NATIONAL CREDIT UNION ADMINISTRATION

NATIONAL CREDIT UNION SHARE INSURANCE FUND

LETTER TO CREDIT UNIONS

LETTER NO. 141

DATE: December 1992

DEAR BOARD OF DIRECTORS:

Enclosed, for your information, are summary statistics of the 1991 Home Mortgage Disclosure (HMDA) data reported by credit unions and all other mortgage lenders and two press releases issued by the Federal Financial Institutions Examination Council (FFIEC).

Enclosure (1) contains a summary of denial rates from 1991 mortgage lending data submitted by credit unions required to report under HMDA. The enclosure contrasts credit union denial rates with denial rates for all mortgage lenders. While denial rates in credit unions are lower in all categories, minority loan applicants are denied loans more than twice as often as white applicants.

The October 9, 1992, press release is the FFIEC response to the findings on fair treatment in mortgage lending by financial institutions, including credit unions, after the completion of a study conducted by the Federal Reserve Bank of Boston. The October 27, 1992, press release announces the availability of the 1991 HMDA information for public inspection.

The National Credit Union Administration would like to remind credit union officials, employees and members that we must all be concerned about intentional and unintentional discrimination in policies and practices. We must also be aware of the need for continuing equal credit opportunity education programs for management, lending personnel, and consumers.

The FFIEC press release states that the initiatives many lenders instituted in response to the 1990 HMDA data cannot be abandoned. NCUA urges credit unions to study the enclosed data and put the credit union spirit to work in improving credit opportunities for all members.

For the National Credit UnionAdministration Board,

Roger W. Jepsen
Chairman

Enclosures FICU

ENCLOSURE (1)

HMDA AGGREGATE STATISTICS

ALL MORTGAGE LENDERS
	% Denial divided by Applications % of White Applications		% Denied divided by % of White
BLACK	33.9%	2.35	37.6	2.17
NATIVE AMERICAN/ ALASKAN	22.4	1.56	27.3	.58
HISPANIC	21.4	1.49	26.6	1.54
JOINT (WHITE AMERICAN)	14.9	1.03	17.5	1.01
WHITE	14.4	1.00	17.3	1.00
ASIAN/PACIFIC ISLANDER	12.9	0.90	15.0	1.87
OTHER	19.0	1.32	19.9	1.15

Credit Unions
	% Denial divided by Applications % of White Applications		% Denied divided by % of White
BLACK	20.5%	2.56	22.6	2.60
NATIVE AMERICAN/ ALASKAN	5.0	0.63	13.0	1.49
HISPANIC	17.1	2.14	17.5	2.01
JOINT (WHITE AMERICAN)	9.3	1.16	14.2	1.17
WHITE	8.0	1.00	8.7	1.00
ASIAN/PACIFIC ISLANDER	9.5	1.19	11.6	1.33
OTHER	14.0	1.75	16.5	1.90

Federal Financial Institutions Examination Council

2100 Pennsylvania Avenue, NW
Suite 200
Washington, DC 20037
(202) 634-6526
FAX (202) 634-6556

Press Release

For immediate releaseOctober 9, 1992

The federal financial institutions regulatory agencies are issuing this statement to emphasize their concerns about fair treatment of applicants for mortgage loans and their continuing efforts to assure such treatment. Increasing evidence indicates that differences in loan approval rates between white and minority home mortgage applicants, apparently unwarranted by economic factors, characterize some lending.

On October 8, 1992, the Federal Reserve Bank of Boston, with the cooperation of the federal financial institutions regulatory agencies and the U.S. Department of Housing and Urban Development (HUD), released a detailed study on mortgage loan denial rates in the Boston area. The study examined whether racial and ethnic disparities in mortgage loan denial rates reflect even handed use of legitimate credit standards.

That question arose because the Home Mortgage Disclosure Act (HMDA) data released in October 1991 revealed disparities in denial rates. Those data showed patterns of relatively low mortgage lending in minority areas and indicated that black and Hispanic

applicants were denied loans two to three times as often as whites. The disparity in denial rates in the Boston metropolitan area was among the highest in the country.

STUDY RESULTS

The study by the Federal Reserve Bank of Boston incorporated data submitted voluntarily by about 130 Boston lenders. The data came from application and loan files for 1,013 minority and 3,123 nonminority applicants and included information from loan applications, credit reports, and lenders' worksheets. The statistical analysis performed for the study took into account an applicant's debt ratios, wealth, credit history, and income stability. The analysis also included loan- to-value ratios, information about private mortgage insurance sought or approved, the neighborhood characteristics of the property being financed, and other factors.

