of the value of the home--is higher than 80 percent.  Most
lenders require mortgage insurance for these loans because they are
more likely to default than loans with lower loan-to-value ratios. 
If a loan with mortgage insurance defaults, the lender may foreclose
on the loan and collect all or a portion of the losses from the
insurer. 

Virtually all single-family mortgage insurance is provided by private
mortgage insurers, FHA, and VA.  In general, private mortgage
insurers operate standard programs for typical borrowers and special
affordable programs for qualified borrowers who have fewer down
payment funds and need increased underwriting flexibility.\2

FHA provides most of its single-family mortgage insurance through the
Section 203(b) program.  The Section 203(b) program has not required
any federal funds to operate because FHA has collected enough revenue
from insurance premiums and foreclosed property sales to cover claims
and other expenses.  FHA also operates some smaller, specialized
single-family mortgage insurance programs.  A primary goal of FHA's
single-family programs is to assist households who may be underserved
by the private market.  VA provides insurance through its Home Loan
Guaranty Program.\3

FHA, VA, and the private mortgage insurers provide lenders with
guidelines for deciding whether or not a mortgage is eligible for
mortgage insurance.  In addition, the Federal National Mortgage
Association (Fannie Mae) and the Federal Home Loan Mortgage
Corporation (Freddie Mac)--government-sponsored enterprises that
provide a secondary market for many home mortgages--establish their
own guidelines for the loans they will purchase.  A borrower's
ability to repay the mortgage is often evaluated by computing ratios
of the borrower's total debt burden and housing expenses to his/her
income (referred to as "qualifying ratios").  The
"total-debt-to-income ratio" compares all of the borrower's long-term
debt payments, including his/her housing expenses, with his/her
income.  The "housing-expense-to-income ratio" compares the
borrower's expected housing expenses with his/her income. 

GAO was unable to distinguish loans insured through FHA's Section
203(b) program from loans insured through FHA's smaller single-family
mortgage insurance programs in the Home Mortgage Disclosure Act
database and other automated data sources.  Consequently, sections of
this report on FHA's market share, the characteristics of FHA
borrowers, and the borrowers who may have qualified for private
mortgage insurance pertain to all single-family loans insured by FHA. 
Other sections of the report provide separate information on FHA's
Section 203(b) program. 


--------------------
\1 This figure is based on mortgages reported by lenders through the
Home Mortgage Disclosure Act.  However, the number of mortgages
written in 1994 is somewhat higher because the Home Mortgage
Disclosure Act collects information on most but not every mortgage. 

\2 Underwriting is the process of analyzing a borrower's willingness
and ability to repay a loan. 

\3 Although VA actually guarantees mortgages rather than insuring
them, this report uses the term "mortgage insurance" to refer to the
mortgage guaranty provided by VA as well as the mortgage insurance
provided by FHA and the private mortgage insurers.  VA's guaranty is
available only to U.S.  veterans and their families. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

The FHA and VA programs permit borrowers to make smaller down
payments and have higher total-debt-to-income ratios than allowed by
private mortgage insurers.  FHA's program differs from both the
private mortgage insurers' and VA's programs in that it allows
closing costs to be financed in the mortgage, insures loans only up
to a maximum amount of $155,250, and provides nearly full insurance
coverage to lenders. 

FHA is a major participant in the single-family housing market.  Of
the approximately 3.5 million home purchase loans made in 1994, FHA
insured 15 percent.  FHA insured 35 percent of the insured home
purchase loans and fulfilled an even larger role in some specific
market segments, particularly low-income home buyers and
minorities.\4 However, more of the home purchase loans to all
borrowers (and to low-income borrowers and minorities) were uninsured
than were insured by either FHA, the private mortgage insurers, or
VA.  In addition to insuring a major portion of loans made nationwide
in 1994, FHA insured more home purchase mortgages than did the
private mortgage insurers or VA in at least nine states.  While about
a third of the loans FHA insured in 1995 might have qualified for
private mortgage insurance, the other two-thirds probably would not
have qualified, on the basis of the loan-to-value and qualifying
ratios of the loans FHA insured. 

Although a number of other federal programs share FHA's mission to
assist households who may be underserved by the private market, none
reach as many households as FHA.  In addition to FHA and VA, the
federal government promotes affordable homeownership through programs
run by HUD, the Department of Agriculture's Rural Housing Service,
the Federal Home Loan Bank System, state housing finance agencies,
and the Neighborhood Reinvestment Corporation.  Unlike FHA's Section
203(b) program, over half of these other programs require direct
federal funds.  Several of these other programs assist home buyers by
combining their assistance with FHA mortgage insurance.  The federal
government also promotes homeownership among home buyers who might
otherwise be underserved through requirements placed upon Fannie Mae,
Freddie Mac, and certain lenders. 


--------------------
\4 "Low-income" refers to a borrower with an income no greater than
80 percent of the median income in the Metropolitan Statistical Area
where the borrower is located. 


   GAO'S ANALYSIS
---------------------------------------------------------- Chapter 0:4


      MORTGAGE INSURANCE TERMS
      OFFERED BY FHA DIFFER FROM
      THOSE OF PRIVATE MORTGAGE
      INSURERS AND VA
-------------------------------------------------------- Chapter 0:4.1

The mortgage insurance programs of FHA, the private mortgage
insurers, and VA differ in terms of maximum loan-to-value ratios and
mortgage amounts, the financing of closing costs, and the amount that
each will pay lenders to cover the losses associated with foreclosed
loans, according to the guidance prepared by the insurers for
lenders.  While both FHA and VA can insure loans with effective
loan-to-value ratios that exceed 100 percent (due to the financing of
closing costs or other fees), the private mortgage insurers do not
offer insurance for loans with loan-to-value ratios greater than 97
percent.  In addition, FHA insures loans only up to a maximum of
$155,250, while the private mortgage insurers and VA permit insurance
of larger loans.  In connection with settlement costs, FHA allows
borrowers to finance most closing costs, but private mortgage
insurers and VA do not.  However, both FHA and VA allow borrowers to
finance their insurance premiums.  Finally, while FHA protects
lenders against nearly 100 percent of the loss associated with a
foreclosed mortgage, the private mortgage insurers and VA limit their
coverage to a portion of the mortgage balance.  Private mortgage
insurers generally cover only 20 to 35 percent and VA covers only 25
to 50 percent of the mortgage balance, even if a loss exceeds that
amount. 

Although GAO found some variation in the qualifying ratios of FHA,
the private mortgage insurers, and VA, the guidance provided by the
insurers showed few other clear differences in the underwriting
standards for borrowers.  Each of the insurers permits the lenders to
consider compensating factors, such as a large down payment, when a
borrower does not meet the qualifying ratios.  In addition, although
lenders must apply established credit standards, each of the insurers
relies on the individual judgment and interpretation of the lenders
in evaluating the credit history of borrowers. 


      FHA WAS THE PRIMARY INSURER
      FOR LOW-INCOME AND MINORITY
      HOME BUYERS
-------------------------------------------------------- Chapter 0:4.2

GAO's analysis of data available through the Home Mortgage Disclosure
Act and through the Mortgage Insurance Companies of America showed
that FHA insured 15 percent of all home purchase loans made in 1994
and an even larger portion of the loans made to low-income home
buyers and minorities.  FHA served 20 percent of all the low-income
home buyers and 24 percent of all the minority home buyers in 1994. 
Furthermore, mortgages for low-income and minority home buyers
constituted a greater portion of FHA's 1994 business than they did
for private mortgage insurers or VA, as shown in table 1. 



                                Table 1
                
                 Low-Income and Minority Home Buyers in
                  1994, by Type of Mortgage Insurance


                                    Percentage              Percentage
                         Number of          of   Number of          of
                        households   insurer's  households   insurer's
                          served\a  business\b    served\a  business\b
----------------------  ----------  ----------  ----------  ----------
FHA                        220,000          42     147,000          28
Private mortgage           178,000          25     114,000          16
 insurers
VA                          74,000          34      48,000          22
----------------------------------------------------------------------
\a Number of households rounded off to the nearest thousand. 

\b Figures are based on all home purchase loans insured by either
FHA, private mortgage insurers, or VA during 1994. 

Source:  GAO's analysis of data obtained through the Home Mortgage
Disclosure Act and Mortgage Insurance Companies of America. 

In addition, FHA insured more home purchase loans than the private
mortgage insurers combined in at least nine states--Arkansas,
Maryland, Minnesota, Montana, Nevada, North Dakota, Oklahoma,
Tennessee, and Utah--even though private mortgage insurance companies
insured far more home purchase loans overall than FHA in 1994 (about
725,000 and 519,000, respectively). 

Although FHA was a major participant in the housing market in 1994,
58 percent of all home buyers used no mortgage insurance.  In
addition, a similarly large portion of the low-income home buyers and
the minority home buyers in that year used no mortgage insurance (57
percent and 49 percent, respectively). 

On the basis of the loan-to-value and qualifying ratios of their
FHA-insured mortgages and the maximum ratios generally permitted by
private mortgage insurers, about 34 percent of the home buyers who
used FHA insurance in 1995 may have qualified for private insurance
for the loans they received.  Conversely, about 66 percent of the
loans insured by FHA in 1995 would probably not have qualified for
private mortgage insurance.  These loans had loan-to-value ratios
above the private mortgage insurance maximum of 97 percent, had
housing-expense-to-income ratios above the private mortgage insurance
maximum of 33 percent, or had total-debt-to-income ratios above the
private mortgage insurance maximum of 38 percent.  Looking
exclusively at the first-time home buyers who took out FHA loans in
1995, 77 percent would not have met the private mortgage insurers'
loan-to-value and qualifying ratio standards.  An even greater
portion (85 percent) of the low-income FHA borrowers would not have
met the private mortgage insurers' standards. 


      OTHER FEDERAL ACTIVITIES
      PROMOTE AFFORDABLE
      HOMEOWNERSHIP
-------------------------------------------------------- Chapter 0:4.3

The federal government is involved in many other efforts to make
homeownership affordable.  HUD operates three grant programs--the
Community Development Block Grant program, the HOME Investment
Partnership program, and Housing Opportunities for People
Everywhere--that promote affordable homeownership.  The Federal Home
Loan Bank System has its Affordable Housing Program and Community
Investment Program, which provide subsidies, subsidized advances, or
other advances to member institutions to be used to fund affordable
housing projects and loans to home buyers.  The Department of
Agriculture's Rural Housing Service operates a subsidized direct loan
program for low- and very-low-income rural Americans and a guaranteed
loan program for moderate-income rural Americans.  The state housing
finance agencies, through the use of tax-exempt mortgage revenue
bonds, provide financing for affordable homeownership.  The
Neighborhood Reinvestment Corporation, through its network of local
development organizations and its secondary market organization,
promotes affordable homeownership primarily through second mortgages
and home buyer education.  These programs provide assistance in the
form of grants, direct loans, guaranties, interest subsidies, and
other forms. 

There are several important distinctions between FHA's single-family
mortgage insurance programs and these other federal programs.  First,
FHA serves more homeowners than the other programs combined.  In
1995, almost 570,000 homeowners took out insured loans through FHA's
programs.  In addition, at least half of the other programs require
federal funds, while FHA's Section 203(b) program does not. 
Furthermore, the programs are generally targeted at borrowers with
low incomes or at borrowers who are otherwise underserved by the
private market to a greater extent than FHA's program.  FHA's Section
203(b) program is not restricted to low-income or otherwise
underserved borrowers.  In fact, diversifying risk by serving a wide
variety of borrowers may have actually helped the program operate
without federal funds, according to industry officials. 

Several of the other federal programs assist low- and moderate-income
home buyers by combining their assistance with FHA mortgage
insurance.  A substantial portion of the mortgages made through state
housing finance agencies and HUD's Housing Opportunities for People
Everywhere program were insured by FHA in 1994.  Similarly, private
mortgage insurance may also be combined with assistance from federal
housing programs.  For example, one private mortgage insurer that GAO
reviewed provided insurance for mortgages assisted through a
Neighborhood Reinvestment Corporation program. 

The federal government also promotes homeownership by requiring major
housing finance players to address housing finance needs. 
Specifically, Fannie Mae and Freddie Mac have legislatively-set goals
for affordable homeownership.  In addition, banks and thrifts are
encouraged to lend in all areas of the communities they serve,
including low- and moderate-income areas, through the Community
Reinvestment Act.  In addition to these requirements targeted at
improving the affordability of homeownership for underserved
households, the federal government also promotes homeownership for
the entire general public through federal tax provisions, such as the
home mortgage interest deduction. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

GAO is making no specific recommendations because its work was
limited to collecting and presenting factual information about FHA
and the single-family finance system. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:6

GAO provided a draft of this report to HUD, VA, Fannie Mae, Freddie
Mac, and the Mortgage Insurance Companies of America for their review
and comment.  GAO also provided excerpts of the draft report which
pertained to their homeownership activities to the National Council
of State Housing Agencies, the Federal Housing Finance Board, the
Neighborhood Reinvestment Corporation, and the U.S.  Department of
Agriculture's Rural Housing Service.  Three of the agencies--Freddie
Mac, the Federal Housing Finance Board, and the U.S.  Department of
Agriculture's Rural Housing Service--commented that the portions of
the report that discuss their agencies, with their suggested changes,
were accurate.  The remaining six agencies did not comment on the
overall accuracy of the draft report.  However, all nine agencies
provided comments consisting primarily of suggested changes to
technical information provided in the report and updating figures
with more recent information.  GAO incorporated the comments, as
appropriate, throughout the report. 


INTRODUCTION
============================================================ Chapter 1

Purchasing a home is one of the greatest financial undertakings of
most American families.  In 1994, about 3.5 million families and
individuals bought homes.  Another 2.5 million families and
individuals refinanced the mortgages on their existing homes.  A
variety of public- and private-sector institutions are involved in
helping borrowers to obtain the mortgage credit they need to purchase
homes.  These institutions include mortgage insurers, who insure
lenders against all or some losses on home mortgages.  The primary
mortgage insurers are private mortgage insurers (PMI), the Federal
Housing Administration (FHA), and the U.S.  Department of Veterans'
Affairs (VA).\1 \2 Mortgage insurance is required primarily for
borrowers with limited down payment funds. 


--------------------
\1 A small number of mortgages are insured each year through the
Department of Agriculture's Rural Housing Service.  In 1994, the
Rural Housing Service insured less than 1 percent of all home
purchase loans made. 

\2 Although VA actually guarantees mortgages rather than insuring
them, this report uses the term "mortgage insurance" to refer to the
mortgage guaranty provided by VA as well as the mortgage insurance
provided by FHA and the private mortgage insurers. 


   MOST MORTGAGES ARE NOT INSURED
---------------------------------------------------------- Chapter 1:1

Of the approximately 6 million mortgages reported through the Home
Mortgage Disclosure Act (HMDA) for purchasing homes and refinancing
existing mortgages in 1994, most (4 million) were not insured.  In
fact, 55 percent or more of the mortgages originated\3 annually in
1984 through 1994 were uninsured.  Uninsured mortgages reached their
peak level in 1992 at 80 percent of all mortgages originated, as
shown in figure 1.1. 

   Figure 1.1:  Insured and
   Uninsured Mortgages as a
   Percentage of All Mortgages
   Originated, 1984-94

   (See figure in printed
   edition.)

Source:  GAO's analysis of HUD's data. 


--------------------
\3 Mortgage origination is the process by which a lender makes a
mortgage secured by real property. 


   MANY HOME PURCHASE MORTGAGES IN
   1994 WERE INSURED
---------------------------------------------------------- Chapter 1:2

Many home buyers (42 percent) who financed the purchase of a home in
1994 required mortgage insurance.  This percentage is substantially
higher than the percentage of all mortgages taken out for refinancing
in 1994 that were insured (19 percent).  Mortgage insurance is
generally used when a borrower makes a down payment of less than 20
percent of the value of the home (when the mortgage has a
loan-to-value (LTV) ratio greater than 80 percent).  If a borrower
does not repay an insured mortgage loan as agreed, the lender may
acquire the property through foreclosure and file a claim with the
mortgage insurer for all or a portion of its total losses (the unpaid
mortgage balance and interest, along with the costs of foreclosure
and other expenses).  If a borrower does not have mortgage insurance
and fails to repay a mortgage, the lender may acquire the property
through foreclosure and is responsible for the full amount of losses
it incurs. 

While FHA and VA are federal entities, PMIs are privately-owned
companies regulated at the state level.\4

FHA is a government corporation operated within the U.S.  Department
of Housing and Urban Development (HUD).  However, its primary
single-family mortgage insurance program, the Section 203(b) program,
is supported by the Mutual Mortgage Insurance Fund, which requires no
federal funds to operate.  The Mutual Mortgage Insurance Fund is
required by law to meet or endeavor to meet statutory capital ratio
requirements; that is, it must contain sufficient reserves and
funding to cover estimated future losses resulting from the payment
of claims on defaulted mortgages and administrative costs.  Cash
flows into the fund from insurance premiums and from the sale of
foreclosed property.  The cash reserves in the fund have always been
more than enough to cover the expenses incurred.  In 1995, the fund
had a negative credit subsidy of $309 million.  Negative credit
subsidies occur when the present value of estimated cash inflows to
the government exceeds the present value of estimated cash outflows. 
However, if the fund were to deplete its reserves, the U.S.  Treasury
would have to directly cover lenders' claims and administrative
costs.  VA's single-family mortgage guaranty program does require
federal funds each year.  In 1995, this program received a credit
subsidy of $684 million.\5

Many insured single-family mortgages are ultimately sold to investors
through the secondary mortgage market.  In 1995, virtually all FHA-
and VA-insured mortgages were sold to investors with the help of the
Government National Mortgage Association (Ginnie Mae), which, like
FHA, is also a part of HUD.  Ginnie Mae guarantees securities backed
by pools of FHA- and VA-insured mortgages.  Specifically, Ginnie Mae
guarantees that investors in Ginnie Mae securities will receive
timely principal and interest payments.  Most privately insured
mortgages are sold by lenders to the Federal National Mortgage
Association (Fannie Mae) or the Federal Home Loan Mortgage
Corporation (Freddie Mac)--two government-sponsored enterprises (GSE)
that sell securities backed by pools of mortgages to investors and
hold other mortgages as investments.  Like Ginnie Mae, Fannie Mae and
Freddie Mac guarantee investors in their securities that they will
receive their expected principal and interest payments.  Fannie Mae
and Freddie Mac's charters require that loans they purchase with LTV
ratios greater than 80 percent have some form of credit enhancement. 
Private mortgage insurance is the most common enhancement used for
these high LTV loans.  For an insured loan they have purchased,
Fannie Mae and Freddie Mac assume the responsibility for foreclosure
losses, if any, that a lender incurs above the amount of the claim
paid by the PMI. 

Lenders use guidelines provided by the PMIs, FHA, and VA to determine
whether a borrower is eligible for mortgage insurance through any of
the insurers.  These guidelines include maximum allowable LTV and
qualifying ratios.  Although some of the guidelines pertaining to FHA
and VA mortgage insurance are set by the agencies, many are set by
the Congress through legislation.  Similarly, although some of the
requirements for private mortgage insurance originate with the PMIs,
others are set by Fannie Mae and Freddie Mac.  When determining if a
mortgage is eligible for private mortgage insurance, the requirements
set by Fannie Mae and Freddie Mac are considered because many lenders
want to sell insured mortgages through the secondary market.  The
requirements set by Fannie Mae and Freddie Mac include underwriting
standards, insurance coverage requirements, and a maximum loan
amount. 


--------------------
\4 The private mortgage insurance industry consists essentially of
eight companies located across the U.S.:  Amerin Guaranty
Corporation, Commonwealth Mortgage Assurance Company, GE Capital
Mortgage Insurance Corporation, Mortgage Guaranty Insurance
Corporation, PMI Mortgage Insurance Company, Republic Mortgage
Insurance Company, Triad Guaranty Insurance Corporation, and United
Guaranty Corporation. 

\5 About $14 million of this amount is a subsidy for direct loans. 
The remaining $670 million is for loan guaranties. 


   PMIS FOLLOWED FHA AND VA'S LEAD
   IN PROVIDING MORTGAGES WITH
   VERY LOW DOWN PAYMENTS
---------------------------------------------------------- Chapter 1:3

In 1994, PMIs joined FHA and VA in offering insurance for mortgages
with LTV ratios greater than 95 percent by creating special
"affordable" programs.  The affordable programs are for loans with
LTV ratios of up to 97 percent.  Generally, these affordable programs
have more flexible underwriting than the standard program.  Some PMIs
also offer products through very specialized programs, sometimes
administered in conjunction with another player, such as a state
housing finance agency (HFA), with even more flexible terms than
those available through the companies' affordable programs. 


   FHA'S SHARE OF THE INSURED
   MARKET HAS VARIED
---------------------------------------------------------- Chapter 1:4

Between 1984 and 1994 (the latest year for which data are available),
FHA's share of all loans insured each year both for purchasing homes
and refinancing existing mortgages has fluctuated between a low of 18
percent in 1984 and a high of 51 percent in 1987.  PMIs' share during
the same period fluctuated between a low of 29 percent in 1987 to a
high of 69 percent in 1984.  VA's share during this period stayed
between 13 and 20 percent.  The relative market shares of the
mortgage insurers are shown in figure 1.2. 

   Figure 1.2:  All Insured
   Mortgages, by Insurer, 1984-94

   (See figure in printed
   edition.)

Source:  GAO's analysis of HUD's data. 

Between 1986 and 1990, FHA was the largest insurer.  The factors
contributing to FHA's large market share during these years may
include an increase in FHA's maximum loan limit in 1988 and economic
downturns in some areas of the country.  Except for FHA's loan
limit,\6 the terms, such as maximum LTV ratio, under which FHA and VA
mortgage insurance is available do not generally vary across
different geographic locations, according to program guidelines. 
However, PMI companies may change the conditions under which they
will provide new insurance in a particular geographic area to reflect
the increased risk of losses in an area experiencing economic
hardship.  By tightening up the terms of the insurance they would
provide, PMIs may have decreased their share of the market in
economically stressed regions of the country. 

From 1990 through 1992, the share of the market insured by FHA fell. 
This decrease may be a result, in part, of increased premiums for FHA
insurance implemented as a result of the Omnibus Budget
Reconciliation Act of 1990 (P.L.  101-508).  In 1992, the insurance
premium for FHA mortgages decreased, which may have contributed to
the rise in FHA's market share that took place in 1992 through 1994. 
Throughout the period from 1991 through 1994, the PMIs had a greater
share of all insured single-family mortgage originations than FHA or
VA.  In 1994, the PMIs' share of all insured single-family mortgage
originations was 48 percent, FHA's was 35 percent, and VA's was 17
percent. 


--------------------
\6 FHA's loan limit may differ among geographic areas to reflect
differentials in the cost of housing. 


   OTHER FEDERAL HOMEOWNERSHIP
   TOOLS
---------------------------------------------------------- Chapter 1:5

The federal government uses a variety of tools to promote
homeownership, in addition to providing mortgage insurance through
FHA and VA.  As mentioned above, the secondary market institutions
(Ginnie Mae, Fannie Mae, and Freddie Mac) help make capital available
for mortgage lending.\7 The Federal Home Loan Bank (FHLB) system also
helps provide liquidity for lenders.  The federal government insures
deposits and provides advances to the thrifts and savings
institutions that are members of the Federal Home Loan Bank system. 

