Credit Unions: Financial Condition Has Improved, but		 
Opportunities Exist to Enhance Oversight and Share Insurance	 
Management (27-OCT-03, GAO-04-91).				 
                                                                 
Recent legislative and regulatory changes have blurred some	 
distinctions between credit unions and other depository 	 
institutions such as banks. The 1998 Credit Union Membership	 
Access Act (CUMAA) allowed for an expansion of membership and	 
mandated safety and soundness controls similar to those of other 
depository institutions. In light of these changes and the	 
evolution of the credit union industry, GAO evaluated (1) the	 
financial condition of the industry and the deposit (share)	 
insurance fund, (2) the impact of CUMAA on the industry, and (3) 
how the National Credit Union Administration (NCUA) had changed  
its safety and soundness processes.				 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-91						        
    ACCNO:   A08764						        
  TITLE:     Credit Unions: Financial Condition Has Improved, but     
Opportunities Exist to Enhance Oversight and Share Insurance	 
Management							 
     DATE:   10/27/2003 
  SUBJECT:   Bank examination					 
	     Banking regulation 				 
	     Credit unions					 
	     Federal law					 
	     Funds management					 
	     Insurance						 
	     Lending institutions				 
	     Performance measures				 
	     Regulatory agencies				 
	     Financial condition				 
	     NCUA Large Credit Union Pilot Program		 
	     NCUA National Credit Union Share			 
	     Insurance Fund					 
                                                                 
	     NCUA Prompt Corrective Action Program		 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-04-91

United States General Accounting Office

GAO

Report to the Ranking Minority Member,

         Committee on Banking, Housing, and Urban Affairs, U.S. Senate

October 2003

CREDIT UNIONS

 Financial Condition Has Improved, but Opportunities Exist to Enhance Oversight
                         and Share Insurance Management

                                       a

GAO-04-91

Highlights of GAO-04-91, a report to the Ranking Minority Member,
Committee on Banking, Housing, and Urban Affairs, U.S. Senate.

Recent legislative and regulatory changes have blurred some distinctions
between credit unions and other depository institutions such as banks. The
1998 Credit Union Membership Access Act (CUMAA) allowed for an expansion
of membership and mandated safety and soundness controls similar to those
of other depository institutions. In light of these changes and the
evolution of the credit union industry, GAO evaluated (1) the financial
condition of the industry and the deposit (share) insurance fund, (2) the
impact of CUMAA on the industry, and (3) how the National Credit Union
Administration (NCUA) had changed its safety and soundness processes.

With respect to the share insurance fund, GAO recommends that the Chairman
of NCUA explore developing a risk-based funding system, improve the
process for allocating overhead expenses, and refine the process for
estimating future losses. To improve reporting, the Chairman should also
use tangible indicators to determine whether credit unions are serving
people in underserved areas. To help ensure safety and soundness, Congress
may wish to consider making credit unions subject to internal control
reporting and attestation requirements applicable to banks and thrifts and
providing NCUA legislative authority to examine third-party vendors.

www.gao.gov/cgi-bin/getrpt?GAO-04-91.

To view the full product, including the scope and methodology, click on
the link above. For more information, contact Richard J. Hillman at (202)
512-9073 or [email protected].

October 2003

CREDIT UNIONS

Financial Condition Has Improved, but Opportunities Exist to Enhance Oversight
and Share Insurance Management

The financial condition of the credit union industry has improved since
GAO's last report in 1991, and the federal share insurance fund appears
financially stable. However, a growing concentration of industry assets in
large credit unions creates the need for greater risk management on the
part of NCUA. The question of who benefits from credit unions' services
has also been widely debated. While it has been generally accepted that
credit unions have a historical emphasis on serving people of modest
means, our analysis of limited available data suggested that credit unions
served a slightly lower proportion of low- and moderate-income households
than banks.

CUMAA and subsequent NCUA regulations enabled federally chartered credit
unions to expand their membership, serve larger geographic areas, and add
underserved areas. According to NCUA officials, these changes were
necessary to maintain the competitiveness of the federal charter with
respect to state-chartered credit unions. While NCUA has stated its
commitment to ensuring that credit unions provide financial services to
all segments of society, NCUA has not developed indicators to determine if
credit union services have reached the underserved.

In response to the growing concentration of industry assets and increased
services offered by credit unions, NCUA recently adopted a risk-focused
examination and supervision program but still faces a number of
challenges, including lack of access to third-party vendors that are
providing more services to credit unions. Further, credit unions are not
subject to internal control and attestation reporting requirements
applicable to banks and thrifts. GAO also found that the insurance fund's
rate structure does not reflect risks that individual credit unions pose
to the fund, and NCUA's estimation of fund losses is based on broad
historical analysis rather than a current risk profile of insured
institutions.

Mortgages Made by Credit Unions and Banks, by Income Level of Purchaser,
2001

Contents

  Letter

Results in Brief
Background
Financial Condition of the Credit Union Industry Has Improved

Since 1991 Limited Comprehensive Data Are Available to Evaluate Income of
Credit Union Members CUMAA Authorized NCUA to Continue Preexisting
Policies That Expanded Field of Membership NCUA Adopted Risk-Focused
Examination and Supervision Program, but Faces Challenges in
Implementation

NCUSIF's Financial Condition Appears Satisfactory, but Methodologies for
Overhead Transfer Rate, Insurance Pricing, and Estimated Loss Reserve Need
Improvement

System Risk That May Be Associated with Private Share Insurance

Appears to Have Decreased, but Some Concerns Remain Conclusions
Recommendations for Executive Action Matters for Congressional
Consideration Agency Comments and Our Evaluation

1 4 8

10

16

29

42

56

66 80 82 83 84

Appendixes

                          Appendix I: Appendix II: Appendix III: Appendix IV:

           Appendix V: Appendix VI: Appendix VII: Appendix VIII: Appendix IX:
                                                     Appendix X: Appendix XI:

Appendix XII:

Objectives, Scope, and Methodology 90

Status of Recommendations from GAO's 1991 Report 101

Financial Condition of Federally Insured Credit Unions 113

Comparison of Bank and Credit Union Distribution of
Assets 121

Credit Union Services, 1992-2002 128

Characteristics of Credit Union and Bank Users 140

Key Changes in NCUA Rules and Regulations, 1992-2003 144

NCUA's Budget Process and Industry Role 146

NCUA's Implementation of Prompt Corrective Action 150

Accounting for Share Insurance 159

Comments from the National Credit Union
Administration 162

Comments from American Share Insurance 168

                                    Contents

Appendix XIII:	GAO Contacts and Staff Acknowledgments 174 GAO Contacts 174
Staff Acknowledgments 174

Tables	Table 1: Table 2:

Table 3: Table 4: Table 5:

Table 6:

Table 7:

Table 8:

Table 9:

Regulatory Definitions of Local Community, 2000 and
2003
Federally Insured Credit UnionsWere Similar to Banks and
Thrifts with Respect to Capital Categories, as of December
31, 2002
Peer Group Definitions
Definition of Income Categories
Status of GAO Recommendations to NCUA and Congress,
as of August 31, 2003
Federally Insured Credit Union Growth in Assets and
Shares, 1992-2002
Distribution of Credit Unions by Asset Size, 1992 and
2002
Asset Composition of Credit Unions as a Percentage of
Total Assets, 1992-2002
Comparison of the Loan Portfolios of Federally Insured
Credit Unions with Peer Group Banks and Thrifts, as of
2002

34

54 92 93

103

115

116

117

118

144

151 153

154

Table 10: Timeline of Key Changes to NCUA Rules and Regulations, January
1992-September 2003

Table 11: CUMAA Mandates and NCUA Actions on PCA Regulation Implementation

Table 12: Discretionary Supervisory Actions

Table 13: Net Worth Category Classification for New Credit Unions

Figures	Figure 1: Figure 2: Figure 3:

Figure 4: Figure 5:

Comparison of Credit Union and Bank Capital Ratios, 1992-2002 11 Credit
Union Industry Size and Total Assets Distribution, as of December 31, 2002
15 Income Characteristics of Households Using Credit Unions versus Banks,
Low and Moderate Income versus Middle and High Income 21 Income
Characteristics of Households Using Credit Unions versus Banks, by Four
Income Categories 22 Mortgages Made by Credit Unions and Banks, by Income
Level of Purchaser, 2001 24

Contents

Figure 6: Loans Made by Credit Unions and Banks, by Average
Income in the Purchased Home's Census Tract, 2001 25
Figure 7: Percentage of Federally Chartered Credit Unions, by
Charter Type, 2000-2003 31
Figure 8: Actual and Potential Members in Federally Chartered
Credit Unions, by Charter Type, 2000-2003 36
Figure 9: Actual and Potential Members in Federally and
State-chartered Credit Unions, 1990-2003 37
Figure 10: Underserved Areas Added before and after CUMAA, by
Federal Charter Type, 1997-2002 40
Figure 11: Credit Union Mortgage Loans Have Grown Significantly

Since 1992 43
Figure 12: NCUSIF's Equity Ratio, 1991-2002 57
Figure 13: Equity to Insured Shares or Deposits of the Various

Insurance Funds 58
Figure 14: Net Income of NCUSIF, 1990-2002 59
Figure 15: Financing Sources of NCUSIF and NCUA's Operating

Fund 60
Figure 16: Share Payouts and Reserve Balance, 1990-2002 65
Figure 17: States Permitting Private Share Insurance (March 2003)

and Number of Privately Insured Credit Unions
(December 2002) 68
Figure 18: Capital Ratios in Federally Insured Credit Unions, 1992-
2002 114
Figure 19: Profitability of Federally Insured Credit Unions, 1992-
2002 119
Figure 20: Federally Insured Credit Unions, by CAMEL Rating,
1992-2002 120
Figure 21: Total Assets of All Credit Unions and All Banks, as of
2002 121
Figure 22: Total Assets of Credit Unions and Banks with Less Than
$100 Million in Assets, as of 2002 122
Figure 23: Total Assets of Credit Unions with Less Than $5 Million in
Assets, as of 2002 123
Figure 24: Percentage of All Credit Unions and All Banks Holding
Various Loans, as of 2002 124

Figure 25: Percentage of Credit Unions and Banks with Assets of
$100 Million or Less Holding Various Loans, as of
2002 125

Figure 26: Percentage of Credit Unions and Banks with Assets
between $1 Billion and $18 Billion Holding Various Loans,
as of 2002 126

Contents

Figure 27: Percentages of Credit Unions and Banks Holding Various
Loans, by Institution Size, as of 2002 127
Figure 28: Percentage of Credit Unions Holding Various Loans,
1992-2002 129
Figure 29: Percentage of Assets Held in Various Loans by All Credit
Unions, 1992-2002 131
Figure 30: Percentage of Credit Unions Offering Various Accounts,
1992-2002 133
Figure 31: Credit Union Employees and Number of Credit Unions,
1992-2002 134
Figure 32: Percentage of Credit Unions, Smallest versus Largest,
Holding Various Loans, 1992-2002 136
Figure 33: Percentage of Assets Held in Various Loans, Smallest
versus Largest Credit Unions, 1992-2002 137
Figure 34: Differences among Services Offered by Smaller and
Larger Credit Unions, as of 2002 138
Figure 35: Credit Union Size and Offerings of More Sophisticated
Services, as of 2002 139
Figure 36: Households Using Credit Unions and Banks, by
Education Level, 2001 140
Figure 37: Households Using Credit Unions and Banks, by Age
Group, 2001 141
Figure 38: Households Using Credit Unions and Banks, by Race and
Ethnicity, 2001 142
Figure 39: Mortgages Made by Credit Unions and Banks, by Race

and ethnicity, 2001 143
Figure 40: NCUA Budget Levels, 1992-2004 148
Figure 41: NCUA-authorized Staffing Levels, 1992-2003 149

Abbreviations

ASI American Share Insurance
ATM Automatic Teller Machines
BIF Bank Insurance Fund
BSA Bank Secrecy Act
CLF Central Liquidity Facility
CPA Certified Public Accountant
CRA Community Reinvestment Act
CUIC Credit Union Insurance Corporation
CUMAA Credit Union Membership Access Act of 1998

Contents

CUNA Credit Union National Association
CUSO Credit Union Service Organization
FCUA Federal Credit Union Act
FDIA Federal Deposit Insurance Act
FDIC Federal Deposit Insurance Corporation
FDICIA Federal Deposit Insurance Corporation Improvement Act of

1991 FFIEC Federal Financial Institutions Examination Council FRC
Financial Risk Committee FTC Federal Trade Commission HMDA Home Mortgage
Disclosure Act HUD Department of Housing and Urban Development IRPS
Interpretive Ruling and Policy Statement LAR Loan Application Records MCIC
Metro Chicago Information Center MSA Metropolitan Statistical Area NCUA
National Credit Union Administration NCUSIF National Credit Union Share
Insurance Fund NFCDCU National Federation of Community Development Credit

Unions OCC Office of the Comptroller of the Currency OTS Office of Thrift
Supervision PCA Prompt Corrective Action RISDIC Rhode Island Share and
Depositors Indemnity

Corporation SAIF Savings Association Insurance Fund SCF Survey of Consumer
Finances

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

A

United States General Accounting Office Washington, D.C. 20548

October 27, 2003

The Honorable Paul S. Sarbanes

Ranking Minority Member

Committee on Banking, Housing, and Urban Affairs United States Senate

Dear Senator Sarbanes:

Credit unions have historically occupied a unique niche among depository
institutions. Credit unions are not-for-profit, member-owned cooperatives
that are exempt from paying federal income taxes on their earnings. Unlike
banks, credit unions are subject to limits on their membership because
members must have a "common bond"-for example, working for the same
employer or living in the same community. However, over the years, these
membership requirements have loosened considerably and credit unions have
received expanded powers, which have raised questions about the extent
that credit unions remain unique and serve a different population than
banks. We last conducted a comprehensive review of the credit union
industry, including the National Credit Union Administration (NCUA), in
1991.1 Since that time, the credit union industry has experienced
substantial growth and expansion of activities. In addition, recent
legislative and regulatory changes have blurred some distinctions between
credit unions and other depository institutions-banks and thrifts. For
example, the 1998 Credit Union Membership Access Act (CUMAA) expanded the
definition of common bond and provided for reforms intended to strengthen
the safety and soundness of credit unions, including instituting
procedures for prompt corrective action (PCA) when credit unions' capital
levels fall below a certain threshold.2

In 2002, there were about 10,000 credit unions with approximately 82
million members. Credit unions, like banks and thrifts, are chartered by
both the federal government and state governments, also referred to as the
dual-chartering system. NCUA has oversight authority for federally

1U.S. General Accounting Office, Credit Unions: Reforms for Ensuring
Future Soundness, GAO/GGD-91-85 (Washington, D.C.: July 10, 1991). This
report contained a variety of recommendations to Congress and NCUA. See
appendix II for information on the implementation of these
recommendations.

2See Pub. L. No. 105-219 (Aug. 7, 1998).

chartered credit unions and requires its credit unions to obtain federal
share (deposit) insurance for their members' deposits from the National
Credit Union Share Insurance Fund (NCUSIF). This fund, administered by
NCUA, also provides share insurance to most state-chartered credit unions.
Some states permit their credit unions to purchase private share insurance
as an alternative to federal insurance.

In light of the evolution of the credit union industry and the passage of
CUMAA, you asked us to review a variety of issues involving the credit
union industry and NCUA. In response, we provided your staff information
on how NCUA responded to recommendations made in our 1991 report and
conducted preliminary research on the industry and NCUA.3 After discussing
this information with your staff, we agreed that the objectives of this
study were to evaluate (1) the financial condition of the credit union
industry; (2) the extent to which credit unions "make more available to
people of small means credit for provident purposes";4 (3) the impact, if
any, of CUMAA on credit union field of membership requirements for
federally chartered credit unions; (4) how NCUA's examination and
supervision processes have changed in response to changes in the industry;
(5) the financial condition of NCUSIF; and (6) the risks associated with
the use of private share insurance. You also asked us to review issues
associated with corporate credit unions, which we plan to address in a
separate report.5

3In addition, we recently completed a separate review of private insurance
issues. See U.S. General Accounting Office, Federal Deposit Insurance Act:
FTC Best Among Candidates to Enforce Consumer Protection Provisions,
GAO-03-971 (Washington, D.C.: Aug. 20, 2003).

4This quotation is taken from the title of the Federal Credit Union Act of
June 26, 1934. In addition, in CUMAA the congressional findings stated
among other things that credit unions "have the specified mission of
meeting the credit and savings needs of consumers, especially persons of
modest means (Pub. L. No. 105-219 S: 2 (1998)). While these statutes have
used "small means" and "modest means" to describe the type of people who
credit unions might serve, in this report we used "low- and
moderate-income," as defined by banking regulators.

5A corporate credit union is one whose members are credit unions, not
individuals. Corporate credit unions provide credit unions with services,
investment opportunities, loans, and other forms of credit should credit
unions face liquidity problems. See 12 C.F.R. Part 704 (2003).

To evaluate the financial condition of the credit union industry we
performed quantitative analyses on credit union call report data for 1992-
2002.6 Since NCUA lacked readily available data to assess the extent to
which credit unions serve people of low and moderate incomes, we analyzed
data from the 2001 Federal Reserve Survey of Consumer Finances (SCF) to
identify the characteristics of credit union members. This survey is the
only comprehensive source of publicly available data on financial
institutions and consumer demographics that we could identify that is
national in scope. We also analyzed 2001 mortgage data from the Home
Mortgage Disclosure Act (HMDA) database, which allowed us to categorize
the income levels of households receiving mortgages from credit unions and
banks, and reviewed other industry studies. To determine how CUMAA
affected field of membership requirements for federally chartered credit
unions, we analyzed NCUA regulations and obtained data on field of
membership trends from NCUA. In addition, we surveyed state regulators to
obtain information about their chartering provisions, particularly for
credit unions serving geographic areas. To determine how NCUA's
examination and supervision process has changed, we reviewed NCUA
documentation on its risk-focused program and conducted structured
interviews of NCUA regional directors and examiners, as well as selected
state credit union supervisors. We also analyzed NCUA data on examiner
resources provided to states and progress in implementing PCA. To
determine the financial condition of NCUSIF, we obtained and analyzed key
financial data about the fund from NCUA's annual audited financial
statements for 1991-2002. Finally, to assess the risks associated with the
use of private share insurance, we identified and analyzed relevant
federal and state statutes and regulations and surveyed the 50 state
credit union regulators to determine which states permitted private share
insurance. In addition, we conducted interviews with state supervisors
from states where credit unions are permitted to choose private
insurance-Alabama, California, Idaho, Illinois, Indiana, Maryland, Nevada,
and Ohio. We also interviewed and obtained relevant documentation from
representatives of American Share Insurance (ASI)-the remaining provider
of private share insurance. Appendix I provides additional details on our
scope and

6We only reviewed federally insured credit unions-about 98 percent of all
credit unions- because they were all required to submit call report data
to NCUA, while not all privately insured credit union call report data
were reported to NCUA. Call reports are submitted by credit unions to NCUA
and contain data on a credit union's financial condition and other
operating statistics. Throughout the report, when we use the term
"industry," we are referring to federally insured credit unions and
exclude the 212 privately insured credit unions.

methodology. We conducted our review from August 2002 through September
2003 in accordance with generally accepted government auditing standards.