The study found that economic differences account for a substantial part of the disparities in the HMDA data, but that they do not explain those differences entirely. Minority applicants in the study had, on average, lower income, fewer liquid assets, more late and delinquent credit accounts, and higher debt to income ratios than white applicants. However, even after controlling for differences in relevant economic and financial variables, black and Hispanic mortgage applicants were more likely to be turned down than similarly situated whites. Specifically, on a statistical basis, minority applicants with credit-related characteristics identical to those of an average white applicant would experience a 17 percent denial rate, as compared to the white denial rate of 11 percent.

The study also found that the vast majority of minority applicants obtained loans, and that loan approval rates for well-qualified and clearly unqualified applicants of all races were essentially what would be expected in a nondiscriminatory system. Disparities appeared to be most common among applicants who have some imperfections or flaws in credit qualifications. Many such applicants can -- and do -- qualify for a loan if a lender acts diligently, creatively, and flexibly on their behalf. However, the study's results suggest that applicants of different races do not all benefit equally from such discretionary efforts in the lending process. This tentative explanation of the study's findings is also consistent with investigation-based allegations recently reported by the U. S. Department of Justice.

SUPERVISORY INITIATIVES

This recent evidence suggests that lenders and regulators should continue their efforts to ensure that minorities have equal access to credit. Several actions by the agencies to improve enforcement of fair lending laws and prevent discrimination already have been taken or are underway.

Enforcement-oriented analysis. The agencies are analyzing the Boston data at the level of individual lenders. If statistical analysis -- or follow-up review of applications and loan files -yields reliable evidence indicating loan decisions by particular lenders were based on race or ethnicity, that information will be forwarded to the Department of Justice or HUD, as appropriate.

HMDA-based targeting. The agencies are continuing to

emphasize the need for examiners to use HMDA data, as well as other information, to identify institutions that should be subjected to follow-up procedures to ensure fair lending practices. Where appropriate, examiners are conducting detailed reviews and comparisons of loan and application files as part of fair lending examinations.

New computer programs and analytical tools are being developed to make HMDA data more useful to examiners. This software, when fully developed and available to examiners, will help them identify geographic and other discrepancies in mortgage lending that may suggest further inquiry. The agencies are also continuing their efforts to develop additional software capabilities to assist examiners and institutions to analyze HMDA data.

Revision of examination strategies. The computerized statistical analysis used in the Boston study (and by the Department of Justice in its investigation) is of very limited use in most bank examinations. very few institutions have a sufficient volume of denials for minority applicants to reliably document whether there is a "double standard" in loan qualification criteria. Nevertheless, these findings make it imperative that the agencies explore every alternative to make bank examinations as reliable as possible in identifying possible discrimination.

Under the auspices of the Federal Financial Institutions Examination Council (FFIEC), the agencies have initiated a comprehensive, coordinated review of their examination policies, procedures, and examiner training to determine whether they can be strengthened to help deter and detect racially-based disparate

treatment of applicants by financial institutions. A key part of this process will be the use of a third-party consultant who can provide an objective review of existing agency efforts and make recommendations for changes to policies and procedures.

Industry education. Lending personnel of financial institutions may need additional information and training about fair lending laws, and greater sensitivity to cultural and racial differences among applicants for credit. Consumers need information about the credit granting process -- how it works and what is expected in order to be a successful applicant for mortgage credit. In this regard, the agencies believe that additional efforts to expand and intensify educational programs for both lenders and consumers will be required.

The agencies themselves will continue to develop informational materials and participate in conferences and seminars for financial institution representatives concerning their responsibilities. The agencies have distributed a guide for financial institutions, Home Mortgage Lending and Equal Treatment. The guide highlights lending standards and practices that may adversely impact the capacity of institutions to serve minorities and other protected classes of borrowers, and alerts them to subtle forms of discriminatory behavior that should be avoided. Copies are available from each of the agencies. For mortgage applicants, a brochure, Home Mortgages: Understanding the Process and Your Rights, is available from the agencies to assist borrowers in knowing what to expect in applying for a loan and what their rights are if they suspect mistreatment.

LENDING PRACTICES

The agencies call on financial institutions and their trade associations to intensify their equal credit opportunity education programs for management, lending personnel and consumers. This should include efforts to identify and promote examples of successful techniques used by institutions to ensure equal treatment of loan applicants.

These techniques might include use of internal systems to conduct independent second reviews of applications from minorities initially recommended for denial to assure equal treatment with nonminorities. One lender recently reported making loans to 35 percent of initially rejected applicants after instituting such a review procedure. A similar kind of review underwriter is also an element of the Department of Justice settlement.

