oversight

Homeownership: Problems Persist With HUD's 203(k) Home Rehabilitation Loan Program

Published by the Government Accountability Office on 1999-06-14.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                  United States General Accounting Office

GAO               Report to the Chairman, Subcommittee
                  on Housing and Community Opportunity,
                  Committee on Banking and Financial
                  Services, House of Representatives

June 1999
                  HOMEOWNERSHIP
                  Problems Persist With
                  HUD’s 203(k) Home
                  Rehabilitation Loan
                  Program




GAO/RCED-99-124
      United States
GAO   General Accounting Office
      Washington, D.C. 20548

      Resources, Community, and
      Economic Development Division

      B-280930

      June 14, 1999

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

      Dear Mr. Chairman:

      The 203(k) Home Rehabilitation Mortgage Insurance program was
      established to help promote the rehabilitation and repair of the nation’s
      housing stock through a program that combines, in one insured mortgage,
      the funds needed to purchase and rehabilitate a single-family home. These
      loans are made by banks and other private lenders from their own funds
      and are insured by the Department of Housing and Urban Development’s
      (HUD) Federal Housing Administration (FHA). If a borrower defaults and the
      lender subsequently forecloses on the loan, the lender can file an
      insurance claim with HUD for nearly all of its losses. Over the past 4 years,
      reports of abuses within the 203(k) program have raised serious concerns
      about HUD’s management of it. Collectively, these reports—by HUD’s
      Inspector General and others—indicate that the 203(k) program is highly
      vulnerable to waste, fraud, and abuse.

      Because of your concerns with HUD’s management of the program, you
      asked us to determine (1) the risk the 203(k) program poses to FHA’s
      insurance fund1 relative to the 203(b) program,2 (2) HUD’s efforts to correct
      program deficiencies identified by HUD’s Inspector General and others, and
      (3) weaknesses, if any, in HUD’s oversight of the 203(k) program and the
      extent to which lenders are complying with HUD’s underwriting guidelines
      for making program loans. We have also provided information on the
      growth in the program, the performance of the program’s loans overall and
      by borrower type, and the customers served by the program. (See app. I.)

      To address these issues, we focused on the activities of HUD’s headquarters
      and its four Homeownership Centers located in Santa Ana, California;

      1
       Loans under the 203(k) Home Rehabilitation Mortgage Program are insured by FHA’s General
      Insurance Fund, which provides for a number of specialized mortgage insurance programs.
      2
       FHA provides most of its single-family mortgage insurance through the 203(b) program, which covers
      loans for purchasing a new or existing one-to four-family home. Loans under the 203(b) program are
      insured by FHA’s Mutual Mortgage Insurance fund. In fiscal year 1998, loans worth over $90 billion
      were insured under the 203(b) program.



      Page 1                                   GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                   B-280930




                   Denver, Colorado; Atlanta, Georgia; and Philadelphia, Pennsylvania. Each
                   of these centers is responsible for the general management of the program
                   in its region. We also reviewed reports and recommendations by HUD’s
                   Inspector General, KPMG Peat Marwick, and HUD management. We
                   compared the 203(k) program with the 203(b) mortgage insurance
                   program because the 203(b) program is FHA’s largest single-family
                   mortgage insurance program and HUD oversees both programs in the same
                   manner.


                   The 203(k) loan program is more risky than HUD’s largest single-family loan
Results in Brief   program—the 203(b) program—because it combines the risk of a
                   traditional mortgage with the risk of a construction loan. For loans
                   endorsed3 from fiscal years 1994 through 1996,4 the claim rate5 for loans
                   made under the 203(k) program is almost double that of loans made under
                   the 203(b) program. In addition, HUD projects that while loans made under
                   the 203(b) program in fiscal year 1994 through fiscal year 1998 will make
                   money for the Mutual Mortgage Insurance Fund, it projects that 203(k)
                   loans will cost the General Insurance Fund—that is, claims and other costs
                   will exceed premiums and other income by over $25 million.6 This cost
                   represents approximately .7 percent of the total amount insured by the
                   program, as of the end of fiscal year 1998. HUD stated that it finds this loss
                   rate to be expected for a home rehabilitation program. Figure 1 illustrates
                   the expected losses from loans endorsed in fiscal years 1994 through 1998.7




                   3
                    After making a loan to a borrower, a lender seeks FHA’s approval to insure the loan. When FHA
                   formally approves mortgage insurance for the loan, it is considered “endorsed.”
                   4
                    These are the most recent years with sufficient data to establish claim rates.
                   5
                    A claim rate is the percentage of loans endorsed in a specific year that result in a claim being filed
                   against the insurance fund.
                   6
                    This figure represents the net present value of current and future projected losses.
                   7
                    Program losses are expressed in fiscal year 1999 dollars.



                   Page 2                                      GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                       B-280930




Figure 1: Projected Losses to the
General Insurance Fund for 203(k)       $2      Projected Losses (in millions of dollars)
Loans Made in Fiscal Year 1994
Through Fiscal Year 1998, Dollars in
Millions
                                         0




                                        -$2




                                        -$4




                                        -$6




                                        -$8




                                       -$10




                                       -$12

                                                1994      1995      1996     1997       1998
                                                Fiscal Year of Endorsement



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




                                       Figure 1 shows that anticipated losses will increase from approximately
                                       $2 million for loans endorsed in fiscal year 1995 to over $11 million for
                                       loans endorsed in fiscal year 1997.

                                       Despite the recognized risk associated with the 203(k) program and the
                                       potential for mounting losses to the General Insurance Fund, HUD has done
                                       little to address the problems identified by its Inspector General and
                                       others. In the past 4 years, reports by HUD’s management, HUD’s Inspector
                                       General, and others have repeatedly cited problems and weaknesses in the
                                       management of this program, indicating that it is highly vulnerable to
                                       waste, fraud, and abuse. For example, HUD’s Inspector General reported
                                       that the program design encourages risky property deals and overstated
                                       property appraisals and does not adequately safeguard HUD’s interests.



                                       Page 3                                       GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
             B-280930




             Furthermore, in its review of a sample of loans made to investors and
             nonprofit organizations, the Inspector General found that many of the
             homes had not been properly rehabilitated. Although HUD’s management
             has, for the most part, agreed with the findings as reported, it has done
             little to address the problems. The most recent study of the
             program—completed in October 1998 by outside contractors—found that
             the Department had done little to address the long term viability of the
             program and recommended that HUD radically redesign or eliminate it.

             HUD is not adequately overseeing key aspects of the 203(k) program. With
             respect to lenders, HUD’s Homeownership Centers do not adequately
             ensure that lenders are complying with the program’s guidelines. At the
             four Homeownership Centers, we found that 203(k) loans were not
             targeted for review to ensure that the lenders are properly administering
             them. In one center, HUD management made a conscious decision not to
             review 203(k) loans because it lacked trained staff. Furthermore, HUD does
             not properly train and monitor 203(k) home inspectors and consultants,
             who are responsible for designing and overseeing the home rehabilitation
             process. We also found cases in which the agency failed to address
             consultants’ abuses or incompetence. For example, we found incidences
             in which a consultant who was also an inspector approved payments for
             work by contractors that was either not completed or was completed
             improperly. Furthermore, HUD still does not adequately ensure that
             nonprofit organizations comply with HUD guidelines for participating in the
             program.