Results in Brief	The overall financial condition of the credit union
industry, as measured by capital ratios, asset growth, and regulatory
ratings, has improved since our last report in 1991. An example of the
improved condition of the credit union industry is the decline in the
number of credit unions identified by NCUA as being in weak or
unsatisfactory condition-578 (about 5 percent of all credit unions) in
1992 compared with 211 (about 2 percent of all credit unions) in 2002.7
While credit union profitability, as measured by the return on assets
ratio, generally declined between 1992 and 1999, it has since stabilized.
The number of credit unions declined between 1992 and 2002 while total
industry assets have grown. This has resulted in two distinct groups of
credit unions-larger credit unions, which are fewer in number and provide
a wider range of services that more closely resemble those offered by
banks, and smaller credit unions, which are greater in number and provide
more basic financial services. Credit unions with over $100 million in
assets represented about 4 percent of all credit unions and 52 percent of
total credit union assets in 1992 compared with about 11 percent of all
credit unions and 75 percent of total credit union assets in 2002. These
larger credit unions were more likely to provide sophisticated financial
services, such as Internet banking and electronic loan applications, and
engage in mortgage lending than smaller credit unions.

As credit unions have become larger and expanded the range of services
they offer, the question of who receives services from credit unions has
been widely debated. While it has been generally accepted that credit
unions have a historical emphasis on serving people of modest means,
limited data exist that can be used to assess the income characteristics
of credit union members. Our analysis of available data suggested that the
income of credit union members is similar to that of bank customers;
although credit unions may serve a slightly lower proportion of low- and
moderate-income households than banks. Our analysis of the Federal
Reserve's 2001 Survey of Consumer Finances indicates that 36 percent of
households that primarily or only used credit unions had low and moderate

7NCUA rates credit unions using the CAMEL system, which stands for capital
adequacy, asset quality, management, earnings, and liquidity. The ratings
are 1 (strong), 2 (satisfactory), 3 (flawed), 4 (poor), and 5
(unsatisfactory).

incomes compared with 42 percent of households that used banks. Our
analysis of HMDA 2001 loan application records indicated that credit
unions provided a slightly lower percentage of their mortgages to low-and
moderate-income households than banks-27 percent compared with 34
percent-of comparable asset size. However, relying on HMDA data to
evaluate credit union service to low-and moderate-income households has
limitations because most credit unions are (1) small and, therefore, not
required to report HMDA data and (2) generally make more consumer loans
(for example, for cars) than residential mortgage loans. An analysis of
consumer loans or other services by household income would provide a more
complete picture of credit union service to low- and moderate-income
households. Other industry studies concluded that credit union members
tended to have higher incomes than nonmembers, but indicated that this was
likely due to credit union membership being primarily occupationally
based.

CUMAA authorized preexisting NCUA policies that had enabled federally
chartered credit unions to expand their membership over the last two
decades. In response to a Supreme Court decision, Congress enacted
provisions of CUMAA permitting federally chartered credit unions to form
multiple-bond credit unions-consisting of groups, such as for employment,
each with their own distinguishing characteristics-and permitted these
credit unions to add communities underserved by financial institutions to
their membership. NCUA permitted single- and community-bond, federally
chartered credit unions to add underserved communities to their field of
membership as well. CUMAA also amended a chartering provision authorizing
community credit unions by specifying that the area in which their members
are located should be "local." However, NCUA regulations have made it
easier for credit unions to qualify to serve larger geographic areas (for
example, entire cities). According to NCUA officials, these changes were
necessary to maintain the competitiveness of the federal charter with
respect to what they perceived as less restrictive field of membership
requirements allowed for state-chartered credit unions in some states.
While CUMAA permitted multiple-bond credit unions to add underserved
areas, and NCUA has stated its commitment to ensuring that credit unions
provide financial services to all segments of society, NCUA has not
developed indicators to determine if credit union services have reached
the underserved. Instead, NCUA uses "potential membership," the number of
people who could join credit unions, as an indirect measure of credit
union success in penetrating these areas.

In response to the growing concentration of assets in the credit union
industry and increased services and activities offered by credit unions,
NCUA adopted a risk-focused examination and supervision program similar to
that of other depository institution regulators. While NCUA has taken a
number of steps to ensure the successful implementation of its
risk-focused program, it faces a number of challenges. NCUA has met with
the other depository institution regulators, such as the Federal Deposit
Insurance Corporation (FDIC), to learn about how they implemented their
risk-focused programs. However, opportunities exist to further leverage
the experiences of other depository institution regulators to more
effectively deal with ongoing challenges such as ensuring that examiners
have sufficient training and expertise to evaluate the more sophisticated
activities of credit unions, such as Internet banking and member business
lending. Furthermore, unlike the other depository institution regulators,
NCUA lacks authority to review the operations of third-party vendors,
which credit unions increasingly rely on to provide services such as
Internet banking. However, these third-party arrangements present risks
such as threats to security of information systems, availability and
integrity of systems, and confidentiality of information. In addition,
credit unions are not subject to the internal control reporting
requirements that the Federal Deposit Insurance Corporation Improvement
Act of 1991 (FDICIA) imposed on banks and thrifts. NCUA implemented PCA,
in 2000 as mandated by CUMAA, as another control for safety and soundness
of the industry. To date, there have been very few credit unions subject
to PCA partially because of a generally favorable economic climate for
credit unions.

Indicators of the financial condition and performance of NCUSIF have
generally been stable over the past decade. For example, the ratio of fund
equity to insured shares-a measure of the fund's equity available to cover
losses on insured deposits-was within statutory requirements at December
31, 2002, as it has been over the past decade. While NCUSIF's net income
has remained positive through 2002, it experienced significant declines in
2001 and 2002 due to decreased yields from the investment portfolio,
increases in the amount paid to NCUA's Operating Fund for administrative
expenses (overhead transfer rate), and increasing insurance losses on
failed credit unions. NCUA's external auditors reviewed the basis on which
the transfer rate was determined and made several recommendations for
improvement that, according to NCUA officials, are being assessed and
implemented. While financial indicators have generally remained
satisfactory, NCUSIF is the only share or deposit insurance fund that has
not adopted a risk-based insurance structure. Currently, credit

unions are assessed a flat rate that does not reflect the risk that
individual credit unions pose to the fund. Moreover, NCUA's process for
estimating anticipated losses to the fund lacks precision, as it does not
identify specific historical failure rates and related loss rates for the
group of credit unions that have been identified as being in troubled
condition. As a result, NCUA may be over or underestimating probable
losses to the fund.

The overall system risk to the credit union industry that may be created
by private primary share insurance appears to have decreased since 1990,
although some concerns remain. The number of privately insured credit
unions and providers of private primary share insurance have declined
significantly since 1990. Specifically, in 1990, there were 1,462
privately insured credit unions-with $18.6 billion in insured
shares-compared with 212 privately insured credit unions-with about $10.8
billion in insured shares, as of December 2002. This represented a 42
percent decrease in privately insured shares. Moreover, during the same
period the number of private primary share insurers decreased from 10 to
1-ASI. Although the use of private share insurance has declined, some
circumstances of the remaining private insurer raise concerns. First,
ASI's insured risks are overly concentrated in a few large credit unions
and in certain states. Second, ASI may have a limited ability to absorb
catastrophic losses because it does not have the backing of any
governmental entity and its lines of credit are limited. However, ASI has
implemented a number of risk-management strategies, including increased
monitoring of its largest credit unions to help mitigate concentration
risk. In addition, state regulation of ASI and the privately insured
credit unions it insures provides some additional assurance that ASI and
the credit unions operate in a safe and sound manner. One additional
concern, as we recently reported, is that many privately insured credit
unions failed to make required disclosures about not being federally
insured and, therefore, the members of these credit unions may not have
been adequately informed that their deposits lacked federal deposit
insurance.

This report contains recommendations to NCUA and matters for congressional
consideration that, if implemented, would better ensure NCUA's ability to
achieve its goal of ensuring that credit unions can safely provide
financial services to all segments of society, promote greater consistency
in federal oversight of depository institutions, and enhance share
insurance management.

We requested comments on a draft of this report from the Chairman of the
National Credit Union Administration and the President and Chief

Executive Officer of American Share Insurance. We received written
comments from NCUA and ASI that are discussed in this report and reprinted
in appendixes XI and XII respectively. NCUA generally agreed with most of
the report's assessment regarding the challenges facing NCUA and credit
unions since 1991 and planned to implement the majority of the report's
recommendations. In commenting on a draft of the private share insurance
section, ASI stated that this report did not adequately assess the private
share insurance industry and objected to our conclusions that ASI's risks
are concentrated in a few large credit unions and a few states; it has
limited ability to absorb large losses because it does not have the
backing of any governmental agency; and its lines of credit are limited in
the aggregate as to amount and available collateral. In response, we
considered ASI's positions and materials provided, including ASI's
actuarial assumptions and ASI's past performance, and believe our report
addresses these issues correctly as originally presented.

Background	Credit unions differ from other depository institutions because
of their cooperative structure and tax exemption. Credit unions are
member-owned cooperatives run by boards elected by their members. They do
not issue capital stock; rather, they are not-for-profit entities that
build capital by retaining earnings. However, like banks and thrifts,
credit unions have either federal or state charters. Federal charters have
been available since 1934 when the Federal Credit Union Act was passed.
States have their own chartering requirements. As of December 2002, the
federal government chartered about 60 percent of the nearly 10,000 credit
unions, and about 40 percent were chartered by their respective states.
Both federally and state-chartered credit unions are exempt from federal
income taxes, with federally chartered and most state-chartered credit
unions also exempt from state income and franchise taxes.

Another distinguishing feature of credit unions is that they may serve
only an identifiable group of people with a common bond. A common bond is
the characteristic that distinguishes a particular group from the general
public. For example, a group of people with a common profession or living
in the same community could share a common bond. Over the years,
common-bond requirements at the state and federal levels have become less
restrictive, permitting credit unions consisting of more than one group

having a common bond to form "multiple-bond" credit unions.8 The term
"field of membership" is used to describe all the people, including
organizations, that a credit union is permitted to accept for membership.
As previously noted, the loosening of common-bond restrictions, as well as
expanded powers, have brought credit unions into more direct competition
with other depository institutions, such as banks. In addition, credit
unions can offer members additional services made available by third-party
vendors and by certain profit-making entities with which they are
associated, referred to as credit union service organizations (CUSO).9

CUMAA was the last statute that enacted major provisions affecting, among
other things, how federally chartered credit unions could define their
fields of membership and how federally insured credit unions demonstrate
the safety and soundness of their operations. In February 1998, the
Supreme Court ruled that NCUA lacked authority to permit federal credit
unions to serve multiple membership groups.10 In response, CUMAA
authorized multiple-group chartering, subject to limitations NCUA must
consider when granting charters. Also, the act limited new community
charter applications to well-defined "local" communities. Moreover, CUMAA
placed several additional restrictions on federally insured credit unions.
It tightened audit requirements, established PCA requirements when capital
standards were not met, and placed a cap on the percentage of funds that a
credit union could expend for member business loans.

NCUA has oversight responsibility for federally chartered credit unions
and has issued regulations that, among other things, guide their field of
membership and the scope of services they can offer. NCUA also has
responsibility for overseeing the safety and soundness of federally
insured credit unions through examinations and off-site monitoring. In
addition,

8See GAO/GGD-91-85 for additional background on the history of NCUA and
state field of membership regulatory policies.

9A CUSO is a corporation, limited liability corporation, or limited
partnership that provides services such as insurance, securities, or real
estate brokerage, primarily to credit unions or members of affiliated
credit unions. NCUA specifies which types of activities a CUSO may
undertake. Credit unions can invest up to 1 percent of paid-in and
unimpaired capital and surplus in CUSOs. Credit unions can loan up to an
aggregate of 1 percent of paid-in and unimpaired capital and surplus to
CUSOs. The CUSO must maintain a separate identity from the credit union.
See 12 C.F.R. Part 712 (2003).

10National Credit Union Administration v. First National Bank & Trust Co.,
522 U.S. 479 (1998).

NCUA administers NCUSIF, which provides primary share (deposit) insurance
for 98 percent of the nation's credit unions.11 NCUA, in its role as
administrator of NCUSIF, is responsible for overseeing federally insured,
state-chartered credit unions to ensure that they pose no risk to NCUSIF.

State governments have responsibility for regulating state-chartered
credit unions. State regulators oversee the safety and soundness of
state-chartered credit unions; although, as mentioned above, NCUA also has
responsibility for ensuring that state-chartered credit unions that are
federally insured pose no risk to NCUSIF. States set their own rules
regarding field of membership and the services credit unions can provide.
In addition, some states allow the credit unions in their states the
option of obtaining private primary share insurance. Currently, 212 credit
unions in eight states have primary share insurance from a private
company, ASI, located in Ohio. Primary share insurance for these privately
insured credit unions covers up to $250,000.

Financial Condition of the Credit Union Industry Has Improved Since 1991

Between 1992 and 2002, the capital ratios of federally insured credit
unions improved and remained higher than those of other depository
institutions. The industry's assets also grew over this period, coincident
with an increased emphasis on mortgage loans. Credit union industry
profitability, after declining from 1992 to 1999, has since stabilized. In
addition, since 1991 there has been a significant drop in the number of
problem credit unions as measured by regulatory ratings. Consolidation in
the industry has continued while total industry assets have grown, which
has in part resulted in two distinct groups of federally insured credit
unions-larger credit unions, which are fewer in number and provide a wider
range of services that more closely resemble those offered by banks, and
smaller credit unions, which are larger in number and provide more basic
financial services.

11Generally, primary deposit insurance covers the first portion of
members' deposits up to a specified amount. For example, NCUSIF provides
primary deposit insurance up to $100,000 per member per qualifying
account. In contrast, excess deposit insurance is optional coverage above
the amount provided by primary deposit insurance that credit unions may
purchase from private insurers.

Credit Union Capital Ratios Have Improved Since 1991 and Remain Higher
Than Those of Banks

The capital of federally insured credit unions as a percent of total
industry assets-the capital ratio-increased steadily between 1992 and 1997
and has since remained mostly level. As shown in figure 1, the capital
ratio of the industry was 8.1 percent in 1992, increased to 11.1 percent
in 1997, and was 10.9 percent in 2002. As a point of comparison, the
capital ratio of credit unions has remained higher than that of banks and
thrifts since 1992.12 As a result, credit unions have a greater proportion
of assets available to cover potential losses than banks and thrifts. This
may be appropriate since credit unions, unlike banks, are unable to raise
capital in the capital markets but must instead rely on retained earnings
to build and maintain their capital levels.

Figure 1: Comparison of Credit Union and Bank Capital Ratios, 1992-2002
Capital ratio

12

10

8

6

4

2

0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Credit unions

Banks and thrifts Source: Call report data.

Note: Bank and thrift data are from all FDIC-insured institutions filing
call reports, excluding insured branches of foreign institutions.

12Throughout the report we use the terms "banks," "banks and thrifts," and
"FDIC-insured institutions" interchangeably.

Industry Assets Have Grown and Asset Composition Has Changed

Total loans as a percent of total assets of federally insured credit
unions grew between 1992 and 2002. In 1992, 54 percent of credit union
assets were made up of loans and 16 percent were in U.S. government and
agency securities, while in 2002 loans represented 62 percent of industry
assets, and U.S. government and agency securities represented 14 percent
of total assets. The largest category of credit union loans was consumer
loans (a broad category consisting of unsecured credit card loans, new and
used vehicle loans, and certain other loans to members, but excluding real
estate loans such as mortgage or home equity loans), followed by real
estate loans. For example, in 2002, 31 percent of credit union total
assets were classified as consumer loans and 26 percent were classified as
real estate loans.

However, over time, holdings of real estate loans have grown more than
holdings of consumer loans. For example, real estate loans grew from 19
percent of total assets in 1992 to 26 percent in 2002, while consumer
loans grew from 30 percent to 31 percent over the same period. Despite a
larger increase in real estate lending relative to consumer lending,
credit unions still had a significantly larger percentage of consumer
loans relative to total assets compared with their peer group banks and
thrifts: consumer loan balances of peer group banks and thrifts were less
than 8 percent of total assets in 2002. To provide context, in terms of
dollar amounts, credit unions had $175 billion in consumer loans while
peer group banks and thrifts had $190 billion in consumer loans. However,
these banks and thrifts held a greater percentage of real estate loans
than credit unions. See appendix III for additional details.

Credit Union Profitability Has Been Relatively Stable in Recent Years

The profitability of credit unions, as measured by the return on average
assets, has been relatively stable in recent years.13 The industry's
return on average assets was higher in the early to mid-1990s than in the
late 1990s and early 2000s. While declining from 1.39 in 1993 to 0.94 in
1999, the return on average assets has since stabilized. It has generally
hovered around 1, which, by historical banking standards, is a performance
benchmark, and it was reported at 1.07 as of December 31, 2002. For
comparative purposes, the return on average assets for peer group banks
and thrifts was 1.24 in 2002. Earnings, or profits, are an important
source of capital for financial

13The return on average assets is calculated as the current period's net
income divided by the average of current period assets and prior year-end
assets.

institutions in general and are especially important for credit unions, as
they are mutually owned institutions that cannot sell equity to raise
capital. As previously mentioned, credit unions create capital, or net
worth, by retaining earnings. Most credit unions begin with no net worth
and gradually build it over time.

Regulatory Ratings Have Improved

Since we last reported on the financial condition of credit unions, there
has been a significant drop in the number of problem credit unions as
measured by the regulatory ratings of individual credit unions. Regulatory
ratings are a measure of the safety and soundness of credit union
operations, and credit unions with an overall CAMEL rating of 4 (poor) or
5 (unsatisfactory) are considered problem credit unions. The number of
problem credit unions declined by 63 percent from 578 (5 percent of all
credit unions) in 1992 to 211 (2 percent of total) in 2002.