Federal tax incentives, such as the home mortgage interest deduction
that is available to homeowners of all income levels, also are
designed to encourage homeownership.  Federal programs and
requirements that are designed to make homeownership affordable,
particularly among households who may be underserved by the private
market, are discussed in chapter 4. 


--------------------
\7 Fannie Mae and Freddie Mac are not federal agencies but are
congressionally-chartered, shareholder-owned corporations. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:6

To obtain more information about the role of FHA's single-family
program in today's housing finance system, the Chairman of the
Subcommittee on Housing and Community Opportunity, House Committee on
Banking and Financial Services, asked us to provide information on
(1) the terms of products available through FHA's Section 203(b)
program in comparison with the terms of products available through
the programs of the PMIs and VA; (2) FHA's share of the home purchase
mortgage market, the characteristics of home buyers using FHA
mortgage insurance in comparison with other home buyers, and the
portion of FHA borrowers who met certain qualifying ratios for
private mortgage insurance; and (3) other federal programs and
activities, besides FHA and VA, that promote affordable
homeownership. 

To compare the terms available through FHA's Section 203(b) program
with the terms offered by PMIs and VA, we reviewed FHA's and VA's
program regulations, FHA's mortgagee letters, and the underwriting
guidelines and marketing materials published by the PMIs.  Although
the private mortgage insurance industry is composed of eight
companies, we limited our review to the guidelines of the six
companies that accounted for 97 percent of all new mortgages that
were privately insured in 1994.\8 We also collected information on
the insurance offered by PMIs through interviews with officials from
four PMIs and from the Mortgage Insurance Companies of America
(MICA).  Because we wanted to compare the terms of private mortgage
insurance that are most similar to the terms of FHA mortgage
insurance, we focused on the affordable programs administered by the
PMIs that may be used for loans with LTV ratios as high as 97
percent.  We restricted our review of the terms of FHA single-family
mortgage insurance to the Section 203(b) program because it is FHA's
primary single-family mortgage insurance program. 

We computed FHA's share of the home purchase mortgage market and
compared the home buyers using FHA insurance to other home buyers by
using the following data sources:  (1) data reported through the HMDA
on the volume of mortgages made by lenders in 1994 and the income,
race, and location of borrowers; (2) data reported by MICA on the
volume of mortgages insured by PMIs in 1994 and the income, race, and
location of borrowers; (3) data from the 1993 American Housing Survey
(AHS) conducted by the U.S.  Department of Commerce on the age of
home buyers; (4) data from HUD on the LTV ratio of the loans insured
by FHA in 1994; (5) data from the Monthly Interest Rate Survey
conducted by the Federal Housing Finance Board on the LTV ratios of
conventional loans made in 1994; (6) data from VA on the LTV ratios
of loans insured by VA in 1994; and (7) data from the Mortgage
Bankers' Association (MBA) on first-time home buyers who took out
mortgages in 1994.  We used the data from the American Housing
Survey, FHA, the Federal Housing Finance Board, VA, and the MBA
because the HMDA and MICA data do not include data on the age of
borrowers, LTV ratio, and first-time home buyers.  We used 1993 and
1994 data because these were the most recent years for which data
comparing FHA-insured loans with other types of loans were available. 
We did not attempt to examine trends over time in borrowers'
characteristics because our objective was limited to describing how
FHA's current clientele compares with other home buyers.  We also
limited our examination of borrowers' characteristics to home
purchase mortgages, excluding mortgages taken out to refinance
existing mortgages. 

The data we used that were collected through HMDA do not include
every mortgage made in 1994 for several reasons.  First, not all
lenders are required to report under HMDA.  A depository institution
is required to report if it has an office in a metropolitan area and
its assets are at least $10 million.  A mortgage company is required
to report if it processes 100 or more mortgage applications.  Second,
the HMDA data we used do not include any mortgages made by lenders
who do not lend in a Metropolitan Statistical Area (MSA).  Third, the
data do not include mortgages for non-owner-occupied homes. 
According to a Federal Reserve official, the HMDA data for 1994
include 77 percent of all home purchase mortgage loans. 

Single-family mortgages insured through FHA's Section 203(b) program
are not distinguished from mortgages insured through FHA's other
smaller single-family mortgage insurance programs in the HMDA data. 
However, 60 percent of all single-family loans insured through FHA in
1995 were made through FHA's Section 203(b) program.  Information on
how FHA's 203(b) mortgages compare to mortgages insured through the
other FHA programs is presented in chapter 4. 

The data from MICA that we used pertain to nearly all of the
mortgages insured by PMIs in 1994 through both standard and
affordable programs.  Although the MICA data pertain only to
privately insured loans, the HMDA data consist of FHA-insured loans
and all others.  To compare FHA-insured mortgages to those insured by
PMIs and to uninsured mortgages, in some cases we subtracted the
number of loans reported to MICA from the non-government-insured
loans in the HMDA data to obtain information about those loans that
were uninsured.  To take into account underreporting in the HMDA
data, we reduced the MICA data by 23 percent when comparing the
relative shares of the market for privately insured, FHA-insured, and
uninsured loans. 

We also used AHS data, which were derived from a 1993 survey of
approximately 65,000 households.  The data we used are from the
subset of these households that acquired a mortgage in 1993.  We also
reviewed existing studies conducted by officials at HUD, the Federal
Reserve System, and the MBA. 

We looked at borrowers' income, race, and first-time home buyer
status because these characteristics have been highlighted by studies
showing that certain types of borrowers, such as low-income
borrowers, have difficulty obtaining mortgage credit.  In addition,
we looked at LTV ratios to compare the types of loans insured by FHA
and by the PMIs, given differences in maximum allowable LTV ratios
and the associated risk.  We also looked at the state in which the
loan was insured to identify geographic differences in the use of
mortgage insurance.  We could not compare the claim rates or loss
rates of FHA-insured loans with other loans because data for
privately insured and uninsured loans were unavailable. 

We determined how many of the loans that FHA insured would have met
PMIs' requirements through analysis of FHA home buyers'
characteristics and PMIs' guidelines.  Unlike the analyses described
above for which 1994 and 1993 data were the most recent available, we
were able to use 1995 data for this analysis because it only required
information about FHA-insured mortgages.  We asked FHA program staff
to use their automated data to determine the percentage of 1995 FHA
home buyers who reported on their loan application qualifying ratios
and LTV ratios below the maximum levels generally allowed by PMIs
(mortgages with LTV ratios no greater than 97 percent, total
debt-to-income ratios no greater than 38 percent, and
housing-expense-to-income ratios no greater than 33 percent). 

We obtained most of our information on other ways in which the
federal government helps to provide affordable homeownership
opportunities by reviewing published information and by discussing
program features with program officials at the various departments
and agencies involved.  We did not verify the accuracy or the
completeness of the data provided by program officials.  We also
reviewed program regulations and budget submission information. 

We provided a draft of this report to HUD, VA, Fannie Mae, Freddie
Mac, and the MICA for their review and comment.  We also provided
excerpts from the draft report that pertained to their homeownership
activities to the National Council of State Housing Agencies, the
FHFB, the Neighborhood Reinvestment Corporation, and the U.S. 
Department of Agriculture's Rural Housing Service.  All nine agencies
provided comments.  We incorporated these comments, as appropriate,
throughout the report.  Some specific comments from HUD, MICA, and
the FHFB are discussed at the end of chapters 3 and 4.  Comments from
MICA, the National Council of State Housing Agencies, the FHFB, and
the Neighborhood Reinvestment Corporation are reproduced in
appendixes III-VI.\9

We performed our work from March 1995 through July 1996 in accordance
with generally accepted government auditing standards. 


--------------------
\8 The six companies we looked at are were Commonwealth Mortgage
Assurance Company, GE Capital Mortgage Insurance Corporation,
Mortgage Guaranty Insurance Corporation, PMI Mortgage Insurance
Company, Republic Mortgage Insurance Company, and United Guaranty
Corporation. 

\9 The other five agencies provided their comments orally. 


SINGLE-FAMILY MORTGAGE INSURANCE
TERMS OFFERED BY FHA, PMIS, AND VA
ARE DIFFERENT
============================================================ Chapter 2

The single-family mortgage insurance programs of FHA, PMIs, and VA
protect private lenders against all or some of the losses that might
result from foreclosure.  However, the products offered by these
organizations differ in terms of the (1) maximum mortgage amounts and
LTV ratios allowed; (2) underwriting standards for borrowers, such as
the income-to-expense qualifying ratio requirement; (3) funds
required at loan closing for such items as down payment and closing
costs; and (4) dollar amount or percent of loss that each
organization will pay lenders to cover the losses associated with
foreclosed loans. 

While FHA's maximum loan amount, effective January 1, 1996, is
$155,250, PMIs can insure and VA can guarantee loans that exceed this
amount.  On the other hand, PMIs can insure loans with a maximum LTV
ratio of 97 percent, while both FHA and VA can insure/guarantee loans
with ratios exceeding 100-percent.  In addition, both FHA and VA
generally require borrowers to pay less cash at loan closing than
PMIs require.  Finally, FHA provides lenders with essentially 100
percent protection against losses from foreclosed loans, while both
PMIs and VA protect against a portion of the losses.  While this
chapter does not discuss all of the differences that exist in the
terms offered by the three organizations, the terms described were
cited as the most significant by industry and FHA officials we
interviewed. 

FHA offers a number of specialized insurance programs; however,
because almost 60 percent of its home purchase loans in 1995 were
made under the Section 203(b) program, our analysis of FHA addresses
only this program. 

Similarly, PMIs insure loans under a variety of different programs. 
In general, PMIs insure mortgages using what they refer to as either
standard or affordable housing loans.\1 Affordable insurance programs
differ from standard ones in that they offer more flexible
underwriting guidelines than the PMIs' standard programs in several
areas such as LTV ratios, qualifying ratios, and reserve
requirements.\2 PMIs offer two types of affordable loans:  (1) a
maximum LTV ratio of 95 percent and (2) a maximum LTV ratio of 97
percent.  The underwriting guidelines for the 97-percent LTV ratio
are more restrictive than the 95-percent affordable program for
things such as credit ratios.  To compare the PMI programs that were
most like FHA's Section 203(b) program and VA's loan guaranty
program, we used the published underwriting guidelines for the PMI
affordable loans and limited our review to the insurance terms for
these loans. 

While both FHA's and PMIs' programs can provide insurance to any
borrowers who meet the programs' underwriting guidelines, VA's loan
guaranty program is limited to qualified veterans and their
survivors. 


--------------------
\1 We focus on the six largest PMIs, which together insured about 97
percent of all loans receiving private mortgage insurance in calendar
year 1994. 

\2 Five of the six PMIs had different underwriting guidelines for
their standard and affordable housing initiatives; the sixth had one
set of guidelines that encompassed both types of products. 


   GENERAL LOAN REQUIREMENTS
---------------------------------------------------------- Chapter 2:1

When deciding whether to finance a home with a loan insured by FHA or
a PMI, or guaranteed by VA, the prospective home buyer must consider
both the amount of money needed to purchase the home and the amount
of the down payment or other cash needs required by the lender.  The
three organizations differ in the maximum loan amounts and cash
requirements for closing a loan. 


      MAXIMUM LOAN AMOUNTS
-------------------------------------------------------- Chapter 2:1.1

While FHA is legislatively constrained by the dollar amount of loans
it can insure, PMIs and VA are not.  FHA's maximum loan amount for a
single family home is legislatively set at the lesser of 95 percent
of the median house price in the area or 75 percent of the conforming
loan limit for Freddie Mac.  As of January 1, 1996, FHA's maximum
loan amount in the highest-cost areas was $155,250.\3 Many areas are
not at the highest cost; these areas' maximum loan amounts range from
$78,660 up to the high-cost limit. 

While PMIs can insure loans of any size, they have established loan
limits for loans insured under their affordable housing programs. 
This limit differs depending on the company.  For loans with a 97
percent LTV ratio, four of the PMIs specified $203,150 as their
maximum loan limit when this was the conforming loan limit; one
stated that the limit was $250,000; and the remaining PMI did not
specify a loan limit. 

VA places no limit on the maximum loan that may be guaranteed, except
that the mortgage may not exceed the home's appraised value plus the
VA funding fee,\4 if it is financed.  As a rule, however, lenders
generally limit VA loans to four times the VA guaranty amount.  Since
the maximum VA guaranty is currently legislatively set at $50,750, VA
loans will rarely exceed $203,000.\5


--------------------
\3 In certain high-cost areas such as Hawaii and Alaska, the maximum
FHA loan can be higher. 

\4 For VA loans, a nonrefundable "funding fee," calculated as a
percentage of the original loan amount, is charged to the borrower
when a VA loan is originated. 

\5 The current conforming loan limit is $207,000. 


      MAXIMUM LTV RATIOS
-------------------------------------------------------- Chapter 2:1.2

FHA, PMIs, and VA differ in terms of their maximum allowable LTV
ratios and how they calculate this ratio.  FHA and VA allow higher
LTV ratios than PMIs.  The LTV ratio represents the ratio of the
unpaid principal balance of the loan to the lesser of the appraised
value or the sales price of the property.  LTV ratios are important
because of the direct relationship that exists between the amount of
equity a borrower has in his/her home and the likelihood or risk of
default.\6 The higher the LTV ratio, the less cash a borrower is
required to pay out of his/her own funds.  However, the higher the
LTV ratio, the less cash the borrower will have invested in the home
and the more likely it is that he/she may default on the mortgage
obligations, especially during times of economic hardship.  Thus,
while FHA and VA's higher LTV ratios allow a home buyer to purchase a
higher-priced home with less money, these loans have a greater risk
of defaulting than PMI loans. 

The Omnibus Budget Reconciliation Act of 1990 (P.L.  101-508),
enacted in November 1990, established LTV limits for FHA loans of
98.75 percent if the home value is $50,000 or less, or 97.75 percent
if the home value is in excess of $50,000.  However, because FHA
allows financing of the up-front insurance premium, borrowers can in
effect receive loans with LTV ratios that exceed 100 percent.  The
method of determining the maximum FHA mortgage amount requires two
steps, as shown in table 2.1.  The example assumes a home with a
purchase price of $100,000 and closing costs of $2,300.  We also
assume that the purchase price is equal to or less than the appraised
value of the property. 



                               Table 2.1
                
                  Calculation Example of FHA's Maximum
                            Mortgage Amount

                                           Step 1:
                                  acquisition cost       Step 2: value
------------------------------  ------------------  ------------------
Purchase price\a                          $100,000            $100,000
Closing costs                              2,300\b               N/A\c
Mortgage basis                            $102,300            $100,000
Step 1\d
97 percent first $25,000                    24,250
95 percent the remainder                    73,435
Step 2                                                          97,750
 97.75 percent of value
======================================================================
Total                                      $97,685             $97,750
Mortgage allowed                           $97,685
 (lesser of 1 or 2)
----------------------------------------------------------------------
\a Calculations of maximum mortgage amount are based on the lesser of
the appraised value or purchase price of the property. 

\b In this example, we assume that closing costs of $2,300 are
financed in the mortgage.  In effect, 95 percent of closing costs may
be financed in the final mortgage. 

\c Not applicable. 

\d Calculating the maximum LTV ratio in this manner for loans over
$50,000 insures compliance with the legislative LTV limits. 

In the above example, the lesser of the two amounts, $97,685, becomes
the maximum mortgage allowed.  However, this is not the final
mortgage amount if the borrower also finances the up-front,
2.25-percent insurance premium in the mortgage ($2,198 in the example
above).  The total mortgage in this case is $99,883 ($97,685 plus
$2,198), or an LTV ratio of 99.9 percent of the purchase price. 

PMIs, on the other hand, establish maximum LTV ratios for loans they
insure, which means that any cost above this amount must be paid at
closing.  Essentially, all six of the PMIs allow LTV ratios up to a
maximum of 97 percent.  In general, loans having a maximum LTV of 97
percent fall under the PMIs' affordable housing loan programs.\7 The
PMIs began to insure loans with a 97-percent LTV ratio in 1994. 
Loans with a 95-percent LTV ratio can be made under the affordable
loan guidelines also, as well as under the companies' regular loan
programs. 

For VA loans, the maximum LTV ratio is 100 percent of the lesser of
the amounts shown on the "certificate of reasonable value" issued by
VA, or the "notification of value" issued by the lender if processed
under VA's Lender Appraisal Processing Program, or the selling price,
plus the VA funding fee.  The certificate or notification is issued
in response to an appraisal request for the determination of
reasonable value from a veteran, lender, builder, or owner.  The
certificate or notification is used to notify the requester of the
maximum amount VA will guarantee. 

Table 2.2 illustrates how each entity calculates its LTV ratio and
summarizes the LTV calculations for each of the three entities.  The
example assumes a $100,000 purchase price (appraisal value) and a
30-year fixed-rate loan at 7.5 percent interest. 



                               Table 2.2
                
                 LTV Calculations for FHA, VA, and PMIs

                                                   PMI (95     PMI (97
                                                   percent     percent
                               FHA          VA        LTV)        LTV)
----------------------  ----------  ----------  ----------  ----------
Purchase price            $100,000    $100,000    $100,000    $100,000
Plus allowable closing       2,300         -0-         -0-         -0-
 costs
======================================================================
Total                   $102,300\a    $100,000    $100,000    $100,000
Less down payment         $4,615\b         -0-      $5,000      $3,000
Loan amount before        97,685\c    $100,000      95,000      97,000
 up-front insurance
 premium
Plus up-front              2,198\d       2,000         -0-         -0-
 insurance premium/
 fee
======================================================================
Total mortgage             $99,883    $102,000     $95,000     $97,000
Actual loan-to-value          99.9       102.0          95          97
 ratio, in percent
----------------------------------------------------------------------
\a FHA allows most closing costs to be financed. 

\b In this example, funds needed to close = acquisition cost less
maximum mortgage amount:  $102,300 - $97,685 = $4,615. 

\c See table 2.1 for calculation. 

\d Up-front insurance premium = $97,685 x 2.25% = $2,198. 

As shown in table 2.2, the VA loan has an LTV ratio of 102 percent,
the FHA loan has an LTV ratio of 99.9 percent, and the two PMI loans
have LTV ratios of 95 percent and 97 percent--the lowest of the LTV
ratios.  These results reflect differences between the three
organizations in their maximum allowable LTV ratios as well as their
requirements for down payment and the financing of closing costs and
insurance premiums. 


--------------------
\6 See Mortgage Financing:  FHA Has Achieved Its Home Mortgage
Capital Reserve Targets (GAO-RCED-96-50, Apr.  12, 1996). 

\7 Five of the six PMIs considered loans with a 97-percent LTV ratio
as part of their affordable housing programs and have specific
underwriting guidelines for these loans.  The remaining PMI
considered these loans as part of its standard program. 


   LOAN UNDERWRITING STANDARDS
---------------------------------------------------------- Chapter 2:2

When underwriting mortgage loans, FHA, PMIs, and VA all require that
lenders examine a borrower's ability and willingness to repay the
mortgage debt by examining the borrower's qualifying ratios and
credit history.  Differences exist in the qualifying ratios allowed
among FHA, PMIs, and VA.  However, differences in the written
requirements for credit history examination among insurers were
minimal. 


      QUALIFYING RATIOS
-------------------------------------------------------- Chapter 2:2.1

Both FHA and PMIs use two qualifying ratios to determine whether a
borrower will be able to meet the expenses involved in homeownership. 
The "housing-expense-to-income ratio" examines a borrower's expected
monthly housing expenses as a percentage of his or her monthly
income; and the "total-debt-to-income ratio" looks at a borrower's
expected monthly housing expenses plus long-term debt as a percentage
of his or her monthly income. 

VA's underwriting standards are different in that they use the
total-debt-to-income ratio in combination with an estimate of
adequate monthly "residual income" when determining borrowers'
qualifications for a home loan.  VA defines residual income as gross
monthly income less federal taxes and other monthly expenses.  In
qualifying a borrower, VA's underwriting guidelines establish a
maximum total-debt-payment-to-income ratio and a minimum monthly
residual income requirement.  The monthly debt-payment-to-income
ratio for VA borrowers is set at 41 percent.  To qualify a borrower
under VA's residual income method, housing (including mortgage
payments) and other monthly payments are subtracted from the
borrower's net take-home pay.  Net take-home pay is gross income less
federal income taxes.  The remaining value is the residual monthly
income for family support.  VA provides a table of residual monthly
incomes by region based on the Department of Labor's consumer
expenditure surveys.  VA provides the residual income tables as a
guide to qualify borrowers; however, VA states that these figures
should not automatically trigger approval or rejection of a loan. 

Table 2.3 summarizes the housing-expense-to-income ratios and total
debt-to-income ratios acceptable to the three organizations. 



                               Table 2.3
                
                Comparison of Qualifying Ratios for FHA,
                              PMIs, and VA

                                Housing-expense-    Total-debt-to-
Organizations                   to-income ratio     income ratio
------------------------------  ------------------  ------------------
FHA                             29                  41

PMIs\a                          28 or 33 for loans  36 or 38 for loans
                                with a 97-percent   with a 97-percent
                                LTV ratio           LTV ratio

VA                              None                41
----------------------------------------------------------------------
\a Guidance for three of the PMIs stated qualifying ratios of 28
percent and 36 percent, while the other three stated ratios of 33
percent and 38 percent for fixed-rate, fixed-payment loans with a
maximum LTV ratio of 97 percent.  In addition, two of the PMIs stated
that the total-debt-to-income ratio should not exceed 40 percent. 

Each of the three organizations give examples of compensating
factors, which may allow the borrower to exceed the maximum
qualifying ratios or residual income figures in the case of VA. 
Examples of the compensating factors provided in the various
underwriting guidelines include

  -- a large down payment;

  -- the demonstrated ability of the borrower to devote a greater
     portion of income to housing expense;

  -- substantial cash reserves;

  -- the borrower's net worth is substantial enough to evidence an
     ability to repay the mortgage regardless of income;

  -- evidence of an acceptable credit history or limited credit use;

  -- less-than-maximum mortgage terms;

  -- funds provided by a health, welfare, or community service
     organization for unusual services, house repairs, etc.; and

  -- a decrease in monthly housing expenses. 


      CREDIT STANDARDS
-------------------------------------------------------- Chapter 2:2.2

In addition to the use of qualifying ratios to determine a borrower's
ability to repay the mortgage debt, FHA, PMIs, and VA also require
that a borrower's credit history be evaluated to determine his or her
willingness to handle financial obligations in a timely manner.  For
these organizations, past credit performance serves as the most
useful guide in determining a borrower's attitude toward credit.  A
borrower who has made payments on previous or current obligations in
a timely manner represents reduced risk.  Conversely, if the credit
history, despite adequate income to support obligations, reflects
continuous slow payments and delinquent accounts, the organizations
require that strong offsetting factors should exist for the loan to
be approved. 

The guidelines of FHA, PMIs, and VA are very similar in their
approach and requirements for determining satisfactory credit.  For
all three, it is the overall pattern of credit behavior that must be
examined rather than isolated occurrences of unsatisfactory or slow
payments.  A period of financial difficulty in the past does not
necessarily make the risk unacceptable if a good payment record has
been maintained since.  For any derogatory items found, the PMI or
lender must determine whether the late payments were due to a
disregard for, or an inability to manage financial obligations, or to
factors beyond the control of the borrower.  All three organizations
allow a good deal of judgment and interpretation on the part of the
underwriter in determining the creditworthiness of the prospective
borrower. 