             This report makes recommendations designed to improve HUD’s
             management and oversight of the 203(k) program.


             The 203(k) Home Rehabilitation Mortgage Insurance program8 is HUD’s
Background   primary program for rehabilitating and repairing single-family homes. A
             single loan insured under the 203(k) program can be used to both
             purchase and rehabilitate a home. The 203(k) program, like other FHA
             programs, insures mortgage loans to encourage lenders to make loans
             available to borrowers who would not otherwise qualify for conventional
             loans on affordable terms. Eligible borrowers may include the
             owner/occupant, nonprofit organizations, and until October 1996,




             8
             The Rehabilitation Home Mortgage Insurance program was authorized by section 203(k) of the
             National Housing Act, as amended.



             Page 4                                  GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
B-280930




investors.9 The General Insurance Fund supports the 203(k) program as
well as other specialized housing programs.10 Unlike the Mutual Mortgage
Insurance Fund, which supports the 203(b) program, the General
Insurance Fund is not intended to be self-sustaining. When the General
Insurance Fund incurs losses, funds are appropriated by the Congress. As
of fiscal year 1998, HUD had about $3.6 billion in insurance on 44,000
outstanding 203(k) loans.

The 203(k) program allows borrowers to finance both the purchase or
refinancing of a house and the cost of its rehabilitation through a single
mortgage or to finance the rehabilitation of their existing home. When
buying a house that is in need of repair or modernization, homebuyers not
using the 203(k) program often must follow a complicated and costly
process: They must obtain financing to purchase the property, get
additional financing to rehabilitate it, and find a permanent mortgage after
rehabilitation is completed to pay off the interim loans. The interim
acquisition and improvement loans often have relatively high interest rates
and short repayment terms. However, the 203(k) program helps both
borrowers and lenders by insuring a single, long-term, fixed- or
adjustable-rate loan to cover the costs of both the acquisition and the
rehabilitation. A loan insured under the 203(k) program also protects a
lender by allowing it to have the loan insured for the full value of the
rehabilitated home before the rehabilitation process begins.

Compared with other FHA mortgage loan programs, such as the 203(b)
program, the structure of the 203(k) program is far more complex. Like the
203(b) program, the 203(k) program provides mortgage insurance to
protect lenders against the risk of default on loans to qualified buyers and
may be used to finance the purchase of single-family housing as well as to
refinance debt. The completion of a 203(k) loan, however, involves
multiple entities and estimates. At closing, a lender must set aside the
estimated funds to pay for the rehabilitation in an escrow account. A
HUD-approved consultant is often needed to determine the extent of work
that must be done to rehabilitate a property and the estimated cost of that
work. A HUD-approved inspector is needed to monitor the progress of the
rehabilitation and to co-sign with the borrower any request of escrow
funds. HUD guidelines allow the consultant to also function as the
inspector. The procedures for the loan application and rehabilitation
process are discussed in appendix II.

9
 Because of abuses by investors in the program, a moratorium on investor participation was
implemented in October 1996.
10
 The General Insurance Fund provides for a large number of specialized mortgage insurance
programs, including insurance for loans for condominiums, land development, and nonprofit hospitals.


Page 5                                   GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                   B-280930




                                   For fiscal year 1994 through fiscal year 1998, HUD endorsed 62,757 loans
                                   for 203(k) insurance, with more than 80 percent of the endorsements made
                                   in fiscal years 1996 through 1998. (See fig. 2.) Although the 203(k) program
                                   was established in its present form in 1978, HUD began promoting and
                                   streamlining it in 1994 to make it more acceptable to lenders and
                                   borrowers, and by 1995, program use had noticeably increased. Since
                                   March 1994, HUD has made a number of changes to the 203(k) program to
                                   make it more user-friendly for lenders and borrowers. These changes were
                                   also made to streamline and shorten the processing time for 203(k) loans.
                                   For example, in an effort to streamline the process, HUD eliminated the
                                   requirement for lenders to submit change orders to the HUD field office for
                                   an extension of time to complete improvements. In 1995, HUD made
                                   additional changes to the program that were recommended by a working
                                   group that consisted of HUD offices, lenders, nonprofit organizations, and
                                   government agencies. The changes included (1) making the HUD offices
                                   responsible for ensuring that consultants are properly trained and
                                   (2) allowing consultants to serve as inspectors in order to reduce the
                                   processing time of a 203(k) loan.


Figure 2: Number of 203(k) Loans
Endorsed, Fiscal Years 1990-98     20,000 Number of loans endorsed

                                   18,000

                                   16,000

                                   14,000

                                   12,000

                                   10,000

                                    8,000

                                    6,000

                                    4,000

                                    2,000

                                       0

                                              1990   1991    1992     1993      1994    1995    1996    1997   1998
                                              Year



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




                                   Page 6                                    GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                       B-280930




                       The management and oversight of the 203(k) program is distributed among
                       headquarters staff and the Department’s four regional Homeownership
                       Centers in Philadelphia, Pennsylvania; Atlanta, Georgia; Denver, Colorado;
                       and Santa Ana, California. Each regional Homeownership Center is
                       responsible for the operation of the 203(k) program in its territory. These
                       responsibilities include processing and underwriting; lender monitoring;
                       marketing; and customer assistance. In 1994, FHA’s Office of Housing
                       began consolidating some of its mortgage insurance processes from its 81
                       field offices into the four Homeownership Centers in order to provide
                       faster processing. With the exception of a few field offices that have not
                       transferred to the Homeownership Centers, the reorganization is
                       complete.


                       The 203(k) program’s design, as well as its past performance, indicate that
Program Design and     future claims from defaulted loans may lead to losses to the General
Performance            Insurance Fund. As a result of HUD’s streamlining of the program in 1994,
Demonstrate a High     the value of insured 203(k) loans grew from just under $400 million in
                       fiscal year 1994 to almost $3.6 billion in fiscal year 1998, an increase of
Degree of Risk         about 835 percent. Because of the program’s complex design and relatively
                       high claim rate, this increase in the dollar amount of loans insured has
                       increased the risk of loss to the federal government. After deducting
                       premiums paid and other income, HUD projects that net losses to the
                       General Insurance Fund will exceed $25 million from loans made during
                       fiscal years 1994 through 1998. These losses represent approximately
                       .7 percent of the total amount insured by the program, as of the end of
                       fiscal year 1998. HUD stated that this loss rate is to be expected for a home
                       rehabilitation program.