Consolidation in Industry Has Widened the Gap between Larger and Smaller
Credit Unions

Total assets in federally insured credit unions grew from $258 billion in
1992 to $557 billion in 2002, an increase of 116 percent. During this same
period, total member shares in these credit unions grew from $233 billion
to $484 billion, an increase of 108 percent. At the same time, the number
of federally insured credit unions fell from 12,595 to 9,688. As a result
of the increase in total assets and the decline in the number of federally
insured credit unions, the credit union industry has seen an increase in
the average size of its institutions and a slight increase in the
concentration of assets. At year-end 1992, credit unions with more than
$100 million in assets represented 4 percent of all credit unions and 52
percent of total assets; at year-end 2002, credit unions with more than
$100 million in assets represented about 11 percent of all credit unions
and 75 percent of total assets. From 1992 to 2002, the 50 largest credit
unions by asset size went from holding around 18 percent of industry
assets to around 23 percent of industry assets. Despite the slight
increase in concentration of assets in the credit union industry, it was
neither as concentrated as the banking industry, nor did it witness the
same degree of increased concentration. From 1992 to 2002, the 50 largest
banks by asset size went from holding around 34 percent of industry assets
to around 58 percent of industry assets. Appendix IV has additional
information on assets in federally insured credit unions and banks.

This consolidation in the credit union industry has in part widened the
gap between two distinct groups of federally insured credit unions-larger
credit unions, which are relatively few in number and provide a wider

range of services, and smaller credit unions, which are greater in number
and provide more basic banking services. Figure 2 illustrates institution
size and asset distribution in the credit union industry as of 2002, with
institutions classified by asset ranges; smaller credit unions are
captured in the first category, while credit unions with assets in excess
of $100 million are separated into additional asset ranges for
illustrative purposes. For example, as of December 31, 2002, the 8,642
smaller credit unions-those with $100 million or less in total
assets-constituted nearly 90 percent of all credit unions but held only 25
percent of the industry's total assets (see right-hand axis of fig. 2).
Conversely, the 71 credit unions with assets of between $1 billion and $18
billion, held 27 percent of total industry assets (see right-hand axis of
fig. 2) but represented less than 1 percent of all credit unions.14

14There were 68 credit unions with assets between $1 billion and $5
billion, which held 21 percent of industry assets, and three credit unions
with assets in excess of $5 billion, which held 6 percent of industry
assets. As of December 31, 2002, the largest credit union held $17.6
billion in assets.

Figure 2: Credit Union Industry Size and Total Assets Distribution, as of
December 31, 2002

Number of credit unions

                           Percentage of total assets

10,000 30

8,642

25 8,000

20

6,000

                                       15

4,000

                                       10

2,000 5

00 $100 $250 $500 $1,000 $18,000 Total assets (in millions)

Number of credit unions

Percentage of total assets

Source: Call report data.

Note: This figure depicts credit union industry distribution both in terms
of the number of federally insured institutions in a particular size
category as well as the percentage of industry assets that are held by
institutions in that category. Group I credit unions had assets of $100
million or less; Group II credit unions had assets greater than $100
million and less than or equal to $250 million; Group III credit unions
had assets greater than $250 million and less than or equal to $500
million; Group IV credit unions had assets greater than $500 million and
less than or equal to $1 billion; and Group V credit unions had assets
greater than $1 billion and less than or equal to $18 billion, which is
the asset size, rounded up to the nearest billion dollars, of the largest
credit union as of December 31, 2002. Thus, Group I represents smaller
credit unions and Groups II, III, IV, and V represent larger credit
unions.

We observed that larger credit unions tended to hold a wider variety of
loans than did smaller credit unions, and larger credit unions emphasized
different loan types than smaller credit unions. For example, new and used
vehicle loans have represented a relatively greater proportion of total
assets for smaller credit unions, and nearly all smaller credit unions
held such loans. However, while nearly all of the larger credit unions
held new and used car loans, first mortgage loans represented a relatively
greater proportion of total assets for larger credit unions. In fact,
nearly all larger credit unions held first mortgage loans, junior mortgage
and home equity

loans, and credit card loans, while in general less than half of the
smaller credit unions held these loans. Larger credit unions also tended
to be more likely to provide more sophisticated services, such as
financial services through the Internet and electronic applications for
new loans. While nearly all larger credit unions offered automatic teller
machines, less than half of smaller credit unions did. In fact, when
compared with similarly sized peer group banks and thrifts, larger credit
unions tended to appear very similar to their bank peers in terms of loan
holdings. Appendixes IV and V provide further details.

Limited Comprehensive Data Are Available to Evaluate Income of Credit
Union Members

As credit unions have become larger and offer a wider variety of services,
questions have been raised about whether credit unions are more likely to
serve households with low and moderate incomes than banks. However,
limited comprehensive data are available to evaluate income of credit
union members. Our assessment of available data-the Federal Reserve's 2001
SCF, 2001 HMDA data, and other studies-provided some indication that
credit unions served a slightly lower proportion of households with low
and moderate incomes than banks. Industry experts suggested that credit
union membership characteristics-occupationally based fields of membership
and traditionally full-time employment status-could have contributed to
this outcome. However, limitations in the available data preclude drawing
definite conclusions about the income characteristics of credit union
members. Additional information, especially with respect to the income
levels of credit unions' members receiving consumer loans, would be
required to assess more completely whom credit unions serve.

Data Lacking on Income Characteristics of Credit Union Members and Users

It has been generally accepted, particularly by NCUA and credit union
trade groups, that credit unions have a historical emphasis of serving
people with modest means. However, there are currently no comprehensive
data on the income characteristics of credit union members, particularly
those who actually receive loans and other services. As credit unions have
become larger and expanded their offerings of financial services, industry
groups, as well as consumer advocates, have debated which economic groups
benefit from credit unions' services. Additionally, questions have been
raised about credit unions' exemption from federal income taxes. As stated
in our 1991 report, and still true, none of the common-bond criteria
available to federally chartered credit unions refers to the economic
status of their members or potential members.

Information on the extent to which credit unions are lending and providing
services to households with various incomes is scarce because NCUA,
industry trade groups, and most states (with the exception of
Massachusetts and Connecticut) have not collected specific information
describing the economic status of credit union members who obtain loans or
benefit from other credit union services.15 Credit unions, even those
serving geographic areas, are not subject to the federal Community
Reinvestment Act (CRA), which requires banking regulators to examine and
rate banks and thrifts on lending and service to low-and moderate-income
neighborhoods in their assessment area.16 As a consequence, credit unions
are not required by NCUA or other regulators to maintain data on the
extent to which loans and other services are being provided to households
with various incomes.

However, two states-Massachusetts and Connecticut-collect information on
the distribution of credit union lending by household income and the
availability of services because their state-chartered credit unions are
subject to examinations similar to those of federally regulated
institutions. Modeled on the federal examination procedures for large
banks, the state regulators apply lending and service tests to assess
whether credit unions are meeting the needs of the communities they have
set out to serve, including low-and moderate-income neighborhoods.
Massachusetts established its examination procedures in 1982, and

15The Credit Union National Association (CUNA) collects information about
the characteristics (for example, income, race, and age) of credit union
members but not specifically the income levels of members who actually
receive mortgage and consumer loans or use other services.

16The CRA requires all federal bank and thrift regulators to encourage
depository institutions under their jurisdiction to help meet the credit
needs of the local communities in which they are chartered, consistent
with safe and sound operations. See 12 U.S.C. S:S: 2901, 2903, and 2906
(2000). CRA requires that the appropriate federal supervisory authority
assess the institution's record of meeting the credit needs of its entire
community, including low- and moderate-income areas. Federal bank and
thrift regulators perform what are commonly known as CRA examinations to
evaluate services to low- and-moderate income neighborhoods. Assessment
areas, also called delineated areas, represent the communities for which
the regulators are to evaluate an institution's CRA performance.

Connecticut in 2001.17 All credit unions in Massachusetts are subject to
these examinations, including those whose field of membership is
community-based.18 In contrast, in Connecticut, only state-chartered
credit unions serving communities with more than $10 million in assets are
subject to the examination. According to a Connecticut state official, the
Connecticut legislature established its examination due to an increasing
trend of multiple-bond credit unions to convert to community-chartered
bonds, and the $10-million threshold was chosen because the legislature
believed credit unions of that size would normally have the personnel and
technological resources to appropriately identify and serve their market.
In May 2003, Connecticut started to examine community-chartered credit
unions with assets of more than $10 million.

Consumer and industry groups have debated if information that demonstrates
whether credit unions serve low- and-moderate income households is
necessary. Some consumer groups believe that credit unions should supply
information that indicates they serve all segments of their potential
membership. The Woodstock Institute-an organization whose purpose is to
promote community reinvestment and economic development in lower-income
and minority communities-recommended, among other things, that the CRA
requirement should be extended to include credit unions, based on a study
they believe demonstrated that credit unions are not adequately serving
low-income households.19 Woodstock Institute officials noted that they
would prefer to see CRA requirements applied to larger credit unions,
those with assets over $10 million. The National Federation of Community
Development Credit Unions (NFCDCU) has recommended that credit unions
whose fields of

17Overall, State officials reported that credit union examination ratings
have been similar to those of banks, except that credit unions have
received a somewhat lower percentage of "outstanding," the highest rating.
As of July 2003, no Massachusetts credit union had a rating lower than
"satisfactory" for Massachusetts's version of the CRA examination. The
officials also noted that analysis of HMDA data by itself is inadequate
because loan application records do not capture all the information
available in an application.

18The State of Massachusetts permits a credit union not serving geographic
areas to designate its membership as its assessment area. For example, one
credit union, serving a major communications company, designated its
membership as those who are employees or retired employees of the credit
union itself; retirees and employees of other communication companies,
including their affiliates and subsidiaries; and family members of
eligible employees and retirees.

19Woodstock Institute, "Rhetoric and Reality: An Analysis of Mainstream
Credit Unions' Record of Serving Low Income People" (Chicago: February
2002).

membership cover large communities should be affirmatively held
accountable for providing services to all segments of those communities,
and that NCUA publish annual reports on the progress and status of these
expanded credit unions.20 In contrast, NCUA and industry trade groups have
opposed these and related requirements largely because they state that no
evidence suggests that credit unions do not serve their members.21

Federal Reserve Board Data Suggest That Credit Unions Serve a Slightly
Lower Proportion of Low- and Moderate-income Households

Our analysis of the Federal Reserve Board's 2001 SCF suggested that credit
unions overall served a lower percentage of households of modest means
(low- and moderate-income households combined) than banks.22 More
specifically, while credit unions served a slightly higher percentage of
moderate-income households than banks, they served a much lower percentage
of low-income households. The SCF is an interview survey of U.S.
households conducted by the Federal Reserve Board that includes questions
about household income and specifically asks whether households use credit
unions or banks. Our analysis of the SCF indicated the following
percentages for those households that used a financial institution:23

o  8 percent of households only used credit unions,

20NFCDCU represents and provides, among other things, financial
assistance, technical assistance, and human resources to about 215
community development credit unions for the purpose of reaching low-income
consumers.

21In 2000, NCUA required that any type of application related to
expanding, converting, or chartering a community credit union include
information known as a "community action plan," which described the credit
union's plan for serving the entire community. In interim rules issued in
December 2001 and final rules adopted in May 2002, NCUA repealed this
requirement. In discussion of the final rule, NCUA stated: "It is an
unreasonable practice to require only certain credit unions to adopt
specific written policies addressing service to the entire community,
without any evidence that these credit unions are failing to serve their
entire communities." CUNA and the National Federation of Credit Unions
concurred with this decision. CUNA further noted that the imposition of
this requirement could encourage federally chartered credit unions to
convert to a state charter.

22The SCF is conducted every 3 years and is intended to provide detailed
information on the balance sheet, pension, income, and other demographic
characteristics of U.S. households and their use of financial
institutions. See appendix I for details.

23These percentages reflect the percent of households using financial
institutions as a percent of all financial institution users and does not
include those households that are sometime referred to as unbanked.

o  13 percent of households primarily used credit unions,24

o  17 percent of households primarily used banks, and

o  62 percent of households only used banks.

To provide a more consistent understanding of our survey results, we used
the same income categories used by financial regulators-low, moderate,
middle, and upper-in their application of federal CRA examinations.25

To determine the extent to which credit unions served people of "modest
means," we first combined households with low or moderate incomes into one
group and combined households with middle or upper incomes into another
group. We then combined the SCF data into two main groups- households that
only and primarily used credit unions versus households that only and
primarily used banks. As shown in figure 3, this analysis indicated that
about 36 percent of households that only or primarily used credit unions
had low or moderate incomes, compared with 42 percent of households that
used banks. Moreover, our analysis suggested that a greater percentage of
households that only and primarily used credit unions were in the middle-
and upper-income grouping than the proportion of households that only and
primarily used banks.

24Those who "primarily" used credit unions placed more than 50 percent of
their assets in credit unions and those who "primarily" used banks placed
more than 50 percent of their assets in banks. The term "use" refers to a
household's placement of assets in a checking, savings, or money market
account. Our methodology for determining these classifications was based
on work performed by Dr. Jinkook Lee, a professor and researcher at Ohio
State University. See Jinkook Lee and William A. Kelly Jr., in "Who Uses
Credit Unions?" (Prepared for the Filene Research Institute and the Center
for Credit Union Research, 1999, 2001).

25See appendix I for the income category definitions.

Figure 3: Income Characteristics of Households Using Credit Unions versus
Banks, Low and Moderate Income versus Middle and High Income Percentage of
households

70

64

60

50

40

30

20

10

0

Households Households only and primarily only and primarily using credit
unions using banks

Low and moderate income

Middle and upper income

Source: 2001 SCF.

To better understand the distribution of households by income category, we
also looked at each of the four income categories separately. As shown in
figure 4, this analysis suggested that the percentage of households that
only and primarily used credit unions in the low-income category was lower
than the percentage of households that used banks in the same category (16
percent versus 26 percent). In contrast, households that only and
primarily used credit unions were more likely to be moderate- and
middle-income (19 percent and 22 percent) than those that only and
primarily used banks (16 and 17 percent). Given that credit union
membership has traditionally been tied to occupational- or employer-based
fields of membership, the higher percentage of moderate- and middle-income
households served by credit unions is not surprising.

Figure 4: Income Characteristics of Households Using Credit Unions versus
Banks, by Four Income Categories

Percentage

50

43

40

30

20

10

0 Households only and Households only and primarily using credit unions
primarily using banks

Low income

Moderate income

Middle income

Upper income

Source: 2001 SCF.

Note: We found no statistical difference in the percentage of upper-income
households when the only and primarily using credit union group and the
only and primarily using bank group were compared.

We also attempted to further explore the income distribution of credit
unions' members by separately analyzing households that only used credit
unions or banks from those that primarily used credit unions or banks.
However, the results were subject to multiple interpretations due to
characteristics of the households in the SCF database. For example, when
user groups are combined and compared, the results may look different than
when the groups are separated and compared. Because such a high percentage
of the U.S. population only uses banks (62 percent), the data obtained
from the SCF is particularly useful for describing characteristics of bank
users but much less precise for describing smaller population groups, such
as those that only used credit unions (8 percent).

In addition to assessing the income characteristics of households using
credit unions and banks, we also performed additional analysis by
education, race, and age. The results of these analyses can be found in
appendix VI.

Credit Unions Made a Slightly Lower Proportion of Mortgage Loans to
Households with Low and Moderate Incomes Than Banks

As an indicator of the income levels of households that utilize credit
union services, we used 2001 HMDA loan application records to analyze the
income of households receiving mortgages for the purchase of one-to-four
family homes from credit unions and peer-group banks.26 Our analysis
indicated that credit unions reporting HMDA data made a lower proportion
of mortgage loans to households with low and moderate incomes than peer
group banks reporting HMDA data-27 percent compared with 34 percent.27
More specifically, credit unions made 7 percent of their loans to
low-income households compared with 12 percent for banks, and credit
unions made 20 percent of their loans to moderate-income households
compared with 22 percent for banks (see fig. 5).28

26HMDA, 12 U.S.C. S:S: 2801-2811 (2000), was enacted to provide regulators
and the public with information on home mortgage lending so that both
could determine whether institutions were serving the credit needs of
their communities. As required by the Federal Reserve Board's Regulation C
(12 C.F.R. Part 203), lenders subject to HMDA are required to collect data
containing information about the loan and the loan applicant. This
information is submitted on files known as loan application registers
(loan records). HMDA-reportable mortgages include those for home purchase,
home improvement, and refinancing of home purchase loans, but we analyzed
only those made for home purchases because these loans are a gateway to
homeownership and other loans are easier to obtain. See appendix I for
more information.

27We created a bank peer group that consisted of financial institutions
with less than $16 billion in assets because the largest credit union held
assets between $15 billion and $16 billion as of December 2001. We
excluded financial institutions that only made mortgages. Our analysis
included 4,195 peer group banks.

28To categorize the home purchaser's income, we used the 2001
HUD-estimated median income estimates for each Metropolitan Statistical
Area (MSA) based on the 1990 U.S. Census, as supplied by the Federal
Reserve Board. Results may have been more accurate if these estimates were
based on the 2000 U.S. Census. In 2003, HUD must begin basing their median
income estimates on the 2000 U.S. Census.

Figure 5: Mortgages Made by Credit Unions and Banks, by Income Level of
Purchaser, 2001 Percentage of loans

50

44

40

30

20

10

0 Credit unions Peer group banks

Low-income purchasers

Moderate-income purchasers

Middle-income-purchasers

Upper-income purchasers

Source: 2001 HMDA database.

Note: About 16 percent of all credit union and peer group bank loans
reported to HMDA were excluded from this analysis because their loan
records did not identify the MSA of the purchased property. Because we did
not know the MSA, we could not calculate a MSA median income to categorize
the loan. HMDA reporting requirements allow for the omission of the MSA
when the property is not located in an MSA where the institution has a
home or branch office. Also, percentages of loans made by credit unions do
not add up to 100 percent due to rounding.