The use of information from national credit reporting agencies is
required by all three organizations.  However, they also allow
lenders to use alternative methods of establishing credit histories
for borrowers who do not have the type of credit history that would
appear on a credit report.  Other types of information that can be
used include histories on the payment of utilities and rent. 

Given the similarity between the three entities' credit standards and
the fact that the standards are applied using judgment and
interpretation, it was not possible, when comparing stated credit
requirements, to determine which, if any, of these entities
requirements are more, or less, stringent than the others.  Such a
determination would require a number of individual case studies to
determine how specific borrowers would be judged when applying for a
loan insured by a PMI versus FHA or VA. 


   FINANCING LOAN CLOSING COSTS
---------------------------------------------------------- Chapter 2:3

All three organizations require prospective home buyers to pay
certain costs at the time of loan closing.  Funds required to close a
loan include down payment, closing costs, and premium/fee charges. 
In addition, five of the six PMIs require the home buyer to have cash
reserves of 1 or 2 months' principal-interest-taxes-insurance after
loan closing if it is a 97-percent LTV ratio loan.  These differences
are important because the amount of cash needed by the borrower at
loan closing, to meet either closing costs or reserve requirements,
represents a major barrier to homeownership for lower-income and
first-time home buyers. 

Table 2.4 shows the money a borrower will need at closing, including
reserves, if needed, to purchase a $100,000 home assuming $2,300 in
closing costs and a minimum down payment.  As can be seen, VA
requires the least amount of cash at closing ($2,300) while a PMI 95
percent LTV loan requires the most ($7,362). 



                               Table 2.4
                
                  Total Cash Required at Closing on a
                             $100,000 Home

                                                   PMI (95     PMI (97
                                                   percent     percent
                               FHA          VA        LTV)        LTV)
----------------------  ----------  ----------  ----------  ----------
Closing cost                $115\a      $2,300      $2,300      $2,300
First month's mortgage          41         -0-          62          73
 insurance premium in
 advance
Down payment                 4,500         -0-       5,000       3,000
======================================================================
Total cash to close\b       $4,656      $2,300      $7,362      $5,373
Plus reserves needed           -0-         -0-         -0-        $751
 in bank
======================================================================
Total cash required         $4,656      $2,300      $7,362      $6,124
----------------------------------------------------------------------
\a Assumes the remainder is financed in the mortgage. 

\b In addition, the borrower will need cash to pay for taxes,
insurance, and the required escrow balance.  Since this amount varies
widely from jurisdiction to jurisdiction and would be the same for
each organization, we did not include an estimate in our example. 

As stated earlier, the down payment needed for an FHA loan depends on
the calculation of the maximum mortgage amount, which in our example
discussed previously is $97,685.  FHA allows the entire down payment
to be a gift.  In addition, up to 95 percent of the closing costs on
an FHA loan can be financed through the mortgage. 

Under their affordable programs, all of the PMIs require a minimum
down payment of 3 percent from a borrower's own funds.  Additional
funds to be used for a larger down payment or for closing costs can
come from a variety of sources, such as gifts or grants from family
members, nonprofit organizations or public agencies; unsecured loans;
or secured loans. 

VA does not require a down payment, and all of the closing costs must
be paid in cash at closing.  However, the borrower can include the VA
funding fee in the mortgage. 


      PREMIUMS/FEES
-------------------------------------------------------- Chapter 2:3.1

FHA, PMIs, and VA all charge the borrower an insurance premium, or
guaranty fee, to cover potential losses on mortgage loans that go
into foreclosure.  These organizations differ in the amount of
premiums they charge to borrowers, the type of premium plans they
offer, and whether or not these costs can be financed in the
mortgage.  FHA charges both a single, up-front premium as well as an
annual premium for all mortgages over 15 years.  The up-front premium
is equal to 2.25 percent of the maximum mortgage allowed.  It also
can be financed as part of the mortgage and is partially refunded if
the loan is paid in full during the first 7 years.  The annual
premium is equal to .5 percent of the outstanding mortgage balance
and is charged for a time period that depends on the LTV ratio of the
loan.\8 If the LTV calculation is less than 90 percent of the
property's assessed value, the annual premium is charged for 11
years.  If the LTV exceeds 90 percent, the premium is charged for a
full 30 years or the full length of the loan, whichever is less. 
FHA's premium schedule was established by the Omnibus Budget
Reconciliation Act of 1990. 

PMIs charge different premiums to individual borrowers on the basis
of the risk posed by that borrower.  Premium rates will differ on the
basis of factors such as the type of mortgage instrument that the
borrower selects (i.e., fixed rate/adjustable rate), the purpose of
the loan (i.e., home purchase/refinance), the LTV ratio, the length
of the loan (30/25/15 years), and the amount of coverage that is
required.  PMIs also offer borrowers several different ways to pay
premiums.  However, PMI company representatives we interviewed stated
that most borrowers choose to pay premiums under a monthly premium
program because it allows them to pay less cash at closing.  Under a
monthly premium plan, only 1 month of the mortgage insurance premium
is due at closing rather than for a year or more, as in other PMI
plans.  Since the PMI insurance premium is not paid up-front, there
is no refund due if the insurance is canceled.  On the other hand,
FHA and VA borrowers pay an up-front premium.  While the FHA mortgage
insurance and VA loan guaranty remain in effect over the life of the
mortgage, PMI mortgage insurance can be canceled if the unpaid
principal balance of the mortgage has been paid down to either 80
percent of the original value of the property or 80 percent of the
current appraised value of the property. 

The amount of the funding fee charged by VA at loan origination
depends on whether the veteran is a first-time or repeat borrower,
the amount of the down payment, and whether or not the borrower is a
reservist.  Currently, for first-time use, the fee for loans with
less than 5 percent down is 2 percent; with at least 5 percent down,
1.5 percent; and with at least 10 percent down, 1.25 percent.  For
eligible reservists, 0.75 percent is added to the above amounts. 
Repeat borrowers must pay a 3-percent fee for loans with less than 5
percent down.  In addition, borrowers must pay the entire funding fee
at loan closing.  However, the fee can be financed as part of the
mortgage.\9


--------------------
\8 For the purpose of determining how long the premium will be paid,
the LTV is computed using the base loan amount excluding the up-front
premium divided by the value of the property excluding closing costs. 

\9 Veterans receiving VA compensation for service-connected
disabilities, or who, but for receipt of retirement pay, would be
entitled to such compensation, and surviving spouses of veterans who
died in service or from service-connected disabilities are exempt
from the funding fee requirement. 


      RESERVES
-------------------------------------------------------- Chapter 2:3.2

Reserves represent the amount of monthly
principal-interest-taxes-insurance that a borrower must have
accumulated in savings at the time of loan closing.  For loans with a
97-percent LTV ratio, the amount of reserves required by the PMIs
differed, depending on the company.  One PMI required 2 months
reserves, four required 1 month, and one did not require any
reserves.  In addition, two of the PMI companies that required 1
month in reserves stated that these reserves could be waived under
certain circumstances, such as if the property had a satisfactory
mechanical and structural inspection and/or a homeowner's warranty. 
For loans with a 95-percent LTV ratio, PMIs generally do not require
any reserves.  FHA does not require a reserve, and VA's guidelines
make no mention of them. 


   INSURANCE COVERAGE
---------------------------------------------------------- Chapter 2:4

FHA, PMIs, and VA differ in the amount of insurance or guaranty they
provide to protect lenders against the losses associated with loans
that go to foreclosure.  Losses generally include the unpaid
principal balance and delinquent interest due on the loan, legal
expenses incurred during foreclosure, the expense of maintaining the
home, and any advances the lender made to pay taxes or insurance. 

While FHA essentially protects against almost 100 percent of the
losses associated with a foreclosed loan, PMIs and VA protect only
against a portion of the loss.  For PMIs, the type and amount of
coverage selected by the lender determine how much the private
mortgage insurer will pay if the borrower defaults and the lender
must foreclose.  Typically, this amount is limited to between 20
percent and 30 percent of the losses but can go as high as 35
percent. 

The amount that the VA guarantees against loss depends on the
original loan amount, is set by law, and has been periodically
increased by the Congress.  Currently, the VA guaranty is as follows: 
(1) for loans up to $45,000, the VA will guarantee 50 percent of the
loan; (2) for loans greater than $45,000, but not more than $56,250,
the guaranty will not exceed $22,500; (3) for loans of more than
$56,250 and not more than $144,000, the guaranty will be the lesser
of 40 percent of the loan or $36,000; and (4) for loans of more than
$144,000, the guaranty is the lesser of 25 percent of the loan or
$50,750. 

Table 2.5 illustrates the amount that FHA, PMIs, and VA would have to
pay for a loan that is foreclosed on at the end of the fourth year of
a mortgage.  As shown in the table, FHA suffers a loss of $44,500 on
the foreclosed loan as opposed to $36,000 for VA and $30,449 to
$31,068 for PMIs.  The differences in losses are primarily related to
the limits on insurance or guaranty coverage provided by each
organization.  Those losses not incurred by the mortgage insurers
become the responsibility of the lenders. 



                               Table 2.5
                
                  Comparison of Foreclosure Losses for
                          Lenders and Insurers

                                                   PMI (95     PMI (97
                                                   percent     percent
                               FHA          VA        LTV)        LTV)
----------------------  ----------  ----------  ----------  ----------
Home purchase price       $100,000    $100,000    $100,000    $100,000
Original loan amount       $99,883    $102,000     $95,000     $97,000
Percent of insurance           100          35          30          30
 coverage

Loan foreclosed at end
of 4th year
----------------------------------------------------------------------
Outstanding loan           $95,749     $97,778     $91,068     $92,985
 amount
Delinquent interest          7,181       7,333       6,830       6,974
 (12 months)
Legal and filing fees        2,500       2,500       2,500       2,500
Property taxes                 600         600         600         600
Hazard insurance               300         300         300         300
Preservation cost\a            200         200         200         200
======================================================================
Total cost                $106,530    $108,711    $101,498    $103,559
Revenues from sale of      $60,000     $60,000     $60,000     $60,000
 property
======================================================================
Total loss paid/         $44,500\b   $36,000\c   $30,449\d   $31,068\d
 incurred by insurer
Net loss to lender/         $2,030     $12,711     $11,049     $12,491
 mortgage holder
----------------------------------------------------------------------
Note:  For this analysis, we assume that the process of acquiring,
maintaining, and disposing of properties is equally efficient among
FHA, VA and PMIs. 

\a This covers the expenses of maintaining the property until it is
resold. 

\b Loss includes everything except what FHA does not allow--one-third
of the administrative costs ($833) and 2 months of delinquent
interest ($1,197). 

\c The actual claim amount that VA pays when a loan is terminated
depends on the total costs to acquire the property minus the amount
that VA determines to be the net value of the property at the time of
foreclosure.  VA computes net value by subtracting a fixed percent
(currently 15.11 percent) from the fair market (appraised) value.  In
this example, we assume that the net value is $60,000, and therefore
VA pays the maximum claim amount--$36,000. 

\d For both PMI examples, the loss is calculated by multiplying the
total cost by 30 percent. 

FHA's additional losses of about $9,000 to $14,000 reflect the
additional risk exposure that FHA assumes from insuring the lender
for close to 100 percent of the loss. 


   OBSERVATIONS
---------------------------------------------------------- Chapter 2:5

PMIs have recently begun to offer a number of specialized or
affordable housing programs with more flexible underwriting
guidelines and higher LTV ratios than their standard mortgage
insurance programs.  However, the terms offered by FHA and VA still
differ in important ways from those offered by PMIs.  These
differences affect the size of mortgage loans that borrowers can
obtain, the amount of cash needed by borrowers at loan closing, and
the exposure to risks assumed by these organizations. 

PMIs can insure and VA can guarantee mortgage loans that exceed those
that can be insured by FHA.  More importantly, however, FHA's and
VA's underwriting standards reduce the amount of cash needed to
purchase a home to a greater extent than the PMIs' new affordable
standards.  While the additional funds needed to purchase a
PMI-insured home would not eliminate the borrower's ability to
purchase a home at some point in time, it can delay the purchase date
substantially or require the borrower to purchase a less costly home. 
Primarily because PMIs and VA both limit their losses to a portion of
the loss and FHA does not, losses on FHA-insured homes that enter
foreclosure are greater than the losses experienced by VA and PMI on
similar homes, all other things being equal.  The additional loss for
FHA reflects the additional risk exposure that it assumes from
insuring the lender for close to 100 percent of the loss.  It should
be stressed, however, that under FHA's Section 203(b) program, FHA
borrowers' premiums pay for these losses, not the U.S.  Treasury. 


FHA HAS A MAJOR ROLE IN THE
HOUSING MARKET
============================================================ Chapter 3

FHA is a major participant in the housing market.  It insured nearly
15 percent of all reported home purchase loans made in 1994 and 35
percent of the insured loans and fulfills a larger role in some
specific market segments, particularly low-income, first-time home
buyers and minorities.  On the basis of the loan-to-value and
qualifying ratios of the FHA loans made in 1995, most of the
FHA-insured loans would probably not have been made by private
mortgage insurance companies.  However, the largest number of home
purchase loans made during 1994, including the loans to low-income
and minority borrowers, were not insured.  More than twice as many
uninsured loans were made to low-income and minority borrowers as
were made by FHA. 

This chapter compares the statistics for FHA, the PMIs, VA, and
uninsured mortgages in the housing market.  It provides information
on FHA's share of loans made to low-income, first-time home buyers
and minority borrowers, as reported in the Home Mortgage Disclosure
Act (HMDA) data.  To the extent that data are available, a comparison
of the characteristics of insured borrowers (FHA, PMIs, and VA) to
uninsured borrowers in the housing market is also presented.  Besides
income, race, and first-time home buyers, the characteristics
discussed include the borrowers' age, the location of the home, and
the LTV ratio of the loan.  This discussion covers all single-family
home purchase loans originated in 1994. 


   FHA INSURED A LARGE NUMBER OF
   ALL HOME PURCHASE LOANS MADE IN
   1994
---------------------------------------------------------- Chapter 3:1

FHA insured 14.7 percent (519,102) of all home purchase loans made in
1994 and included in the HMDA data.\1 HMDA recorded 6.1 million loans
made in 1994, including about 4.1 million loans that were not
insured.  A large portion of the 6.1 million loans made were
refinanced loans (42 percent, or 2.5 million) and are not included in
the analysis in this chapter because we wanted to focus on the
characteristics of home purchase mortgage borrowers. 

Figure 3.1 shows the type of insurance obtained on the 3.5 million
home purchase loans included in the HMDA data.  We estimate that the
PMIs insured 20.5 percent, or 725,188,\2 of these loans made in 1994. 
VA guaranteed another 6.2 percent of the mortgage market.  Rural
Housing Service (RHS) guaranteed less than 1 percent of home loans in
1994.  The largest share of the home purchase loans, which we
estimate to be about 58.5 percent or 2.1 million, was uninsured. 
(Table I.3 in app.  I lists total home purchase loans reported by
HMDA, by type of insurance.)

   Figure 3.1:  Market Share of
   Home Purchase Loans Made in
   1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from HMDA and MICA. 


--------------------
\1 As stated in chapter 1, the HMDA data cover less than the total
number of home mortgages because some institutions are not required
to report and some institutions underreport loans made. 

\2 We reduced the MICA data on all home purchase loans made by PMIs
by 23 percent to reflect a reasonable estimate of the extent to which
the HMDA data are incomplete.  In doing this, we are implicitly
assuming that the share of conventional loans insured by PMIs is the
same for loans included in the HMDA data as it is for those not
included in those data. 


   FHA INSURED ONE-FIFTH OF ALL
   HOME PURCHASE LOANS MADE TO
   LOW-INCOME BORROWERS IN 1994
---------------------------------------------------------- Chapter 3:2

We estimate that FHA insured about one-fifth of the approximately 1.1
million home purchase loans\3 in the 1994 HMDA data made to
low-income borrowers.\4 We estimate that FHA insured more of these
loans (20.1 percent) than the PMIs (16.2 percent) or VA (6.7
percent), as shown in figure 3.2.  FHA's share of home purchase loans
made to low-income borrowers was higher than its 14.7 percent share
of the housing market.  However, more than half of the home purchase
low-income borrowers were uninsured; the percentage of uninsured home
purchase loans made to low-income borrowers is about the same as for
all borrowers.  (A break down of all loans made in 1994 by income
classification is contained in table I.4 of app.  I.)

   Figure 3.2:  Market Share of
   Home Purchase Loans Made to
   Low-Income Borrowers in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from HMDA and MICA. 

Figure 3.3 shows the proportion of each insurer's 1994 home purchase
loans that were made to low-income borrowers.  Forty-two percent of
the home purchase loans that FHA made were to low-income borrowers. 
For PMIs, a smaller percentage (24.5 percent) of the home purchase
loans they insured were for low-income borrowers.  Thirty-four
percent of VA home purchase loans and 30 percent of the uninsured
home purchase loans were made to low-income borrowers. 

   Figure 3.3:  Proportion of Home
   Purchase Loans Made to
   Low-Income Borrowers in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from HMDA and MICA. 

As discussed previously, FHA generally does not insure loans over the
FHA loan limit (maximum amounts were $151,725 for most of 1994 and
$155,250 in 1996 for the areas with the highest housing costs), while
PMIs and VA can insure higher-value loans.  Consequently, it is
useful to know how the PMIs would compare with FHA when facing
similar constraints.  When comparing the home purchase loans that the
PMIs made under the FHA loan limit, the share of the PMIs' business
that is composed of low-income borrowers increases.  According to
data obtained from the Federal Reserve Board,\5 in 1994, 33 percent
of the PMI home purchase loans that were less than the FHA loan limit
were to low-income borrowers.  These data also indicate that 45
percent of FHA's business in 1994 was to low-income borrowers,
similar to the 42 percent we found in the HMDA data. 

Figure 3.4 shows the percent of each insurer's home purchase loans
made at various borrowers' income ranges.  As shown, FHA and VA loans
are more concentrated in the lower income ranges compared with PMI
and uninsured loans.  The Mortgage Bankers Association\6 also reports
that the average income of FHA borrowers for calendar year 1993 was
lower than the average income of the PMIs' borrowers.  (Table I.5 in
app.  I lists the percentage of each group's loans that fall within
the listed income ranges.)

   Figure 3.4:  Proportion of Home
   Purchase Loans, by Income
   Group, in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from HMDA and MICA. 


--------------------
\3 We were able to use only about three quarters of the HMDA data in
assigning income categories, primarily because of missing information
on location and income.  We applied the percentage we found for these
data to the total number of loans in the HMDA data, as well as to our
estimates of loans in each insurance category (FHA, VA, PMI, and
uninsured).  In doing this, we are implicitly assuming that the loans
for which we do not have the data needed to assign an income category
would be divided in the same way among income categories and, for the
conventional loans, between PMIs and uninsured, as the loans for
which we can assign income categories. 

\4 Low-income borrowers are defined as borrowers whose income is less
than or equal to 80 percent of the Metropolitan Statistical Area's
median family income. 

\5 Glenn Canner and Wayne Passmore, "Credit Risk and the Provision of
Mortgages to Lower-Income and Minority Home Buyers", The Federal
Reserve Bulletin, Vol.  81, No.  11, November 1995. 

\6 FHA Single-Family Insurance Program:  A Primer, MBA, Washington,
D.C., 1995.  This primer gives many statistics, historical and
current, on FHA loan characteristics and compares FHA's and PMIs'
insured loans. 


   FHA INSURED NEARLY ONE-FOURTH
   OF ALL LOANS MADE TO MINORITIES
   IN 1994
---------------------------------------------------------- Chapter 3:3

According to GAO's estimate of the HMDA data, about 613,550\7 home
purchase loans were made to minorities in 1994.\8 FHA insured more
loans for minority borrowers in 1994 than the PMIs and substantially
more than VA.  FHA insured about 147,423 minority loans in 1994
compared with about 114,197 insured by PMIs.  Figure 3.5 shows the
relative share of the insured and uninsured minority market. 

   Figure 3.5:  Market Share of
   Home Purchase Loans Made to
   Minorities in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from HMDA and MICA. 

Although the majority (2.9 million, or 83 percent) of all home
purchase loans made in 1994 were to white borrowers, FHA's 1994 loans
consisted of a larger share of minority loans than any other insurer. 
As shown in figure 3.6, 28 percent of FHA loans were to minorities. 
For the PMIs, 16 percent of their loans were to minorities,\9

while of the VA loans made in 1994, 22 percent were to minorities. 
Even though only 15 percent (303,477) of 2.1 million uninsured loans
were made to minority borrowers, 49 percent of all minority loans
made were uninsured.  (Table I.6 in app.  I lists the number of
minority and nonminority home purchase loans made, by insurance
type.)

   Figure 3.6:  Proportion of Home
   Purchase Loans Made to
   Minorities in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of data from HMDA and MICA. 

In addition, the Federal Reserve reported that in 1994, home purchase
loans below the FHA loan limit made by PMIs were about 21 percent
minority, while FHA home purchase loans were about 26 percent
minority.  When considering loans made under the FHA loan limit, FHA
is still more concentrated with minority borrowers than the PMIs. 


--------------------
\7 As with low-income borrowers, the HMDA data were adjusted for
unequal reporting of FHA, VA, and conventional loans.  The HMDA data
were also adjusted to take into account that a higher percentage of
PMI loans did not indicate race.  We made the assumption that the
relative percent of PMI to uninsured loans remained the same in the
nonreporting population as in the reporting population. 

\8 "Minorities" include African Americans, Asians, Hispanics, and
American Indians. 

\9 Of the 941,566 PMI loans reported by the Mortgage Insurance
Corporation of America (MICA), 20 percent (190,635) were classified
as "other" in the data.  In addition to when "other" was indicated on
the application, the term "other" was coded when the information was
not provided by the applicant in a mail or telephone application. 
These data were removed before the calculation of the minority
percentage.  The actual distribution of these loans can affect the
ratios presented for PMIs. 


   FHA INSURED MORE THAN ONE-FIFTH
   OF ALL FIRST-TIME HOME BUYERS
   IN 1994
---------------------------------------------------------- Chapter 3:4

FHA insured a higher percentage of loans for first-time home buyers
than its share of the market in 1994.  However, non-FHA-loan
providers made about four times as many loans (79 percent) as FHA (21
percent) to first-time home buyers in 1994.  According to the
Mortgage Bankers Association (MBA), in 1994, about 4.6 million home
purchase mortgage loans were made in the home mortgage market.\10

Of these loans, FHA insured 15 percent (686,487).  Furthermore, MBA
reported that about 2.2 million loans were made to first-time home
buyers in 1994; FHA insured 21 percent of these loans (see fig. 
3.7).  The MBA report does not distinguish between private insured
and uninsured mortgages.  These are combined with VA loans and make
up the Non-FHA group shown in figure 3.7.  (Table I.7 in app.  I
lists the number of first-time home buyers for FHA and non-FHA home
purchase mortgages.)

   Figure 3.7:  Market Share of
   Home Purchase Loans Made to
   First-Time Home Buyers in 1994

   (See figure in printed
   edition.)

Note:  Non-FHA includes PMIs, uninsured mortgages, VA, and the RHS. 

Source:  GAO's analysis of Mortgage Bankers Association's data. 