Program Design Poses   The process for approving and disbursing 203(k) loans is far more
Risk to the General    complex than the processes for other FHA-insured single-family lending
Insurance Fund         programs. The complexity stems from the rehabilitation component of the
                       program, which (1) places far greater emphasis on estimates, reports, and
                       opinions; (2) has more individual steps required in the underwriting and
                       funding process; (3) involves multiple participants; and (4) requires a
                       greater sophistication on the part of the borrower. Both internal and
                       outside reviews of the 203(k) program have found that the program’s
                       design presents a high degree of risk to the Department and the General
                       Insurance Fund. In a February 1997 report on the program,11 HUD’s

                       11
                          Audit Report: Section 203(k) Rehabilitation Mortgage Insurance Program (HUD/OIG
                       97-AT-121-0001,Feb. 6, 1997).



                       Page 7                                   GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                              B-280930




                              Inspector General concluded that the program design does not adequately
                              safeguard HUD’s interests. The report stated that the program design
                              encourages risky property deals, overstated property appraisals, and
                              phony or excessive fees. Furthermore, an internal HUD study of the 203(k)
                              program reported that because the loan program is designed to let
                              borrowers purchase and rehabilitate property, it poses multiple inherent
                              risks, including the failure to complete work in an acceptable manner and
                              to accurately estimate the cost of rehabilitation. In an October 1998 draft
                              study,12 HUD contractors asserted that, by virtue of its complexity, the
                              203(k) program poses a high risk of loss to the Department and that this
                              risk has been reflected in high default and claim rates.


Past Performance of           The performance of loans made under the 203(k) program is poorer than
203(k) Program Indicates      that of loans made under HUD’s 203(b) program. For loans endorsed during
It Poses an Increasing Risk   fiscal year 1994 through fiscal year 1996—the most current years with
                              sufficient data to determine meaningful claim rates—the cumulative claim
of Loss                       rates through 1998 for 203(k) loans are almost double the rates for 203(b)
                              loans made in the same years. (See fig. 3.)




                              12
                                KPMG/META’s Front End Risk Assessment (FERA)-Section 203(k) Rehabilitation Mortgage
                              Insurance Program, Oct. 23, 1998. HUD officials told us they consider this study to be a draft until they
                              issue a written response to it.



                              Page 8                                     GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                         B-280930




Figure 3: Cumulative Claim Rates
Through 1998 for the 203(k) and 203(b)   7.0 Percent
Programs for Loans Endorsed in
Fiscal Years 1994-96
                                         6.0



                                         5.0



                                         4.0



                                         3.0



                                         2.0



                                         1.0



                                         0.0


                                               1994                  1995                    1996

                                                        203(k) Cumulative Claim Rate

                                                        203(b) Cumulative Claim Rate



                                         Note: Cumulative claim rates are total claims to date for a specific year of loan endorsements, as
                                         of Sept. 30, 1998.

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




                                         Figure 3 shows that cumulative claim rates for the 203(k) program are
                                         consistently higher than those for the 203(b) program for the years shown.
                                         For example, as of September 30, 1998, the cumulative claim rate for
                                         203(k) loans endorsed in fiscal year 1994 was 6.1 percent, while the claim
                                         rate for 203(b) loans endorsed in the same year was 2.4 percent.

                                         Furthermore, HUD projects that the 203(k) program will incur net losses for
                                         loans endorsed in each of fiscal years 1995 through 1998, while the 203(b)
                                         program is projected to incur net gains for the same period. For example,
                                         the projected net loss, as a percent of endorsements, for 203(k) loans
                                         made in fiscal year 1997 is .75 percent of the total amount of endorsements



                                         Page 9                                        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                           B-280930




                           in that year. The projected net gain for 203(b) loans made in that year, as a
                           percent of endorsements, is 1.7 percent.

                           According to HUD, the projected net loss/gain for the General Insurance
                           Fund for the 203(k) program is expected to go from a surplus of $810,000
                           for loans endorsed in fiscal year 1994 to a deficit of $11.6 million for loans
                           endorsed in fiscal year 1997. This is in contrast to the 203(b) program, for
                           which HUD projects each year’s loans will add economic value to the
                           Mutual Mortgage Insurance Fund.


Performance Indicators     The cumulative claim rates of loans made under the 203(k) program will
Suggest Cumulative Claim   most likely increase. The recent growth in the 203(k) program, its
Rates May Increase         historical pattern of claim rates, and an increase in early default rates,13
                           suggest that the percentage of loans resulting in claims for the 203(k)
Significantly              program will increase in the next few years.

                           At least 75 percent of the loans endorsed under the 203(k) program since
                           fiscal year 1994 were made in the past 3 years. According to an external
                           review of historical and projected claim rate data for HUD-insured loans,14
                           during the 1990s the percentage of 203(b) loans resulting in claims is at its
                           highest in the fourth and fifth year after the loans were endorsed. If this
                           pattern holds for 203(k) loans, and it has so far, claim rates for loans
                           endorsed in fiscal year 1996 through fiscal year 1998 will continue to
                           increase over the next 4 years, resulting in increasing claim outlays. The
                           review also found that the claim rate begins to decrease in the sixth year
                           and beyond. For instance, as of fiscal year 1995, only 5 of the
                           approximately 3,700 loans made under the 203(k) program in fiscal year
                           1994 had resulted in a claim against the General Insurance Fund, a
                           cumulative claim rate of .1 percent. By fiscal year 1998, 235 additional
                           claims were filed against the fund from loans made in fiscal year 1994, and
                           the cumulative claim rate for these loans grew to approximately
                           6.5 percent.

                           While the 203(k) program was growing by over 800 percent, the early
                           default rate was increasing. HUD tracks the early default rates of 203(k)
                           loans by lender to determine how well lenders are underwriting them. This
                           rate was about 2.3 percent for 203(k) loans made in fiscal year 1994,

                           13
                            Early defaults are all loans that have defaulted within 1 year of loan closing. A mortgage loan is
                           considered in default when the borrower misses three consecutive monthly payments.
                           14
                            Price Waterhouse, An Actuarial Review for Fiscal Year 1998 of the Federal Housing Administration’s
                           Mutual Mortgage Insurance Fund, Final Report, Mar. 1, 1999.



                           Page 10                                    GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                           B-280930




                                           5.5 percent for fiscal year 1995 loans, and 5.2 percent for fiscal year 1996
                                           and fiscal year 1997 loans.


Figure 4: Early Default Rates for 203(k)
Program for Loans Endorsed in Fiscal       12 Percent of early defaults
Years 1994 Through 1997

                                           10



                                            8



                                            6



                                            4



                                            2



                                            0

                                                   1994             1995           1996       1997
                                                Fiscal year of closing

                                                            Within 1 Year

                                                            Within 2 Years

                                                            Volume



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




                                           According to HUD, an early default rate is a good indication of a lender’s
                                           underwriting quality and the likelihood that a loan will result in a claim
                                           against the insurance fund. A HUD official added that loans that default
                                           within 12 months after they are endorsed frequently have serious
                                           underwriting problems associated with them.