We also analyzed and compared the proportion of mortgage loans reported by
peer group banks and credit unions for the purchase of homes by the median
family income of the census tracts in which the homes were located. We
found that credit unions made roughly the same proportion of loans for the
purchase of homes, by census tract income category, as banks. For example,
we found that both credit unions and banks made 1 percent of their loans
for the purchase of homes in low-income census tracts and that credit
unions made 9 percent of their loans for the purchase of properties in
moderate-income census tracts compared with 10 percent by banks (see fig.
6). In addition, we found that both credit unions and banks made 54
percent of their loans for the purchase of homes in middle-

income census tracts, and that credit unions made about 37 percent of
their loans in upper-income census tracts compared with 35 percent by
banks. This analysis is a measure of whether all neighborhoods (census
tracts within an assessment area) are receiving financial services,
including low-and moderate-income ones.

Figure 6: Loans Made by Credit Unions and Banks, by Average Income in the
Purchased Home's Census Tract, 2001 Percentage of loans

                                       60

                                       50

                                       40

                                       30

                                       20

                                       10

                                       0

                                     54 54

Credit unions Peer group banks

Low-income tract Moderate-income tract Middle-income-tract Upper-income
tract

Source: 2001 HMDA database.

Note: About 16 percent of the credit union and peer group bank loans
reported to HMDA were excluded from this analysis because their loan
records did not identify the census tract of the purchased property.
Because we did not know the census tract, no census tract median income
was available to categorize the loan. HMDA reporting requirements allow
for the omission of the census tract locations under certain conditions;
for example, when the property did not have an identified census tract for
the 1990 census or was located in a county with a population of 30,000 or
less. Also, percentages of loans made by credit unions do not add up to
100 percent due to rounding.

Because each HMDA loan record identified the income of the mortgage loan
recipient and the location of the property, the HMDA database allowed us
to determine the proportion of mortgages made within the four income

categories-low, moderate, middle, and upper-used by financial regulators
for CRA examinations. However, not all financial institutions are required
to report HMDA data-for example, depository institutions were exempt from
reporting data in 2001 if they had assets less than $31 million as of
December 31, 2000, and if they did not have a home or branch office in an
MSA. Further, not all credit unions, including those that had more than
$31 million in assets, made home purchase loans.29 As a result, most
credit unions did not meet HMDA's reporting criteria-only about 14 percent
of all credit unions submitted data included in our analysis.30 On the
other hand, the credit unions that did report their loans to HMDA held
about 70 percent of credit union assets and included about 62 percent of
all credit union members.

HMDA Analysis Has Certain Limitations

Our analysis of HMDA data allowed us to determine the overall proportion
of mortgage loans credit unions and peer group banks made to households
and neighborhoods with low and moderate incomes. However, we would need
information on the proportion of low-and moderate-income households within
credit union fields of membership to actually make an evaluation of
whether credit unions, collectively or individually, have met the credit
needs of their entire field of membership. Similar to analyses used in
federal CRA lending tests, this information could then be used as a
baseline from which to evaluate an individual credit union's actual
lending record.31 In addition, information on factors (for example, a
community's economic condition, local housing costs) that could affect the
ability of a

29Our analysis of NCUA call report data indicated that 93 percent of
credit unions with more than $31 million in assets, as of December 31,
2000, made first mortgage loans, loans that include home purchase loans,
compared with only 34 percent of credit unions with fewer assets.

30In total, for 2001, 1,717 credit unions reported data to HMDA, but our
analyses only included the 1,446 that made mortgage loans that met our
criteria. For example, we only included mortgage loans for home purchases
rather than refinancing.

31For larger institutions, those with more than $250 million in assets,
CRA examinations generally consist of three parts-a lending test, a
service test, and an investment test. The lending test entails a review of
an institution's lending record, including originations and purchases of
home mortgages, small business, small farm, and, at the institution's
option, consumer loans throughout the institution's assessment area,
including low- and moderate-income areas. The lending test is weighted
more heavily than the investment and service tests in the institution's
overall CRA rating. The service test requires the examiner to analyze an
institution's system for delivering retail banking services and the extent
and innovativeness of its community development services. The investment
test evaluates an institution's investment in community development
activities.

credit union to make loans consistent with safe and sound lending would be
necessary to evaluate an institution's lending record. If regulators were
to make these types of evaluations for credit unions, they would be easier
to implement for those serving geographic areas because demographic
information (for example, on census tract median income levels) would be
available to describe credit union field of membership. For credit unions
with an occupational or associational membership, other ways of
characterizing their field of membership would need to be determined.

In addition, as previously mentioned, using HMDA data to analyze credit
union mortgage lending to members does not provide any information on
smaller credit unions, because in 2001 credit unions with less than $31
million in assets as of December 31, 2000, were not required to report
HMDA data. Because smaller credit unions did not report HMDA data, one
group of credit unions-the roughly 3,800 credit unions that qualified for
NCUA's Small Credit Union Program in December 2002-were largely excluded
from our HMDA analysis. Credit unions qualifying for assistance from this
program must have less than $10 million in assets or have received a
"low-income" designation from NCUA.32 In addition, low-income credit
unions must demonstrate that more than half of their current members meet
one of NCUA's low-income criteria.33 Further, smaller credit unions are
more likely than larger credit unions to make consumer loans than
mortgages, making an evaluation of mortgage lending more relevant to
larger credit unions than smaller ones. Because most credit unions can be
classified as small, analyzing the distribution of consumer loans by
household income would provide a more complete picture of credit union
lending.34

32These credit unions receive special help from NCUA regional staff,
including assistance in completing business plans and maintaining
financial records. Low-income credit unions also qualify for low-interest
loans and technical assistance grants and are permitted to accept
nonmember deposits and secondary capital accounts. According to NCUA
estimates, as of December 31, 2002, the median asset level of these credit
unions was about $3.4 million. About 107 of these credit unions had more
than $32 million in assets, the threshold for reporting lending data to
HMDA in 2003.

33As of December 31, 2002, there were 907 low-income credit unions. Credit
unions can use a number of methods to document their low-income
eligibility, such as reviewing loans to identify members' wages or
household incomes, or written membership surveys that request the members'
total household income and annual wages.

34See appendix V for more detailed information on credit union services by
asset size.

Other Studies Indicate That Credit Unions Serve Households with Higher
Incomes Than Banks

Other recently published studies-CUNA and the Woodstock Institute-
generally concluded that credit unions served a somewhat higher-income
population. The studies noted that the higher income levels could be due
to the full-time employment status of credit union members.

The CUNA 2002 National Member Survey reported that credit union members
had higher average income households than nonmembers- $55,000 compared
with $46,000.35 The report provided several reasons for the income
differential, including the full-time employment status of credit union
members, credit union affiliation with businesses or companies, and weak
credit union penetration among some of the lowest-income age groups-18 to
24 and 65 and older. However, the report noted that additional analyses,
specifically those grouping consumers based on the extent to which they
rely on banks and credit unions as their primary provider should also be
considered.36 In addition, a study sponsored by the Woodstock Institute,
based on an analysis of 1999 and 2000 survey responses obtained from
households in the Chicago, Illinois, metropolitan area concluded that
credit unions in the Chicago region served a lower percentage of
lower-income households than they did middle- and upper-income ones.37 For
example, while 40 percent of surveyed households with incomes between
$60,000-$70,000 contained a credit union member, only 23 percent of
households earning between $30,000-$40,000 contained a

35CUNA 2002 National Member Survey and research and information from CUNA
and affiliates. CUNA based its statistics on average household income on a
survey of 1,000 randomly selected households conducted in February 2002.
The data from this survey were weighted to accurately represent U.S.
consumers age 18 and older.

36CUNA supplemented its average income analysis of members and nonmembers
with one that divided consumers into four institution user groups-as
similarly done by Jinkook Lee, in "Who Uses Credit Unions" in her analysis
of the SCF and in our previous analysis-and calculated the average
household income of each institution user group. CUNA determined that
consumers who only used banks and only used credit unions had a lower
average income than consumers who used both institutions. In addition,
when comparing the average income of consumers who used both institutions,
the analysis concluded that those who primarily used credit unions had a
slightly lower average income than those who primarily used banks.

37The study cited is "Rhetoric and Reality: An Analysis of Mainstream
Credit Unions' Record of Serving Low Income People" (February 2002). To
determine the characteristics of credit union members, the Woodstock
Institute analyzed 1999 and 2000 survey data collected by the Metro
Chicago Information Center (MCIC). MCIC surveyed roughly 3,000 households
in the Chicago area and asked respondents whether they were members of a
credit union. However, the survey did not specifically ask whether the
respondents held accounts at a bank or credit union.

credit union member. The study also noted that household members working
for larger firms, and those who were members of a labor union, were
significantly more likely to be credit union members.

Officials from NCUA and the Federal Reserve Board also noted that credit
union members were likely to have higher incomes than nonmembers because
credit unions are occupationally based. An NFCDCU representative noted
that because credit union membership is largely based on employment,
relatively few credit unions are located in low-income communities.
However, without additional research, especially on the extent to which
credit unions with a community base serve all of their potential members,
it is difficult to know whether full-time employment is the sole
explanatory factor.

CUMAA Authorized NCUA to Continue Preexisting Policies That Expanded Field
of Membership

The Credit Union Membership Access Act of 1998 authorized preexisting NCUA
policies that had allowed credit unions to expand field of membership. In
1998, the Supreme Court ruled against NCUA's practice of permitting
federally chartered credit unions to consist of more than one common
bond.38 In CUMAA, Congress specifically permitted credit unions to form
multiple-bond credit unions and allowed these credit unions to serve
underserved areas.39 CUMAA also specified that community-chartered credit
unions serve a "local" area.40 However, after the passage of CUMAA, NCUA
revised its regulations to make it easier for credit unions to serve
communities larger than before CUMAA. To some extent, these NCUA policies
appear to have been triggered by concerns about competing with the states
to charter credit unions. While CUMAA permitted multiple-bond credit
unions to add underserved areas to their membership, the impact of this
provision will be difficult to assess because NCUA does not track credit
union progress in extending service to these communities.

38National Credit Union Administration v. First National Bank & Trust Co.,
522 U.S. 479 (1998).

39Pub. L. No. 105-219 S: 101(2).

40Id.

CUMAA Permitted NCUA Policies Expanding Field of Membership

CUMAA authorized several preexisting NCUA field of membership policies
that had enabled federally chartered credit unions to expand their fields
of membership. These policies had allowed credit unions to consist of more
than one membership group and expand their membership to include
underserved areas. In addition, CUMAA permitted credit unions to retain
their existing membership.

Specifically, CUMAA affirmed NCUA's 1982 policy of permitting credit
unions to form multiple-bond credit unions, allowing these credit unions
to retain their current membership and authorizing their future
formation.41 A credit union with a single common bond has members sharing
a single characteristic, for example, employment by the same company. In
contrast, multiple-bond credit unions consist of more than one distinct
group. 42 Congressional affirmation of NCUA's policy of permitting
multiple-bond credit unions was important because earlier in 1998 the
Supreme Court had ruled that federally chartered, occupationally based
credit unions were required to consist of a single common bond.43 Figure 7
provides additional information since 2000 on the percent of federally
chartered credit unions by charter type.44

41CUMAA permitted the following common bonds: (1) the single common bond,
defined as one group with a common bond of occupation or association; (2)
the multiple common bond, defined as including more than one group, each
with a common bond of occupation or association; and (3) the community
bond, defined as persons or organizations within a well-defined local
community, neighborhood, or rural district. Formation of multiple
common-bond credit unions is limited to groups having fewer than 3,000
members unless NCUA grants an exception based on criteria contained in
CUMAA. See 12 U.S.C. S: 1759(b), (d), as amended.

42According to NCUA officials, single-bond credit unions are more
susceptible to failure because they are reliant on one type of
occupational group. For example, if an occupational group were subject to
layoffs, the credit union could lose its membership base or experience a
decline in assets.

43National Credit Union Administration v. First National Bank & Trust Co.,
522 U.S. 479 (1998).

44Although single-bond credit unions included about 38 to 40 percent of
all federally chartered credit unions between 2000 and March 2003, during
this time period they only held about 18 percent of all assets of
federally chartered credit unions. In contrast, federally chartered
multiple-bond credit unions held about 70 percent of federal assets in
March 2000, and this percentage dropped to about 65 percent in 2003.
Federally chartered community credit unions held about 13 percent of
federal assets in 2000, and this percentage increased to about 17 percent
of assets in March 2003.

Figure 7: Percentage of Federally Chartered Credit Unions, by Charter
Type, 2000-
2003
Percentage of credit unions

50 48 48 48 47

40

30

20

10

0

2000 2001 2002 2003 (through March)

Community

Single bond

Multiple bond

Source: NCUA.

Note: With the exception of the statistics provided for multiple-bond
credit unions for 1996, NCUA cannot provide us data on federal chartering
trends before 2000. However, NCUA was able to report that by 1996, about
half of all federally chartered credit unions were multiple-bond credit
unions.

In addition, CUMAA affirmed other preexisting NCUA policies. For example,
CUMAA authorized multiple-bond credit unions to add individuals or
organizations in "underserved areas" to their field of membership. This
provision was similar to an NCUA policy that permitted multiple-bond
credit unions, as well other federally chartered, single-bond, and
community-chartered credit unions, to add low-income communities

to their field of membership.45 In addition, CUMAA affirmed NCUA's "once a
member, always a member policy," which had been in effect since 1968.
CUMAA authorized this policy such that credit union members may retain
their membership even after the basis for the original bond ended.46
However, CUMAA still contained provisions encouraging the creation of new
credit unions whenever possible.47

NCUA Eased Requirements for Permitting Credit Unions to Serve Larger
Geographic Areas

Despite the qualification in CUMAA that a community-chartered credit
union's members be within a well-defined "local" community, neighborhood,
or rural district, NCUA eased requirements for permitting credit unions to
serve larger geographic areas. CUMAA added the word "local" to the
preexisting requirement that community-chartered credit unions serve a
"well-defined community, neighborhood, or rural district," but provided no
guidance with respect to how the word "local" or any other part of this
requirement should be defined.48

45In 1994, NCUA's Interpretive Ruling and Policy Statement (IRPS) 94-1
authorized all federally chartered credit unions, regardless of bond, to
include in their membership, without regard to location, communities and
associational groups satisfying the definition of low income. This program
should not be confused with NCUA's "low-income designated program," which
permits credit unions who exclusively serve low-income areas to maintain
secondary capital and accept nonmember deposits.

46Pub. L. No. 105-219 S: 101 (2), 12 U.S.C. S: 1759 (e)(2), as amended.
Under this provision, once a person becomes a member of a credit union,
that person or organization may remain a member of that credit union until
the person or organization chooses to withdraw from membership in the
credit union.

47The Federal Credit Union Act requires NCUA to encourage the formation of
separately chartered credit unions instead of approving the inclusion of
an additional common-bond group within the field of membership of an
existing credit union. 12 U.S.C. S: 1759(f)(1). From 1991 to March 2003,
only 143 new federally insured credit unions were chartered, an average of
about 11 to 12 new credit unions per year. NCUA said that small groups are
generally not economically sustainable and prefer to join multiple-bond
credit unions.

48Pub. L. No. 105-219 S: 101; See 12 U.S.C. S: 1759(c)(2), as amended.

Following passage of CUMAA, NCUA expanded the ability of credit unions to
serve larger geographic areas through its regulatory rulings.49
Interpretive Ruling and Policy Statement (IRPS) 99-1, issued soon after
CUMAA, was the first regulation to set standards for what could be
considered a "local" area. It required credit unions to document that
residents of a proposed community area interact or have common interests.
Credit unions seeking to serve a single political jurisdiction (for
example, a city or a county) with more than 300,000 residents were
required to submit more extensive documentation than jurisdictions with
fewer than 300,000 residents.50 However, IRPS 03-1, which replaced IRPS
99-1, eliminated these documentation requirements, regardless of the
number of residents. Further, IRPS 03-1 allowed credit unions to propose
MSAs with less than 1 million residents for qualification as local areas.
See table 1 for changes in "local" requirements. NCUA adopted these
definitions of local community based on its experience in determining what
constituted a local community charter.

49Prior to CUMAA, NCUA regulations did not limit the size of the community
a credit union could serve. However, NCUA required extensive documentation
to establish the existence of a community. For example, up until March 1,
1998, credit unions were required to provide written evidence of community
support for their applications, such as letters of support, petitions, or
surveys. In March 1998, in IRPS 98-1, NCUA deleted the information
requirement but noted that credit unions still had to demonstrate that
residents of the proposed community interacted.

50For example, in IRPS 99-1, if the population of a single political
jurisdiction was less than 300,000, the credit union was only obligated to
submit a letter describing how the area met standards for community
interaction or common interests. However, if the population exceeded
300,000, the credit union would have to submit additional documentation;
demonstrating, for example, the existence of major trade areas or shared
facilities (such as educational).

Table 1: Regulatory Definitions of Local Community, 2000 and 2003

IRPS 99-1, effective in November 2000,
(as amended by IRPS 00-1) IRPS 03-1, effective in May 2003

1. Areas in single political jurisdictions (for example, counties or
cities) qualified as a local community if the number of residents did not
exceed 300,000.

2. States, noncontiguous jurisdictions, and MSAs did not meet the
definition of a local community.

3. Contiguous political jurisdictions qualified as a local community if
they contained 200,000 or fewer residents.

4. A letter describing community interaction or common interests was
required for conditions (1) and (3) above. Otherwise, the credit union had
to provide additional documentation.

1. Any city, county, or political equivalent in a single political
jurisdiction, regardless of population size, automatically meets the
definition of a local community.

2. MSAs may meet the definition of local community provided the population
does not exceed 1 million.

3. Contiguous political jurisdictions qualify as a local community if they
contain 500,000 or fewer residents.

4. A letter describing community interaction or common interests is
required for conditions (2) and (3) above. Otherwise, the credit union
must provide additional documentation.

Source: IRPS 99-1 and IRPS 03-1.

Note: NCUA amended IRPS 99-1, the first field of membership regulation
issued by NCUA after CUMAA, several times (IRPS 00-1 on Oct. 27, 2000;
IRPS 01-1 on March 2001; and IRPS 02-2 on April 24, 2002.) This table only
highlights key changes pertaining to the geographic and population
criteria used by NCUA to approve community charters.

Specifically, NCUA officials said that they decided single political
jurisdictions should automatically qualify as "local" areas based on their
review of applications by credit unions for community charters. They
reported that they came to this conclusion because credit unions
converting to a community charter or expanding their service areas had
generally been able to successfully supply the documentation required by
NCUA. We asked NCUA officials what kind of relationships
community-chartered credit union members could have if, for example, a
local community were to be defined as all of New York City. NCUA officials
said that the defining factors for them were that people lived in the same
political jurisdiction-thus providing, for example, a common government
and educational system-and noted that credit unions applying to serve
these larger jurisdictions still had to meet other requirements related to
safety and soundness. The officials also said that had CUMAA not
introduced the word "local," NCUA could have considered providing credit
unions permission to expand their field of memberships statewide.