Other initiatives might include development of training programs to ensure fair treatment of prospective borrowers, credit counseling education for groups of prospective loan applicants, participation on mortgage review boards, and the use of "shoppers" hired by the institution itself to test its personnel's adherence to its own lending procedures.

Despite the many commendable efforts of lenders, the data from the Boston Study demonstrate that unequal treatment of applicants still occurs in some instances. The message from Boston is that the initiatives many lenders instituted in response to the release of the HMDA data cannot now be abandoned. For lenders who have not initiated such efforts, the message is that they must do so to assure that all prospective borrowers are considered fairly.

Federal Financial Institutions Examination Council

2100 Pennsylvania Avenue, NW
Suite 200
Washington, DC 20037
(202) 634-6526
FAX (202) 634-6556

Press Release

For use in AMS
Wednesday
October 28, 1992October 27, 1992

Reports of 1991 mortgage lending activity in metropolitan statistical areas (MSAs) will be available for public inspection at a central depository in each MSA this week, the Federal Financial Institutions Examination Council (FFIEC) announced today. These reports include individual disclosure statements for each lender covered by the Home Mortgage Disclosure Act (HMDA) and aggregate data by MSA.

Last year, the HMDA reports of 1990 lending activity contained--for the first time--data on the race, sex, and income level of borrowers and applicants and data on home loan applications denied or withdrawn (as well as continuing to report originated and purchased loans). The expanded 1990 data showed significant differences in loan denial rates among racial and ethnic groups. The data also showed that black and Hispanic applicants were more likely to apply for government-backed loans. overall, the HMDA data for 1991 continue to show these differences.

An article prepared for the November issue of the Federal Reserve Board Bulletin, containing a preliminary analysis of national aggregates from the 1991 data, was also made available today. It provides an overview of some of the responses within the public and private sectors to the HMDA data released in October 1991, including initiatives undertaken to promote affordable housing in low- and moderate-income areas.

The FFIEC has prepared reports of 1991 lending activity for 9,358 institutions on behalf of its member agencies- -the Comptroller of the Currency, the Federal Deposit Insurance Corporation, the Office of Thrift Supervision, the National Credit Union Administration, and the Federal Reserve System--and the U.S. Department of Housing and Urban Development. The number of reporters is about the same as for 1990, while the volume of reported loans and applications increased substantially, from 6.9 million to more than 7.8 million. The 1991 volume of loans for refinancing more than doubled from 1990, reflecting last year's decline in mortgage interest rates. Approximately three-quarters of the applications for home purchase loans were for conventional mortgage loans, while the remainder were government-backed loans-- loans insured or guaranteed by the Federal Housing Administration (FHA), the Veterans Administration (VA), or the Farmers Home Administration (FmHA).

In 1991, lenders approved more than 70 percent of applications for conventional and government-backed home purchase loans. The

denial rate for conventional home loans and FHA-insured loans was somewhat higher in 1991 than in 1990. Several factors may account for this change. Lenders may have tightened underwriting criteria in response to rising delinquencies and weak housing markets in many parts of the country. The change may reflect an increased tendency for loan originators who sell loans in the secondary market to use secondary market guidelines to assess applications, rather than apply more flexible under-writing standards for nonconforming loans. Additionally, the increase in the loan denial rate may indicate that more loan applicants had less stable incomes and weaker credit histories due to the recession.

The data for 1991 show a significant increase in the number of low-income applicants. This may reflect the results of an expanded marketing and outreach efforts by lenders, as well as a greater proportion of first-time homebuyers (whose income tends to be lower than the income of current owners purchasing new homes) in 1991 than in 1990.

FAIR LENDING ENFORCEMENT

To assess compliance with fair lending laws, the expanded HMDA data must be augmented by specific information on individual applicants and the underwriting standards used by the lender. Recent investigative efforts have examined factors that are not included in the HMDA data- -such as credit history, financial assets, and employment experience--to determine to what extent these factors account for differences in loan approval rates among different racial groups.

In September 1992, the Department of Justice filed a consent decree resolving claims that a major mortgage lender in the Atlanta area had discriminated against prospective black home-buyers when marketing home loan products and granting mortgage loans. After controlling for differences in credit history, and other underwriting variables, analysis of over 4,000 loan application files indicated that race was a factor in loan decisions. The case, which was resolved without any admission of wrongdoing by the lender, requires remedial action that includes $1 million in payments to black applicants whose home mortgage loan applications were rejected.