                                           Page 11                                  GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                             B-280930




                             Despite the risk of loss the 203(k) program poses to the General Insurance
HUD Management               Fund, HUD management has done little to address problems with the
Has Done Little to           program identified by both internal and external reviews over the past 4
Address Recognized           years. These reviews indicate that the program is highly vulnerable to
                             fraud, waste, and abuse and make a number of recommendations to
Problems in the              strengthen the program. Although HUD’s management has agreed with
203(k) Program               these reports and indicated a willingness to implement most of the
                             recommendations, it has failed to do so in many cases.


HUD Management Fails to      Reporting on the results of a 3-year review in February 1997,15 HUD’s
Adequately Address           Inspector General stated that the involvement of investors and nonprofit
Problems Identified by the   organizations in the 203(k) program increased HUD’s risk of loss to the
                             General Insurance Fund from defaults. Many of the homes covered by the
Inspector General            loans the Inspector General reviewed had not been properly rehabilitated.
                             The report concluded that the program design encourages risky property
                             deals and overstated property appraisals and does not adequately
                             safeguard HUD’s interests. The Inspector General made several
                             recommendations for improving the program, such as permanently barring
                             investors from the program to curb their abuses. The Inspector General
                             added that anything less than a permanent ban was unacceptable because
                             of the substantial waste, fraud, and abuse that the review had discovered.

                             In addition, in an August 1997 report, the Inspector General concluded that
                             HUD  had no effective criteria for approving consultants16 and consultant
                             trainers and recommended that HUD develop uniform criteria.

                             In May 1998, the Inspector General commented positively on
                             owner-occupant participation in the program but expressed concerns
                             about lenders’ performance. Auditors found problems with the
                             underwriting in 40 of the 50 loans they reviewed. For example, several of
                             the lenders were approving borrowers who did not meet the eligibility
                             criteria for 203(k) loans. The Inspector General found that HUD needed to
                             take steps to ensure more effective lender performance in originating
                             loans and administering the rehabilitation process.

                             Although it formally agreed to implement most of the Inspector General’s
                             recommendations in these reports, HUD has failed to do so. For example,

                             15
                                Audit Report: Section 203(k) Rehabilitation Mortgage Insurance Program (HUD/OIG 97-AT-121-0001,
                             Feb. 6, 1997).
                             16
                              Consultants are responsible for preparing work write-ups and cost estimates and generally
                             overseeing the rehabilitation process to protect the interests of the borrower.



                             Page 12                                  GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                            B-280930




                            the Inspector General had recommended that HUD develop written criteria
                            for approving and training 203(k) program consultants. In response, in an
                            October 1997 memorandum, HUD stated that it would develop a national
                            certification examination for consultants. According to a senior HUD
                            official, HUD has no plans in the near future to contract for, or otherwise
                            develop, this examination.

                            HUD management has taken some steps in response to the Inspector
                            General’s findings to minimize the program’s exposure to fraud, waste, and
                            abuse. For instance, in October 1996, HUD placed a moratorium on investor
                            participation in the program. HUD has issued several other policy directives
                            clarifying for lenders the Department’s policy related to, among other
                            things, lender underwriting and the disbursal of funds. HUD has also
                            eliminated certain provisions, such as providing FHA insurance on 203(k)
                            second mortgages, which placed HUD at unacceptable risk.

                            In addition, in 1997, HUD developed a draft policy directive designed to
                            help curb fraud and abuse by program participants. The draft directive
                            proposes several changes to the program, including limiting the number of
                            properties that nonprofit organizations can purchase under the program
                            and imposing stricter requirements for approving these organizations.
                            Although the policy was to be effective in October 1997, the directive was
                            never issued because the Acting Housing Commissioner was hesitant to
                            make changes to the program until a permanent Housing Commissioner
                            was appointed. The permanent Housing Commissioner was appointed in
                            October 1998.


HUD Fails to Take Action    In October 1998, the contractors hired by HUD to study the 203(k) program
on Problems Identified by   reported that the Department has done little to reduce the risks associated
Independent Studies         with the program, and hence, its long-term viability.17 Specifically, the
                            draft report identified six major risks associated with the 203(k) program:
                            (1) program complexity, (2) insufficient lender monitoring, (3) inadequate
                            guidance concerning consultants, (4) hesitant management direction,
                            (5) increased loss potential from nonprofit organizations, and (6) low
                            program volume. The contractors reported that continued inaction on the
                            part of HUD management will likely contribute to a downward trend in the
                            performance of the program, putting it in jeopardy. The contractors
                            recommended that HUD either eliminate the program or radically redesign


                            17
                              KPMG/META, Front End Risk Assessment (FERA)- Section 203(k) Rehabilitation Mortgage
                            Insurance Program, Oct. 23, 1998. HUD officials told us they consider this study to be a draft until they
                            issue a written response to it.



                            Page 13                                    GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                          B-280930




                          it. They added that unless HUD can devote the resources to the program to
                          properly manage it, the program should be eliminated.

                          Responding to the contractors’ study, HUD indicated that it would not take
                          immediate action to correct the problems identified. According to the
                          Acting Director of Single-Family Housing, the new FHA Housing
                          Commissioner, who took office in October 1998, has been devoting most
                          of his attention to addressing issues in programs that are much larger in
                          scope and volume than the 203(k) program. According to this official,
                          when resources are freed from addressing these other programs,
                          management would probably prepare a comprehensive plan to improve
                          the 203(k) program.


                          HUD provides little direct oversight over the 203(k) program. We identified
HUD’s Oversight of        three areas of particular concern. HUD is not (1) targeting the 203(k)
the 203(k) Program Is     program for oversight, (2) properly training and overseeing consultants
Inadequate                and home inspectors, and (3) adequately monitoring nonprofit
                          organizations’ participation in the program.


HUD Does Not Target the   Although HUD recognizes the program’s unique risks and potential for
203(k) Program for        abuse, it has not targeted the 203(k) program to minimize those risks.
Oversight                 Instead, HUD oversees the 203(k) program using the same approach it uses
                          to oversee the less-risky single-family insurance programs, such as HUD’s
                          203(b) Home Mortgage Insurance Program. For these programs, HUD’s four
                          Homeownership Centers conduct post-endorsement technical reviews and
                          quality assurance reviews, which are HUD’s primary means to oversee
                          lenders.

                          HUD  has not used post-endorsement technical or quality assurance reviews
                          effectively to minimize the unique risks posed by the 203(k) program.
                          Post-endorsement technical reviews are desktop audits of a sample of loan
                          cases that have already been endorsed for insurance by FHA. These reviews
                          are designed to determine the quality of a lender’s underwriting of an
                          endorsed loan. For example, the reviewer will ensure that all required loan
                          application documents were obtained or completed by the lender and that
                          these documents were properly signed. Quality assurance reviews occur at
                          a lender’s place of business. Trained HUD staff perform in-depth reviews of
                          troubled loans and evaluate a lender’s internal control system for
                          originating FHA-insured loans.