The regulatory changes in IRPS-03-1 pertaining to the definition of local
community have made it easier for federally chartered credit unions to
serve larger communities. Under IRPS-03-1, NCUA approved the largest
community yet-the 2.3 million residents of Miami-Dade County, Florida.51
NCUA had disapproved this same credit union's request about 2 years
earlier, under IRPS 99-1, as amended by IRPS 01-1. Prior to IRPS-03-1,
some of the largest community field of memberships approved by NCUA
included service to 836,231 residents on Oahu, Hawaii, and service to
710,540 residents in Montgomery County and Greene County, Ohio.52 In
addition, over the last 3 years, potential membership--an estimate of the
maximum number of members that could join a credit union--in
community-chartered credit unions has come to exceed that in multiple-bond
credit unions.53 According to NCUA estimates, in March 2003,
community-chartered credit unions had 98 million potential members
compared with multiple-bond credit unions with 92 million potential
members (see fig. 8).

51This multiple-bond credit union, located in Miami, Florida, originally
applied to serve Miami-Dade County, Florida, in April 2001. However, NCUA
officials denied both the original application and subsequent appeal on
the grounds that the residents of this area (including two large cities
and 28 other municipalities) did not have common interests or
interactions. As required by IRPS 99-1 (as amended by IRPS 01-1), the
credit union was required to supply documentation that residents within
this area interacted but the evidence, while described as "voluminous" by
NCUA officials, did not meet with their approval. Under the new rule (IRPS
03-1), approved in May 2003, this level of evidence was no longer
required.

52While the examples in this paragraph represent some of the largest
community-charter field of memberships approved by NCUA, the population
sizes of these communities can vary tremendously. For example, in 2002,
NCUA field of membership approvals ranged from a population of 695 in
Delta County, Colorado, to a population of 1.1 million residents in the
area surrounding Maple Grove, Minnesota. Since 1999, the average
population of approved communities has increased-in 1999, this average was
134,000 and as of June 25, 2003, 357,000.

53Federally insured credit unions are required to report their potential
membership on NCUA's call report. This number is expected to include
current membership as well as potential members. While the instructions
require that the estimates must be reasonable and supportable, no further
instructions are provided. Two or more credit unions whose field of
membership overlaps can count the same person as a potential member.

Figure 8: Actual and Potential Members in Federally Chartered Credit
Unions, by Charter Type, 2000-2003

                      Number of members in millions 100 80

                                       60

                                       40

                                      20 0

According to NCUA, a major reason for NCUA's recent regulatory changes was
to maintain the competitiveness of the federal charter in a dual
chartering system. They also characterized NCUA's field of membership
regulations as more restrictive than those in some states. Officials in
three of the states in which we conducted interviews-California, Texas,
and Washington-said that the ability to expand field of membership more
readily under state rules was a reason that federally chartered credit
unions had converted to state charters.

Consistent with this assertion, we found that state-chartered credit
unions have experienced greater membership growth, although federally
chartered credit unions still had more members. Between 1990 and March
2003, state-chartered credit union membership increased by 88 percent,

                                      2000

                                      2001

2002
2003 (through March)

                                 Source: NCUA.

Dual Chartering System May Have Created Pressure for Less Restrictive
Field of Membership Regulations

from 19.5 million to 36.6 million, while membership in federally chartered
credit unions increased by 24 percent, from 36.2 million to 44.9 million.
In addition, if estimates of potential membership serve even as an
approximation of future membership, state-chartered credit unions could be
positioned to experience greater growth (see fig. 9). In March 2003,
state-chartered credit unions had about 405 million potential members,
almost twice the 208 million for federally chartered credit unions.

Figure 9: Actual and Potential Members in Federally and State-chartered
Credit

Unions, 1990-2003

Number of members in millions

450

400

350

300

250

200

150

100

50

0

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
(through March)

State actual membership

Federal actual membership

State potential membership

Federal potential membership

Source: NCUA.

Note: In 2001, the total population of the United States was about 285
million people. In contrast, between 2001 and 2003, the total number of
potential credit union members ranged from 446 million to about 613
million. The total number of potential members exceeds the total
population of the United States because credit unions can count the same
individuals as potential members when their field of membership overlaps.

We also found that states had chartered a higher percentage of their
credit unions to serve geographic areas (communities) than NCUA.54 In
2002, we estimated that about 1,146 state-chartered credit unions, 30
percent of all state-chartered credit unions, served geographic areas
compared with 848 federally chartered credit unions, 14 percent of all
federally chartered credit unions.55 However, this number increases to
1,096, 18 percent of all federally chartered credit unions, once federally
chartered credit unions serving underserved areas are included.
State-chartered credit unions serving geographic areas held about 59
percent of state-chartered credit unions assets compared with 17 percent
held by federally chartered credit union serving geographic areas, or 29
percent when the assets of credit unions with underserved areas were
included.56

Credit Unions Have Added Underserved Areas, but No Information Available
to Evaluate Actual Service

An NCUA objective is to ensure that credit unions provide financial
services to all segments of society, including the underserved, but NCUA
has not developed indicators to evaluate credit union progress in reaching
the underserved.57 This type of evaluation could require information
similar

54We use the term "serving geographic areas" because some states (for
example, California and Texas) permit their credit unions to serve a mix
of occupational and associational groups and communities. Because NCUA
could not provide us information on the number of state-chartered credit
unions serving communities, we surveyed state regulators to obtain this
information.

55The number of credit unions serving geographic areas varied by state.
For example, in California, state-chartered credit unions serving
geographic areas represented about 48 percent of state-chartered credit
unions and held about 82 percent of state-chartered assets. In comparison,
in New York, state-chartered credit unions serving geographic areas
represented about 5 percent of state-chartered credit unions and held
about 11 percent of state-chartered assets.

56Because chartering provisions among the states and the federal
government vary, we would like to emphasize that these numbers are
estimates only. For example, we had no way of knowing, short of contacting
individual credit unions, whether state-chartered credit unions relied
more extensively on a community or an occupational group for their
membership. In addition, some state-chartered credit unions were excluded
from our calculations, including those that were privately insured,
because we could not identify them in the NCUA call report data.

57Part of NCUA's vision statement, included as part of its 2003-2008
Strategic Plan, is: "Ensure the cooperative credit union movement can
safely provide financial services to all segments of American society."
Further, in NCUA's 2003 Annual Performance Plan, NCUA states as a specific
goal that it plans to "Facilitate credit union efforts to increase credit
union membership and accessibility to continue to serve the underserved,
and enhance financial services."

to that provided as part of CRA examinations-for example, information on
the distribution of loans made by the income levels of households
receiving mortgage and consumer loans-and provide comprehensive
information on how credit unions have utilized opportunities to extend
their services to underserved areas, including low- and moderate-income
households. 58 CUMAA had specifically provided that multiple-bond credit
unions could serve underserved areas, and NCUA permitted single-bond and
community-bond credit unions to add them as well. However, neither CUMAA
nor NCUA required that credit unions report on services to these areas
once they had been added. Figure 10 shows the number of underserved areas
added before and after CUMAA.

58The Federal Credit Union Act, as amended by CUMAA, provides NCUA
criteria to use to determine if an area is "underserved." See 12 U.S.C. S:
1759 (c)(2). Among other things, these areas must qualify as "investment
areas" as defined by section 103 (16) of the Community Development Banking
and Financial Institutions Act of 1994 (12 U.S.C. S: 4703(16)). Areas
could qualify, for example, by having at least 20 percent of their
population living in poverty. Second, areas must qualify as underserved
based on data from the NCUA board and the federal banking agencies. NCUA
officials, however, apply only the first criterion, presuming that areas
qualifying as an investment area automatically qualify as underserved.

Number of underserved areas

350

300

250

200

150

100

50

0 1997 1998 1999 2000 2001 2002

Single bond

Multiple bond

Community

Source: NCUA.

Note: Between 1997 and 1999, credit unions were adding communities under
NCUA's low income standards. While CUMAA did not specifically permit
single-bond and community-chartered credit unions to add underserved
areas, NCUA permitted them to do so.

Instead of developing indicators to evaluate credit union progress in
reaching the underserved, NCUA officials have claimed success based on the
increase in the number of potential members added by credit unions in
underserved areas and, recently, on the membership growth rate of
federally chartered credit unions that have added underserved areas. As of
March 2003, credit unions had added 48 million potential members in
underserved areas. As noted previously, potential membership is an
estimate of the maximum number of people who could be eligible to join a
credit union. However, NCUA officials believe that potential membership is
an appropriate measure because they view NCUA's role as expanding
membership opportunities for credit unions as opposed to the credit

unions' role of actually extending services to new members.59 In addition,
in June 2003, NCUA claimed success based on estimates indicating that
annual membership growth in credit unions that expanded into underserved
areas has been higher than that of all federally chartered credit
unions-4.8 percent compared with 2.49 percent. However, they could not
identify whether the increase in membership actually came from the
underserved areas or provide any descriptive information (for example, the
income level) about the new members.

Because NCUA does not collect information on credit union service to
underserved areas, it would be difficult for NCUA or others to demonstrate
that these credit unions are actually extending their services to those
who have lower incomes or do not have access to financial services.60 As
the number of credit unions adding underserved areas increases, this
question becomes more important. For example, in 1999, the year after
CUMAA, 13 credit unions added 16 underserved areas to their membership. In
2002, 223 credit unions added about 424 underserved areas. Further, the
size of these communities can be substantial. For example, in May 2003,
NCUA permitted one multiple-bond credit union to add an additional 300,000
residents within Los Angeles County, California, for a total of almost 1
million added residents in the last 2 years. In the same month, NCUA also
approved a multiple-bond credit union's (headquartered in Dallas, Texas)
addition of 600,000 residents in underserved communities in Louisiana.

59To promote adoption of these areas, NCUA developed a public relations
program called "Access to America" that promotes awareness of NCUA
programs that provide resources, or other support, to credit unions to
expand their financial services to the underserved.

60CUNA published a study, 2003 "Serving Members of Modest Means" Survey
Report, on how credit unions served consumers having annual household
incomes of $40,000 or less. While the survey findings cannot be
generalized to all credit unions, the survey results indicated that most
credit unions responding to the survey targeted at least one service (for
example, money orders, check-cashing services) to lower-and
moderate-income members, and that credit unions with underserved areas
were likely to offer more of these services. About 35 percent of the
credit unions responding to the survey indicated they would grant a loan
for $100 or less and about 30 percent indicated they would open a
certificate account for less than $100. The study noted that credit unions
had difficulty responding to questions that asked them to estimate
members' or potential members' income distributions.

NCUA Adopted Risk-focused Examination and Supervision Program, but Faces
Challenges in Implementation

Industry consolidation and changes in products and services offered by
credit unions prompted NCUA to move from an examination and supervision
approach that was primarily focused on reviewing transactions to an
approach that focuses NCUA resources on high-risk areas within a credit
union. Prior to implementing its risk-focused program in August 2002, NCUA
sought guidance from other depository institution regulators that had
several years of experience with risk-focused programs. While this
consultative approach helped NCUA, it still faces a number of challenges
that create additional opportunities for NCUA to leverage off the
experience of the other depository institution regulators. These
challenges include ensuring that examiners have sufficient expertise in
areas such as information systems, monitoring the risks posed by expansion
into nontraditional credit union activities such as business lending, and
monitoring the risks posed to the federal deposit (share) insurance fund
by institutions for which states are the primary regulator. Moreover,
unlike other depository institution regulators, NCUA currently lacks
authority to inspect third-party vendors, which credit unions increasingly
rely on to provide services such as electronic banking. Further, credit
unions are not subject to the internal control reporting requirements that
banks and thrifts are subject to under FDICIA.61 NCUA adopted prompt
corrective action, a system of supervisory actions tied to the capital
levels of an institution, in August 2000, as required by CUMAA; few
actions have been taken to date due to a generally favorable economic
climate for credit unions.

Changes in the Credit Union Industry Prompted NCUA to Revise Its Approach
to Examination and Supervision

The credit union industry has undergone a variety of changes that prompted
NCUA to revise its approach to examining and supervising credit unions. As
described earlier, the credit union industry is consolidating, and more
industry assets are concentrated in larger credit unions, those with
assets in excess of $100 million. For example, in December 1992, credit
unions with over $100 million in assets held 52 percent of total industry
assets, but by December 2002, they held 75 percent of total industry
assets. Furthermore, credit unions are providing more complex electronic
services such as Internet account access and on-line loan applications to
meet the demands of their members. Thirty-five percent of the industry
offered financial services through the Internet as of December 2002;
however, the rate increased to over 90 percent for larger credit unions.
In addition, the composition of credit union assets has changed over time,
with credit

61Pub. L. No. 102-242 S: 112, 12 U.S.C. S: 1831m (2000).

unions engaging in more real estate loans (see fig. 11). For example, the
number of first mortgage loans about doubled from 589,000 loans as of
December 1992 to 1.2 million loans as of December 2002. During this same
period, the amount of first mortgage loans more than tripled from $29
billion to $101 billion. From 1992 to 2002, the percentage of real estate
loans to total assets grew from 19 percent to 26 percent, a greater rate
of growth than that of consumer loans over the same time period. The
longer-term real estate loans introduced a greater level of interest rate
risk than that introduced through the shorter-term consumer loans credit
unions traditionally made.62

Figure 11: Credit Union Mortgage Loans Have Grown Significantly Since 1992
Number of loans (in millions) Amount of loans held (dollars in billions)
1.2 100

1.0 80

0.8

60

0.6

40

0.4

20 0.2

0.0 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Number of loans

Total dollar amount of loans Source: Call report data.

Note: Only first mortgage loans represented.

62Interest rate risk is the risk that changes in market rates will have a
negative impact on capital and earnings. In September 2003, NCUA issued
Letter to Credit Unions 03-CU-15, which discusses the interest rate risk
for credit unions with large concentrations of fixed-rate mortgages.

As a result of these changes, NCUA found that its old approach of
reviewing the entire operation of credit unions and conducting extensive
transaction testing no longer sufficed, particularly for larger credit
unions, because of the number of transactions in which they engaged and
the variety of products and services they tended to provide. In contrast,
under the risk-focused approach, NCUA examiners are expected to identify
those activities that pose the highest risk to a credit union and to
concentrate their efforts on those activities. For example, as credit
unions engage in more complex electronic services, examiners are to focus
their efforts on reviewing information systems and technology to ensure
that credit unions have sufficient controls in place to manage operations
risk.63 In addition, as credit unions engage in more real estate lending,
examiners are to focus on ensuring that these credit unions have
sophisticated asset-liability management models in place to properly
manage interest rate risk.64 When transaction testing is used under the
risk-focused approach, it is used to validate the effectiveness of
internal control and other risk-management systems. Further, the
risk-focused approach places more emphasis on preplanning and off-site
monitoring of credit union activities, which helps ensure that once
examiners arrive on site, they already will have identified those areas of
the greatest risk in a credit union and where to focus their resources.

To compliment the risk-focused approach and allow NCUA to better allocate
its resources, the agency adopted a risk-based examination program in July
2001. This program eliminates the requirement to perform annual
examinations on low-risk credit unions, replacing annual exams with two
examinations in a 3-year period.65

63Operations risk is the risk that fraud or operational problems could
result in an inability to deliver products, remain competitive, or manage
information.

64Asset-liability management is the process of evaluating balance sheet
risk (interest rate and liquidity risk) and making prudent decisions,
which enable a credit union to remain financially viable as economic
conditions change.

65These credit unions are defined as those with a CAMEL rating of 1 or 2
for the prior two examinations, and those exhibiting additional
characteristics, such as having been in operation for at least 10 years,
having a positive return on assets, having adequate internal controls, and
having added no recent high-risk programs.

NCUA Took Various Steps to Ensure Successful Implementation of the
Risk-focused Program

NCUA consulted with its Office of Corporate Credit Unions to inquire about
their experiences with their risk-focused program that was implemented in
1998. As a result of this consultation, NCUA incorporated a greater level
of examiner judgment in its risk-focused approach, specifically allowing
examiners to determine the appropriate level of on-site versus off-site
supervision. For example, if an examiner discovered a problem during
off-site monitoring of a credit union, the examiner might adjust the
schedule of the on-site examination to directly address the problem. In
addition, in recognition that examiners would be required to assess the
future risks that credit unions might be undertaking, NCUA, after
consulting with its Office of Corporate Credit Unions, required that
examiners review information beyond the financial statements. For example,
under the risk-focused program, examiners might analyze due diligence
reviews by management for new and existing products and services, internal
controls, and measurements of actual performance against forecasted
results, to determine what future risks a particular credit union might be
undertaking.

NCUA's consultations with FDIC and its review of two FDIC Inspector
General reports prompted NCUA to develop programs to address challenges
that FDIC experienced in implementing its risk-focused program. For
example, according to NCUA, FDIC did not conduct much training for its
examiners prior to implementing its risk-focused program. NCUA, on the
other hand, held training for all examiners, including state examiners,
and once the risk-focused program was implemented, NCUA also provided
additional training to help examiners assess risks more effectively.
NCUA's review of the FDIC Inspector General reports found that some FDIC
examiners resisted the move to the risk-focused program. NCUA's response
was to develop a quality control program to ensure that examiners and
supervisors were adopting the risk-focused approach and that documentation
was completed consistently across NCUA's regions. Under the quality
control program, NCUA officials reviewed a sample of examinations from
each region for scope, conciseness of reports, appropriateness of
completed work papers and application of risk-focused concepts. NCUA's
development of the quality control program was timely and appropriate,
because we found some NCUA examiners and state supervisors were reluctant
to move to the risk-focused program. The examiners and supervisors were
concerned that they would be blamed if a credit union later had a problem
in an area they had not initially identified as high-risk.