On October 8, 1992, the Federal Reserve Bank of Boston, with the cooperation of the federal financial regulatory agencies and HUD, issued a study on the disposition of mortgage loan applications in the Boston area. The study was conducted in response to the large difference in the rates of loan denial among blacks, Hispanics, and whites in the Boston area revealed by the 1990 HMDA data. When economic factors other than income were included in the analysis of HMDA data, the disparity between minority and white approval rates was reduced. After controlling for these factors, it was evident that black and Hispanic applicants were still denied mortgage loans more often than white applicants.

On October 9, 1992, the FFIEC issued an interagency statement on disparate treatment in mortgage lending. The statement emphasized the agencies' concerns that all prospective borrowers be

treated fairly and outlined steps that can be taken by individual institutions and their trade associations to identify and eliminate discriminatory lending policies and practices. While the extensive statistical analysis used in Boston and in Atlanta is generally not feasible in examinations, the agencies noted that it is imperative that examinations be as thorough and reliable as possible. The FFIEC has initiated a comprehensive review of agencies' examination policies, procedures, and examiner, training to detect bias in lending procedures.

LENDING PRACTICES

Financial institutions and trade associations are seeking to address concerns about discriminatory lending practices and to improve access to home mortgages for minorities and lower-income applicants. Lenders are developing and improving fair lending education programs for management, lending personnel, and consumers. To ensure that credit policies are applied fairly, many lenders are taking a second look at mortgage applications from minorities initially recommended for denial. Some are hiring "shoppers" to evaluate staff Is compliance with credit policies and procedures. Lenders are also expanding community outreach programs to provide credit counseling and to help potential applicants improve their chances of qualifying for a mortgage.

HMDA DATA: PUBLIC ACCESS

The FFIEC makes the HMDA data publicly available in a variety of formats, including disclosure statements for individual institutions and aggregate reports for each MSA (in hard copy,

microfiche, and computer tape) , and an edited version of loan application register (LAR) records for the nation (on data tape). New formats now available include tables showing nation-wide aggregates and tables providing key information for each MSA. The FFIEC's HMDA data order form is attached.

The FFIEC will offer additional new HMDA reports in December. A reporter directory will list all HMDA reporters within an MSA, and provide information such as their location and asset size, as well as loans originated and applications received by state and county. A three-report package will provide analyses of individual institutions' home-lending activity by MSA, including:

- disposition of applications by income and race,

- distribution of loan originations for home purchase loans, home improvement loans and refinancings within an MSA by racial and income characteristics of the census tract, and

- summaries of loan activity by census tract, including the percentage of minority population, the number of owner-occupied units, and the denial rate.

The FFIEC expects that LAR records by MSA also will be available on PC diskette in early 1993.

FEDERAL FINANCIAL INSTITUTIONS EXAMINATION COUNCIL HMDA DATA ORDER FORM ITEM DESCRIPTIONS

Aggregation Report: The Aggregation Reports of MSAs (340 for 1990, 341 for 1991 and 1992), sorted by MSA, distributed as either hardcopy or on tape. Printed MSAs range from 30 to 604 pages each, with an average of approximately 60 pages. Tape distribution is used in lieu of hardcopy as a convenience to those sites that can print the tape locally.

For hardcopy distribution, enter the MSA number for each MSA desired on the report in the space provided on the form. If additional space is required, use the back of the form. The charge for hardcopy distribution is $50.00 for each MSA requested.

For tape distribution, all MSAs will be included on the tape. Also, a listing will be provided that shows the relative position and number of pages for each MSA on the tape. Data will be in EBCDIC and formatted as follows: record length = 166; block size = 24,900; and record format = fixed block. Printing these records requires a printer featured for 165 character output. The charge for tape distribution is $500.00.

Census Information: This is the nationwide census data that is used as input to HMDA processing. Data is distributed as a hardcopy report or on tape. The hardcopy Census Tract Listing contains valid census listings and does not have demographic information; whereas the data on the tape does contain demographic information for each tract. Tape distribution is available on either reels or cartridges; file descriptions are included with each order. The charge for hardcopy distribution is $50.00; for tape distribution, the charge is $250.00.

It is important to note that 1980 census tracts have been used in processing 1990 and 1991 HMDA data; 1990 census tracts for processing beginning with 1992 HMDA data. The census data distributed will be consistent with the HMDA year specified. For example, if HMDA year 1991 is selected, 1980 census tract data with 1991 MSA definitions will be issued.

Disclosure Statement: The Disclosure Statement summarizes mortgage and home improvement lending information from data that is prepared yearly by individual institutions. The Disclosure Statement is distributed in hardcopy only at a cost of $50.00 for each institution requested.