Compared with others in the home mortgage market, on average, FHA
made a higher proportion of loans to first-time home buyers.  MBA's
data show that about 67 percent of all 1994 FHA home purchase
borrowers were first-time home buyers and 44 percent of non-FHA home
purchase borrowers were first-time home buyers.  Figure 3.8 shows the
share of FHA and non-FHA loans made to first-time home buyers. 

   Figure 3.8:  Proportion of Home
   Purchase Loans Made to
   First-Time Home Buyers in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of MBA's data. 

In addition to the study by MBA, HUD's Policy Development and
Research Division also analyzed the characteristics of loans.  The
HUD study reported that for the period 1989-91,\11 66 percent of FHA
home purchase borrowers were first-time home buyers, while 56 percent
of PMIs' home purchase borrowers were first-time home buyers.  They
also report that 87 percent of FHA's first-time home buyers are 40
years old or younger. 


--------------------
\10 FHA Single-Family Insurance Program:  A Primer, p.  21.  The
number of mortgage loans made as reported by MBA is different than
the number discussed elsewhere in this chapter because the data
reported to HMDA are only about 77 percent of all home purchase
mortgages issued.  Also, sources of information are different. 

\11 An Analysis of FHA's Single-Family Insurance Program, HUD's
Office of Policy Development and Research, Oct.  1995. 


      FHA BORROWERS ARE YOUNGER
      THAN CONVENTIONAL BORROWERS
-------------------------------------------------------- Chapter 3:4.1

According to GAO's analysis of the 1993 American Housing Survey, FHA
insured 9 percent of home purchase mortgages received in 1993 prior
to the survey.\12 This is lower than its 14.7 percent share of the
home purchase mortgage market in 1994 in the HMDA database.  The age
distribution of borrowers, as indicated in the 1993 AHS data, shows
that FHA borrowers tend to be younger than other borrowers. 
Sixty-two percent of FHA home purchase borrowers are less than 40
years old, while only 38 percent of the conventional home purchase
loans were obtained by borrowers under 40.  According to GAO's
analysis of the AHS data, of the 2.9 million home purchase loans made
in 1993 prior to the survey, 41 percent were made to people under the
age of 40 and 14 percent of them were insured by FHA.  This is a
larger share of this market segment than FHA's share of the entire
market for 1993.  Conventional loans (private insurers and uninsured
groups combined) had 82 percent of the younger-than-40 submarket, and
VA provided the remaining 4 percent.  (Table I.8 in app.  I lists, by
borrowers' age, the number of FHA, VA, and conventional loans made
prior to the 1993 AHS.)


--------------------
\12 The AHS' 1993 national survey of housing (conducted between July
and December 1993) covers a sample of 64,998 homes throughout the
country.  When the sample is weighted, the AHS shows that there were
2.903 million home purchase mortgages obtained in 1993 prior to the
survey.  Of this universe, FHA's share is 9.36 percent and has a
sampling error of 1.6.  In other words, using the AHS data, FHA's
share of home purchase mortgages reported by AHS with 95 percent
likelihood is between 10.96 and 7.76 percent. 


   FHA WAS THE PRIMARY INSURER IN
   AT LEAST NINE STATES
---------------------------------------------------------- Chapter 3:5

FHA's relative share of the insurance market varied from state to
state.  According to the HMDA data for 1994, although PMIs insured
more home purchase loans than FHA, FHA made more home purchase loans
than the six PMIs combined in at least nine states\13 --Arkansas,
Maryland, Minnesota, Montana, Nevada, North Dakota, Oklahoma,
Tennessee, and Utah.  In all except 4 of the 50 states, FHA's share
is between 20 and 50 percent.  In Iowa, Massachusetts, and Wisconsin,
FHA's business was less than 20 percent of the market.  VA's share of
the insured market was the highest only in the state of Alaska. 
Figure 3.9 shows which of the three insurers--FHA, PMIs, or VA--made
the greatest number of home purchase loans in each state during 1994. 
(Table I.9 in app.  I lists FHA's relative share of the insurance
market in each state.)

   Figure 3.9:  Each Insurer's
   Relative Share of the Insured
   Home Purchase Mortgage Market,
   by State, in 1994

   (See figure in printed
   edition.)

   Source:  GAO's analysis of data
   from HMDA and MICA.

   (See figure in printed
   edition.)


--------------------
\13 Within the MICA data, 204,516 of all PMI loans and 72,319 FHA
loans within the HMDA data did not have a state code or contained
edit failures.  When the adjustments were made for under count in the
HMDA data and to account for missing state codes and edit failures,
we estimated that an additional four states may have been primarily
insured by FHA (Arizona, Colorado, Mississippi, and South Dakota). 
Two other states (Louisiana and Virginia) are estimated to have the
same percentage of home purchase loans insured by PMI companies and
FHA. 


   FHA INSURED NEARLY HALF OF ALL
   LOANS WITH A LOAN-TO-VALUE
   RATIO GREATER THAN 90 PERCENT
---------------------------------------------------------- Chapter 3:6

Of all home purchase loans made in 1994 with an LTV ratio of at least
90 percent, FHA insured 43 percent of them.  The PMIs insured 37
percent and VA guaranteed 19 percent.\14 This occurred even though
the PMIs made 49 percent of the insured home purchase loans in 1994,
FHA made about 35 percent, and VA made only 15 percent.  Generally, a
borrower is required to have mortgage insurance if the LTV ratio is
above 80 percent.  The LTV ratios of uninsured loans are generally
below 80 percent.  Figure 3.10 compares FHA, PMI, and VA in the high
LTV ratio market. 

   Figure 3.10:  Market Share of
   Home Purchase Loans Made With a
   Loan-to-Value Ratio of at least
   90 Percent in 1994

   (See figure in printed
   edition.)

Notes:  The VA loans used in this graph are those loans with an LTV
ratio greater than 91 percent, as VA's data are accumulated in this
manner.  We assumed that conventional loans with LTV ratios below 80
percent are uninsured. 

Source:  GAO's analysis of data from FHA, Federal Housing Finance
Board, VA, HMDA, and MICA. 

In addition to the number of loans made with high LTV ratios, there
was also a difference between the proportion of such loans made by
FHA and PMIs.  This difference is demonstrated by figure 3.11, which
shows that 88 percent of FHA's loans had LTV ratios of at least 90
percent, compared with 55 percent of the PMI loans with such LTV
ratios.  In addition, up to 94 percent of VA loans guaranteed in 1994
had an LTV ratio greater than 91 percent. 

   Figure 3.11:  Proportion of
   Home Purchase Loans Made With a
   Loan-to-Value Ratio of at Least
   90 Percent in 1994

   (See figure in printed
   edition.)

Source:  GAO's analysis of FHA's, VA's, and the Federal Housing
Finance Board's data. 

Furthermore, as shown in figure 3.12, 65 percent of FHA-insured loans
had LTV ratios of 95 percent or greater, while only about 8 percent
of PMI loans had LTV ratios greater than 95 percent.  VA, shown
separately in figure 3.13, has the vast majority of its guaranteed
loans concentrated in the greater than 97 percent LTV range.  (Tables
I.10 and I.11 in app.  I list the percentage of FHA, VA, and PMI home
purchase loans within selected income ranges.)

   Figure 3.12:  Proportion of FHA
   and PMI Home Purchase Loans, by
   Loan-to-Value Ratio, in 1994

   (See figure in printed
   edition.)

Note:  GT means greater than, and LE is less than or equal to. 

Source:  GAO's analysis of FHA's and the Federal Housing Finance
Board's data. 

   Figure 3.13:  Proportion of VA
   Home Purchase Loans, by
   Loan-to-Value Ratio, in 1994

   (See figure in printed
   edition.)

Note:  GT means greater than, and LE is less than or equal to. 

Source:  GAO's analysis of VA's data. 


--------------------
\14 Loan to value is defined as the loan amount divided by the
property's appraised value.  FHA adjusts the LTV calculation by (1)
reducing the loan amount by the amount of the mortgage insurance
premium that is financed and (2) including closing costs with the
appraised value. 


   MOST FHA-INSURED HOME PURCHASE
   LOANS WOULD NOT HAVE BEEN
   INSURED BY PRIVATE MORTGAGE
   INSURERS
---------------------------------------------------------- Chapter 3:7

While some FHA-insured home purchase loans might qualify for private
mortgage insurance, most might not have been written under the same
terms by private mortgage insurers.  Specifically, on the basis of
the PMIs' most liberal standards for (1) maximum LTV, (2)
housing-expense-to-income, and (3) total-debt-to-income ratios alone,
about two-thirds of FHA's 1995 home purchase borrowers would not
qualify for private mortgage insurance on the loans they received. 
That is, these borrowers had loans with LTV ratios greater than 97
percent, had housing-expense-to-income ratios greater than 33
percent, or had total-debt-to-income ratios greater than 38 percent. 

Conversely, about one-third of FHA's single-family home purchase
borrowers met the most liberal private mortgage insurance guidelines
for LTV and total-debt-to-income ratios.  These borrowers had loans
with LTV ratios of 97 percent or lower, had ratios of
housing-expense-to-income of 33 percent or lower, and had
total-debt-to-income ratios of 38 percent or lower.  This potential
overlap in FHA-insured borrowers that may qualify for private
mortgage insurance is shown as the shaded area in figure 3.14.  In
addition, relatively fewer FHA first-time home buyers and borrowers
with low incomes met all three of these ratios.  That is, while 34.1
percent of all home purchase FHA borrowers met all three ratios, only
22.6 percent of FHA's first-time home buyers and 14.5 percent of
FHA's low-income home purchase borrowers met all three ratios. 

   Figure 3.14:  Proportion of FHA
   Home Purchase Loans Made in
   1995 That Meet Private
   Insurers' Guidelines for All
   Three Ratios

   (See figure in printed
   edition.)

Source:  GAO's analysis of FHA's data. 

We cannot say with certainty that if an FHA borrower meets all three
of the PMIs' guidelines, a PMI would insure that borrower's mortgage. 
Similarly, with the possible exception of LTV ratio, we cannot say
categorically that an FHA borrower that does not meet any one ratio
would not qualify for private mortgage insurance.  Also, this
analysis does not consider the credit history of a borrower, which
lenders and insurers must consider when underwriting a loan. 
Furthermore, the process of underwriting mortgage insurance requires
some judgment on the part of the lender and insurer, and the
debt-to-income ratios we employ in this analysis may be exceeded if
there are compensating factors. 

Finally, there are other features of FHA and private mortgage
insurance that may influence a borrower's choice of mortgage
insurance.  A borrower may have sufficient financial resources to
qualify for private mortgage insurance and choose FHA insurance
instead so that he or she may invest the funds saved in an asset
other than his or her home.  Therefore, some FHA borrowers whom we
have identified as not being able to qualify for private mortgage
insurance on the loan they received may have been able to increase
their down payment, thereby lowering their LTV and
total-debt-to-income ratios and qualifying for private mortgage
insurance on a smaller loan.  In its October 1995 report,\15 HUD's
Office of Policy and Development found that most of FHA's loans would
not have been insured by PMIs because of differences in LTV or
noncredit factors, even before the companies considered differences
in personal credit history. 

About 68 percent of FHA borrowers in 1995 were within the most
liberal PMIs' guidelines for LTV ratio (had an LTV ratio less than
97).  Conversely, about 32 percent of FHA borrowers in 1995 had LTV
ratios that exceeded the maximum allowable LTV ratio under the PMIs'
most liberal guidelines.  These borrowers with high LTV ratios may
not have qualified for private mortgage insurance for the loans they
received on the basis of their high LTV ratio alone.  Under the PMIs'
recently initiated affordable programs, private mortgage insurers
will insure loans with LTV ratios of up to 97 percent.  Under their
standard programs, private mortgage insurers will insure loans with
LTV ratios of up to 95 percent.  Only 37 percent of FHA borrowers in
1995 had LTV ratios of 95 percent or less, and an additional 31
percent had LTV ratios greater than 95 percent, but not greater than
97 percent.  The share of FHA borrowers that had LTV ratios of 95
percent and below and 97 percent and below are shown by the shaded
area in figure 3.15. 

   Figure 3.15:  Proportion of FHA
   Home Purchase Loans Made in
   1995 That Are Within Private
   Insurers' Guidelines for
   Loan-to-Value Ratio

   (See figure in printed
   edition.)

Source:  GAO's analysis of FHA's data. 

For the ratio of total-debt-to-income, about 60 percent of FHA
borrowers in 1995 could meet the most liberal guidelines established
by private mortgage insurers.  That is, these borrowers had monthly
payments for all debt that was not greater than 38 percent of their
monthly income.  The PMIs' standard programs include guidelines of 36
percent for this ratio.  Almost half of the mortgages insured by FHA
in 1995 would have met this more restrictive ratio.  The share of
FHA-insured loans made in 1995 that went to borrowers with ratios of
total-debt-to-income no greater than the maximums published by the
PMIs is shown as the shaded areas in figure 3.16.  As discussed
previously, those FHA borrowers that did not meet the PMIs' most
liberal guideline for this ratio would not necessarily be precluded
from obtaining private mortgage insurance. 

   Figure 3.16:  Proportion of FHA
   Home Purchase Loans Made in
   1995, by Ratio of
   Total-Debt-to-Income

   (See figure in printed
   edition.)

Source:  GAO's analysis of FHA's data. 

Almost all borrowers that received an FHA-insured mortgage for the
purchase of a house in 1995 had ratios of housing-expense-to-income
that were within the published guidelines of private mortgage
insurers.  That is, under the affordable programs of the PMIs, a
borrower may have a monthly housing debt up to 33 percent of his or
her monthly income.  Over 90 percent of FHA borrowers had housing
debt that was within this guideline.  Under their standard programs,
the PMIs' guidelines generally call for a ratio of
housing-expense-to- income of no more than 28 percent. 
Three-quarters of FHA borrowers in 1995 met this guideline. 

That nearly all FHA borrowers meet the PMIs' guidelines for
housing-expense-to-income is not surprising because the guidelines
established by the PMI companies are nearly the same as or more
liberal than those of FHA for this particular ratio.  The shaded
areas of figure 3.17 show those borrowers that received an
FHA-insured mortgage in 1995 who would meet the ratio for
housing-expense-to-income found in the guidelines of the PMIs.  As
with the ratio for total debt to income, those FHA borrowers who did
not meet the PMIs' most liberal guideline for
housing-expense-to-income would not necessarily be precluded from
obtaining private mortgage insurance. 

   Figure 3.17:  Proportion of FHA
   Home Purchase Loans Made in
   1995, by Ratio of
   Housing-Expense-to-Income

   (See figure in printed
   edition.)

Source:  GAO's analysis of FHA's data. 


--------------------
\15 An Analysis of FHA's Single-Family Insurance Program, pp.  ES-3
and 6-26. 


   OBSERVATIONS
---------------------------------------------------------- Chapter 3:8

FHA is a prominent player in the home mortgage loan
market--particularly in certain market segments.  The loans it
insured in 1994 were concentrated to a greater extent on low-income
and minority borrowers, first-time home buyers, and borrowers with
high LTV ratios than the loans made by the PMIs.  FHA was also the
primary insurer in at least nine states.  In addition, solely on the
basis of the LTV and qualifying ratios of borrowers who obtained
loans in 1995, most FHA borrowers might not have qualified for
private mortgage insurance for the loans they received. 
Consequently, many FHA borrowers in 1995 may not have been able to
obtain or could have been delayed in obtaining a home mortgage
without the more lenient terms offered by FHA. 

While FHA is a prominent participant in the home purchase mortgage
loan market, it is not the major source of loans to home buyers, nor
is it the major source of loans to low-income and minority home
buyers.  The uninsured market, with about three times the number of
loans that FHA had, made about twice as many loans to such borrowers
as FHA did in 1994. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 3:9

An official in HUD's Office of Policy Development and Research
suggested revising the methodology used for some of the analyses
described in this chapter.  In response, we adjusted our analyses of
home purchase loans made in 1994, which used data from HMDA and MICA,
to reflect that the HMDA data pertained to about 77 percent of all
loans insured in 1994, while the MICA data pertained to all privately
insured loans.  We also adjusted our analyses to recognize that some
loan records in the two data sets were missing geographical location
codes and consequently were not being drawn into some analyses. 

In response to comments from an Executive Vice President of MICA, we
added explanations to this chapter about the federal liability
associated with FHA's Section 203(b) program and differences in the
way FHA and private mortgage insurers calculate loan-to-value ratios. 
This official also commented that our report underestimates the (1)
percentage of FHA borrowers who would qualify for a privately insured
loan because compensating factors may enable a borrower to qualify
even if he does not meet the ratios we considered and (2) importance
of the role of Fannie Mae and Freddie Mac because it does not present
those organizations' criteria for purchasing loans.  We disagree with
these two comments.  First, this chapter describes the limitations of
the analyses presented on FHA borrowers who might qualify for private
insurance.  This analysis was not intended to determine with
certainty how many of FHA's borrowers would have qualified for a
privately insured loan.  To make such a determination would require
considering many more factors than loan-to-value and qualifying
ratios.  Rather, as pointed out in this chapter, our analysis is
intended to determine how many of FHA's borrowers might have
qualified for private mortgage insurance on the basis of the ratios
alone.  We point out further in this chapter that we cannot say
definitively that an FHA borrower who does not meet all three private
mortgage insurance guidelines would not qualify for private
insurance.  Similarly, if an FHA borrower meets all three private
mortgage insurance guidelines, it cannot be said categorically that a
private mortgage insurer would insure the borrower's mortgage.  In
connection with the roles of Fannie Mae and Freddie Mac, we point out
in this report that many guidelines pertaining to private mortgage
insurance are set by the two secondary market institutions.  We also
point out that these requirements include underwriting standards,
insurance coverage requirements, and maximum loan amount. 


THE FEDERAL GOVERNMENT PROMOTES
AFFORDABLE HOMEOWNERSHIP IN MANY
WAYS
============================================================ Chapter 4

Besides the FHA Section 203(b) and VA single-family loan programs
described in chapter 2, the federal government promotes affordable
homeownership through a complex web of at least 10 programs, through
the requirements that it places upon the lenders and purchasers of
mortgages and through individual tax incentives.  Although these
tools differ in their scope and technique, the federal government
uses these and other tools to promote homeownership.  In comparison
with FHA's Section 203(b) program, over half of the other 10 programs
require direct federal funds, all reach fewer persons, and they
generally direct a greater proportion of assistance to low-income
home buyers (income less than/equal to 80 percent of an area's median
income).  These programs provide home buyers with grants, direct
loans, guaranties, interest subsidies, and other assistance in
financing a home purchase, and in some instances they rely heavily
upon FHA for mortgage insurance.  These homeownership programs are
run by the Departments of Agriculture, Housing and Urban Development,
and Veterans Affairs; the Federal Home Loan Banks; state housing
agencies; the Neighborhood Reinvestment Corporation, a
government-funded corporation; and Neighborhood Housing Services of
America, a government-funded nonprofit organization.\1

The federal government also promotes affordable homeownership by
placing upon certain lenders and the Federal National Mortgage
Association (Fannie Mae) and the Federal Home Loan Mortgage
Corporation (Freddie Mac) special requirements for meeting housing
finance needs.  Specifically, the Community Reinvestment Act (CRA)
encourages depository institutions and other lenders to meet the
housing credit needs of the communities they serve, and the Federal
Housing Enterprises Financial Safety and Soundness Act of 1992 places
upon Fannie Mae and Freddie Mac numerical goals for the loans they
purchase that are made to low- and moderate-income persons and are
made in underserved areas.\2 Because of the difficulties experienced
in implementing the CRA and the relative newness of the goals set for
Fannie Mae and Freddie Mac, it may be too soon to judge the effect of
these special requirements. 

Finally, through the mortgage interest deduction, one-time exclusion
of capital gains, and other tax provisions, the federal government
provides incentives for individuals to be homeowners.  The Joint
Committee on Taxation estimates that, for 1995, the mortgage interest
deduction alone was the second largest tax expenditure that the
government provides to individuals, totaling an estimated $53.5
billion--exceeding the total tax expenditures given to
corporations.\3

This chapter describes the federal programs that promote affordable
homeownership and the applicable requirements recently placed upon
Fannie Mae, Freddie Mac, banks, and thrifts.  An analysis of the
impact on homeownership of making it affordable through tax
incentives given to individuals is beyond the scope of this study. 
Also, this chapter does not describe efforts that support
homeownership in general, such as those of the secondary market, the
bank system, and the fair lending requirements. 


--------------------
\1 In addition, the Federal Deposit Insurance Corporation has a small
affordable housing program and administers the former Resolution
Trust Corporation's affordable housing program. 

\2 In addition, the FHLBank System operates a Community Support
Program that, among other things, requires FHLBank members to meet
standards of community investment or service in order to maintain
continued access to long-term FHLBank System advances.  Among other
information, a member is required to provide the public disclosure
portion of the member's most recent CRA evaluation and a description
of how the member assists first-time home buyers. 

\3 A "tax expenditure" is a reduction in individual and corporate
income tax liabilities that result from special tax provisions or
regulations that provide benefits to particular taxpayers. 


   DESCRIPTION OF AFFORDABLE
   HOMEOWNERSHIP PROGRAMS
---------------------------------------------------------- Chapter 4:1

In addition to the FHA and VA programs, many federal programs aim at
affordable homeownership.  The state housing finance agencies (HFA),
through the use of tax-exempt mortgage revenue bonds (MRB), may
provide subsidized financing for affordable homeownership.  The
Federal Home Loan Bank (FHLBank) System has its Community Investment
Program (CIP) and Affordable Housing Program (AHP), which provide
subsidies and subsidized or otherwise below-market-rate advances to
member institutions to be used to fund affordable housing projects
and loans to home buyers.  The Department of Agriculture, through the
Rural Housing Service (RHS), operates a subsidized direct loan
program for low- and very-low income rural Americans and a guaranteed
loan program for moderate-income rural Americans.  The Department of
Housing and Urban Development operates three grant programs--the
Community Development Block Grant (CDBG) program, Home Investment
Partnership Program (HOME), and Homeownership and Opportunity for
People Everywhere that promote affordable homeownership.  The
Neighborhood Reinvestment Corporation (NRC), through its network of
local development organizations called NeighborWorksï¿½ organizations
(NWO) and its secondary market organization--Neighborhood Housing
Services of America (NHSA)--promotes affordable homeownership
primarily through second mortgages and home buyer education.  (See
app.  II for detailed descriptions of each program included in our
analysis.)

Even within FHA, there are homeownership programs other than the
Section 203(b) program.  For example, FHA also offers mortgages for
individual condominium units under Section 234(c), rehabilitation
mortgages under the Section 203(k) program, home equity conversion
mortgages under Section 255, and homeownership counseling.  The
Section 203(b) program, however, is FHA's principal means of
promoting affordable single-family homeownership.  In 1995, about 60
percent of all FHA single-family mortgages were made under the
Section 203(b) program.  For the purposes of this chapter, we provide
data for all of the federal programs, including the FHA Section
203(b) program, which we use as a guide for describing the other
programs. 


      TYPE OF ASSISTANCE PROVIDED
      BY FEDERAL PROGRAMS AND
      THEIR RESTRICTIONS
-------------------------------------------------------- Chapter 4:1.1

These programs assist homeowners by providing loans, guarantees,
interest subsidies, help with down payments and closing costs, or
other forms of assistance.  This assistance may go directly to the
homeowner or through an intermediary, such as a local government or
nonprofit organization.  A homeowner may benefit from more than one
program.  For example, HOPE3 funds may be used to help with closing
costs on a loan made by a state HFA.  The state HFA may obtain
funding from MRBs as well as FHLB System advances.  The loan may be
insured by FHA and securitized by the Government National Mortgage
Association (Ginnie Mae). 