NCUA's consultations with the Office of the Comptroller of the Currency
(OCC) enabled NCUA to consider a different approach to improve its
oversight of large credit unions under the risk-focused program. OCC had
implemented a large bank program in recognition of the need for an
alternative approach to oversight of large and sophisticated banks. NCUA
likewise found the need for an alternative approach to oversight of large
credit unions because its examiners traditionally examined a large number
of small credit unions and very few larger ones and, thus, had been unable
to gain sufficient comfort and expertise in examining the larger, more
complex institutions. As a result of consultations with OCC, NCUA
implemented its Large Credit Union Pilot Program in January 2003 to, among
other things, develop a core of examiners with experience overseeing these
larger credit unions. Under this program, NCUA has also experimented with
different examination approaches, including targeted examinations, which
focus on certain aspects of credit union operations such as the loans,
investments, or asset-liability management. NCUA officials told us that
they received some preliminary feedback from credit unions that found the
pilot to be beneficial. However, because the pilot ended recently, NCUA
officials stressed that it was too early to tell how effective this
program will be in helping NCUA improve its examinations of large credit
unions.

In recognition that the risk-focused program was a significant departure
from NCUA's old approach to examination and supervision, NCUA also sought
feedback from the industry on the risk-focused program by developing a
survey for credit unions to complete once they had gone through their
first risk-focused examination. NCUA reported that it had received
preliminary results from the survey that indicated that the risk-focused
program has been well received. Specifically, NCUA received the highest
marks for examiners' courteous and professional conduct, effective overall
examination process, and effective communication with management and
officials throughout the examination. Officials from some of the large
credit unions we interviewed were pleased with the program because they
felt that the examination was focused on the high-risk areas that credit
union officials needed to monitor. Likewise, examiners with whom we spoke
told us that adopting a risk-focused approach had made a bigger difference
in their oversight activities at the larger credit unions because they
could focus their resources on the high-risk areas of these institutions.
In contrast, the examiners relied on the old approach of extensive
transaction testing at the smaller credit unions that lacked sufficient
resources to implement robust internal control structures and

tended to limit their activities to the basic or traditional services
offered by credit unions.

NCUA Has Further Opportunities to Leverage the Experiences of Other
Regulators to Address Existing Challenges

NCUA Faces Challenges in Ensuring That Examiners Are Adequately Trained to
Assess Changing Technology

NCUA faces a number of challenges in implementing its risk-focused
approach that create additional opportunities for it to leverage the
experiences of the other regulators that have been using risk-focused
programs for several years. These challenges include ensuring that
examiners have sufficient training to keep pace with changes in industry
technologies and methods, adequately preparing for monitoring credit
unions as they expand more heavily into nontraditional credit union
activities such as business lending, and overseeing state-chartered
institutions in states that lack sufficient examiner resources and
expertise.

According to NCUA examiners who had recently implemented the risk-focused
program, NCUA faces challenges in training its examiners in specialized
areas such as information systems and technology. Likewise, as we found in
prior reviews, other depository institution regulators also faced these
challenges in implementing risk-focused programs.66 Some NCUA examiners
with whom we spoke indicated that NCUA's formal and on-the-job training of
subject matter examiners, particularly in the areas of information systems
and technology, payment systems, and specialized lending, was insufficient
and did not help them keep pace with the changing technology in the
industry.67 As a result, some examiners were not confident that they could
assess the adequacy of information systems that were vital to the
operations of some credit unions.

NCUA officials sought to address concerns about specialist training by
modifying their training manual to more clearly state what classes were
appropriate for the different specialized areas. Further, as a member of
the Federal Financial Institutions Examination Council (FFIEC), NCUA was
aware of specialized training offered by other depository institution
regulators under the auspices of FFIEC, and encouraged NCUA examiners

66U.S. General Accounting Office, Risk-focused Bank Examinations:
Regulators of Large Banking Organizations Face Challenges, GAO/GGD-00-48
(Washington, D.C.: Jan. 24, 2000).

67NCUA's subject matter examiner program was created in February 2002 to
train experienced and knowledgeable examiners in specialized areas, such
as capital markets and information systems, to help examiners assess risks
more effectively. The program also was designed to augment NCUA's existing
core of specialist examiners.

to take advantage of this training.68 However, NCUA had not specifically
consulted with other depository institution regulators on how these
regulators addressed the challenge of training their specialists as banks
and thrifts had become more complex over time.

NCUA Faces Challenge of NCUA's revised regulation on member business loans
also presents NCUA Ensuring That It Is Adequately with the challenge of
ensuring that it is adequately prepared to monitor this Prepared to
Monitor Credit growing area of lending. A recent NCUA final rule on member
business Unions as They Expand into loans relaxed certain requirements
(allowing well-capitalized, federally Nontraditional Activities insured
credit unions to offer unsecured business loans) and introduced a

new risk area for NCUA to monitor.69 (Appendix VII provides a detailed
description of changes to this and other NCUA rules and regulations since
1992.)

While member business loans are still a relatively small percentage of
credit union loans (2 percent) and there are statutory limits placed on
these loans, NCUA's recently revised rules could result in credit unions
making more of these loans.70 The Department of the Treasury has raised
concerns that allowing credit unions to engage in unsecured member
business loans

68FFIEC is a formal interagency body empowered to prescribe uniform
principles, standards, and report forms for the federal examination of
financial institutions by the Board of Governors of the Federal Reserve
System, the Federal Deposit Insurance Corporation, the National Credit
Union Administration, the Office of the Comptroller of the Currency, and
the Office of Thrift Supervision. FFIEC also serves to make
recommendations to promote uniformity in the supervision of financial
institutions.

69See 68 Fed. Reg. 56537 (Oct. 1, 2003). Under NCUA's prior regulations,
all business loans to members had to be secured by collateral. Under the
revised rule, NCUA now allows well-capitalized credit unions that have
addressed unsecured loans in their member business loan policies to make
unsecured business loans to members. These loans are subject to the limit
that (1) the aggregate unsecured business loans to one borrower not exceed
the lesser of $100,000 or 2.5 percent of a credit union's net worth and
(2) the aggregate of all unsecured business loans not exceed 10 percent of
a credit union's net worth. The revised rule also permits the exclusion of
participation interests-credit union purchases of an interest in a loan
originated by another credit union-in member business loans from the
aggregate business loan limit, provided that the loan was for a nonmember
of the purchasing credit union. However, the total of nonmember and member
business loans may only exceed the aggregate business loan limit if
approved by NCUA regional directors. Finally, the revised rule expands
preapproved CUSO activities to include business loan originations.

70Under CUMAA, credit unions had an aggregate business loan limit of the
lesser of 1.75 times the credit union's net worth or 12.25 percent of the
credit union's total assets.

Variability in State Oversight May Constrain NCUA's Ability to Monitor
Risks to NCUSIF Posed by Federally Insured, State-chartered Credit Unions

would increase risks to safety and soundness.71 Since member business
loans constitute only a small percentage of credit union lending, most
NCUA examiners will not have significant experience looking at this type
of lending activity. In contrast, banks and thrifts offer these loans to a
much greater extent than credit unions and their regulators do have
experience in this area.

Due to variability in levels of state oversight and resources, NCUA may
face challenges in implementing the risk-focused program at the state
level. Lack of examiner resources and expertise in some states, high state
examiner turnover, and weakness of enforcement by some state regulators
may affect oversight of federally insured, state-chartered credit unions,
according to NCUA officials.

While state officials with whom we met had adopted NCUA's risk-focused
program and indicated they were generally pleased with NCUA's support,
some of these officials indicated that they faced challenges related to
oversight of their credit unions. For example, they indicated that budget
problems had made it difficult to hire additional staff. In addition, some
state officials indicated that they could not compete on pay with the
industry, which led to high examiner turnover. A state official from a
large state indicated that the increase in credit unions converting from
federal to state charters had stretched her examiner resources.

The challenges faced by states are of particular concern given that state
supervisors have primary responsibility for examining federally insured,
state-chartered credit unions, which as of December 31, 2002, held 46
percent of industry assets. Inadequate oversight of these state-chartered
institutions could have a negative impact on the financial condition of
NCUSIF. The FDIC and Federal Reserve share oversight responsibility with
state supervisors for state-chartered banks, and these regulators also
face challenges similar to those faced by NCUA with regard to variability
in state oversight.

71Department of the Treasury comment letter concerning NCUA's proposed
rule on member business lending, dated June 2, 2003. Further, Treasury
stated that excluding business participation loans and business loans
originated by CUSOs from member business loan limits would undermine the
intent of congressional limitations on credit union business loans
established in CUMAA.

In commenting on how it addressed some of the issues facing states, NCUA
officials told us that in cases where states lacked examiner resources or
expertise, NCUA provided its own staff to ensure that federally insured,
state-chartered credit unions were adequately examined. In addition, NCUA
conducted joint examinations with state supervisors on selected federally
insured, state-chartered credit unions to assess the risk they posed to
NCUSIF. Some state officials with whom we met raised concerns over joint
examinations, claiming that NCUA examiners tried to impose federal
regulations on these state-chartered credit unions. These state officials
also expressed concern over NCUA's process for developing its overhead
transfer rate, which they claimed was not transparent.72 We discuss the
overhead transfer rate more fully later in this report.

NCUA Lacks Authority to Examine Third-party Vendors

As we reported in July 1999, NCUA does not have the third-party oversight
authority provided to other federal banking regulators, and the lack of
such authority could limit NCUA's effectiveness in ensuring the safety and
soundness of credit unions.73 Credit unions are increasingly relying on
third-party vendors to support technology-related functions such as
Internet banking, transaction processing, and funds transfers. While these
third-party arrangements can help credit unions manage costs, provide
expertise, and improve services to members, they also present risks such
as threats to security of systems, availability and integrity of systems,
and confidentiality of information. With greater reliance on third-party
vendors, credit unions subject themselves to operational and reputation
risks if they do not manage these vendors appropriately. Although NCUA
received authority to examine third-party vendors as part of the year 2000
readiness effort, this authority was temporary and expired on December 31,
2001.

While NCUA has issued guidance regarding due diligence that credit unions
should be applying to third-party vendors, NCUA must ask for permission to
examine third-party vendors. Without vendor examination authority, NCUA
has no enforcement powers to ensure full and accurate disclosure. For
instance, in one case NCUA was denied access by a third-party vendor that
provides record-keeping services for 99 federally insured credit unions

72The overhead transfer rate is the percentage of NCUA's share insurance
fund (NCUSIF) that is transferred to support the agency's expenses
(operating fund).

73U.S. General Accounting Office, Electronic Banking: Enhancing Oversight
of Internet Banking Activities, GAO/GGD-99-91 (Washington, D.C.: July 6,
1999).

with $1.4 billion in assets. NCUA notified the credit unions to heighten
their due diligence to ensure that appropriate controls were in place at
the third-party vendor. In another case, NCUA was given access to a
third-party vendor, but the vendor withheld financial statements from NCUA
examiners. The third-party vendor served 113 credit unions representing
almost $750 million in assets.

Credit Unions Not Subject to Internal Control Reporting Requirements of
FDICIA

Credit unions with assets over $500 million are required to obtain an
annual independent audit of financial statements by an independent
certified public accountant, but unlike banks and thrifts, these credit
unions are not required to report on the effectiveness of their internal
controls for financial reporting. Under FDICIA and its implementing
regulations, banks and thrifts with assets over $500 million are required
to prepare an annual management report that contains

o 	a statement of management's responsibility for preparing the
institution's annual financial statements, for establishing and
maintaining an adequate internal control structure and procedures for
financial reporting, and for complying with designated laws and
regulations relating to safety and soundness; and

o 	management's assessment of the effectiveness of the institution's
internal control structure and procedures for financial reporting as of
the end of the fiscal year and the institution's compliance with the
designated safety and soundness laws and regulations during the fiscal

74

year.

Additionally, the institution's independent accountants are required to
attest to management's assertions concerning the effectiveness of the
institution's internal control structure and procedures for financial
reporting. The institution's management report and the accountant's
attestation report must be filed with the institution's primary federal
regulator and any appropriate state depository institution supervisor and
must be available for public inspection. These reports allow depository
institution regulators to gain increased assurance about the reliability
of financial reporting.

74See 12 U.S.C. S: 1831m; 12 C.F.R. Part 363 (2003).

Banks reporting requirements under FDICIA are similar to the reporting
requirement included in the Sarbanes-Oxley Act of 2002. Under
Sarbanes-Oxley, public companies are required to establish and maintain
adequate internal control structures and procedures for financial
reporting and the company's auditor is required to attest to, and report
on, the assessment made by company management on the effectiveness of
internal controls. As a result of FDICIA and Sarbanes-Oxley, reports on
management's assessment of the effectiveness of internal controls over
financial reporting and the independent auditor's attestation on
management's assessment have become normal business practice for financial
institutions and many companies. Extension of the internal control
reporting requirement to credit unions with assets over $500 million could
provide NCUA with an additional tool to assess the reliability of internal
controls over financial reporting.

NCUA Implemented PCA as Mandated by CUMAA, but Few Actions Taken to Date

In August 2000, NCUA initially implemented PCA for credit unions. CUMAA
mandated that NCUA implement a PCA program in order to minimize losses to
NCUSIF. Under the program, credit unions and NCUA are to take certain
actions based on a credit union's net worth.75 Other depository
institution regulators were required to implement PCA in December 1992.
PCA was intended to be an additional tool in NCUA's arsenal and did not
preclude NCUA from taking administrative actions, such as cease and desist
orders, civil money penalties, conservatorship, or liquidation of credit
unions.

CUMAA requires credit unions to take up to four mandatory supervisory
actions-an earnings transfer, submission of an acceptable net worth
restoration plan, a restriction on asset growth, and a restriction on
member business lending-depending on their net worth ratios.76 Credit
unions that are adequately capitalized (net worth ratio from 6.0 to 6.99
percent) are

75A credit union's net worth represents the sum of the various reserve
accounts-undivided earnings, regular reserves, and any other
appropriations designated by management or regulatory authorities-and
reflect the cumulative net retained earnings of the credit union since its
inception.

76The net worth ratio is defined as net worth divided by total assets.

required to take an earnings transfer.77 Credit unions that are
undercapitalized (net worth ratio from 4.0 to 5.99 percent), significantly
undercapitalized (net worth ratio from 2.0 to 3.99 percent), or critically
undercapitalized (net worth ratio of less than 2 percent) are required to
take all four mandatory supervisory actions.78

CUMAA also required NCUA to develop discretionary supervisory actions,
such as dismissing officers or directors of an undercapitalized credit
union, to complement the prescribed actions under the PCA program. CUMAA
also authorized NCUA to implement an alternative system for new credit
unions in recognition that these credit unions typically start off with
zero net worth and gradually build their net worth through retained
earnings.79 Appendix IX provides more detail on NCUA's implementation of
PCA.

To date, NCUA has taken few actions against credit unions under the PCA
program due to a generally favorable economic climate for credit unions.
As of December 31, 2002, NCUA took mandatory supervisory actions against
2.8 percent (276 of 9,688) of federally insured credit unions. Of these
credit unions, the vast majority-92 percent or 253-had under $50 million
in assets. Further, 41 percent (113 of 276) of these credit unions were
required to develop net worth restoration plans. However, it is too early
to tell how effective these plans will be in improving the condition of
the credit unions or minimizing losses to NCUSIF.

Credit unions were similar to banks and thrifts with respect to PCA
capital categorization with 97.6 percent of credit unions considered
well-capitalized compared to 98.5 percent of banks and thrifts (see table
2). However, a slightly higher percentage of credit unions were
undercapitalized, significantly undercapitalized, and critically
undercapitalized than banks and thrifts.

77Credit unions that are less than well-capitalized-that is, have less
than a 7.0 percent net worth ratio-are required to increase the dollar
amount of their net worth quarterly by transferring at least 0.1 percent
of their total assets to the regular reserve account. These credit unions
must meet applicable risk-based net worth requirements if they are
complex, which under PCA is defined as a credit union having more than $10
million in assets and a risk-based net worth ratio that exceeds 6.0
percent. The ratio is a calculation that assigns risk weightings to
different types of assets and investments.

78The net worth restoration plan is a blueprint for credit union officials
and staff for restoring the credit union's net worth ratio to 6.0 percent
or higher.

79Credit unions are defined as new if they have been in operation for less
than 10 years and have less than $10 million in assets.

Table 2: Federally Insured Credit Unions Were Similar to Banks and Thrifts
with Respect to Capital Categories, as of December 31, 2002

                                            Credit           Banks/   
                                            unions            thrifts 
                      Capital categorya  (number)b Percent  (number)  Percent 
                       Well-capitalized      9,363     97.6     9,210    98.5 
                 Adequately capitalized        153      1.6       134 
                       Undercapitalized         61      0.6         6 
         Significantly undercapitalized         10      0.1         2 
            Critically undercapitalized         10      0.1         2 
                                  Total      9,597  100.0       9,354   100.0 

Sources: NCUA and FDIC.

Note: Does not include new credit unions.

aAlthough the categories triggering PCA actions are the same for both the
bank regulators and NCUA, the capital requirements underlying these
categories are different.

bNumbers reported by NCUA as of May 2003.

Some NCUA, state, and industry officials claimed that PCA was beneficial
because it provided standard criteria for taking supervisory actions and
was a good way to restrain rapid growth of assets relative to capital.
However, many state officials expressed concern over PCA due to the
limited ability of credit unions to increase their net worth quickly,
because they can only do so through retained earnings. They indicated that
if a credit union were subject to PCA, it would be difficult for that
credit union, particularly a smaller one, to increase capital and graduate
out of PCA. In contrast, other financial institutions are able to raise
capital more quickly through the sale of stock.

Some of these state officials raised the issue of whether credit unions
should likewise have a means to raise capital quickly by allowing credit
unions to use secondary capital toward their capital requirement under
PCA.80 Texas allowed its state-chartered credit unions to raise secondary
capital even though the secondary capital could not count towards PCA.81

80Secondary capital can take the form of investments in an institution by
nonmembers. The investments are subordinated to all other credit union
debt. Currently, only credit unions designated as "low-income" by NCUA are
eligible to raise secondary capital.

81This secondary capital must be in accordance with generally accepted
accounting principles.

According to the Texas credit union regulator, no credit unions had taken
advantage of the state's secondary capital provision. Currently there is a
debate in the industry on whether secondary capital is appropriate for
credit unions. While some in the industry favor secondary capital as a way
to help credit union avoid actions under PCA, others have raised the
concern that allowing credit unions to raise secondary capital (for
example, in the form of nonmember deposits) could change the structure and
character of credit unions by changing the mutual ownership. As of
September 2003, NCUA had not taken a position on secondary capital.