                          Page 14                        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                          B-280930




                          We found that 203(k) loans, although recognized as being high-risk, are not
                          targeted for these reviews. HUD officials told us that they plan to start
                          targeting 203(k) loans for review this year. We also found that the detailed
                          results of the post-endorsement reviews are placed in the loan file; they
                          are not given to the lender. Twice a year, HUD sends the ratings from the
                          post-endorsement reviews to the lenders but does not specify why they
                          were assigned a given rating. As a result, HUD and the lenders cannot take
                          action on any problems identified in the review process in a timely
                          manner.

                          HUD officials told us that while they do not target 203(k) loans for review,
                          they do perform quality assurance reviews on lenders that make these
                          loans. Since January 1, 1996, HUD has performed 52 quality assurance
                          reviews on lenders that make 203(k) loans. Seventeen of these reviews
                          were on lenders that made over 100 203(k) loans during that period.
                          However, HUD could not provide information on how many 203(k) loans
                          may have been reviewed.

                          HUD has emphasized increasing the number of 203(k) loans endorsed by
                          streamlining the program, but it has not committed the resources needed
                          to adequately oversee the program. Officials at two Homeownership
                          Centers told us that they do not have staff qualified to do quality assurance
                          reviews for 203(k) loans. According to an official at a Homeownership
                          Center responsible for 17 states, the Center did not have sufficient travel
                          funds to adequately oversee any of its single-family programs, including
                          the 203(k) program, during the first quarter of fiscal year 1999.


HUD Is Not Properly       We found that HUD has no uniform criteria for training, approving, and
Training and Overseeing   evaluating the performance of consultants and inspectors who participate
Consultants and Home      in the 203(k) program. Consultants and inspectors can be used,
                          interchangeably, to perform home inspections, identify health and safety
Inspectors                problems, and provide descriptions of the work to be performed and cost
                          estimates for the buyers. In addition to having at least 3 years of
                          specialized experience, consultants/inspectors must receive training in the
                          203(k) program. However, at two of the four centers we visited, HUD had
                          not provided any training to 203(k) consultants/inspectors and had not
                          taken any action to evaluate the performance of approved
                          consultant/inspectors. One of the other centers provided minimal training,
                          such as holding training sessions via teleconference, but had not taken any
                          action to evaluate the performance of approved consultants. According to




                          Page 15                         GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
B-280930




a HUD official, many consumer complaints about the program result from
the work of consultants and inspectors.

We also found cases in which the agency failed to address consultants’
abuses or incompetence. For example, according to customer complaints
we reviewed, a Chicago consultant/ inspector allowed a contractor to
receive thousands of dollars for work the contractor either did not do or
did not do adequately. In a request for payment, the consultant/inspector
approved payments to the contractor for a ceiling installation that was
never completed. Figure 5 shows the incomplete ceiling.




Page 16                       GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                         B-280930




Figure 5: Incomplete Ceiling and Light
Fixture Installation




                                         Source: GAO.




                                         The same inspector reported that plumbing work had been completed.
                                         Figure 6 shows that this was not the case.




                                         Page 17                      GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                     B-280930




Figure 6: Incomplete Plumbing Work




                                     Source: GAO.




                                     Figures 7 through 9 show incomplete electrical work in another Chicago
                                     home that was recorded as completed.




                                     Page 18                       GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                    B-280930




Figure 7: Exposed Breaker Box and
Wires Connected to an Old Breaker
Box




                                    Source: GAO.




                                    Page 19        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                        B-280930




Figure 8: Hanging Electrical Wires in
Closet




                                        Source: GAO.




                                        Page 20        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                          B-280930




Figure 9: Electrical Wiring Strung
Along Outside of Wall Instead of Inside
the Wall




                                          Source: GAO.




                                          Similarly, in a recent report, HUD’s Inspector General found that a Michigan
                                          HUD field office failed to take action against a consultant who the office
                                          knew was either incompetent or was abusing the program. This consultant
                                          was the exclusive consultant/inspector for four of the largest 203(k)
                                          lenders in the area.


HUD Is Not Adequately                     HUD has failed to properly monitor the participation of nonprofit
Monitoring Nonprofit                      organizations in the 203(k) program. HUD guidelines require that the
Organizations’                            Homeownership Centers recertify nonprofit organizations every 2 years.
                                          At three of the four centers we visited, we did not find evidence that the
Participation in the                      centers had recertified any of the approved nonprofit organizations for the
Program                                   program.




                                          Page 21                        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
              B-280930




              While they represent only about 5 percent of the 203(k) program’s loan
              volume, nonprofit organizations’ participation is just as problematic as
              investors’ participation. Nonprofit organizations’ claim rate, on average, is
              more than double that of any other borrower type in the 203(k) program,
              including investors. For fiscal years 1994 through 1998, the claim rate for
              nonprofit organizations was 6.3 percent, whereas the claim rate for
              investors and owner/occupants was 2.7 percent and 1.2 percent,
              respectively.

              In 1997, the Inspector General reported that nonprofit organizations
              presented the same level of risk to the program as investors and that their
              participation in the program has failed to achieve HUD’s objectives.
              Nonprofit organizations’ 203(k) loans are so risky and problematic that
              one of the largest lenders of funds for the program decided to no longer
              make or purchase 203(k) loans. According to a senior official at this
              lender, most of the nonprofit loans the bank acquired have eventually gone
              into default.

              To illustrate how nonprofit organizations can abuse the program, a HUD
              official told us that a group of investors agreed to pay a nonprofit
              organization to obtain mortgages under the 203(k) program on their
              behalf. The nonprofit organization was to be paid for each mortgage it
              obtained and for each property the investors rehabilitated and sold. When
              the investors were unable to successfully rehabilitate these properties, the
              organization defaulted on all of the mortgages it had taken out and the
              properties went into foreclosure. When we attempted to determine
              whether the nonprofit organization met HUD’s criteria for participation in
              the program, we discovered that the local field office could not locate its
              records on approved nonprofit organizations in its area. Therefore, HUD
              officials were uncertain as to whether the organization was even approved
              for the program.


              The 203(k) program’s design, coupled with its rapid growth and past
Conclusions   performance, pose a risk of loss to the government. Because of the
              program’s complexity, it is important that HUD aggressively oversee the
              program. HUD, however, has been in a reactive mode since the program
              began to grow rapidly. Because of reports issued over the last 4 years by
              HUD’s Inspector General and others, HUD is fully aware of the weaknesses
              in the program but has done little to correct them. The program’s recent
              performance as well as management’s continued inaction suggest that the
              program will continue to pose a risk of loss to HUD’s $3.6 billion portfolio



              Page 22                         GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                  B-280930




                  of 203(k) loans. Outside contractors have recently suggested that HUD
                  either eliminate the program entirely or radically redesign it to mitigate
                  future losses to the General Insurance Fund.

                  HUD has not used its primary oversight methods—post-endorsement
                  technical reviews and quality assurance reviews—to target 203(k) lenders
                  or loans. Although HUD has identified the 203(k) program as one of its
                  riskiest, it has failed to devote sufficient resources to oversee the program.
                  This failure to target the 203(k) program has allowed those participants
                  who pose a risk to remain in the program.