Each of the non-FHA/VA homeownership programs include some form of
targeting--typically, the income of the borrower.  In some cases, the
programs also include restrictions on the location of the property,
such as with rural loans, or repayment by the borrower of federal
subsidies.  Only the FHA, VA, and Rural Housing Service single-family
loan programs are restricted by the size of the loans that may be
insured/guaranteed/made.\4 Table 4.1 describes the type of assistance
provided to homeowners and the restrictions imposed by federal
homeownership programs. 



                                              Table 4.1
                               
                               Type of Assistance and Restrictions for
                                 FHA and Other Homeownership Programs


                                    Down                        First-
                           Interes  payment                     time                Recaptur
                           t        closing                     home      Property  e of
Program            Loan    subsidy  help        Other   Income  buyer     location  benefits  Other
-----------------  ------  -------  ----------  ------  ------  --------  --------  --------  ------
FHA 203(b)                                      X\a                                           X\b

FHA                X                            X\a,e                                         X\i
non
203(b)

VA                 X                            X\c                                           X\d,k

State MRB/         X       X                    X\a     X       X                   X         X\j
MCC

FHLBs' CIP         X       X                            X

FHLBs' AHP         X       X        X                   X                           X

NRC's NWOs         X                X           X\e,f                     X

NHSA                                            X\g                       X

CDBG               X       X        X                   X

HOME               X       X        X                   X                           X

HOPE3                     X        X           X\e     X       X                   X

RHS                X       X                    X\c     X                 X                   X\b,h
----------------------------------------------------------------------------------------------------
\a Mortgage insurance. 

\b Maximum mortgage amount. 

\c Mortgage guarantee. 

\d Limited to veterans. 

\e Home buyer counseling. 

\f Lender referral. 

\g Purchase of loans. 

\h For direct loan, applicant must be unable to secure needed credit
from other sources. 

\i Section 255 Home Equity Conversion mortgages are limited to
borrowers that are at least 62 years of age. 

\j Principal residence.  Purchase price not to exceed 90 percent of
the average home price for the area. 

\k Lenders generally limit VA loans to four times the VA guaranty
amount, which is set at $50,750.  As a result, VA loans rarely exceed
$203,000. 

Source:  GAO's analysis of data from HUD, VA, the National Council of
State Housing Agencies, the Federal Housing Finance Board, NRC, and
RHS. 


--------------------
\4 As noted in chapter 2, lenders generally limit VA loans to four
times the VA guaranty amount.  Since the maximum VA guaranty is
currently legislatively set at $50,750, VA loans will rarely exceed
$203,000. 


      FUNDING SOURCES
-------------------------------------------------------- Chapter 4:1.2

Except for the programs of the Federal Home Loan Banks and the state
HFAs--which, like FHA's Section 203(b) loan program, require no
direct federal funds--all of the other homeownership programs use
federal funds.\5 These federal funds are used to pay for the
subsidies and assistance provided and the programs' administration. 
For example, the VA received $684 million in budget authority for
fiscal year 1995 for the subsidy and administrative costs of its
direct and guaranteed loan programs.  For the same year, the Congress
appropriated $50 million for the HOPE programs, of which HUD
allocated $20 million to HOPE3.  It appropriated $1.4 billion for
the HOME program, of which about $238 million was used for
homeownership activities.\6 For the CDBG program, the Congress
appropriated $4.8 billion; 70 percent of this amount is for the
entitlement cities program.  Of this, we estimate that seven-tenths
of 1 percent, or about $24 million, may go toward homeownership
assistance.\7

Even the exceptions listed above are not necessarily without costs to
the federal government.  For example, the AHP and CIP programs of the
FHLBank System are paid for through the system's earnings, and
according to the Finance Board, no FHLBank has ever suffered losses
on its advances.  However, the federal government has paid for
liquidating insolvent member institutions that had benefited from the
use of system advances, and the cost of liquidation may have been
higher where advances permitted a troubled institution to incur
larger losses than it may have otherwise incurred.  Furthermore, the
government's past willingness to assist troubled government-sponsored
enterprises means that it may bear the costs of most of the losses
that such enterprises may suffer in the future.  Also, there is a
cost to the federal government of the state HFAs' mortgage revenue
bond program if one considers the lost revenues resulting from the
tax-exempt status of the securities issued by these organizations to
fund housing activities.  The Joint Committee on Taxation estimates
that the tax expenditure for the tax-exempt mortgage revenue bonds
for owner-occupied housing was $1.4 billion for fiscal year 1995. 

Some of these programs also use nonfederal sources of funds.  For
example the Neighborhood Housing Services of America (NHSA) receives
funding from private-sector, institutional investors through the sale
of secondary market notes backed by loans purchased by NHSA.  Both
the HOME and HOPE3 programs require matching contributions of 25
percent from nonfederal sources.  Through 1995, the Federal Housing
Finance Board (FHFB) has approved 23 state HFAs as nonmember
mortgagees, which would allow them to obtain advances from the
FHLBank System.  The source of funds for each homeownership activity,
including federal funds where appropriate, is shown in table 4.2. 



                               Table 4.2
                
                   Sources of Funds for FHA and Other
                     Federal Homeownership Programs

                      Actual budget
                      authority for
                      fiscal year 1995
                      (dollars in
Program               millions)\          Other sources of funds\a
--------------------  ------------------  ----------------------------
FHA 203(b)            None                Insurance premiums

FHA non-              None                Insurance premiums
203(b)

VA                    $684                Funding fee, loan sales,
                                          principal and interest
                                          payments

State MRBs            None                Tax-exempt mortgage revenue
                                          bonds sold to investors,
                                          FHLBank System advances

FHLBs' CIP            None                FHLBank System advances

FHLBs' AHP            None                FHLBank System advances and
                                          grants, HOME, HOPE 3, CDBG,
                                          state and local government,
                                          grants, foundations

NRC's NWOs            $39                 CDBG, HOME, HOPE, and state
                                          and local government;
                                          private-sector
                                          contributions; NHSA; also,
                                          NWOs may obtain liquidity by
                                          selling loans to NHSA

NHSA                  \b                  Investors, NRC, private-
                                          sector contributions

CDBG                  24\c                Communities may also use
                                          funds from other sources

HOME                  238\d               Requires 25 percent matching
                                          funds from nonfederal
                                          sources

HOPE3                20\e                Requires 25 percent matching
                                          funds from nonfederal
                                          sources

RHS                   315\f               Insurance premiums,
                                          principal and interest
                                          payments
----------------------------------------------------------------------
\a May not include sources such as interest on revolving loan funds
and net proceeds from property disposition activities. 

\b From Neighborhood Reinvestment Corporation funds. 

\c Estimate based on $4.8 billion appropriation, 70 percent for
entitlement cities, seven-tenths of 1 percent for homeownership
assistance.  The share going toward homeownership assistance is based
on the share for program year 1992, in which grantees reported that
seven-tenths of 1 percent of expenditures went to homeownership
assistance ($19.4 million out of $2.6 billion). 

\d Actual fiscal year 1995 budget authority for the entire HOME
program was $1.4 billion.  HUD advises that $238 million was spent on
homeownership. 

\e Based on HUD's allocation of $20 million for the HOPE3 program
from the $50 million appropriation for all HOPE programs. 

\f Does not include administrative expenses associated with
single-family programs because administration costs for single-family
programs are not recorded separately

Source:  GAO's analysis of data from fiscal year 1997 Budget
Appendix, HUD, VA, the National Council of State Housing Agencies,
the Federal Housing Finance Board, NRC, and RHS. 


--------------------
\5 The Section 203(b) program is supported by the Mutual Mortgage
Insurance fund, which requires no federal funds to operate.  In 1995,
the fund had a negative credit subsidy of $309 million.  That is, the
present value of estimated cash inflows to the federal government
exceed the estimated cash outflows.  While the Fund's cash reserves
have historically covered expenses, if the fund were to deplete its
reserves, the U.S.  Treasury would have to directly cover lenders'
claims and administrative costs. 

\6 HOME funds may also be used for rental assistance. 

\7 CDBG funds may be used for a variety of eligible activities,
including homeownership assistance.  We base our estimate of 1995
homeownership funds on the share of expenditures that went to
homeownership assistance for program year 1992. 


      NUMBER OF HOME BUYERS
      ASSISTED AND AMOUNT OF
      ASSISTANCE
-------------------------------------------------------- Chapter 4:1.3

The amount of homeownership assistance provided by other federally
supported programs varied widely between programs in terms of the
dollars involved and the number of homeowners assisted.  In all
instances, FHA assisted a greater number of homeowners.  In total, in
a given year, these other programs may reach over 500,000
homeowners.\8

In 1995, almost 570,000 homeowners received mortgage insurance
through FHA insurance programs.\9 During fiscal year 1995, 263,130
homeowners were assisted through the VA's guaranteed loan program. 
The next greatest number of homeowners assisted was through the state
HFAs, which made over 92,000 loans and issued almost 12,000 mortgage
credit certificates in 1994, totaling over $9 billion.  In contrast,
Neighborhood Housing Services of America purchased 1,133 first and
second mortgages in fiscal year 1995 totaling $47.7 million; and the
HOPE3 program had assisted 1,396 homeowners as of December 1995. 

The characteristics of the home buyers who were assisted were not
always available for each of the programs.  In general, where data
were available, these other programs were more heavily concentrated
in assistance provided to homeowners who were low-income, minority,
and first-time buyers than was the case for FHA, which balances its
risk by insuring a broad range of borrowers and thereby operates
without federal subsidies. 

  -- For example, 30 percent of the borrowers insured under FHA's
     Section 203(b) program had incomes no greater than 80 percent of
     the area's median income.  Sixty-four percent of the homeowners
     assisted through state HFA mortgages and mortgage credit
     certificates and 69 percent of new homeowners assisted by the
     Neighborhood Reinvestment Corporation programs had low incomes. 
     For the entitlement cities part of the Community Development
     Block Grant program and the HOME programs, the percentage of
     homeowners assisted who had low incomes was 94 and 100,
     respectively.  For the Federal Home Loan Banks' Affordable
     Housing Program and the HOPE 3 program, all homeowners assisted
     must have incomes no greater than 80 percent of the area's
     median income. 

  -- In connection with the race of the homeowners assisted through
     these various programs, all of the programs with available data
     served proportionately more minorities than did FHA, with the
     exception of the state and RHS programs.  For example, while
     about 30 percent of the borrowers insured under the FHA Section
     203(b) program were minorities, the state and RHS programs'
     percentages were about 22 and 27, respectively, and the NRC,
     CDBG, HOME, and HOPE3 programs' percentages were 61, 65, 50,
     and 62, respectively. 

  -- There were very few data on the percentage of assisted
     homeowners who were first-time home buyers.  With the exception
     of the VA program, the other programs for which data were
     available reported higher percentages of first-time home buyers. 
     The NRC reported that 97 percent of the homeowners assisted by
     NWOs were first-time home buyers.  For the HOPE3 program, all
     assisted homeowners must be first-time home buyers, and for the
     state MRB/MCC programs, applicants may not have owned a home in
     the last 3 years.  For FHA's Section 203(b) program, about 61
     percent of the borrowers insured were first-time home buyers. 
     Just over half of VA borrowers were first-time home buyers. 
     (See table 4.3)



                                     Table 4.3
                      
                      Number and Characteristics of Homeowners
                         Assisted Through the FHA and Other
                               Homeownership Programs


                      Amount of
                      loans and
                          other
                     assistance
Pr                   homeowners
og      Number of      received
ra     homeowners   (dollars in                                First-time
m        assisted   millions)\a  Low income\b      Minority         buyer      Year
--  -------------  ------------  ------------  ------------  ------------  --------
FH        334,079     $25,950.3          29.9          29.4          61.3     CY 95
 A
 2
 0
 3
 (
 b
 )
FH        234,766      19,118.9          42.9          30.5          65.5     CY 95
 A
 n
 o
 n
 2
 0
 3
 (
 b
 )
VA        263,130      25,340.9            NA            NA          51.1     FY 95
St        104,520       9,168.8          64.3          21.9         100\c     CY 94
 a
 t
 e
 M
 R
 B
 /
 M
 C
 C
FH         82,907       3,800.0          NA\d            NA            NA     CY 95
 L
 B
 s
 '
 C
 I
 P
FH         10,241          48.0         100\c            NA            NA     CY 95
 L
 B
 s
 '
 A
 H
 P
NR      7,184\e,f         295.1          69.0          61.0          97.0   FY 95\g
 C
 '
 s
 N
 W
 O
 s
NH        1,133\e          47.7            \h            \h            \h     FY 95
 SA
CD     18,325\e,i          19.4          93.9          65.1            NA   PY 92\j
 BG
HO         18,898         237.9           100          50.2            NA     FY 95
 ME
HO          1,396          23.0         100\c          62.0         100\c     CY 95
 PE
 3
RH       32,082\k       1,982.7          61.8          26.6            NA     FY 95
 S
-----------------------------------------------------------------------------------
Notes:  NA - not available; CY - calendar year; FY - fiscal year; PY
- program year. 

\a Amounts here may differ from those reported in the fiscal year
1997 Budget Appendix. 

\b Income not exceeding 80 percent of the area's median. 

\c Based on program requirement. 

\d Borrowers assisted under the CIP program may not have income
exceeding 115 percent of the area's median. 

\e Some data on number of homeowners assisted may over-count where a
homeowner receives more than one type of assistance. 

\f Includes 2,854 rehabilitation loans and 4,330 first mortgages. 
Over half of first mortgages would also have a rehabilitation loan
(averaging between $7,000 to $8,000). 

\g Data on homeowners' characteristics are for the 36-month period
ending December 1995. 

\h Characteristics should be similar to those for NRC/NWOs, whose
loans NHSA purchases. 

\i Includes 8,000 persons and 10,325 households that obtained
homeownership assistance directly from the grantee and/or indirectly
from a subrecipient of the grantee. 

\j Program year. 

\k Includes 15,405 direct loans and 16,677 guaranteed loans. 

Source:  GAO's analysis of data from HUD, VA, the National Council of
State Housing Agencies, the Federal Housing Finance Board, NRC, and
RHS. 


--------------------
\8 It is difficult to estimate the total number of homeowners
assisted because individuals may benefit from more than one program. 
The number presented here is overstated by an unknown amount. 

\9 Includes refinancings.  However, over 90 percent of the
single-family mortgages insured by FHA in 1995 were home purchase
mortgages.  In this chapter, we report total loans. 


      USE OF MORTGAGE INSURANCE
-------------------------------------------------------- Chapter 4:1.4

The extent to which each of the non-FHA programs utilizes mortgage
insurance is not completely known.  Three programs provide mortgage
insurance or a similar enhancement:  VA and the Rural Housing Service
guarantee mortgages, and seven state HFAs self-insure mortgages. 
Loans made by state HFAs are almost always insured, mostly by FHA. 
In 1994, FHA insured over 55 percent of the loans made by state HFAs. 
VA accounted for over 8 percent.  While there were no data on the
extent to which homeowners assisted through the FHLBs' AHP and CIP
programs had mortgage insurance, because member institutions may keep
loans in their portfolio, they may not require mortgage insurance on
these loans.  The use of mortgage insurance on individual loans made
by NWOs was not known, but both GE Capital Mortgage Corporation and
Mortgage Guaranty Insurance Corporation provide mortgage insurance on
special loan products offered through the NeighborWorksï¿½ Campaign for
Home Ownership that allow for higher LTV ratios.  Furthermore, the
PMI Mortgage Insurance Company, provides pool insurance for first
loans purchased from the Neighborhood Housing Services of America by
the World Savings and Loan Association.  The use of mortgage
insurance in relation to two of HUD's grant programs--CDBG and
HOME--is not known.  For the HOPE3 program, about 19 percent of home
buyers financed their home purchases using FHA insurance. 


   DESCRIPTION OF REQUIREMENTS
   PLACED ON MAJOR HOUSING FINANCE
   PARTICIPANTS
---------------------------------------------------------- Chapter 4:2

The Federal government also promotes affordable homeownership through
requirements that it places upon lenders and purchasers of mortgages. 
Specifically, the Community Reinvestment Act encourages certain
lenders to meet the credit needs of the areas that they serve,
including low- and moderate-income areas; and the Federal Housing
Enterprises Financial Safety and Soundness Act of 1992 contains
provisions that require Fannie Mae and Freddie Mac to meet certain
goals related to the purchase of mortgages made to low- and
moderate-income borrowers and in areas of low- and moderate-income. 
Both the lenders and Fannie Mae and Freddie Mac have taken actions to
better meet the credit needs of low- and moderate-income home
purchasers.  However, given the difficulties experienced in
implementing the CRA, and the relative newness of the social goals,
it may be too soon to judge the impact these requirements will have
on affordable homeownership. 


      THE CRA'S REQUIREMENTS
-------------------------------------------------------- Chapter 4:2.1

The Congress in 1977 enacted the CRA to encourage banks to provide
credit to their entire market areas, including low- and
moderate-income areas.  The CRA requires federal bank and thrift
regulators--the Federal Reserve Board, the Office of the Comptroller
of the Currency, the Federal Deposit Insurance Corporation, and the
Office of Thrift Supervision--to evaluate during periodic
examinations the extent to which banks are fulfilling their lending,
investment, and service responsibilities to their areas.  In
connection with lending, the regulator evaluates a bank's record of
helping to meet the credit needs of its area through its lending
activities by looking at such indicators as the geographic
distribution of the bank's loans, including the incomes of areas and
borrowers and the extent to which the bank uses innovative or
flexible lending practices.  On the basis of the results of these
assessments, the regulators assign the banks one of four overall CRA
ratings, ranging from "outstanding" to "substantial noncompliance."
The CRA is limited to depository institutions, such as banks and
thrifts.  These institutions originated about 46 percent of all home
mortgages made in 1994.  Mortgage companies--the primary providers of
mortgages for single-family homes in 1994--are not subject to the
CRA. 

An institution's CRA rating may affect approval by the regulators of
certain types of applications, an institution's access to FHLB
advances, and the public's perception of the institution.  The
regulators are required to take a depository institution's CRA rating
into account when considering applications for expansions, such as
mergers and acquisitions.  In addition, a FHLBank System member's
access to the long-term advances used to finance residential mortgage
lending is tied, in part, to its CRA rating.  An institution's CRA
rating and related information must also be available to the public
for review.  Finally, the CRA affords community groups or other
members of the public the opportunity to protest an institution's
application for establishing a deposit facility. 

The CRA and the fair lending laws have related objectives.  The
primary purpose of the CRA was to prohibit "redlining"--arbitrarily
failing to provide credit to low- and moderate-income neighborhoods. 
The Fair Housing Act and the Equal Credit Opportunity Act prohibit
lending discrimination that is based on certain characteristics of
the potential and actual borrowers.  In addition, the Home Mortgage
Disclosure Act (HMDA) provides regulators and the public with
information on mortgage applications.  In November 1995, we issued a
report analyzing the implementation of the CRA.\10


--------------------
\10 Community Reinvestment Act:  Challenges Remain to Successfully
Implement CRA (GAO/GGD-96-23, Nov.  28, 1995). 


      IMPACT OF THE CRA
-------------------------------------------------------- Chapter 4:2.2

Because of difficulties in implementing the CRA and the relative
newness of reforms intended to address these difficulties, it may be
too soon to judge the impact of the CRA.  Yet even with the
disagreements over the implementation of the CRA, bank and thrift
regulators report some actions taken to better meet the needs of
underserved communities.  In connection with difficulties in
implementing the CRA, we reported in November 1995 that because of
the concerns of those lenders subject to the CRA about the burden it
presents and the concern of community groups about the enforcement of
the CRA, the regulators responsible for enforcing the CRA undertook a
series of public hearings in 1993 and revised the regulations for the
CRA in May 1995.  We reported that some of the difficulties that have
hindered past efforts to implement the CRA--differences in examiners'
training and experience, insufficient information to assess
institutions' CRA performance, and insufficient time for examiners to
complete their responsibilities--will likely continue to challenge
the regulators as they implement the revised regulations. 

According to bank and thrift regulators, despite the difficulties in
implementing the CRA, it has played an increasingly important role in
improving access to credit in communities, and many banks and
thrifts, under the impetus of the CRA, have opened new branches,
provided expanded services, and made substantial commitments to
increase lending to all qualified borrowers within their areas.  As
we reported in November, some bankers may lower the relatively high
transaction costs and perceived credit risks to individual
institutions of community reinvestment loans by sharing those costs
and risks through multi-institution programs.  Regulators found that
bankers who had effective CRA performance had undertaken initiatives
such as borrower education and counseling, community outreach
efforts, flexible underwriting standards or policies, and
participation in government-sponsored lending programs.  In addition,
some major participants in the secondary markets have recently
undertaken initiatives intended to make them more responsive to
community development concerns, as discussed in the following
section. 


   GOVERNMENT-SPONSORED
   ENTERPRISES HAVE REQUIRED
   SOCIAL GOALS
---------------------------------------------------------- Chapter 4:3

The secondary mortgage market is the market in which mortgages and
mortgage-backed securities are bought and sold.  The Federal National
Mortgage Association and the Federal Home Loan Mortgage Corporation
are government-sponsored enterprises that operate a secondary
mortgage market in which they purchase mortgages from lenders in
exchange for cash or mortgage-backed securities/participation
certificates. 

The mortgages that Fannie Mae and Freddie Mac may purchase are
limited to those in amounts less than a legislative limit known as
the conforming loan limit.  This limit is adjusted on the basis of a
formula; for 1996, the limit is $207,000 for single-unit,
single-family residences.  In addition, the GSEs are restricted to
purchasing and securitizing only residential mortgages, are obligated
to be active in the secondary market across the country at all times,
and must comply with capital requirements and safety and soundness
regulations issued by the Office of Federal Housing Enterprise
Oversight.  The GSEs face these restrictions on their activities for
the benefits of their federal charter.  An important indirect benefit
is that investors perceive an implied federal guarantee on their
obligations, which allows Fannie Mae and Freddie Mac to borrow at
near-Treasury rates.\11 Direct benefits include (1) $2.25 billion in
conditional lines of credit with the Department of the Treasury, (2)
exemptions from state and local corporate income taxes, and (3)
exemptions from the Securities and Exchange Commission's registration
requirements for their securities.  In addition to Fannie Mae and
Freddie Mac, the secondary mortgage market is served by the
Government National Mortgage Association (limited to securitizing
federal government insured/guaranteed loans) and private conduits
(private companies that purchase mortgages and sell mortgage-backed
securities).  In the first quarter of 1995, about 48 percent of
outstanding single-family mortgage debt was held in mortgage pools. 
Fannie Mae and Freddie Mac accounted for about 62 percent this debt,
Ginnie Mae for 27 percent, and private conduits for about 11 percent. 


--------------------
\11 FNMA and FHLMC:  Benefits Derived From Federal Ties: 
(GAO/GGD-96-98R, Mar.  25, 1996). 