Another concern raised by NCUA officials is in regard to the most
appropriate measure of the net worth ratio for PCA purposes. NCUA
officials have suggested using risk-based assets, rather than total
assets, to calculate the net worth ratio of credit unions because they
believe risk-based assets more clearly reflect the risks inherent in
credit unions' portfolios. NCUA officials recognize that, similar to
banks, a minimum net worth ratio based on total assets (tangible equity
for banks and thrifts) would still be needed for those institutions that
are critically undercapitalized. For most credit unions, risk-based assets
are less than total assets; therefore, a given amount of capital would
have a higher net worth ratio if risk-based assets were used. While there
may be some merit in using risk-based assets, credit unions have been
subject to PCA programs for a short time, and the advantages and
disadvantages of the current programs are not yet evident.

Finally, some NCUA officials raised the concern that PCA has led to more
liquidations of problem credit unions. In the past, NCUA sought merger
partners for problem credit unions. However, NCUA officials told us that
it was more difficult to find merger partners because stronger credit
unions were concerned that their net worth ratio would be lowered by
merging with problem credit unions, thereby putting them closer to the 7.0
percent net worth ratio that triggers PCA. As a result, the cost of
mergers has increased under PCA because NCUA would have to provide greater
incentives to a potential partner, and that has forced the agency to
liquidate credit unions to a greater extent than prior to PCA. While the
initial costs of liquidations appear to be high, the purpose of PCA is to
reduce the likelihood of regulatory forbearance and protect the federal
deposit (share)

insurance funds through early resolution of problem institutions; thus, in
the long run, the overall costs to NCUSIF should be less because of PCA.82

NCUSIF's Financial Condition Appears Satisfactory, but Methodologies for
Overhead Transfer Rate, Insurance Pricing, and Estimated Loss Reserve Need
Improvement

NCUSIF appears to be in satisfactory financial condition. For most of the
past 10 years, NCUSIF's financial condition has been stable as indicated
by the fund's equity ratio, earnings, and net income. However, while
remaining positive as of December 31, 2002, NCUSIF's net income declined
in 2001 and 2002. Among the factors contributing to the decline was a drop
in investment revenues, a sharp increase in the overhead transfer rate,
which is the amount paid to NCUA's Operating Fund for administrative
expenses, and an increase in losses to the insurance fund. Moreover,
NCUA's methods for pricing NCUSIF insurance and for estimating losses to
the fund did not consider important factors such as current credit union
risk. NCUA's flat-rate insurance pricing does not allow for the fact that
some credit unions are at greater risk of failure than others, and the
historical analysis NCUA uses for determining estimated losses does not
reflect current economic conditions or consider the loss exposure of
credit unions with varying risk. As a result of the current weaknesses in
the methodologies used by NCUA, information reported on the financial
condition of the fund may not accurately reflect the current risks to the
fund.

NCUSIF Has Met Statutory Fund Equity Ratio Requirements, but Concerns
Exist over Transfers of Expenses to the Fund

Indicators of the financial condition and performance of NCUSIF have
generally been stable over the past decade. NCUSIF's fund equity ratio-a
measure of the fund's equity available to cover losses on insured
deposits- was within statutory requirements at December 31, 2002, as it
has been over the past decade.83

CUMAA defines the "normal operating level" for the fund's equity ratio as
a range from 1.20 percent to 1.50 percent. CUMAA has designated the NCUA
board to evaluate and set the specific operating level for the fund equity
ratio. In setting the level, the board considers current industry and fund
conditions, as well as the future economic outlook. For 2002, NCUA's board

82Regulatory forbearance occurs when regulators delay taking corrective
action, assuming that problems will not occur in the short-term, or that
economic conditions may change in a way favorable to the troubled
institution.

83The ratio is calculated as the fund balance (assets minus liabilities)
of NCUSIF divided by the sum of all credit union members' shares insured
by the fund.

set the specific operating level at 1.30 percent. If the equity ratio
exceeds the board's determined operating level, CUMAA requires NCUA to
distribute to contributing credit unions an amount sufficient to reduce
the equity ratio to the operating level. Also, should the equity ratio
fall below the minimum rate of 1.20 percent, under CUMAA, NCUA's board
must assess a premium until the equity ratio is restored to and can be
maintained at 1.20 percent. (See appendix X for a more detailed discussion
of the funding process and accounting for NCUSIF.)

Between 1991 and 2002, the equity ratio has fluctuated between 1.23
percent and 1.30 percent, a rate that has remained in line with legal
requirements (see fig. 12). As of December 31, 2002, the ratio of fund
equity to insured shares for NCUSIF as reported by NCUA was 1.27 percent.

Figure 12: NCUSIF's Equity Ratio, 1991-2002 Equity ratio 1.30

1.28

1.26

1.24

1.22

1.20

1.18 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Statutory minimum

Source: NCUA.

NCUSIF's ratio can be usefully compared with the only other share or
deposit insurance funds in the United States currently-FDIC's Bank
Insurance Fund (BIF), which insures banks, and its Savings Association
Insurance Fund (SAIF), which insures thrifts; and ASI, which insures
state-chartered credit unions that are not federally insured. The NCUSIF
ratio

was comparable with the other share and deposit insurance funds as of
December 31, 2002 (see fig. 13).

1.5

1.0

0.5

0.0 NCUSIF BIF SAIF ASI

Sources: NCUA, FDIC, and ASI.

NCUSIF's earnings-principally derived from its investment portfolio, which
has increased significantly since 1991-have been sufficient to

o  cover operating expenses and losses from insured credit union failures;

o 	make additions to its equity with the net income that is retained by
the fund;

o  maintain its equity in accordance with legal requirements;

o  maintain its allowance for anticipated losses on insured deposits;

o  avoid assessing premiums, except for 1991 and 1992; and

o  make, in some years, distributions to insured credit unions.

NCUSIF's net income has remained positive through 2002 and had generally
been increasing since 1993, until significant declines occurred in 2001
and 2002 (see fig. 14). The declines were due to a combination of

decreased yields from the investment portfolio, an increase in the
overhead transfer rate, and larger insurance losses on failed credit
unions. The investment portfolio of NCUSIF consists entirely of U.S.
Treasury securities. Yields on these securities have declined-for example,
from 6.07 percent in 2000 to 5.10 percent in 2001 and to 3.18 percent in
2002 on its 1-to 5-year maturities-following similar general declines in
market yields for Treasury securities. Of the $40.2 million net income
decline between 2000 and 2001, $22.2 million of the decline was
attributable to increases in the overhead transfer rate, and $15.3 million
was attributable to declines in investment income. Of the $47.5 million
decline in net income between 2001 and 2002, $39.6 million was
attributable to declines in investment income, while $12.5 million was
attributable to provision for insurance losses. At the same time,
operating expenses decreased by $5.1 million. For 2003, interest rates
have continued to decline, which will likely continue to negatively affect
investment earnings.

Figure 14: Net Income of NCUSIF, 1990-2002 Dollars in thousands

250,000

200,000

150,000

100,000

50,000

0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Source: NCUA.

The sharp increase in the overhead transfer rate and its negative impact
on NCUSIF's net income have raised questions about NCUA's process for
determining the transfer rate. The Federal Credit Union Act of 1934
created the Operating Fund for the purpose of providing administration and
service

to the credit union system-for example, the supervision and regulation of
the federally chartered credit unions.

NCUA's Operating Fund is financed through assessment of annual fees to
federally chartered credit unions as well as the overhead transfer from
NCUSIF (see fig. 15). Federally chartered credit unions are assessed an
annual fee by the Operating Fund based on the credit union's asset size as
of the prior December 31. The fee is designed to cover the costs of
providing administration and service, as well as regulatory examinations
to the Federal Credit Union System. NCUA's board reviews the fee structure
annually. The overhead transfer from NCUSIF for administrative services
provides a substantial portion of funding for the Operating Fund. The
annual rate for the overhead transfer is set by NCUA's board based on
periodic surveys of NCUA staff time spent on insurance-related activities
compared with noninsurance-related, or regulatory, activities. An amount
of overhead or administrative expense is transferred to NCUSIF in
proportion to staff time spent on insurance-related activities. The
overhead transfer is intended to account for NCUA staff being responsible
for both insurance and supervisory-related activities.

Source: NCUA.

Between 1986 and 2000, the transfer rate was 50 percent, which, according
to NCUA management, was based on surveys that indicated staff time was
equally split between insurance and regulatory activities. For example, 50

percent of the Operating Fund's $127.6 million, or $63.8 million, in
expenses for 2000 were allocated to and paid by NCUSIF. For 2001, NCUA's
Board of Directors increased the overhead transfer rate to 67 percent on
the basis that Operating Fund staff had increased their insurance-related
activities. This resulted in a $24.7 million increase (almost 40 percent)
from 2000 in the amount being allocated to NCUSIF. For 2002 and 2003, the
NCUA board lowered the 67-percent overhead transfer rate to 62 percent by
adjusting downward its allocation of what it considered "nonproductive"
time factors such as employee administrative and education time used in
the 2001 survey because it was reflective of regulatory rather than
insurance-related activities.

In September 2001, NCUA management engaged its financial audit firm,
Deloitte & Touche, to review the basis on which the transfer rate was
determined. The auditor's report contained several recommendations that
indicated that NCUA's 2001 survey of staff time spent on insurance-related
functions-the primary basis on which NCUA allocates administrative
expenses-may not have resulted in an accurate allocation. The lack of a
clear separation of the insurance and supervisory functions had also been
the focus of a recommendation in our 1991 report (still unimplemented)
that NCUA should establish separate supervision and insurance offices.84
The 2001 recommendations from NCUA's financial audit firm included
improvements in communication with staff on the survey process and
results, frequency and timing of the survey, methods of survey
distribution, and updated documentation of survey definitions and purpose.
The auditors also noted that individuals were allocating time after the
fact, when recollection may have been faulty, rather than tracking their
time concurrently as would be possible if provided the survey and
guidelines prior to an assignment. Additionally, the auditors reported
that, to provide reliable results, the survey should cover a greater
period of time. The limited period used could significantly skew the
resulting proportion of activities devoted to insurance versus regulatory
activities. The auditor's recommendations indicated that the survey's lack
of consistency and reliability may have resulted in a misallocation of
overhead expenses between the operating and insurance funds. Any
misallocation would affect NCUSIF's financial condition because any
increase in the overhead transfer rate results in a decrease of NCUSIF's
net income. Misallocations also can significantly affect the financial
results of the Operating Fund. In addition to the auditor's findings, some
federally insured, state-chartered credit

84GAO/GGD-91-85, p. 197.

unions and trade groups have expressed concerns about NCUA's calculation
of its overhead transfer rate. Primarily, they say that NCUA has not
clearly defined insurance and regulatory functions, and its methodology
for determining the overhead transfer rate is not transparent or
understandable to participating credit unions.

According to NCUA's management, NCUA has begun implementing Deloitte &
Touche's recommendations. For example, selected field examiners are now
completing surveys in a timely manner for periods covering a full year.
However, headquarters staff are not required to complete the surveys as
management asserts the split of their time mirrors that of field
examiners. In addition, the transfer rate is calculated and approved by
management every few years. However conditions can change that may result
in the transfer rate not representing the current condition. Changing
workloads and conditions can also cause a significant change in future
rates.

Federal Credit Union Insurance Pricing Is Not Based on Risk to Insurer

The Federal Credit Union Act requires all federally insured credit unions
to allocate 1 percent of their insured shares to NCUSIF. This flat rate
does not take into consideration variations in risk posed by individual
credit unions. Although FDIC had implemented a version of risk-based
pricing in 1993, FDIC has continued to study options for improving deposit
insurance funding. FDIC's suggestions for improvement were issued in a
2001 report that noted the cost of insurance, regardless of type
(property, casualty, or life), in the private sector is priced based upon
the risk assumed by the insurer.85 Premiums and loss experience are
generally actuarially determined, such that increased risk equates to
increased cost. Since passage of FDICIA in 1991, deposit insurance for
banks and thrifts are adjusted for some risk, and since December 31, 2000,
private-sector insurance for credit union shares has been adjusted for
risk. (See appendix X for additional information on accounting for
insurance.) While BIF and SAIF are adjusted for some risk, FDIC has made
additional proposals for enhancing the risk-based nature of its insurance
pricing. For instance, the current BIF and SAIF funding does not require a
fast-growing institution to pay premiums if it is well capitalized and
CAMEL-rated 1 or 2. As a result, FDIC has proposed that the pricing
structure for BIF and SAIF

85Federal Deposit Insurance Corporation, Keeping the Promise:
Recommendations for Deposit Insurance Reform (Washington, D.C.: April
2001).

be amended so that fast-growing institutions would be required to pay
premiums.

NCUSIF is the only share or deposit insurer that has not adopted a
risk-based insurance structure. Therefore, some credit unions could be
overpaying while others could be underpaying if their current rates were
compared to their risk profiles-with the cost of insurance not being
equitable based on the level of risk posed to NCUSIF by individual credit
unions. In contrast, FDIC's BIF and SAIF and ASI currently operate on a
risk-based capitalization structure. Depository institutions insured by
BIF and SAIF pay a premium twice a year based upon their capital levels
and supervisory ratings, with institutions with the lowest capital levels
and worst supervisory ratings paying higher premiums. ASI's insurance fund
requires its insured credit unions to maintain deposits between 1.0 and
1.3 percent of their insured shares. The amount for each credit union is
determined based upon its supervisory rating, with lower-rated credit
unions maintaining higher deposits.

The risk-based structure has certain advantages. First, by varying pricing
according to risk, more of the burden is distributed to those members that
put an insurance fund at greater risk of loss. Second, risk-based pricing
provides an incentive for member owners and managers of credit unions to
control their risk. Finally, risk-based pricing helps regulators focus on
higher-risk credit unions by enabling them to allocate their insurance
activities in proportion to the price charged. During our review, members
of NCUA's management told us that they believe that risk-based pricing
would adversely affect small credit unions and suggested that an option
would be to add risk-based pricing only for credit unions over a certain
size. By not having risk-based insurance structure, NCUSIF puts a
disproportionate share of the pricing burden on less-risky credit unions
and does not provide an incentive through pricing for owners and managers
to control their risk.

Management's Estimation of Insured Share Losses Does Not Reflect Specific
Loss Rates

NCUA's process for determining estimated losses from insured credit
unions-the largest potential liability of the fund-does not reflect
current economic conditions and loss exposures of credit unions with
varying risk. The estimated liability balance is established to cover
probable and estimable losses as a result of federally insured credit
union failures. The estimated liability balance is reduced when the
insurance claims are actually paid. NCUSIF's estimated liability for
losses was $48 million at December 31, 2002.

In 2002, NCUA's management analyzed historical loss trends over varying
periods of time in order to assess whether the estimated liability for
losses was adequate. It analyzed historical rates of insurance payouts for
the past 3-year, 5-year, 10-year, and 15-year averages. The 15-year
analysis encompassed an economic period of dramatic losses, which
management contends may be cyclical and indicative of future exposure,
although not necessarily indicative of current economic conditions. As a
result of this analysis, in July 2002, management began building the
estimated losses account balance by $1.5 million a month to $60 million
(from $48 million at December 31, 2002), the amount the analysis
determined would be needed to cover identified and anticipated losses.

NCUA's estimation method does not identify specific historical failure
rates and related loss rates for the group of credit unions that had been
identified as troubled, but instead specifically calculates expected
losses for each problem credit union, if it is determined that a
particular credit union is likely to fail. This methodology essentially
assigns a probability of failure of either zero or 100 percent to each
individual credit union considered to be troubled. By not considering
specific historical failure rates and loss rates in its methodology, NCUA
is using an over-simplified estimation method. As a result, NCUA may not
be achieving the best estimate of probable losses. Therefore, NCUA may be
over or underestimating its probable losses because it does not apply more
targeted and specific loss rates to currently identified problem
institutions, but instead, makes a determination that essentially selects
from two probabilities: zero or 100-percent probability of failure.

From 2000 to 2002, the amount of insured shares in problem credit unions
doubled, going from $1.5 billion insured shares in 2000 to nearly $3
billion insured shares in 2002. The increase in insured shares of problem
credit unions may be an indicator of larger future losses to the fund,
since problem credit unions are more likely to fail. In addition, recent
increases the share payouts show that the insurance fund is suffering from
increasing losses that totaled $40 million in 2002. At the same time, the
estimated loss reserve, which is intended to cover actual losses, has been
declining since 1994. As a result the cushion between payouts for
insurance losses and the reserve balance became increasingly smaller
between 2001 and 2002 (see fig. 16). Given the recent trends, it is
especially important to utilize specific data on failure rates for
troubled institutions.

0.15

0.12

0.09

0.06

0.03

0.00 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Share payouts

Reserves Sources: Call report data.

In contrast to NCUA's method, FDIC's method records estimated bank and
thrift insurance losses based on a detailed analysis of institutions in
five risk-based groups. The first group consists of institutions
classified as having a 100-percent expected failure rate. This
determination is based on the scheduled closing date for the institution,
the classification of the institution as "critically undercapitalized," or
identification of the institution as an imminent failure. The remaining
four risk groups are based on federal and state supervisory ratings and
the institutions' projected capitalization levels. Every quarter, FDIC
meets with representatives from other federal financial regulatory
agencies to discuss these groupings and ensure that each institution is
appropriately grouped based on the most recent supervisory information.
Also on a quarterly basis, FDIC's Financial Risk Committee (FRC), an
interdivisional committee, meets to discuss and determine the appropriate
projected failure rates to be applied to each of

the four remaining risk-based groups.86 The projected failure rate for
each risk-based group is multiplied by the assets of each institution in
that group, which results in expected failed assets. Expected failed
assets are then multiplied by an expected loss experience rate, the
product of which results in the loss estimate for anticipated failures.
The projected failure rates for the remaining four risk-based groups are
based on historical failure rates for those categories. However, FRC has
the responsibility for determining if the historical failure rates for
each group are appropriate given the current and expected condition of the
industry and may adjust failure rates, if necessary. The expected loss
experience rates have been based on asset size and reflect FDIC's
historical loss experience for banks of different sizes. FRC may also use
loss rates based on institution-specific supervisory information rather
than the historical rates. This process, as implemented by FDIC, results
in a more targeted estimation process that specifically captures current
changes in the risk profile of insured institutions.