                  Furthermore, HUD does not have a uniform training program for
                  consultants and inspectors. Most of the complaints that HUD receives from
                  203(k) borrowers are the result of consultants/inspectors’ not adequately
                  ensuring that the rehabilitation work performed by contractors is
                  adequate. The borrower often depends on the consultant/inspector to
                  ensure that the work is properly completed and meets contractual
                  requirements. Incomplete or improper rehabilitation can leave the home
                  uninhabitable and can ultimately lead to foreclosure.

                  Finally, serious concerns have been raised by HUD’s Inspector General
                  concerning the continued participation of nonprofit organizations,
                  particularly the problems associated with their loan underwriting and loan
                  administration. Waste, fraud, and abuse in the 203(k) program by
                  nonprofit organizations will most likely continue unabated if HUD
                  continues to fail to scrutinize their qualifications and operations.


                  In order to provide the necessary management and oversight of the 203(k)
Recommendations   program, we recommend that the Secretary of Housing and Urban
                  Development direct the Assistant Secretary for Housing—Federal Housing
                  Commissioner to take the following actions:

                  Improve the post-endorsement technical review to identify lenders’
                  underwriting violations before the program incurs losses as a result of
                  poor underwriting. Lenders should also be notified immediately of
                  underwriting violations and be required to rectify the violation or risk
                  losing HUD’s insurance on the loan.

                  Recognizing the risk inherent in the program, target high-risk 203(k)
                  lenders for routine Quality Assurance Reviews to minimize the risks that
                  these lenders pose to the General Insurance Fund.



                  Page 23                         GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                  B-280930




                  Establish strict criteria to ensure that consultants/inspectors are well
                  versed in residential construction/rehabilitation and cost estimating to
                  protect the interests of the borrower.

                  Establish strict criteria for qualifying and recertifying nonprofit
                  organizations for their continued participation in the program to ensure
                  they have the resources and the expertise to rehabilitate properties.


                  We provided a draft of this report to the Department of Housing and Urban
Agency Comments   Development for its review and comment. In commenting on a draft of this
                  report, the Department agreed that lenders with poor underwriting
                  practices should be targeted for enforcement actions, but it did not agree
                  with our approach of improving the post endorsement review process to
                  mitigate losses. The Department believes that its new Credit Watch system
                  will identify lenders with unusually high loan default and claim rates. The
                  Credit Watch system is designed to terminate poor-performing lenders or
                  place them on probation, which results in heightened scrutiny of their
                  performance. We do not agree that the Credit Watch system will improve
                  the post endorsement technical review process, which identifies
                  underwriting violations early in the life of the loans to mitigate potential
                  losses before they occur. Although the Credit Watch system may identify
                  poor-performing lenders, it does not address how the Department would
                  better communicate the results of the post endorsement technical review
                  to lenders to ensure that underwriting violations are corrected
                  expeditiously. Therefore, we made no change to this recommendation.

                  The Department also agrees that high-risk lenders should be targeted for
                  extra monitoring activities. However, it did not comment on our
                  recommendation of targeting 203(k) lenders for routine quality assurance
                  reviews. In response to this recommendation, HUD stated that its new
                  Credit Watch system will result in a heightened level of review and
                  monitoring for 203(k) lenders. Accordingly, we did not change this
                  recommendation.

                  In response to our recommendations to establish criteria for consultants,
                  home inspectors, and nonprofit organizations participating in the 203(k)
                  program, HUD agreed that the consultants’ role needs to be reassessed and
                  will be considered as part of the Department review of the 203(k)
                  program. HUD will also take into consideration our recommendation to
                  improve qualifying standards for nonprofit organizations.




                  Page 24                        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
B-280930




In addition to its comments on our recommendations, HUD took exception
to our comparison of the 203(k) program with the Federal Housing
Administration’s 203(b) loan insurance program, stating that the 203(k)
program is known to be inherently riskier. We compared the two programs
to highlight this inherent risk and to demonstrate the need to devote
resources to oversee the 203(k) program. Furthermore, we recognize in
the report that, while the 203(k) program is inherently risky because of the
construction component of the loan, HUD oversees both programs in
essentially the same way. In addition, HUD states that it includes 203(k)
lenders in its targeting of lenders for review. We do not disagree that the
Department has reviewed 203(k) lenders, but HUD could not tell us the
extent to which the lenders’ 203(k) activity was reviewed.

HUD’s   comments and our detailed response are presented in appendix IV.


We conducted our review from May 1998 through May 1999 in accordance
with generally accepted government auditing standards. Our scope and
methodology are discussed in detail in appendix III.

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 15 days from the
date of this letter. At that time, we will send copies of this report to
Representative Barney Frank, Ranking Minority Member, House
Subcommittee on Housing and Community Opportunity, Committee on
Banking and Financial Services. We will also send copies of this report to
the Honorable Andrew M. Cuomo, the Secretary of Housing and Urban
Development; and the Honorable Jacob J. Lew, Director, Office of
Management and Budget. We will make copies available to others on
request.




Page 25                         GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
B-280930




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

Sincerely yours,




Stanley J. Czerwinski
Associate Director, Housing and
  Community Development Issues




Page 26                         GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Page 27   GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Contents



Letter                                                                                              1


Appendix I                                                                                         30

203(k) Program Data,
Fiscal Years 1994
Through 1998
Appendix II                                                                                        33

203(k) Application
and Rehabilitation
Process
Appendix III                                                                                       34

Scope and
Methodology
Appendix V                                                                                         36

Comments From the
Department of
Housing and Urban
Development
Appendix VI                                                                                        41

Major Contributors to
This Report
Table                   Table I.1: Value and Number of Loans Endorsed in the 203(k)                31
                          Program, Fiscal Years 1994-98

Figures                 Figure 1: Projected Losses to the General Insurance Fund for 203            3
                          (k) Loans Made in Fiscal Year 1994 Through Fiscal Year 1998,
                          Dollars in millions
                        Figure 2: Number of 203 (k) Loans Endorsed, Fiscal Years                    6
                          1990-98




                        Page 28                       GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Contents




Figure 3: Cumulative Claim Rates Through 1998 for the 203 (k)              9
  and 203 (b) Programs for Loans Endorsed in Fiscal Years 1994-96
Figure 4: Early Default Rates for 203 (k) Program for Loans               11
  Endorsed in Fiscal Years 1994 Through 1997
Figure 5: Incomplete Ceiling and Light Fixture Installation               17
Figure 6: Incomplete Plumbing Work                                        18
Figure 7: Exposed Breaker Box and Wires Connected to an Old               19
  Breaker Box
Figure 8: Hanging Electrical Wires in Closet                              20
Figure 9: Electrical Wiring Strung Along Outside of Wall Instead          21
  of Inside the Wall
Figure I.1: Growth in the Number of Loans Insured Under the 203           30
  (k) Program, Fiscal Years 1994-98
Figure I.2: Lenders’ Share of 203 (k) Loans Endorsed in Fiscal            31
  Years 1994-98
Figure I.3: Loans Endorsed in Underserved Areas by Borrower               32
  Type, Fiscal Years 1994-98




Abbreviations

FHA        Federal Housing Administration
HUD        Department of Housing and Urban Development
OIG        Office of the Inspector General


Page 29                      GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix I

203(k) Program Data, Fiscal Years 1994
Through 1998

                                      This appendix presents information on the 203(k) Home Rehabilitation
                                      Mortgage Insurance program for fiscal years 1994 through 1998. It
                                      provides data on the growth in the program, the value and number of
                                      loans, types of lenders, and the customers served by the program.