      A DESCRIPTION OF THE
      REQUIREMENTS
-------------------------------------------------------- Chapter 4:3.1

The Congress requires Fannie Mae and Freddie Mac to support mortgage
lending for low- and moderate-income persons and for residents of
areas where home loans may be difficult to obtain.  Their charters
charge the GSEs with providing ongoing assistance to the secondary
market for home mortgages--including the market for mortgages for
low- and moderate-income families.  More recently, the Federal
Housing Enterprises Financial Safety and Soundness Act of 1992
required the Secretary of HUD to establish housing goals for Fannie
Mae's and Freddie Mac's purchases of mortgages for low- and
moderate-income families; housing located in central cities, rural
areas, and other underserved areas; and special affordable housing
meeting the unaddressed housing needs of targeted families.\12 The
act established interim annual goals for the 2-year period beginning
on January 1, 1993.  These annual goals were that (1) 30 percent of
the total number of dwelling units financed by the mortgage purchases
of the enterprise shall be for low- and moderate-income families;\13
(2) 30 percent of the total number of dwelling units financed by the
mortgage purchases of the enterprise shall be mortgages on properties
located in central cities; and (3) the mortgage purchases of Fannie
Mae shall include not less than $2 billion ($1.5 billion for Freddie
Mac) in "special affordable" mortgages, split evenly between
mortgages on single-family and multifamily housing.\14

In October 1993, HUD published interim goals for the GSEs, setting
the low- and moderate-income goal for Fannie Mae at 30 percent for
1993 and 1994.  The goal for Freddie Mac was 28 percent for 1993 and
30 percent for 1994.  HUD set the central cities goal for 1993 at 28
percent for Fannie Mae and 26 percent for Freddie Mac.  Both had a
goal of 30 percent for 1994.  The goals for 1995 were kept at the
level for 1994.  The goals for 1993, 1994, and 1995 are shown in
table 4.4. 



                               Table 4.4
                
                     Fannie Mae's and Freddie Mac's
                Affordable Housing Goals for 1993, 1994,
                                and 1995


Year                          Goal      Actual        Goal      Actual
----------------------  ----------  ----------  ----------  ----------
Low-and moderate-income goals (percent of dwelling units financed)
----------------------------------------------------------------------
1993                            30          36          28          29
1994                            30          46          30          38
1995                            30          46          30          39

Central cities goals (percent of dwelling units financed)
----------------------------------------------------------------------
1993                            28          26          26          24
1994                            30          32          30          25
1995                            30          30          30          23

Special affordable housing goals (dollars in billions)
----------------------------------------------------------------------
1993-94                       12.4        16.7        11.1        12.2
 (single-family)
1993-94                        3.7         4.5         0.8         0.5
 (multifamily)
1995                           3.4         6.2         3.0         4.4
 (single-family)
1995                           1.2         2.2         0.4         1.1
 (multifamily)
----------------------------------------------------------------------
Source:  HUD, Fannie Mae, Freddie Mac. 

In February 1995, HUD proposed goals to increase Fannie Mae's and
Freddie Mac's affordable housing purchase requirements.  HUD issued
the final regulations in December 1995, specifying, among other
things, the goals for 1996.  The goals for 1996 increased to 40
percent the portion of dwelling units for low- and moderate-income
borrowers.  The regulations set at 21 percent the central cities
housing goal for 1996 and expanded the areas to be included in this
goal to include rural and other underserved areas along with central
cities.  The special affordable housing goal for 1996 required that
12 percent of the total number of dwelling units financed by each
GSE's mortgage purchases are to be in mortgages for low-income
families in low-income areas and very-low income families.\15 Table
4.5 shows the affordable housing goals for 1996. 



                               Table 4.5
                
                     Fannie Mae's and Freddie Mac's
                   Affordable Housing Goals for 1996

Fannie Mae                                                 Freddie Mac
----------------------------------------  ----------------------------
Low-and moderate-income goals (percent of dwelling units financed)
----------------------------------------------------------------------
40                                                                  40

Central city, rural, and other underserved areas (percent of dwell
units financed)
----------------------------------------------------------------------
21                                                                  21

Special affordable housing goals (percent of dwell units financed)
----------------------------------------------------------------------
12                                                                  12
----------------------------------------------------------------------

--------------------
\12 Although the authority to establish goals previously existed
under the Charter Act and was implemented under regulations, the 1992
legislation defined and expanded this authority and defines the goals
in terms of borrower/renter income rather than house price. 

\13 In general, low-income is defined as family income not in excess
of 80 percent of an area's median family income.  Moderate income is
defined as family income not in excess of the area's median income. 
In the case of rental units, the Secretary may adjust median income
for smaller and larger families. 

\14 For single-family mortgages to be counted toward the special
affordable target, they must be made to (1) low-income families who
live in census tracts in which the median incomes do not exceed 80
percent of the areas' median incomes or (2) very-low-income families
without regard to where they live.  For multifamily mortgages to be
counted toward the special affordable target, they must meet similar
tests. 

\15 The goal includes mortgage purchases financing dwelling units in
multifamily housing totaling not less than 0.8 percent of the dollar
volume of mortgages purchased by the respective GSE in 1994. 


      IMPACT OF SOCIAL GOALS
-------------------------------------------------------- Chapter 4:3.2

In terms of what loans the GSEs purchase, an increasing proportion
were made to persons in targeted income groups and locations during
the first 2 years of the social goals.  However, it may be too soon
to judge the impact that the social goals ultimately may have. 
According to HUD's data, the GSEs purchased a greater proportion of
loans made to low- and moderate-income persons in 1994 than they did
in 1993--up 10 percentage points for Fannie Mae and 9 percentage
points for Freddie Mac.  The same is true for loans made in central
cities--up 6 percentage points for Fannie Mae and 1 percentage point
for Freddie Mac.  HUD further reports that the increases made in
these goals appear to have been made without significant adverse
impact on the GSEs' financial condition.\16 With the exception of the
1993 goal for central cities, Fannie Mae has exceeded its goals for
1993 and 1994.  Freddie Mac was unable to meet the central cities
goal for both years and was unable to meet the special affordable
housing goal for multifamily housing for the period 1993 through
1994.  For 1995, Fannie Mae met or exceeded each of its housing
goals, and Freddie Mac exceeded the low- and moderate-income and
special affordable housing goals but did not meet the goal for loans
in central cities. 

In recent years, both GSEs have undertaken efforts to make more
flexible their underwriting guidelines and develop new loan products
that require less cash to obtain a home.  For example, Fannie Mae's
Community Home Buyer's Program allows borrowers to make a down
payment of 5 percent from their own funds and to qualify with housing
expense and total debt ratios of 33/38 (or higher with compensating
factors).  Fannie Mae recently added the Fannie 97 mortgage product
to its community lending product line.  Borrowers need only a 3
percent down payment from their own funds; family members, nonprofit
groups, or government agencies are eligible to pay the closing costs. 
For a 30-year term, the qualifying ratios for a Fannie 97 mortgage
product are the same as for the standard product--28/36.  Freddie
Mac's Affordable Gold program provides for 95-percent LTV ratio loans
with what is called a 3/2 option.  Under this option, borrowers need
only 3 percent of the value of the loan from their funds, with the
remaining 2 percent from a gift, a grant, or an unsecured loan.  In
connection with qualifying ratios, Freddie Mac's affordable program
has no maximum housing expense ratio, and the total debt ratio is 38
to 40.  Both GSEs require home buyer counseling for certain
affordable products. 


--------------------
\16 Annual Report to Congress, 1995, Office of Federal Housing
Enterprise Oversight, June 15, 1995. 


      RELATIONSHIP TO MORTGAGE
      INSURANCE
-------------------------------------------------------- Chapter 4:3.3

A lender wishing to sell a loan to either of the GSEs must meet the
GSEs' underwriting standards.  Those standards require credit
enhancement--typically, mortgage insurance--for loans with LTV ratios
greater than 80.  The lender selects a mortgage insurer from those
that are approved by the GSEs.  Any of the loans with high LTV ratios
that are part of the effort to reach low- and moderate-income
borrowers and borrowers located in central cities and other
underserved areas require mortgage insurance or other credit
enhancement.  However, as is the case with Fannie Mae's portfolio in
general, about two-thirds of the loans counted toward the social
goals had LTV ratios of 80 percent or less and therefore generally
would not require mortgage insurance.  Specifically, Fannie Mae
reports that for 1995, 65 percent of the dwelling units that counted
toward the low- and moderate-income goal had LTV ratios of 80 percent
or less.  For the central city goal, the percentage was 64 percent,
and for the special affordable goal, it was 64 percent.  In
comparison, about 65 percent of all mortgages acquired by Fannie Mae
had LTV ratios of 80 percent or less.  For Freddie Mac, the
percentage of loans that counted toward the social goals varied;
there were relatively more loans with LTVs below 80 percent and which
therefore would not require mortgage insurance.  Specifically, for
Freddie Mac the percentage of units that counted toward the three
goals in 1995 and had LTV ratios of 80 percent or less were 74, 67,
and 79.  The percentage of all mortgages acquired by Freddie Mac that
had LTV ratios of 80 percent or below was 70 percent in 1995. 

In comparison with the total mortgages acquired, Fannie Mae had
relatively more of its loans that counted toward its social goals
with LTV ratios above 90 percent, while Freddie Mac had relatively
fewer for two of the social goals.  Overall, 17.9 percent of
single-family mortgages purchased by Fannie Mae in 1995 had LTV
ratios above 90 percent.  For Freddie Mac, the number was 13.5
percent.  While Fannie Mae purchased in 1995 few loans with LTV
ratios above 95 percent, relatively more of these loans counted
toward the social goals.  Specifically, while 2.2 percent of the
loans Fannie Mae purchased in 1995 had LTV ratios greater than 95
percent, the proportion of loans that counted toward the low- and
moderate-income, central cities, and special affordable housing goals
and for which the LTV of the loan was greater than 95 percent, were
5.4, 2.9, and 6.1 percent, respectively.\17 Freddie Mac purchased
nearly no loans with LTV ratios above 95 percent in 1995; none were
counted toward the social goals. 


--------------------
\17 Fannie Mae first announced its program for the purchase of loans
with LTV ratios of 97 percent in November 1994. 


   OBSERVATIONS
---------------------------------------------------------- Chapter 4:4

Of the programs used by the federal government to promote affordable
homeownership, FHA's Section 203(b) mortgage insurance program
reaches more homeowners than does any other program; and in some
instances, FHA's insurance is used in conjunction with other
programs.  Where the use of mortgage insurance is known, two of the
other programs used FHA mortgage insurance--in one instance for 60
percent of the loans made and in another instance for 19 percent of
the buyers assisted.  However, according to available data, FHA's
program in many instances is not as focused on low-income and
minority homeowners and first-time home buyers as are the other nine
programs.  While the other programs are generally more targeted to
these underserved borrowers, they often have a cost to the federal
government.  In contrast, the costs of FHA's Section 203(b) program
are paid by the program's participants and not by the U.S.  Treasury. 
In comparison with all of these programs, the requirements placed
upon certain lenders and purchasers of mortgages may have the
greatest potential for promoting affordable homeownership, although
the extent to which these requirements affect lenders' behavior is
not clear.  Finally, the most pervasive government incentive for
homeownership--though not targeted to low-income home buyers--is the
deduction of the interest on home mortgages from an individual's
taxable income. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 4:5

In response to comments from the Managing Director of the FHFB, we
made a number of revisions, including clarifying our discussion of
the potential federal government liability associated with advances
provided by FHLB to member institutions.  However, in contrast with
the Managing Directors' comments, we continue to believe that there
is a potential federal cost associated with such advances because the
federal government has paid for liquidating insolvent member
institutions.  Although the federal government has incurred no direct
costs due to FHLB advances, the costs the federal government could
incur for liquidating insolvent member institutions may be higher
when member institutions have been provided additional resources for
lending through the advances.  In addition, government sponsorship of
the FHLBank System creates potential liabilities for the federal
government.  For these reasons, we retained the discussion of this
potential cost in our report. 


DATA FOR THE FIGURES IN CHAPTERS 1
AND 3
=========================================================== Appendix I



                               Table I.1
                
                  Insured and Uninsured Home Purchase
                    Mortgages as a Percentage of All
                   Originations, 1984-94 (Figure 1.1)


                          Uninsure                    Uninsure
Year             Insured         d   Total   Insured         d   Total
--------------  --------  --------  ------  --------  --------  ------
1984             $92.027  $111.678  $203.7        45        55     100
                                        05
1985             $94.488  $195.296  $289.7        33        67     100
                                        84
1986            $134.057  $365.355  $499.4        27        73     100
                                        12
1987            $152.473  $354.758  $507.2        30        70     100
                                        31
1988            $102.194  $344.069  $446.2        23        77     100
                                        63
1989             $95.906  $357.001  $452.9        21        79     100
                                        07
1990            $120.660  $337.744  $458.4        26        74     100
                                        04
1991            $116.196  $445.878  $562.0        21        79     100
                                        74
1992            $175.865  $717.817  $893.6        20        80     100
                                        81
1993            $261.247  $758.615  $1,019        26        74     100
                                      .861
1994            $274.485  $494.243  $768.7        36        64     100
                                        28
----------------------------------------------------------------------
Source:  HUD, "U.S.  Housing Market Conditions," August 1995. 



                               Table I.2
                
                All Insured Home Purchase Mortgages, by
                     Insurer, 1984-94 (Figure 1.2)

                                       Percent
                           Percent  insured by     Percent
                        insured By         PMI  insured by
Year                           FHA   companies          VA       Total
----------------------  ----------  ----------  ----------  ----------
1984                            18          69          13         100
1985                            30          53          16         100
1986                            48          34          17         100
1987                            51          29          20         100
1988                            46          39          16         100
1989                            47          39          14         100
1990                            50          32          18         100
1991                            40          46          13         100
1992                            29          57          14         100
1993                            32          52          16         100
1994                            35          48          18         100
----------------------------------------------------------------------
Source:  HUD, "U.S.  Housing Market Conditions," August 1995



                               Table I.3
                
                 Number of Mortgages Issued in 1994, by
                      Type of Insurer (Figure 3.1)

                                                            Percent of
                                          Home                    home
Type of insurer              Total    purchase   Refinance    purchase
----------------------  ----------  ----------  ----------  ----------
Uninsured                4,111,663   2,069,974   2,041,689       58.48
PMI\a                      883,669     725,188     158,481       20.49
FHA                        722,342     519,102     203,240       14.67
VA                         333,926     218,052     115,874        6.16
RHS                          7,723       7,215         508        0.20
======================================================================
Total                    6,059,323   3,539,531   2,519,792         100
----------------------------------------------------------------------
\a Mortgage Insurance Companies of America's data on private mortgage
insurer (PMI) loans were adjusted to compare with data from the Home
Mortgage Disclosure Act (HMDA) data.  HMDA data include approximately
77 percent of all home purchase loans.  MICA's data, however, include
nearly all loans insured by PMIs.  To determine the relative share of
the market of loans in the HMDA database held by FHA and PMIs, MICA's
data were reduced by 23 percent. 

Source:  GAO's analysis of data obtained from HUD and through HMDA
and from MICA. 



                               Table I.4
                
                Number of Home Purchase Mortgages Issued
                in 1994 to Low-, Middle-and High-Income
                 Borrowers, by Type of Insurer (Figures
                              3.2 and 3.3)

                                                              Uninsure
Income type\a          Total     FHA   PMI\b     RHS      VA         d
------------------  --------  ------  ------  ------  ------  --------
High                1,287,40  108,07  269,72     788  57,443   851,374
                           1       2       4
Middle              1,157,33  190,67  277,82   3,223  86,808   598,798
                           1       9       3
Low                 1,094,79  220,35  177,64   3,204  73,801   619,802
                           8       1       1
Percent                30.93   42.45   24.50   44.41   33.85     29.94
 low income
======================================================================
Total               3,539,53  519,10  725,18   7,215  218,05  2,069,97
                           1       2       8               2         4
----------------------------------------------------------------------
\a We defined borrower's income as "low" if it is at or below 80
percent of the Metropolitan Statistical Area's (MSA) median family
income, "middle" if it is greater than 80 percent but at or below 120
percent of the MSA median, and "high" if it is greater than 120
percent of the MSA median. 

\b The relative percentages of PMI loans in the high-, middle-, and
low-income groups as reported in the MICA data were applied to our
estimate of the number of home purchase loans insured by PMIs and
included in the HMDA data. 

Source:  GAO's analysis of data from HMDA, MICA, and HUD. 



                               Table I.5
                
                   Proportion of Insurers' 1994 Home
                  Purchase Mortgages in Various Income
                          Groups (Figure 3.4)

                                Uninsure
                                       d       FHA       PMI        VA
Income                           percent   percent   percent   percent
------------------------------  --------  --------  --------  --------
$0-$20,000                          5.91      4.38      1.24      3.59
$20,001-$30,000                    11.81     19.14      7.51     13.55
$30,001-$40,000                    14.38     27.46     14.82     24.63
$40,001-$50,000                    13.65     22.52     18.28     22.55
$50,001-$60,000                    11.98     13.58     16.99     16.19
$60,001-$80,000                    16.88      9.45     22.58     14.33
$80,001-$100,000                    9.45      2.07     10.00      3.63
$100,001-$200,000                  12.41      1.21      7.91      1.41
$200,001-$300,000                   2.05      0.11      0.48      0.06
$300,000                            1.48      0.08      0.19      0.07
======================================================================
Total                             100.00    100.00    100.00    100.00
----------------------------------------------------------------------
Note:  Conventional loans reported in GAO's analysis of data from
HMDA were divided between the PMIs and the uninsured loans on the
basis of (a)the total number of loans identified with income data and
(b)adjustments to the total number of PMI loans reported by MICA to
account for the under-count in the HMDA data as described in the note
to table I.3. 

Source:  GAO's analysis of data from HMDA and MICA. 



                               Table I.6
                
                 Number of Home Purchase Mortgages Made
                  in 1994, by Race of Borrowers and by
                     Insurer (Figures 3.5 and 3.6)

                      Uninsure
Race of borrower             d       FHA       PMI        VA     Total
--------------------  --------  --------  --------  --------  --------
Minority               303,477   147,423   114,197    48,453   613,550
Nonminority           1,766,49   371,679   610,991   169,599  2,918,76
                             7                                       6
Total                 2,069,97   519,102   725,188   218,052  3,532,31
                             4                                       6
Percent\a minority       14.66     28.40     15.75     22.22     17.37
Percent of minority
                         49.46     24.03     18.61      7.90    100.00
----------------------------------------------------------------------
\a Percent minority is the proportion of each group's loans that is
identified as minority.  Percent of minority is the percent of all
minority loans that are insured by each group. 

Source:  GAO's analysis of data from HMDA and MICA. 




                               Table I.7
                
                 Number of Home Purchase Mortgages Made
                   to First-Time Home Buyers in 1994
                         (Figures 3.7 and 3.8)

                                                  Number of first-time
                                   Total loans             home buyers
----------------------  ----------------------  ----------------------
FHA                                    686,487                 459,851
Non-FHA                              3,888,652               1,695,039
======================================================================
Total                                4,575,139               2,154,890
----------------------------------------------------------------------
Source:  Data from the Mortgage Bankers Association. 



                               Table I.8
                
                 Number of 1993 Home Purchase Mortgages
                Reported in the American Housing Survey,
                   by Age of Borrowers and by Insurer

                                          Conventi
Age                                           onal       FHA        VA
----------------------------------------  --------  --------  --------
30 & below                                 142,757    19,073     5,040
At least 30 & less than 40                 828,055   148,895    38,116
At least 40 & less than 50                 882,682    58,288    19,236
At least 50 & less than 60                 447,353    40,649    16,424
At least 60 & less than 75                 227,390     4,861     4,385
75 & over                                   19,805         0         0
======================================================================
Total                                     2,548,04   271,766    83,201
                                                 2
----------------------------------------------------------------------
Note:  The 1993 AHS survey took place between July and December of
1993.  The data include home purchase mortgages received in 1993
prior to the survey. 

Source:  GAO's analysis of the American Housing Survey's 1993 data. 




                                              Table I.9
                               
                                Number of Home Purchase Mortgages Made
                               in Each State in 1994, and Market Share
                                Held by PMIs, FHA, and VA (Figure 3.9)


State                    PMI         FHA          VA       Total         PMI         FHA          VA
-------------  -------------  ----------  ----------  ----------  ----------  ----------  ----------
AL                     8,820       6,289       3,266      18,375          48          34          18
AK                     1,293       1,214       1,396       3,903          33          31          36
AZ                    17,122      15,273       6,143      38,538          44          40          16
AR                     2,793       4,320       1,659       8,772          32          49          19
CA                    75,741      50,612      17,603     143,956          53          35          12
CO                    18,090      16,716       6,870      41,676          43          40          16
CT                     8,018       5,361         990      14,369          56          37           7
DE                     3,001       1,071         593       4,665          64          23          13
DC                     1,708         944         202       2,854          60          33           7
FL                    48,714      29,982      14,303      92,999          52          32          15
GA                    25,875      15,113       6,233      47,221          55          32          13
HI                     1,475       1,203         327       3,005          49          40          11
ID                     3,074       2,106         846       6,026          51          35          14
IL                    42,972      20,967       4,450      68,389          63          31           7
IN                    19,169      11,003       3,157      33,329          58          33           9
IA                     9,280       1,821         761      11,862          78          15           6
KS                     8,550       3,551       1,844      13,945          61          25          13
KY                     6,936       4,136       1,653      12,725          55          33          13
LA                     7,872       6,641       2,557      17,070          46          39          15
ME                     1,536         729         428       2,693          57          27          16
MD                    16,138      16,531       6,736      39,405          41          42          17
MA                    19,140       4,069       1,884      25,093          76          16           8
MI                    36,302      15,560       3,491      55,353          66          28           6
MN                    14,931      16,067       3,181      34,179          44          47           9
MS                     3,401       3,334       1,183       7,918          43          42          15
MO                    15,418       9,502       2,690      27,610          56          34          10
MT                       999       1,232         423       2,654          38          46          16
NE                     3,948       2,275       1,160       7,383          53          31          16
NV                     5,070       5,633       2,965      13,668          37          41          22
NH                     2,520       1,245         598       4,363          58          29          14
NJ                    23,034      10,910       3,061      37,005          62          29           8
NM                     4,373       2,264       1,440       8,077          54          28          18
NY                    30,067      17,254       3,431      50,752          59          34           7
NC                    20,363      10,327       8,038      38,728          53          27          21
ND                       851       1,283         252       2,386          36          54          11
OH                    37,143      14,664       5,045      56,852          65          26           9
OK                     6,174       6,514       3,234      15,922          39          41          20
OR                    11,539       4,006       1,563      17,108          67          23           9
PA                    28,655      12,662       3,706      45,023          64          28           8
RI                     2,995       1,425         443       4,863          62          29           9
SC                    10,033       4,079       2,569      16,681          60          24          15
SD                     1,096         987         558       2,641          41          37          21
TN                    10,743      14,473       4,601      29,817          36          49          15
TX                    53,252      33,222      19,505     105,979          50          31          18
UT                     6,631       7,181       1,513      15,325          43          47          10
VT                       503         161         146         810          62          20          18
VA                    20,460      17,243      12,696      50,399          41          34          25
WA                    19,016      10,099       7,999      37,114          51          27          22
WV                     2,017         840         433       3,290          61          26          13
WI                    17,304       1,899       1,969      21,172          82           9           9
WY                     1,132         790         411       2,333          49          34          18
====================================================================================================
Total              737,287\a     446,783     182,205   1,366,275          54          33          13
----------------------------------------------------------------------------------------------------
\a 204,516 (941,803 total PMI loans less 737,287 identified in
states) of all PMI loans within the MICA data, and 72,319 (519,102
total FHA loans less 446,783 identified in states) FHA loans within
the HMDA data did not have a state code or contained edit failures. 
When the adjustments for under-count in the HMDA data and to account
for missing state codes and edit failures are made, we find that an
additional four states may have been primarily insured by FHA
(Arizona, Colorado, Mississippi, and South Dakota).  Two other states
(Louisiana and Virginia) are estimated to have the same percentage of
home purchase loans insured by PMIs and FHA. 