Expected Reporter Panel: The Expected Reporter Panel is the universe of all institutions who were expected to report under HMDA. The panel includes the name of the institution, the reporter ID and agency code, and the MSA number(s) of the metropolitan areas for which they were expected to report. The panel is captured from the database at the same time that the final disclosure and aggregate reports are prepared for the Central Depositories, usually in October. The Expected Reporter Panel is distributed in reel or cartridge only at a cost of $150.00. A format description is also included with each order.

Loan Application Records (LAR) and Transmittal sheet (TS) Data: The LAR & TS data is collected by a financial institution as a result of applications for, and originations and purchases of, home-purchase loans (including refinancings) and home- improvement loans for each calendar year. LARs for 1990 total almost 6.6 million records and 9,332 Transmittal Sheets (TSs); the 1991 LARs data total over 7.9 million records and 9,359 TSs. The following should be noted:

- Data is not certified as error free.

- For reasons of privacy, the dates and loan application numbers are omitted from each record.

To form a unique identifier for a reporter, the Respondent ID and single character Agency Code must be used. To form a unique loan record identifier, the Respondent ID, Agency Code and Loan Sequence Number must be used. See record layout enclosed with distribution tape for additional information.

LAR and TS data is distributed on tape only at a cost of $500.00. cartridges may be chosen. Record formats and file descriptions are included with each order.

If reels are ordered, there will be four or five reels containing the LAR data and one reel containing the TS data. If cartridges are ordered, there will be three or four cartridges containing the LAR data and one containing the TS data. (The number of reels/cartridges containing LAR data will vary according to the year requested because of the volume of data). Note that both reels and cartridges containing LAR data will be multi-volume datasets. This means that to process ALL of the LAR data, each of the reels or cartridges must be read successively.

MSA Median Family Income Listing: This Listing shows Median Family Incomes for 19xx HMDA Reports" where xx is the HMDA year (i.e., 90, 91, etc.). If a 1990 Listing is selected, there will be an adjustment factor of 1.6424 applied to the 1980 Median Family Income column for each MSA to arrive at the Inflation Adjusted 1980 Median Family Income column. For 1991 and future Listings, the median family income of the MSA is/will be based on estimates developed by the Department of Housing and Urban Development (HUD); these estimates will be updated annually. The figure for the MSA is the HUD estimate for the fiscal year that corresponds to the year for which the loan/application data are reported.

The Listing is distributed free of charge.

PC Diskettes by MSA: The PC Diskettes by MSA will contain individual loan application records coded for the MSA(S) selected on the form. The availability date of this product is the first quarter of 1993. The cost of this item as well as a firm availability date will be announced later. The data contained for each MSA will be only for institutions that are located in that MSA, i.e., the data reflected in the Aggregate Report for that MSA.

Reporter Directory: The Reporter Directory is a hard copy of all HMDA reporters within a specified MSA, sorted by name of reporter and then by state and county codes. This report:

- Provides ID number, name of institution, city and statelocations, total assets, and parent ID number.

- For each state and county in which an institution hasactivity, the following are provided:

-- the state and county codes;
-- the number of tracts with loans;
-- the number of loans originated;
-- the number of applications;
-- the ratio of loans to applications; and
-- the number of loans purchased

The Reporter Directory will be available beginning in December 1992. The charge for non-profit organizations will be $45.00 per requested MSA and $75.00 per MSA for all others.

Three-Report Package: This package of three reports is available in hardcopy only. The following information is provided for a specified institution/MSA combination:

Report on Disposition of Loans by Income and Race

- Compares applications by disposition and by income levels (as a percentage of the MSA median income-- less than 80%, 80-99%, 100-120%, and more than 120%)

- Is subdivided by race

Report on Distribution of Loan-Applications

- Provides a summary of a specific institution's activity by income and racial characteristics of the census tract within an MSA

- Covers applications for mortgage and home improvement loans and refinancings

- Provides the number of applications received, the number of loans originated, and the number of 1-4 family owner-occupied units within each category of census tract characteristics.

Census Tract Summary

- Provides the number, dollar amounts, and disposition of applications for a specific institution/MSA by census tracts

- For each census tract, the following are also provided:

-- the median housing value
-- the median age of the housing stock
-- the median age of the population
-- the minority population percentage
-- the number of 1-4 family owner-occupied units
-- the number of households
-- the median income
-- the denial rate

- Census tracts that are low- to moderate-income tracts are highlighted

Available in December 1992, these reports for non-profit organizations are priced at $45.00 per requested institution/MSA, and $75.00 per MSA for all others. These reports are ordered as a package, not separately.