203(k) Program Has                    Over the past 5 years, the number of active 203(k) loans has increased
Grown Substantially                   substantially, from 6,183 at the end of fiscal year 1994 to 43,794 at the end
                                      of fiscal year 1998, as figure I.1 shows. The data presented are for all loans
                                      still active at the end of each fiscal year, ending September 30 of each year.


Figure I.1: Growth in the Number of
Loans Insured Under the 203(k)        50,000 Number of loans outstanding
Program, Fiscal Years 1994-98
                                      45,000

                                      40,000

                                      35,000

                                      30,000

                                      25,000

                                      20,000

                                      15,000

                                      10,000

                                       5,000

                                          0

                                                1994   1995   1996   1997   1998
                                                Year



                                      Source: GAO’s analysis of Federal Housing Administration’s (FHA) data.




                                      Table I.1 lists the dollar amount and number of 203(k) loans insured at the
                                      end of each fiscal year.




                                      Page 30                                 GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                         Appendix I
                                         203(k) Program Data, Fiscal Years 1994
                                         Through 1998




Table I.1: Value and Number of Loans
Endorsed in the 203(k) Program, Fiscal   Fiscal                                                                    Number of loans
Years 1994-98                            year                                Value of loans endorsed                  outstanding
                                         1994                                            $383,976,630                         6,183
                                         1995                                             652,188,055                         9,819
                                         1996                                           1,478,857,502                        19,969
                                         1997                                           2,471,945,434                        31,743
                                         1998                                           3,592,081,172                        43,794
                                         Source: GAO’s analysis of Department of Housing and Urban Development (HUD) data.




A Few Lenders Account for                Since October 1993, 289 of the 2,340 lenders participating in the 203
Most of the Program’s                    program have accounted for 80 percent of 203(k) loans, and the top 10
Loans                                    lenders have accounted for 25 percent (15,682 loans). Half of the loans
                                         were issued by only 51 lenders. More than half (1,370) of the participating
                                         lenders had five or fewer loans. (See fig. I.2.)


Figure I.2: Lenders’ Share of 203(k)
Loans Endorsed in Fiscal Years
1994-98

                                                                  25%                    10 lenders



                                             50%                                         2,289 lenders



                                                                  25%                    41 lenders




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




                                         Page 31                                 GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                                      Appendix I
                                      203(k) Program Data, Fiscal Years 1994
                                      Through 1998




Most 203(k) Loans in                  About 55 percent of all 203(k) loans insured between fiscal years 1994 and
Underserved Areas Are                 1998 were for properties in underserved areas.1 By borrower type
Made by Non-Profit                    75 percent of the loans to nonprofit organizations went to underserved
                                      areas, compared with 65 percent of investor loans and 51 percent of
Organizations                         owner-occupant loans. These ratios have remained relatively constant
                                      over time, as shown in figure I.3.


Figure I.3: Loans Endorsed in
Underserved Areas by Borrower Type,   90                 percent of loans in underserved areas
Fiscal Years 1994-98
                                      80


                                      70


                                      60


                                      50


                                      40


                                      30


                                      20


                                      10


                                      0


                                           1994                  1995                  1996             1997              1998
                                           Fiscal year

                                                         Investor
                                                         Owner/Occupant
                                                         Nonprofit

                                            Source: GAO's analysis of data from HUD




                                      1
                                       Underserved areas are those geographic areas that do not have enough mortgage lenders to meet
                                      demand. These areas are usually poor urban areas.



                                      Page 32                                       GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix II

203(k) Application and Rehabilitation
Process

               When a potential borrower locates an eligible property, the borrower
               submits an application to a lender and enters into a sales contract that is
               contingent upon the approval of a 203(k) loan and the borrower’s
               acceptance of any additional required improvements as determined by HUD
               or the lender. Either the borrower or a consultant describes the work to be
               done and provides a cost estimate. The rehabilitation must include at least
               $5,000 of eligible improvements on the existing structure on the property.
               Eligible improvements include any repair that may affect the health and
               safety of the occupants. Minor or cosmetic repairs by themselves cannot
               be included in the first $5,000 but may be added after the $5,000 threshold
               is reached. Following the lender’s acceptance of the proposed work and
               cost estimate, a written appraisal of the expected market value after
               rehabilitation work is completed. In some cases, an as-is appraisal is also
               required.

               The lender reviews the application and the appraisal to determine the
               maximum insurable mortgage amount for the property. The value of the
               maximum mortgage calculation is based on the lesser of (1) the as-is value
               of property plus rehabilitation costs or (2) 110 percent of the expected
               property market value after rehabilitation work is completed. The as-is
               value of the property is usually based on the borrower’s purchase price or,
               for refinance cases, an as-is appraisal. The rehabilitation cost can include
               up to six mortgage payments to assist the borrower when the property is
               not occupied during rehabilitation. The loan is then closed and the lender
               submits copies of the mortgage documents to HUD for review. If the
               documents are found acceptable, HUD issues a Mortgage Insurance
               Certificate to the lender. The borrower then has up to 6 months to
               complete the rehabilitation work. As the rehabilitation work progresses,
               funds are released from the rehabilitation escrow account after a
               HUD-approved inspector reviews the work. Any unused funds in the
               rehabilitation escrow account are applied to the mortgage.




               Page 33                        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix III

Scope and Methodology


                   To assess the risk the program poses to FHA’s insurance funds if
                   deficiencies in the program are not corrected, we reviewed several prior
                   reports on the program and obtained data on loan activity from HUD’s
                   Single-Family Data Warehouse, which is a database on all 203(k) single
                   family loans and others insured by the FHA, for fiscal years 1994 through
                   1998. To determine the performance of the program, we obtained data on
                   defaults and claims, types of borrowers, and lenders. To assess the costs
                   of the program, we obtained from HUD’s Comptroller’s Office data on the
                   total amount of claims paid for the period reviewed and net losses
                   incurred on those claims.

                   To assess HUD’s actions to improve the program, we reviewed reports by
                   HUD management, HUD’s Office of Inspector General, and a HUD contractor,
                   which was hired to study the program. We discussed the content of these
                   reports and HUD’s plans to address the deficiencies cited in the reports
                   with HUD officials and the contractor.

                   To determine how HUD oversees the 203(k) program, we reviewed
                   pertinent HUD regulations and policy guidance. We also interviewed
                   officials from HUD’s Office of Insured Single-Family Housing and each of
                   the four Homeownership Centers located in Santa Ana, California; Denver,
                   Colorado; Atlanta, Georgia; and Philadelphia, Pennsylvania. To determine
                   the nature and extent of HUD’s primary oversight methods,
                   post-endorsement technical reviews and quality assurance reviews, we
                   interviewed cognizant HUD officials and obtained relevant documentation,
                   such as training and policy manuals.