Source:  GAO's analysis of HMDA's and MICA's data. 



                               Table I.10
                
                 Distribution of 1994 FHA and PMI Home
                  Purchase Mortgages, by LTV (Figures
                          3.10, 3.11 and 3.12)

                                                            Percent of
                                                Percent of       PMI\a
LTV level                                        FHA loans       loans
----------------------------------------------  ----------  ----------
Greater than 80 but not more than 85                  2.36        5.33
Greater than 85 but not more than 90                  8.72       40.11
Greater than 90 but not more than 95                 23.16       47.19
Greater than 95 but not more than 97                 33.44        1.68
Greater than 97 but not more than 100                31.87        5.68
----------------------------------------------------------------------
\a Conventional loans with LTVs above 80 percent. 

Source:  GAO's analysis of Federal Housing Finance Board's and FHA's
data. 



                               Table I.11
                
                    Distribution of VA Home Purchase
                 Mortgages, by Loan-to-Value Ratio, in
                           1994 (Figure 3.13)

LTV level                                          Percent of VA loans
----------------------------------------------  ----------------------
Greater than 81 but not more than 86                              1.77
Greater than 86 but not more than 91                              4.54
Greater than 91 but not more than 95                              7.28
Greater than 95 but not more than 97                              5.26
Greater than 97                                                  81.15
----------------------------------------------------------------------
Source:  GAO's analysis of VA's 1994 data. 



                               Table I.12
                
                Number and Percent of FHA Single-Family
                  Home Purchase Mortgages Made in 1995
                 That Are Within PMI Guidelines for LTV
                  Ratio, and Ratios of Total-Debt-to-
                  Income and Housing-Expense-to-Income
                             (Figure 3.14)

                                                     Percentage of all
                                   Number of loans               loans
------------------------------  ------------------  ------------------
Within guidelines                          176,500                34.1
Not within guidelines                      340,412                65.9
======================================================================
Total (all loans)                          516,912               100.0
----------------------------------------------------------------------
Source:  GAO's analysis of HUD's data. 




                               Table I.13
                
                Number and Percent of FHA Single-Family
                  Home Purchase Mortgages Made in 1995
                 That Are Within PMI Guidelines for LTV
                          Ratio (Figure 3.15)

                                                     Percentage of all
                                   Number of loans               loans
------------------------------  ------------------  ------------------
Ratio less than/equal to 95                180,475                37.4
Ratio greater than 95, less                148,630                30.8
 than/equal to 97
Ratio greater than 97                      153,547                31.8
======================================================================
Total (all loans)                          482,652               100.0
----------------------------------------------------------------------
Source:  GAO's analysis of HUD's data. 



                               Table I.14
                
                Number and Percent of FHA Single-Family
                Home Purchase Mortgages Made in 1995, by
                 Ratio of Total-Debt-to-Income (Figure
                                 3.16)

                                                     Percentage of all
                                   Number of loans               loans
------------------------------  ------------------  ------------------
Ratio less than/equal to 36                231,012                49.3
Ratio greater than 36, less                 51,227                10.9
 than/equal to 38
Ratio greater than 38                      186,699                39.8
======================================================================
Total (all loans)                          468,938               100.0
----------------------------------------------------------------------
Source:  GAO's analysis of HUD's data. 



                               Table I.15
                
                Number and Percent of FHA Single-Family
                Home Purchase Mortgages Made in 1995, by
                   Ratio of Housing-Expense-to-Income
                             (Figure 3.17)

                                                  Percentage of all
                           Number of loans              loans
----------------------  ----------------------  ----------------------
Ratio less than/equal          349,813                  74.6%
 to 28
Ratio greater than 28,          89,775                  19.1%
 less than/equal to 33
Ratio greater than 33           29,373                   6.3%
======================================================================
Total (all loans)              468,961                  100.0%
----------------------------------------------------------------------
Source:  GAO's analysis of HUD's data. 


FEDERALLY SUPPORTED PROGRAMS TO
PROMOTE AFFORDABLE HOMEOWNERSHIP
========================================================== Appendix II

Chapter 4 describes federal efforts to promote affordable
homeownership, including federal programs and requirements placed
upon the lenders and purchasers of mortgages.  This appendix provides
a detailed description of each of the federal programs included in
chapter 4 with the exception of the FHA and VA programs, which are
described in chapters 2 and 3.  Generally, program data come from the
National Council of State Housing Agencies, the Federal Housing
Finance Board, HUD, VA, the Neighborhood Reinvestment Corporation,
and the Rural Housing Service of the Department of Agriculture.  We
did not verify the accuracy and completeness of these data. 


   STATE HOUSING FINANCE AGENCIES
-------------------------------------------------------- Appendix II:1

Central to the activities of the state housing finance agencies (HFA)
are programs that promote homeownership.  These programs rely upon
federal incentives provided through mortgage revenue bonds (MRB) and
mortgage credit certificates (MCC) and the Home Investment
Partnership (HOME) program (described below).  Specifically, under
the federal mortgage revenue bond program, state HFAs raise funds by
issuing tax-free mortgage revenue bonds.  These funds are used to
make loans to first-time home buyers.  Funds may also be used as
mortgage credit certificates issued to the home purchaser, which
allows the purchaser to take a tax credit against a part of his or
her home mortgage interest.  The Tax Reform Act of 1986 restricted
tax-exempt bond issues for housing.  The act placed a cap for all
private purpose tax-exempt bonds--of which MRBs are one--for each
state.  In 1994, every state as well as the District of Columbia,
Puerto Rico, and the U.S.  Virgin Islands had a housing finance
agency. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:1.1

HFAs offer first-time home buyers mortgages that may carry subsidized
interest rates.  The loans may be originated by a lender which then
sells the loans to the HFA.  Alternatively, first-time home buyers
may receive a mortgage credit certificate, which provides the
recipient with a nonrefundable federal income tax credit for a
specified percentage of the annual interest paid on the mortgage of a
principal residence.  Homeowners may also receive loans for the
purpose of rehabilitating a property or making home improvements. 

The Congress has set limits on incomes and home purchase prices for
home buyers under the mortgage revenue bond program.  Home buyers may
not have incomes that exceed 100 percent of the median family income
for the area (115 percent for a family of three or more) and may not
have owned a home in the last 3 years.  The price of the property may
not exceed 90 percent of the average purchase price for homes in the
area.  Also, borrowers whose income rises more than 5 percent above
the income limits are subject to pay the federal government up to
half of any profit they make on the sale of the home within the first
9 years. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:1.2

In 1994, the state housing finance agencies collectively used over $9
billion in MRB funds to make 89,288 home purchase loans and 3,531
rehabilitation and home improvement loans and to issue 11,701
mortgage credit certificates.  Across HFAs, the mortgage amount for
home purchases averaged $67,711.\1 Rehabilitation and home
improvement loans were made in nine states and the District of
Columbia and averaged $9,333.  Across the 17 states that reported
issuing mortgage credit certificates, the average mortgage amount
associated with those receiving certificates was $61,127. 

Of those assisted by HFAs through mortgages or mortgage credit
certificates in 1994, the average income was about $30 thousand.\2
About 64 percent of homeowners assisted through mortgages or Mortgage
Credit Certificates, and for which data were available, had incomes
not exceeding 80 percent of the areas' median incomes.\3 Of the
homeowners that received mortgages, 63.3 percent had incomes not
exceeding 80 percent of the areas' median incomes.  About 74 percent
of the recipients of mortgage credit certificates had incomes no
greater than 80 percent of the areas' median incomes.  On average, 22
percent of borrowers were minorities in the District of Columbia and
the 36 states that reported the race of borrowers under their MRB
program and 30 percent in the 11 states that reported the race of
borrowers under their MCC program.  For the 39 HFAs that reported the
location of properties financed with the proceeds of mortgage revenue
bonds, 74 percent were in areas defined by the HFA as urban.  For the
13 states reporting location data for recipients of mortgage credit
certificates, 81 percent were in areas defined by the HFA as urban. 
For the 34 states reporting such data for their MRB program, an
average of 16 percent of home purchasers were single-parent
households.  For the 12 states reporting such data for their MCC
program, an average of 12 percent of recipients were single-parent
households. 


--------------------
\1 Data are for the District of Columbia and all states except
Arizona and Kansas. 

\2 The data on the average income of homeowners assisted with
mortgages are for the District of Columbia and all states, except
Arizona, Kansas, and Nebraska.  The data on the average income of
homeowners receiving mortgage credit certificates are for 17 states. 

\3 We excluded data for the state of Alabama's MCC and MRB programs
and for the state of West Virginia's MCC program because the data on
income distributions were incomplete. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:1.3

State housing finance agencies utilize mortgage insurance from both
FHA and private mortgage insurers (PMI).  In fact, nearly all the
loans made by state housing finance agencies in 1994 were insured. 
For 1994, FHA insured over 55 percent of loans, VA over 8 percent,
and PMIs insured about 20 percent.  In addition, four states and
seven HFAs insured mortgages, and eight HFAs utilized pool insurance. 
With the added protection of mortgage insurance, HFAs may offer more
flexible terms than they might offer otherwise. 


   FEDERAL HOME LOAN BANKS'
   COMMUNITY INVESTMENT PROGRAM
-------------------------------------------------------- Appendix II:2

The Federal Home Loan Bank (FHLBank) System operates the Community
Investment Program (CIP), which was authorized under section 721 of
the Financial Institutions Reform, Recovery, and Enforcement Act of
1989.  Specifically, each FHLBank, through its CIP program, provides
discounted advances to lenders that are members of the FHLBank
System.  These lenders, in turn, use the funds for eligible loans. 
Federal Home Loan Banks reduce lending costs by providing advances
priced at the FHLBs' cost of funds, plus reasonable administrative
costs, resulting in a discount typically amounting to about
one-quarter of 1 percent from the cost of regular advances.  The
discounted rate on the advance helps lenders to hold long-term,
fixed-rate loans in their portfolios or to make loans they may not
otherwise make.  The Federal Housing Finance Board sets annual
targets for the levels of new CIP advances. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:2.1

CIP funds may be used for loans to finance home purchases and to
rehabilitate housing, as well as for commercial and economic
development activities that benefit low- and moderate-income families
or activities that are located in low- and moderate-income
neighborhoods.  In connection with financing single-family
homeownership, the subsidy provided to member institutions may or may
not be passed on to the borrower, according to an FHFB official. 
When CIP advances are used for the purpose of financing the purchase
of a home, the purchaser's income may not exceed 115 percent of the
area's median income.  According to a FHLB official, there are no
other restrictions placed on the use of advances used to finance
single-family homeownership.  The standards for underwriting CIP
loans are left up to the individual lenders. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:2.2

In 1995, FHLBs made 632 advances for homeownership projects under the
CIP program.  Member institutions provided financing for 82,907
owner-occupied units.  This financing--representing about $3.8
billion in advances--was used for the purpose of home purchase and/or
rehabilitation.  According to an official, the Federal Housing
Finance Board did not have data on the characteristics of the
homeowners assisted through the CIP program.  However, home
purchasers assisted through the program may not have incomes greater
than 115 percent of the area's median income. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:2.3

The Federal Housing Finance Board did not have data on the extent to
which CIP loans have mortgage insurance.  However, a program official
noted that many FHLB member lenders do portfolio lending and that CIP
is a tool that would allow these lenders to finance loans that they
keep in portfolio.  These lenders would then have the option of not
requiring mortgage insurance. 


   FEDERAL HOME LOAN BANKS'
   AFFORDABLE HOUSING PROGRAM
-------------------------------------------------------- Appendix II:3

The Federal Home Loan Banks also operate the Affordable Housing
Program (AHP), through which the banks provide subsidies to members
engaged in long-term lending for owner-occupied and rental housing
targeted to households with very-low, low, or moderate incomes. 
Authorized by the Financial Institutions Reform, Recovery, and
Enforcement Act of 1989, the Affordable Housing Program provides
discounted advances (loans) to members that, in turn, lend the funds
at reduced interest rates to specific AHP projects and make direct
subsidies (grants) to members who pass the subsidies directly to
specific projects or who use the subsidies to reduce the interest
rate on loans that the members themselves provide to specific AHP
projects.  These subsidies are awarded on a competitive basis, and
the program is administered by the Federal Housing Finance Board. 
AHP projects can be sponsored by public or private-sector
organizations, with the AHP funding provided by the member
institution that serves the area where the projects are located.  The
organization receiving AHP funding may receive subsidized interest on
loans or a direct subsidy.  More than 97 percent of the projects have
a nonprofit or government sponsor.  Funding for the AHP is derived
from the FHLBank System's net earnings as determined by a statutory
formula.  AHP projects may also use funds from state HFAs, HOME, Home
Ownership and Opportunity for People Everywhere, Community
Development Block Grant, state and local government programs, and
other grants and foundations. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:3.1

Funds used for homeownership opportunities typically go to the
homeowner in the form of a grant, according to an FHFB official.  The
funds are used for down payment assistance and help with closing
costs.  Also, according to the Federal Housing Finance Board, AHP
funds are often used in conjunction with other sources of funds, such
as the FHLBank System's Community Investment Program and other
federal, state, local, or private assistance programs.  In our 1995
report on the program, we found that the AHP has helped member
institutions expand their interest and experience in financing
affordable housing while helping them meet their statutory
requirements on community lending.\4 In September 1995, the Federal
Housing Finance Board approved a program to set aside a limited
portion of available AHP subsidies to assist first-time home buyers. 
The regulations authorize each bank to establish savings plans that
would match household savings with AHP funds at a rate of 1 to 3, up
to $5,000.  This program includes requirements for counseling
borrowers and a "soft" second mortgage. 

Household income is restricted to 80 percent of the median income in
the area where the funds are used to finance the purchase,
construction, and/or rehabilitation of owner-occupied housing. 
According to an FHFB official, standards for underwriting are left up
to individual lenders.  However, the regulations for the AHP include
language that encourages lenders to be flexible, but prudent.  Also,
projects are selected competitively; are oversubscribed by 3 to 1,
according to a Finance Board official; and income targeting and
long-term retention are criteria in the selection of projects. 
Finally, AHP funds are subject to recapture if, during the long-term
retention period, the owner-occupied home assisted with AHP funds is
sold to households that are not income-eligible for AHP assistance. 


--------------------
\4 Housing Finance:  Improving the Federal Home Loan Bank System's
Affordable Housing Program (GAO/RCED-95-82, June 9, 1995). 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:3.2

In 1995, FHLBs approved $48 million in AHP subsidies for
owner-occupied units.  These subsidies supported the purchase,
construction, or rehabilitation of owner-occupied units for 10,241
households.  For 1994, the average subsidy per unit was about $4,000. 
The income of homeowners who received subsidies from the AHP program
may not exceed 80 percent of the area's median income.  Indeed, the
Finance Board reports that for 1994, about 70 percent of
owner-occupied units that received AHP subsidies were owned by
persons that had very-low income--less than 50 percent of the area's
median income. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:3.3

According to an official, the Federal Housing Finance Board did not
have data on the extent to which mortgage insurance is used in
conjunction with projects that receive AHP subsidies. 


   NEIGHBORHOOD REINVESTMENT
   CORPORATION/NEIGHBORWORKS
   ORGANIZATIONS
-------------------------------------------------------- Appendix II:4

The Neighborhood Reinvestment Corporation (NRC) was created by the
Housing and Community Development Act of 1978.  The NRC promotes
investment in communities by helping the formation of and providing
technical assistance to local private and public-sector
organizations--known as NeighborWorksï¿½ organizations (NWO)--and by
providing grants to these organizations.  In 1994, 173 NWOs were
operating nationwide.  NRC also is a primary source of funding for
Neighborhood Housing Services of America, which purchases mortgages
from NWOs.  NRC is under the direction of a board of directors
comprising the Secretary of HUD, a member of the Board of Governors
of the Federal Reserve System, a Director of the Federal Deposit
Insurance Corporation, the Comptroller of the Currency, the Director
of the Office of Thrift Supervision, and the Vice Chairman of the
National Credit Union Administration.  Although NRC is not a federal
agency, about 91 percent of its $42.4 million fiscal year 1995
revenue came from federal appropriations.  In fiscal year 1995, NRC
provided about 50 percent of its appropriation as direct grants to
NWOs, neighborhood preservation projects, and similar programs, in
addition to technical services, training, program monitoring, and
other direct services.  NWOs work with lenders, insurers, and state
and local governments.  They also receive funding from HUD's CDBG,
HOME, and HOPE programs.  NWOs replenish second
mortgage/rehabilitation loans by selling them to Neighborhood Housing
Services of America. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:4.1

NRC helps in the formation of local organizations that provide or
channel investment in communities and is a primary source of funding
for Neighborhood Housing Services of America, which purchases
mortgages from NWOs.  NWOs offer second-mortgages on properties and
channel prospective home purchasers to lenders for first mortgages. 
In some instances, NWOs themselves will originate first mortgages. 
NWOs also offer homeowner education and counseling. 

According to the NRC Executive Director, while NWOs are focused on
low-income communities, there is no limitation on the income of
individuals within those communities that may use NWOs' services. 
Underwriting standards are left up to the individual NWOs and the
lenders with which they work.  Whoever originates the loans sets the
standards.  Because NWOs obtain funds from other sources, they may be
limited in the terms of the products they offer by the standards
established under other programs. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:4.2

In fiscal year 1995, NeighborWorksï¿½ organizations were responsible
for 2,854 second mortgages that were used by borrowers for
rehabilitation--the NWOs originated about 90 percent of these
mortgages.  Lenders associated with the NWOs originated 4,330 first
mortgages in fiscal year 1995.  Historically, 54 percent of these
first mortgages are on properties that also have second mortgages,
according to the Executive Director of the NRC.  Direct investment
for single-family rehabilitation loans was about $37 million in
fiscal year 1995.  First mortgages for homeownership amounted to
approximately $244 million in fiscal year 1995.  While NWOs
originated about 90 percent of the second mortgages reported, they
typically do not originate first mortgages, according to an NRC
official.  With the advent of the Home Mortgage Disclosure Act and
the Community Reinvestment Act, according to this official, there is
a strong incentive for lenders to originate the first mortgage so
that they get credit under the Community Reinvestment Act. 

According to the data for new homeowners for the 36-month period
ending December 31, 1995, 69 percent had household income of less
than 80 percent of the area's median income.  The median family
income for home purchasers was $24,000.  In addition, 61 percent of
new homeowners were minorities, 97 percent were first-time home
buyers, and 44 percent were female-headed households.  In connection
with the location of the home, NWOs are focused on low-income
communities, according to the Executive Director of NRC. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:4.3

The use of mortgage insurance is not well known for the programs of
NWOs.  However, according to the Executive Director of NRC, some NWOs
use FHA insurance extensively, while others avoid FHA.  For example,
one NWO is a direct endorsement lender of FHA-insured loans,
according to the Executive Director.  Also, a number of NWOs use
FHA's Section 203(k) rehabilitation loans.  Private mortgage
insurance companies also provide insurance for first mortgages
originated by the lenders associated with the NWOs. 


   NEIGHBORHOOD HOUSING SERVICES
   OF AMERICA
-------------------------------------------------------- Appendix II:5

Neighborhood Housing Services of America (NHSA) is a state-chartered,
private, nonprofit organization established in 1974.  It is funded by
the Neighborhood Reinvestment Corporation as authorized by Congress
in 1978.  NHSA operates a secondary market for loans--primarily
second mortgages--made by NWOs, thereby replenishing the loan funds
of these organizations.  NHSA also purchases first mortgages from
NWOs.  According to the Executive Director of the Neighborhood
Reinvestment Corporation, as the grantor, NRC has responsibilities
for overseeing NHSA.  NHSA's Board of Directors is composed of
private individuals.  NRC is a primary source of funding for NHSA. 
Grants made by the NRC to NHSA totaled $4.5 million in fiscal year
1995.  NHSA also receives funding from private-sector, institutional
investors (including insurance companies, savings banks, and pension
funds) through the sale of secondary market notes backed by loans
purchased by NHSA. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:5.1

NHSA provides liquidity to NWOs, which make second mortgages and
arrange for lenders to make first mortgages on single-family
properties.  NHSA provides liquidity by purchasing loans that NWOs
originate, thus permitting them to use the resulting funds for
additional lending.  NHSA then pools the mortgages it purchases and
sells to social investors, through private placement, notes backed by
these mortgages.  In addition to these products related to
single-family housing, NHSA also provides permanent financing for
multi-unit or rental housing owned and managed by nonprofit
organizations and short-term loans to NWOs to finance bridge loans
for housing developed by nonprofits. 

According to the NRC Executive Director, loans purchased by NHSA must
be made for a property that is within the geographic area outlined in
the map describing the area served by the NWO.  Also, mortgages
purchased by NHSA must be recorded, be fully amortizing, have at
least quarterly principal and interest payments, and have fixed
rates.  Second mortgages must have full recourse, that is, the NWO
agrees to substitute a new loan for any delinquent loan. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:5.2

In fiscal year 1995, NHSA purchased 663 rehabilitation loans (second
mortgages) and 470 first mortgages.  These amounted to over $6.3
million in second mortgages on single-family homes and over $41
million in first mortgages on single-family homes.  The median income
of borrowers whose loans were purchased by NHSA was $22,800 for
fiscal year 1995.  Because NHSA purchases loans that are originated
by NWOs, the characteristics of the homeowners it assists are the
same as the characteristics of borrowers assisted by NWOs.  As
described earlier, according to data for new homeowners served by
NWOs for the 36-month period ending December 31, 1995, 69 percent had
household incomes of less than 80 percent of the area's median
income.  In addition, 61 percent of the new homeowners were
minorities, 97 percent were first-time home buyers, and 44 percent
were female-headed households.  In connection with the location of
the home, NWOs are focused on low-income communities, according to
the Executive Director of the NRC. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:5.3

In addition to any mortgage insurance on individual loans purchased
by NHSA, PMI Mortgage Insurance Company provides pool insurance for
first loans purchased from NHSA by the World Savings and Loan
Association, according to the NRC Executive Director.  The World
Savings and Loan Association held $72.6 million in NHSA notes as of
September 1995, of which $51.0 million was in notes backed by first
mortgages. 


   COMMUNITY DEVELOPMENT BLOCK
   GRANT PROGRAM
-------------------------------------------------------- Appendix II:6

The Community Development Block Grant program is authorized by title
I of the Housing and Community Development Act of 1974.  The purpose
of the program, which is administered by the Department of Housing
and Urban Development, is the development of viable urban communities
by providing decent housing and a suitable living environment and
expanding economic opportunities, principally for persons of low and
moderate income.  The program allocates grants to local governments
and states on the basis of formulas that consider certain economic
and demographic conditions.  Grant recipients use CDBG funds for
eligible community development activities of their choice.  Among the
activities that may receive CDBG funding are such housing activities
as financing the purchase of housing, subsidizing mortgage payments,
and paying a part of down payment and closing costs.  In addition to
housing activities, CDBG funds may be used for public works, public
services, economic development, acquisition and clearance, and
planning/administrative expenses. 

Until recently, assistance to homeowners could be provided only by a
subrecipient of the CDBG grantee.  With the passage of the
Cranston-Gonzalez National Affordable Housing Act of 1990, CDBG
grantees were allowed for the first time to provide such homeowner
assistance without the use of a subrecipient.  However, at the close
of fiscal year 1995, this authority had subsided and, without
reauthorization, grantees must once again rely upon community based
development organizations--special subrecipients--to deliver this
kind of housing assistance. 

Grants to localities are made from annual appropriations.  The fiscal
year 1995 appropriation for the program was $4.8 billion.  With the
exception of small amounts used for special purpose grants, statutory
set-asides, and Indian tribes, the appropriations are distributed on
the basis of 70 percent to entitlement communities (cities and urban
counties) and 30 percent to states for their nonentitlement
communities (small cities).  The largest share of CDBG expenditures
by entitlement communities in fiscal year 1991--37 percent--went for
housing-related activities. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:6.1

Under the Cranston Gonzalez National Affordable Housing Act of 1990,
CDBG grantees may provide direct assistance to facilitate and expand
homeownership among persons of low and moderate income.  Grantees may
subsidize interest rates and mortgage principal amounts, finance the
acquisition of housing that is occupied by the home buyer, acquire
guaranties for mortgage financing obtained by home buyers from
private lenders, provide up to 50 percent of any down payment, and
pay reasonable closing costs incurred by the home buyer.  If the
local government uses a subrecipient, the above activities are
eligible without the provisions of the Cranston Gonzalez National
Affordable Housing Act. 

Direct home buyer assistance is limited to home buyers of low and
moderate income, that is, persons with incomes below 80 percent of
the area's median income.  Where housing assistance is provided by a
special subrecipient, at least 50 percent of home buyers benefiting
must be of low or moderate income.  Also, regardless of the type of
activity, at least 70 percent of all CDBG funds received by a grantee
must be used for activities that benefit persons of low and moderate
income. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:6.2

Homeownership assistance accounted for $19.4 million of entitlement
grantees' expenditures for program year 1992.\5 For the small cities
programs administered by HUD and the states, housing activities--both
single- and multifamily, including rehabilitation--totaled about $250
million in fiscal year 1993.\6 Homeownership assistance includes (1)
assistance that is provided by a special subrecipient, typically
including cash transfer payments to the home buyer, and (2) direct
assistance that may be provided by the grantee, including financing
the acquisition of property, subsidizing interest rates and mortgage
principal amounts, acquiring guaranties for mortgage financing from
private lenders, providing up to 50 percent of down payments, and
paying reasonable closing costs.  For entitlement grantees, over
4,400 households received homeownership assistance through a special
subrecipient during program year 1992, and almost 6,000 households
received homeownership assistance from the grantee during that
period. 

Over 90 percent of the homeowners assisted by entitlement grantees
during program year 1992 had incomes that were low or moderate--below
80 percent of the area's median income.  Sixty-five percent were
minorities, and 24 percent were female-headed households.  Data were
not available on the characteristics of homeowners assisted through
the small-cities CDBG program. 


--------------------
\5 The 1992 program year is set by the grantee itself.  The year must
begin no earlier than January 1, 1992, but no later than September
30, 1992, and end no later than September 30, 1993. 

\6 All states except Hawaii and New York administer small-cities
programs.  The data reported on activities of the small-cities
programs represent the reported distribution of funds to housing
activities.  According to a HUD official, housing activities include
rehabilitation as well as homeownership activities for both single-
and multifamily housing.  Data were not available on the amount of
assistance provided through the small-cities CDBG program that went
toward single-family homeownership assistance only, or on the
characteristics of those benefiting from homeownership assistance. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:6.3

The extent to which housing assistance includes the use of mortgage
insurance is not known. 


   RURAL HOUSING SERVICE
-------------------------------------------------------- Appendix II:7

The Rural Housing Service of the Department of Agriculture
administers a program that provides direct and guaranteed home loans
to lower-income rural families.  The loan program was first
authorized in the Housing Act of 1949.  The current Rural Housing
Service was created with the Federal Crop Insurance Reform and
Department of Agriculture Reorganization Act of 1994.  The program
was established to provide housing credit for farm households because
private credit sources were either nonexistent or inadequate in rural
areas. 

Today, the program provides loans to rural households that meet
certain income and asset limitations and are unable otherwise to
obtain mortgage credit at terms they can reasonably be expected to
pay.  The direct loans are made by any of the 1,700 county offices. 
Direct loans are subsidized by the Rural Housing Service.  That is,
interest credits are granted annually, which would reduce the monthly
installment on the note to an amount equal to what it would be if the
note were amortized to as low as 1 percent, depending on the loan
amount and the size and income of the family.  The interest credit is
subject to recapture by the government upon liquidation of the
mortgage.  Guaranteed loans are made by private lenders; the interest
rate is negotiated with the lender.  In guaranteeing a single-family
housing loan, the Rural Housing Service agrees, in the event that a
borrower defaults, to reimburse a commercial lender for up to 90
percent of lost principal plus accrued interest and liquidation
costs. 

As of December 31, 1995, the outstanding principal balance on the
696,665 direct single-family housing loans held by the Rural Housing
Service was $18 billion.  In addition, the Rural Housing Service had
guaranties on 39,241 loans, with an outstanding principal of $2.4
billion.  To cover the interest subsidy on direct loans and expected
losses from guaranteed loans, the estimated subsidy for fiscal year
1995 was $245 million.  Lenders that are approved issuers of
mortgage-backed securities guaranteed by the Government National
Mortgage Corporation may use Rural Housing Service-guaranteed loans
to back securities guaranteed by Ginnie Mae. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:7.1

The Rural Housing Service provides direct loans to rural borrowers
and loan guaranties to lenders that provide loans to rural borrowers. 
The direct loans carry an interest subsidy that could reduce the
interest rate to the borrower to as low as 1 percent.  The interest
subsidy is subject to recapture by the government upon liquidation of
the loan.  Direct loans may be made to eligible applicants to buy,
build, repair, renovate, or relocate homes and provide related
facilities, or to refinance home debts, under certain conditions,
when necessary to help a family retain ownership of its home.  Funds
may also be used to improve water and waste disposal systems. 
Guaranteed loans may be made by lenders for the purpose of a
borrower's acquiring or constructing a home. 

The Rural Housing Service's loan program is generally limited to
rural areas, which include communities with populations of less than
10,000 in metropolitan statistical areas (MSA) and communities with
populations between 10,000 and 20,000 in non-MSAs.\7 To be eligible
for a subsidized direct loan, borrowers generally may not have income
in excess of 80 percent of the area's median income.  Borrowers under
the guaranteed loan program may have moderate income--no more than
115 percent of an area's median income.  Borrowers may obtain loans
for up to 100 percent of the value of the property.  The maximum
mortgage amount under the direct and guaranteed loan programs is
$78,660, except in designated high-cost areas.  As with FHA loans,
the maximum amount of Rural Housing Service loans in designated
high-cost areas may be as high as $155,250.  Borrowers must be unable
to secure the necessary credit from other sources at prevailing terms
and conditions for residential-type financing; be a citizen; have
adequate and dependable available income to meet family living
expenses, including taxes, insurance and maintenance, and repayments
on debts, including the proposed loan. 


--------------------
\7 Loans may also be made in areas with a population in excess of
10,000 but less than 20,000 if (1) the area is not included in an MSA
and (2) the Secretaries of Agriculture and Housing and Urban
Development determine that the location has a serious lack of
mortgage credit for low- and moderate-income borrowers. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:7.2

In fiscal year 1995, the Rural Housing Service made 15,405
single-family loans totaling $934 million.  All direct loans carry an
interest subsidy.  In addition, the Rural Housing Service guaranteed
16,677 loans, totaling over $1 billion.  Direct loans were made in
every state, with the greatest activity in California, North
Carolina, Ohio, Pennsylvania, and Texas.  The Rural Housing Service
also guaranteed loans in every state, with the greatest activity in
California, Florida, Georgia, Michigan, and North Carolina.  To meet
program requirements, all borrowers receiving direct loans from the
Rural Housing Service must have incomes that are no greater than 80
percent of the area's median income.  About 27 percent of the
guaranteed loans made in fiscal year 1995 went to borrowers who had
incomes no greater than 80 percent of the area's median income. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:7.3

The guaranty that the Rural Housing Service provides lenders
functions like mortgage insurance.  Under its loan guaranty program,
the Rural Housing Service will pay a lender up to 90 percent of the
lost principal plus accrued interest and liquidation costs in the
event a borrower defaults on a guaranteed loan.  In comparison, FHA
insurance provides lenders with 100 percent coverage of their losses;
private mortgage insurers provide far less coverage.  For its direct
loan program, because the Rural Housing Service holds these
loans--which have interest subsidies--there is no mortgage insurance. 
However, any losses resulting from a borrower's defaulting on a
direct loan are borne entirely by the federal government. 


   HOME INVESTMENT PARTNERSHIPS
   PROGRAM (HOME)
-------------------------------------------------------- Appendix II:8

The HOME program, administered by HUD, provides funds to
participating state and local governments for the development and
support of affordable rental housing and homeownership opportunities
for low-income families.  The program was enacted by the Congress in
the National Affordable Housing Act of 1990.  Using a formula based
on the extent of local housing needs, HUD distributes HOME funds to
participating jurisdictions.\8 A participating jurisdiction may use
its HOME funds for a variety of eligible activities, provided they
are designated in the affordable housing strategy that the program
requires each jurisdiction to prepare.  Each jurisdiction must set
aside at least 15 percent of its HOME allocation for use by
qualified, nonprofit community housing development organizations. 
These organizations use the funds to own, sponsor, or develop
housing. 


--------------------
\8 Before applying the needs formula, some HOME funds are set aside
for Native Americans, insular areas, and technical assistance.  Also,
of the funds allocated by formula, 60 percent are available for
metropolitan cities, urban counties and consortia and 40 percent for
states. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:8.1

Participating jurisdictions are given the flexibility to use HOME
funds in a variety of ways to promote homeownership, depending on
local needs.  They may use HOME funds for equity investments,
interest-bearing and non-interest-bearing loans, interest subsidies,
deferred payment loans, grants, or in some other way.  According to
the program director at HUD, HOME funds for homeownership are
typically used in the form of a second mortgage.  Despite the
flexibility permitted by the HOME program, several restrictions
govern how participating jurisdictions may use funds for
homeownership.  First, all HOME funds used for homeownership
assistance must be targeted at families whose incomes do not exceed
80 percent of the area's median income.  In addition, homes may not
have an initial purchase price higher than 95 percent of the area's
median purchase price and must be the principal residence of the
assisted family. 

Owner-occupied homes for which HOME assistance was provided are also
subject to federally established resale and recapture provisions. 
During a designated recapture period, these homes must be resold to
other low-income families unless the participating jurisdiction
recaptures the full HOME investment from the net proceeds of the
sale.  The recapture period is 5 years for homes in which less than
$15,000 in HOME funds was provided, 10 years for homes in which
$15,000 to $40,000 in HOME funds was provided, and 15 years for homes
in which more than $40,000 in HOME funds was provided. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:8.2

During fiscal year 1995, participating jurisdictions spent $237.9
million in HOME funds on homeownership activities.  These funds were
used to help 18,898 households.  Total budget outlays for fiscal year
1995 were an estimated $1.2 billion. 

All households that receive homeownership assistance through the HOME
program are required to have incomes no greater than 80 percent of
the area's median income.  According to HUD data, of all homeowners
assisted during fiscal year 1995, 8.4 percent had incomes no greater
than 30 percent of the area's median, 23.6 percent had incomes of 31
to 50 percent of the area's median, and 68.0 percent had incomes of
51 to 80 percent of the area's median.  Half of the homeowners
assisted during fiscal year 1995 were minorities. 

Before 1994, regulations required that all HOME funds used for
homeownership assistance go to first-time home buyers.  However,
concern over the slow expenditure of HOME funds prompted several
program changes, including the elimination of this requirement.  As
of August 1994, repeat home buyers also became eligible for HOME
assistance. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:8.3

According to the program director at HUD, FHA mortgage insurance is
sometimes used in conjunction with HOME funds to assist low- and
moderate-income homeowners.  However, these cases are subject to
additional program resale provisions.  For 30 years after HOME funds
are provided, an owner-occupied home with an FHA insured mortgage may
be resold only to other low-income households.  According to a
program official, HUD has issued many waivers to this affordability
period.  There was no information available on the extent to which
HOME activities employ mortgage insurance. 


   HOMEOWNERSHIP AND OPPORTUNITY
   FOR PEOPLE EVERYWHERE, FOR
   HOMEOWNERSHIP OF SINGLE-FAMILY
   HOMES
-------------------------------------------------------- Appendix II:9

Among three Homeownership and Opportunity for People Everywhere
(HOPE) programs, the Homeownership of Single-Family Homes program
which helps low-income families purchase single-family properties
owned by federal, state, and local governments.  The HOPE3 program
was created by the Cranston-Gonzalez National Affordable Housing Act
of 1990.  The program provides grants to private nonprofit
organizations, cooperative associations, or a public body in
cooperation with a private nonprofit organization.  These grants may
be used for planning and implementing homeownership programs designed
to meet the needs of low-income, first-time home buyers.  Planning
grants help applicants develop homeownership programs, and
implementation grants enable applicants to carry out homeownership
programs.  Homeownership programs allow eligible families to acquire
single-family properties owned or held by HUD, the Department of
Veterans Affairs, the Department of Agriculture, the Resolution Trust
Corporation (RTC), a state or local government, or a public housing
authority.  The program is overseen by HUD and carried out by private
nonprofits or by public agencies in cooperation with private
nonprofits.  The criteria for awarding planning grants are the
capability of the applicant, the extent of public/private support,
the need for the homeownership program, and the soundness of the
planning approach.  The criteria for awarding implementation grants
are the capability of the applicant, the extent of public/private
support, the quality of the program's design, the cost-effectiveness
in using federal grant funds, the extent to which the applicant is
committed to promoting the use of minority- and women-owned
businesses, the extent to which the program uses federal properties,
and the degree to which the applicant furthers fair housing choice. 
The fiscal year 1995 appropriation for HOPE3 was $20 million--all
for implementation grants.  In addition, each grant recipient must
ensure that contributions equal to not less than 25 percent of the
implementation grant amounts are provided from nonfederal sources.\9


--------------------
\9 Nonfederal resources may include assistance that an applicant
receives from the Federal Housing Finance Board under its affordable
housing program, so long as the AHP application is approved by the
Finance Board within 30 days of HUD's conditional approval of the
HOPE3 application. 


      PRODUCTS PROVIDED AND THEIR
      RESTRICTIONS
------------------------------------------------------ Appendix II:9.1

HUD provides grant funds to nonprofits that then use these funds to
help eligible families acquire and rehabilitate homes owned or held
by HUD, VA, Agriculture, RTC, state or local governments, or a PHA. 
To make the property affordable to the home buyer, grantees may offer
interest rate reductions, payment of all or a portion of closing
costs, down payments, mortgage insurance premiums, and other
expenses.  Grantees may also use funds to pay for architectural and
engineering work, relocation of residents in eligible properties who
elect to move, temporary relocation of resident home buyers during
rehabilitation, legal fees, reasonable marketing costs, counseling
and training of home buyers, property management and holding costs,
grantee training, economic development directly related to the
homeownership program, and administrative costs.  Mortgages provided
home buyers that are not fully amortizing may not be used. 

To be eligible for homeownership assistance, a family or individual
must have an income that does not exceed 80 percent of the area's
median income and must be a first-time home buyer.  The cost of
acquiring and rehabilitating a property is limited to 80 percent of
the FHA mortgage limit for the area, plus reasonable and customary
closing costs.  The monthly expenditure for principal, interest,
taxes, and insurance must be not less than 20 percent and not more
than 30 percent of the adjusted income of the family (closing costs
included, if financed).  Each eligible family selected must certify
that it intends to occupy the units as its principal residence during
the 6-year period from the date it acquires ownership. 

Restrictions on resale include a limit on the equity the homeowner
may retain, a requirement for the homeowner to execute a
nonamortizing, nonrecourse, non-interest-bearing promissory note for
the difference between the purchase price and fair market value of
the property, and a prior right of the cooperative or PHA to purchase
the property under certain circumstances.  The promissory note is
forgiven over time and is completely forgiven at the end of a 20-year
period.  Each eligible family selected for the program must
participate in counseling and training of home buyers and homeowners
on the general rights and responsibilities of homeownership. 


      ASSISTANCE PROVIDED AND
      CHARACTERISTICS OF
      HOMEOWNERS
------------------------------------------------------ Appendix II:9.2

In 1995, HUD awarded 45 HOPE3 implementation grants totaling $23
million.  From part of these and previous grants, grantees funded
programs that allowed 1,396 families to become homeowners.  Most
properties sold under the HOPE3 program are planned to come from
FHA.\10 A May 1995 report prepared for HUD found that, according to
the plans of the grantees, 52 percent of the properties to be sold
would come from FHA.  The next greatest expected source of properties
was local governments, with 22 percent of properties.  The remaining
properties were planned to come from VA, RTC, RHS, PHAs, and state
governments.  The May 1995 report found that nearly all of the units
sold through the HOPE3 program require some level of rehabilitation. 
According to this study,

     "20 percent of all grantees engage in minor rehab at a cost of
     less than $10,000 per unit.  On the other hand, half (51
     percent) undertake moderate rehab costing between $10,000 and
     $30,000 per unit, and 29 percent engage in extensive rehab in a
     typical unit at a cost of more than $30,000."\11

In connection with the type of financial assistance that homeowners
received, the study surveyed all grantees during January and February
1994 and found that the most common forms of assistance were closing
cost assistance (86 percent of grantees reporting), down payment
assistance (79 percent), below-market interest rate (61 percent),
reduction in sales price (48 percent), deferred loan payment (39
percent), sweat equity (22 percent), and other grants (19 percent). 
In connection with loan sources, 64 percent of grantees reported that
they planned to use conventional loans, 50 percent reported that they
planned to use HOPE3 direct loans, 28 percent reported that they
planned to use FHA, VA, and RHS loans, 21 percent reported that they
planned to use loans from state HFAs, and 17 percent planned to use
loans from the local government.  Finally, each eligible family
selected for the program is required to participate in counseling and
training on the general rights and responsibilities of homeownership. 

The HUD program office did not have data on the characteristics of
persons who became homeowners through the HOPE3 program.  However,
to be eligible for assistance under the HOPE3 program, a family must
have an income that does not exceed 80 percent of the area's median
income and must be a first-time home buyer.  The May 1995 study found
that according to information on the 211 households that had
purchased properties as of May 1994, 35 percent had incomes below 50
percent of the area's median income, 62 percent were minorities, and
nearly half were single-parent families. 


--------------------
\10 With the exception of the period from November 1993 through
September 1994, HOPE3 grantees have since November 1992 been allowed
to bid on FHA properties along with other prospective "owner
occupants" in advance of their being offered to the general public. 
In addition, if a HOPE3 grantee wins the bid, he or she receives a
10-percent discount off the winning price. 

\11 Evaluation of the Hope 3 Program, prepared for HUD's Office of
Policy Development and Research by Abt Associates, Inc., May 1995. 


      USE OF MORTGAGE INSURANCE
------------------------------------------------------ Appendix II:9.3

FHA-insured loans represented about 19 percent of home buyer
financings under the HOPE3 program.  Twelve percent of financings
were loans insured under FHA's Section 203(b) program, and about 2
percent were loans insured under FHA's Section 203(k) rehabilitation
mortgage program or other programs.  About 48 percent of all
financings were conventional mortgages.  Whether those who financed
their home purchase with a conventional loan had private mortgage
insurance is not known.  If a mortgage is insured by FHA, the
requirements of FHA apply with certain exceptions:  the borrower may
obtain a loan for the down payment from a corporation or another
person; a second mortgage may be placed against the property by an
entity that is not a federal, state, or local government agency; and
certain other restrictions on conveyances exist. 




(See figure in printed edition.)Appendix III
COMMENTS FROM THE MORTGAGE
INSURANCE COMPANIES OF AMERICA
========================================================== Appendix II



(See figure in printed edition.)




(See figure in printed edition.)Appendix IV
COMMENTS FROM THE NATIONAL COUNCIL
OF STATE HOUSING AGENCIES
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)




(See figure in printed edition.)Appendix V
COMMENTS FROM THE FEDERAL HOUSING
FINANCE BOARD
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)




(See figure in printed edition.)Appendix VI
COMMENTS FROM THE NEIGHBORHOOD
REINVESTMENT CORPORATION
========================================================== Appendix II



(See figure in printed edition.)



(See figure in printed edition.)


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII

RESOURCES, COMMUNITY, AND ECONOMIC
DEVELOPMENT DIVISION

Leslie Black-Plumeau
DuEwa Kamara
Cheryl Kramer
Donna Lucas
Robert Procaccini
Mathew Scire
Patrick Valentine

RELATED GAO PRODUCTS

Mortgage Financing:  FHA Has Achieved Its Home Mortgage Capital
Reserve Target (GAO/RCED-96-50, Apr.  12, 1996). 

Government Corporations:  Profiles of Existing Government
Corporations (GAO/GGD-96-14, Dec.  13, 1995). 

Community Reinvestment Act:  Challenges Remain to Successfully
Implement CRA (GAO/GGD-96-23, Nov.  28, 1995). 

Rural Housing:  Opportunities Exist for Cost Savings and Management
Improvement (GAO/RCED-96-11, Nov.  16, 1995). 

Property Disposition:  Information on HUD's Acquisition and
Disposition of Single-Family Properties (GAO/RCED-95-144FS, July 24,
1995). 

Housing Finance:  Improving the Federal Home Loan Bank System's
Affordable Housing Program (GAO/RCED-95-82, June 9, 1995). 

Community Reinvestment Act:  Preliminary Results of GAO's Study on
CRA Problems and Proposed Reforms (GAO/T-GGD-95-113, Mar.  8, 1995). 

Rural Housing:  Shift to Guaranteed Program Can Benefit Borrowers and
Reduce Government's Exposure (GAO/RCED/AIMD-95-63, Dec.  21, 1994). 

Housing Finance:  Implications of Alternative Methods of Adjusting
the Conforming Loan Limit (GAO/RCED-95-6, Oct.  5, 1994). 

Tax Policy:  Tax Expenditures Deserve More Scrutiny
(GAO/GGD/AIMD-94-122, June 3, 1994). 

Housing Finance:  Characteristics of Borrowers of FHA-Insured
Mortgages (GAO/RCED-94-135BR, Apr.  6, 1994). 

Government National Mortgage Association:  Greater Staffing
Flexibility Needed to Improve Management (GAO/RCED-93-100, June 30,
1993). 

Community Development:  Neighborhood Reinvestment Corporation Should
Improve Program Management (GAO/RCED-92-174, July 8, 1992). 

Federal Agricultural Mortgage Corporation:  Potential Role in the
Delivery of Credit for Rural Housing (GAO/RCED-91-180, Aug.  7,
1991). 


*** End of document. ***