                   Our reliability assessments of the specific data elements required for this
                   review indicate that the data were reliable enough for our analyses. To
                   assess reliability, we

               •   reviewed existing information about data quality and controls supporting
                   the data systems,
               •   performed electronic testing of the data elements used,
               •   discussed data we analyzed with agency officials to ensure we interpreted
                   the data properly, and
               •   compared a sample of selected data elements to source documents
                   submitted for endorsement.

                   During our review we found data indicating several cases in which
                   multiple owner-occupant loans were made to the same borrower. We
                   determined that many of the borrowers with multiple owner-occupant



                   Page 34                        GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix III
Scope and Methodology




loans were actually investors and nonprofit organizations that were
incorrectly coded as owner-occupants in the database.

We performed our work from May 1998 through May 1999 in accordance
with generally accepted government auditing standards.




Page 35                       GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix V

Comments From the Department of Housing
and Urban Development




See comment 1.




                 Page 36   GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                 Appendix V
                 Comments From the Department of Housing
                 and Urban Development




See comment 2.




See comment 3.




See comment 4.




See comment 5.




                 Page 37                          GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
                 Appendix V
                 Comments From the Department of Housing
                 and Urban Development




See comment 6.




See comment 7.




                 Page 38                          GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix V
Comments From the Department of Housing
and Urban Development




The following are GAO’s comments on the Department of Housing and
Urban Development’s letter dated May 17, 1999.

1. The report clearly indicates that the 203(k) program is inherently more
risky than the 203(b) program. We compared the two programs to
highlight the inherent risk associated with the 203(k) program and to
demonstrate the need to devote resources to oversee the program.
Although the 203(k) program is riskier, FHA oversees the two programs in
essentially the same way.

2. We disagree that HUD’s current monitoring procedures are adequate to
minimize the unique risks posed by the 203(k) program. The report reflects
information provided by the Homeownership Centers that the 203(k)
program is not targeted for review. Since the 203(k) program is much
more risky than the 203(b) program, FHA should target lenders’ 203(k)
activity for oversight. FHA officials were able to provide us with the
number of lenders the agency reviewed that had made one or more 203(k)
loans. However, they could not tell us the extent to which these lenders’
203(k) activity was reviewed. Furthermore, at the time of our review, none
of the Homeownership Centers, which are responsible for overseeing the
203(k) program, targeted lenders for review on the basis of their overall
203(k) loan performance.

3. Since we relied on net loss figures supplied by FHA, we do not provide a
detailed explanation of these figures in the report. We do not draw any
conclusions from the expected increase in net losses other than that FHA
expects the program to incur increasing net losses from the time period
covered by our review. We point to other factors, such as an increase in
the early default rate, as an indication that the quality of loans made under
the program may be getting worse. We also state that the net losses as a
percentage of endorsements are in contrast to the net gains as a
percentage of endorsements expected for the 203(b) program over the
same time period. We use this fact as a further illustration that the 203(k)
program is riskier than the 203(b) program and that FHA should oversee it
differently. The report does note that the program experienced substantial
growth during the period covered by our review.

4. We disagree. FHA did not provide GAO evidence that its staff have been
trained in conducting 203(k) reviews. Our report reflects the additional
information provided by the Deputy Assistant Secretary for Single-Family
Housing regarding the targeting of 203(k) lenders for review. In the report
we include FHA’s information showing that the four Homeownership



Page 39                          GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix V
Comments From the Department of Housing
and Urban Development




Centers performed 52 quality assurance reviews on lenders that made
203(k) loans. However, the information provided does not show how many
203(k) loans have been reviewed. In addition, the information provided by
FHA did not show that every Homeownership Center has quality assurance
staff trained in the 203(k) program. Finally, during our review, officials at
two Homeownership Centers told us that they did not have staff qualified
to do quality assurance reviews.

5. HUD clarified its statement that the 203(k) program default and claim
rates are acceptable. HUD noted that while the losses are to be expected, it
is reviewing the program to determine if some program design changes are
needed. We have incorporated these clarifications where appropriate. In
addition, while our report states that the 203(k) program is complex
because of the rehabilitation component of the program, we did not state
that the program is unnecessarily bound by rules.

6. We do not agree that the Credit Watch system will improve the post-
endorsement technical review process, which identifies underwriting
violations early in the life of the loans to mitigate potential losses before
they occur. HUD announced the Credit Watch program on May 12, 1999.
This program is designed to identify high-risk lenders with high default
and claim rates. As we note in our report, a post-endorsement technical
review is intended to identify loan underwriting violations or weaknesses
before losses have occurred. Credit Watch is not designed to improve the
process for identifying and notifying lenders of violations discovered in
the post-endorsement technical review. The intent of our recommendation
is to promptly resolve deficiencies noted in the reviews in an attempt to
prevent or mitigate losses before they occur. We recognize that the Credit
Watch program may be an effective tool to identify problem lenders.

7. The Credit Watch program is designed to identify lenders with high
default and claim rates. It does not, however, identify lenders based on
specific program loan activity. In most cases, the lenders’ portfolio of
203(k) loans is a small percentage of their overall portfolio of FHA loans.
Consequently, problems with a lender’s 203(k) portfolio would not be
noticed unless they were encountering high default and claim rates with
their non-203(k) loans. The intent of our recommendation is to target
high-risk lenders on the basis of their experience with 203(k) loans, which
are inherently more risky than many other types of FHA-insured loans. If
the Credit Watch program is modified to identify high-risk lenders by the
various FHA programs, HUD may be able to capture lenders’ performance in
the programs identified as high risk.



Page 40                          GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Appendix VI

Major Contributors to This Report


                        Lenny R. Moore
Housing and             LaSonya R. Roberts
Community               Paul J. Schmidt
Development Issue       Rick B. Smith

Area
                        John T. McGrail
Office of the General
Counsel




(385744)                Page 41              GAO/RCED-99-124 Problems With HUD’s 203(k) Loans
Ordering Information

The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
following address, accompanied by a check or money order
made out to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.

Orders by mail:

U.S. General Accounting Office
P.O. Box 37050
Washington, DC 20013

or visit:

Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC

Orders may also be placed by calling (202) 512-6000
or by using fax number (202) 512-6061, or TDD (202) 512-2537.

Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information on
how to obtain these lists.

For information on how to access GAO reports on the INTERNET,
send an e-mail message with "info" in the body to:

info@www.gao.gov

or visit GAO’s World Wide Web Home Page at:

http://www.gao.gov




PRINTED ON    RECYCLED PAPER
United States                       Bulk Rate
General Accounting Office      Postage & Fees Paid
Washington, D.C. 20548-0001           GAO
                                 Permit No. G100
Official Business
Penalty for Private Use $300

Address Correction Requested