System Risk That May Be Associated with Private Share Insurance Appears to
Have Decreased, but Some Concerns Remain

The amount of insured shares and the number of privately insured credit
unions and providers of private primary share insurance have declined
significantly since 1990. Specifically, 1,462 credit unions purchased
private share insurance in 1990 compared with 212 credit unions as of
December 2002. During the same period, the total amount of privately
insured shares decreased by 42 percent ($18.6 billion to about $10.8
billion). Although the use of private share insurance has declined, some
circumstances of the remaining private insurer, ASI, raise concerns.
First, ASI's risks are concentrated in a few large credit unions and in
certain states. Second, ASI has a limited ability to absorb catastrophic
losses because it does not have the backing of any governmental entity and
its lines of credit are limited in the aggregate as to the amount and
available collateral. To mitigate its risks, ASI has implemented a number
of risk-management strategies, such as increased monitoring of its largest
credit unions. State oversight mechanisms of the remaining private share
insurer and privately insured credit unions also provide some additional
assurance that ASI and the

86The Financial Risk Committee consists of representatives from four
divisions within FDIC: Insurance and Research, Resolutions and
Receiverships, Supervision and Consumer Protection, and Finance. FDIC
maintains statistics on the percentage of institutions within different
risk categories that fail based on the ratio of failed institutions'
assets to total assets. For purposes of this report the term "failure
rate" is used to describe this statistic. A 100-percent projected failure
rate is always applied to the first risk-based group.

credit unions it insures operate in a safe and sound manner. One
additional concern, as we recently reported, is that many privately
insured credit unions failed to make required disclosures about not being
federally insured and, therefore, the members of these credit unions may
not have been adequately informed that their deposits lacked federal
deposit insurance.

Few Credit Unions Are Privately Insured

Compared with federally insured credit unions, relatively few credit
unions are privately insured. As of December 2002, 212 credit unions-about
2 percent of all credit unions-chose to purchase private primary share
insurance.87 These privately insured credit unions were located in eight
states and had about 1.1 million members with shares totaling about $10.8
billion, as of December 2002-a little over 1 percent of all credit union
members and 2 percent of all credit union shares. In contrast, as of
December 2002, there were 9,688 federally insured credit unions with about
81 million members and shares totaling $483 billion.

Through a survey of 50 state regulators and related follow-on discussions
with the regulators, we identified nine additional states that could
permit credit unions to purchase private share insurance.88 Figure 17
illustrates the states that permit or could permit private share insurance
as of March 2003 and the number of privately insured credit unions as of
December 2002.

87Our review focuses on primary share insurance. Generally, primary share
(or deposit) insurance is mandatory for all depository institutions and
covers members' deposits up to a specified amount. Excess share (deposit)
insurance is optional coverage above the amount provided by primary share
insurance. NCUSIF provides primary share insurance up to $100,000 per
member; while ASI provides primary share insurance up to $250,000 per
account and excess share insurance. ASI is chartered by Ohio statute.
ASI's coverage is subject to a $250,000 statutory cap under Ohio law. Ohio
Rev. Code Ann. S: 1761.09(A), (Anderson, 2003).

88States that "could permit" private share insurance include those with
state laws permitting credit unions to purchase private share insurance,
but that have no credit unions in the state that currently carry private
share insurance.

mandate.89 As a result of the conversions from private to federal share
insurance, most private share insurers have gone out of business due to
the loss of their membership since 1990; only one company, ASI, currently
offers private primary share insurance.90

In states that currently permit private share insurance, a comparable
number of credit unions have converted from federal to private share
insurance and from private to federal share insurance since 1990-31 and
26, respectively. Most of the conversions from federal to private share
insurance (26 of 31) occurred since 1997. According to management at many
privately insured credit unions, they converted to private share insurance
to obtain higher coverage and avoid federal rules and regulation.
Additionally, management at these credit unions noted that they were
satisfied with the service they received from the private share insurer
and all but one planned to remain privately insured. According to NCUA-in
states that currently permit private share insurance-since 1990, 26 credit
unions converted from private to federal share insurance; the majority did
so in the early 1990s, following the RISDIC failure and widespread concern
over the safety and soundness of private share insurance.91 Most of the 26
credit unions planned to continue to purchase federal share insurance

89Several factors precipitated the closure of RISDIC in 1991. For example,
weaknesses existed in the Rhode Island bank regulator's and RISDIC's
oversight of institutions. Furthermore, some of the institutions insured
by RISDIC engaged in high-risk activities. In 1991, RISDIC depleted its
reserves because of the failure of one institution. As a result, runs
occurred at several other institutions insured by RISDIC; it was not able
to meet its insurance obligations and was forced to call in a conservator.
The Governor of Rhode Island closed all institutions insured by RISDIC and
required institutions to purchase federal deposit insurance. According to
NCUA, it did not insure all Rhode Island credit unions following the
Governor's closure of institutions insured by RISDIC.

90As of December 2002, we identified two companies that provided private
primary share insurance in the 50 states and the District of Columbia-ASI
and Credit Union Insurance Corporation (CUIC) in Maryland. However, CUIC
was in the process of dissolution and, therefore, we did not include it in
our analysis. As of August 2003, of the five credit unions that CUIC
insured, four purchased private share insurance from ASI, and one
converted to federal share insurance.

91Generally, credit unions that converted from federal to private share
insurance since 1990 were larger than credit unions that switched from
private to federal share insurance during the same period. Specifically,
as of December 2002, about a third of the credit unions that converted to
private insurance had shares between $100 and $500 million; on the other
hand, the majority of credit unions that converted to federal insurance
had shares totaling up to $50 million. Only two of the 26 conversions
occurred since 1995-one because the private insurer went out of business
and the other because of a merger with a federally insured credit union.

either because they were reasonably satisfied or because they viewed
having their share insurance backed by the federal government as a
benefit.

Risks Exist at Remaining Private Share Insurer, but Certain Factors Help
to Mitigate Concerns

Risks of Remaining Private Insurer Concentrated in a Few Credit Unions and
States

Although the use of private share insurance has declined, we found two
aspects of the remaining private insurer that raise potential safety and
soundness concerns. First, ASI faces a concentration of risk in a few
large credit unions and certain states. Second, ASI has limited borrowing
capacity and could find it difficult to cover catastrophic losses under
extreme economic conditions because it does not have the backing of any
governmental agency, its lines of credit are limited in the aggregate as
to the amount and available collateral, and it has no reinsurance for its
primary share insurance. To help mitigate these risks, ASI has taken steps
to increase its monitoring of its largest credit unions and is using other
strategies to limit its risks. In addition, as a regulated entity, state
regulation of ASI and the credit unions it insures provides some
additional assurance that ASI and the credit unions operate in a safe and
sound manner.

ASI is chartered in Ohio statute as a credit union share guaranty
corporation.92 As specified in Ohio statute, the purpose of such a
corporation includes guaranteeing payment of all or a part of a
participating credit union share account.93 Although ASI is commonly
referred to as a provider of insurance, it is not subject to all of Ohio's
insurance laws.94 For example, ASI is not subject to Ohio's insurance law
that limits the risk exposure of an insurance company. Specifically, while
Ohio insurance companies are subject to a "maximum single risk"
requirement-"no insured institution's coverage should comprise more than
20 percent of the admitted assets, or three times the average risk or 1

92Ohio Rev. Code Ann. Oh. 1761.

93Ohio Rev. Code Ann. S: 1761.03. Under Ohio law, other purposes of a
credit union share guaranty corporation are to (1) aid and assist any
participating credit union that is in liquidation or incurs financial
difficulty in order that the credit union share accounts are protected or
guaranteed against losses, and (2) cooperate with participating credit
unions, the superintendent of credit unions, the appropriate credit union
supervisory authorities, and the NCUA for the purpose of advancing the
general welfare of credit unions in Ohio and in other states where
participating credit unions operate.

94In Ohio, credit union guaranty corporations are subject to many Ohio
insurance laws; however, they apply only to the extent that such laws are
otherwise applicable and are not in conflict with Ohio laws for credit
union guaranty corporations. See Ohio Rev. Code Ann. 1761.04(A).

percent of insured shares, whichever is greater"-Ohio has not imposed this
requirement on ASI.95 Although ASI is not subject to this requirement, we
found that ASI exceeded this concentration limit. For example, one credit
union made up about 25 percent of ASI's total insured shares, as of
December 2002. In contrast, the largest federally insured credit union
accounted for only 3 percent of NCUSIF's total insured shares. Other
concentration risks exist; for example, we found that 45 percent of ASI's
total insured shares were located in one state (California). Further, all
of ASI's insured credit unions were located in only eight states, with
almost half being located in one state (Ohio), which represents 14 percent
of all ASI-insured shares. In contrast, 14.3 percent of federally insured
credit union shares were located in one state (California). The credit
unions that NCUSIF insures are located in 50 states and the District of
Columbia, with the largest percentage (8 percent) of credit unions located
in one state (Pennsylvania), which represents about 4 percent of NCUSIF's
insured shares.

While we remain concerned about ASI's concentration of risks, ASI employs
a number of risk-management strategies-intended to mitigate its risk
exposure in individual institutions-including being selective about which
credit unions it insures, conducting regular on- and off-site monitoring
of all its insured institutions, implementing a partially adjusted,
risk-based insurance pricing policy, and establishing a 30-day termination
policy. More specifically, ASI employs the following risk-management
strategies:

o 	To qualify for primary share insurance with ASI, a credit union must
meet ASI's insurance eligibility criteria, which include an analysis of
the financial performance of the credit union over a 3-year period and an
evaluation of the institution's operating policies. For example, to
qualify for ASI coverage, a credit union's fixed assets must be limited to
5 percent of the institution's total assets or the amount permitted by its
supervisory authority, whichever is greater, and credit unions must
maintain a minimum net capital-to-asset ratio of 4 percent of total
assets.96 In contrast, federal PCA requirements compel federally insured
credit unions to maintain a minimum capital to assets ratio of 7 percent

95Under Ohio law, insurers licensed by the state are subject to a "maximum
single risk" requirement. See Ohio Rev. Code Ann. S: 3941.06(B).

96According to ASI, the average net capital-to-assets ratio of all ASI's
primary insured credit unions was 10.88 percent, as of December 31, 2002.

of total assets.97 The credit union also must submit its investment,
asset-liability management, and loan policies for ASI's review. In
addition, ASI obtains and reviews the most recent reports from the credit
union's regulator and certified public accountant (CPA) or supervisory
committee. Between 1994 and July 2003, ASI denied share insurance coverage
to eight credit unions while approving coverage for 31 credit unions.98

o 	ASI also regularly monitors all credit unions it insures. ASI routinely
conducts off-site monitoring and conducts on-site examinations of
privately insured credit unions at least once every 3 years. It also
reviews state examination reports for the credit unions it insures and
imposes strict audit requirements. For example, ASI requires an annual CPA
audit for credit unions with $20 million or more in assets, while NCUA
only requires the annual CPA audit for credit unions with more than $500
million in assets. Further, after insuring a large credit union, ASI
implemented a special monitoring plan for its largest credit unions in
light of its increased risk exposure. For larger credit unions (those with
more than 10 percent of ASI's total insured shares or the top 5 credit
unions in asset size), ASI increased its monitoring by conducting
semiannual, on-site examinations, as well as monthly and quarterly
off-site monitoring, which included a review of the credit unions' most
recent audits (monthly) and financial information (quarterly). ASI also
annually reviews the audited financial statements of these large credit
unions. In January 2003, five credit unions with about 40 percent of ASI's
total insured assets qualified for this special monitoring.99 ASI also
began a monitoring strategy intended to increase its oversight of smaller
credit unions, due in part to experiencing larger-than-expected losses at

97For example, federal PCA regulations require supervisory action when
federally insured credit unions' capital to assets ratio is less than 6
percent of total assets.

98Twenty-eight of these credit unions converted from federal insurance,
while two were newly chartered credit unions and one was an uninsured
credit union.

99As of June 2003, the total shares of these credit unions ranged from
$297.6 million to $2.5 billion. Though the plan targeted only ASI's five
largest credit unions, ASI may increase the number of monitored credit
unions at any time so that it continually reviews at least 25 percent of
total insured shares.

a small credit union in 2002.100 ASI determined that 98 smaller credit
unions qualified for increased monitoring, with shares from the largest of
these smaller credit unions totaling about $23 million.

o 	ASI also has implemented a partially adjusted, risk-based insurance
pricing policy, which produces an incentive for the institutions insured
by ASI to obtain a better CAMEL rating, which in turn lowers the risk to
ASI's insurance fund. Like NCUSIF, ASI's insurance fund is deposit-based;
that is, ASI requires credit unions it insures to deposit a specified
amount with ASI.101 As of December 2002, these deposits with ASI totaled
$112 million. Unlike NCUSIF, ASI's insurance fund is partially adjusted
for risk, which acts as a positive, risk-management strategy to mitigate
against losses. Specifically, a credit union with a higher, or worse,
CAMEL rating is required to deposit more into ASI's insurance fund.102
Conversely, NCUA requires federally insured credit unions to deposit 1.0
percent of insured shares into NCUSIF regardless of their CAMEL
ratings.103 According to ASI, it also has the contractual ability to
reassess all member credit unions up to 3 percent of their total assets to
raise additional funds to cover catastrophic loss.

o 	ASI's credit union termination policy provides another risk-mitigating
strategy that ASI can use to manage its risk exposure to an individual
credit union. ASI's insurance contract identifies several circumstances
that would enable ASI to terminate insurance coverage. For example, ASI
may terminate a credit union's insurance with 30 days notice to the

100ASI assigned a risk level to the credit unions it insured (low,
moderate, or high) and then used this assessment to determine the extent
of oversight at the credit union, which might include conducting
face-to-face interviews with the chair of the supervisory audit committee,
confirming checks over $1,000 have cleared, or verifying the value of
loans, investments, and share accounts with credit union members in
writing or over the phone.

101ASI deposit-based insurance fund is funded through capital
contributions to ASI from member credit unions. The member credit unions
record this capital contribution as a deposit (asset) on their financial
statements.

102ASI's insurance fund is funded through the credit unions it insures
depositing between 1.0 and 1.3 percent of a credit union's insured shares
with ASI. The credit unions' CAMEL ratings determine the rate at which
credit unions are assessed (the ratings are 1-strong, 2-satisfactory,
3-flawed, 4-poor, and 5-unsatisfactory). For example, credit unions with a
CAMEL score of 1 must deposit 1.0 percent of total insured shares into
ASI's insurance fund; credit unions with a CAMEL score of 4 or 5 must
deposit 1.3 percent of their total insured shares.

10312 U.S.C. S: 1782a(c).

Remaining Private Insurer Has Limited Borrowing Capacity and May Find It
Difficult to Cover Losses from Its Largest Insured Credit Unions under
Extreme Economic Conditions

credit union and its state regulator, if the credit union fails to comply
with ASI requirements to remedy any unsafe or unsound conditions or remedy
an audit qualification in a timely manner. According to ASI management, it
has not terminated a credit union's share insurance, although ASI has used
its termination policy as leverage to force changes at a credit union.104

When its largest insured credit union applied for primary share insurance,
ASI undertook an assessment of its financial and underwriting
considerations for insuring this institution.105 ASI had previously
provided excess share insurance to the credit union and was familiar with
its financial condition. ASI's independent actuaries determined that the
ASI fund could withstand losses sustained during adverse economic
conditions for up to 5 years, with or without insuring this large credit
union. Ultimately, ASI's assessment concluded that the credit union's
financial condition was strong and, although it would increase ASI's
concentration of risks, insuring the credit union would have a favorable
financial impact on ASI. According to regulators from the Ohio Department
of Commerce, Division of Financial Institutions (Ohio Division of
Financial Institutions), they did not take exception to ASI insuring the
large credit union and had reviewed ASI's underwriting assessment and
asked to be updated periodically.

Unlike federal share insurance, which is backed by the full faith and
credit of the United States, ASI's insurance fund is not backed by any
government entity. Therefore, losses on member deposits in excess of
available cash, investments, and other assets of ASI-insured institutions
would only be covered up to ASI's available resources and its secured
lines of credit, which serve as a back-up source of funds. According to
ASI documents, the terms of ASI's secured lines of credit required
collateralization between 80 and 115 percent of current market value of
the U.S. government or agency

104ASI's involuntary termination procedure, unlike NCUA's, does not
require a credit union to notify its members that its share insurance has
been terminated. According to ASI, because states generally prohibit
credit unions from operating without share insurance, the states would
require notification to credit union members of the change in the credit
union's insured status. NCUA's involuntary termination policy, on the
other hand, requires 30 days notice and also requires a credit union to
issue "prompt and reasonable" notice to its members that it will cease to
be insured. 12 U.S.C. S:S:1786(b), (c).

105According to ASI documents, this credit union would have represented 22
percent of ASI's insured shares; at the time of the assessment, ASI's
largest credit union represented only 6 percent of the fund's insured
shares.

securities ASI holds. As a result, ASI's borrowing capacity is essentially
limited to the securities it holds. ASI officials also explained that due
to the high cost of reinsurance, it has not purchased reinsurance on its
primary share insurance, although it has reinsurance for its excess share
insurance.

ASI has not had large losses since 1975. ASI has expended funds for 118
claims and its loss experience-from the credit unions that have made
claims-has averaged 3.95 percent of the total assets of these credit
unions. If ASI's historical loss average of 3.95 percent was tested and
proved true for a failure at the largest credit union ASI insured, as of
December 2002, the loss amount would be about $119 million.106 While this
would be a major loss, ASI would most likely be able to sustain this loss.
ASI's historical loss rate is nearly 60 percent less than the loss rate
experienced by NCUSIF for the same period. However, under more stressful
conditions, ASI could have difficulty fulfilling its obligations. For
example, ASI's five largest credit unions represent nearly 40 percent of
insured shares, for which a collective loss at 3.95 percent of the assets
of these credit unions would exceed ASI's equity by approximately $30
million. According to ASI, it could raise additional funds to cover
catastrophic loss by reassessing all member credit unions up to 3 percent
of their total assets, which excluding the top five credit unions, would
generate approximately $214 million of additional capital, while
maintaining minimum capital levels at 4 percent of total assets. Further,
by Ohio statute, the Superintendent of the Division of Financial
Institutions can order ASI to reassess its insured credit unions up to the
full amount of their capital, which, excluding the top five credit unions,
would generate approximately $794 million of funds for ASI with which to
pay claims. This recapitalization process is generally similar to that
required of NCUSIF before accessing its Treasury line of credit. However,
if ASI reassessed its member credit unions during a catastrophic failure,
it would further negatively affect these credit unions at a time that they
were already facing stressful economic conditions.

State Oversight of ASI and the State regulation of ASI and the privately
insured credit unions it insures Credit Unions It Insures Provides
provides some additional assurance that ASI and privately insured credit
Additional Assurance unions operate in a safe and sound manner. As a share
guaranty

106This estimate is based on using December 2002 financial data on the
largest credit union insured by ASI. According to a capital adequacy
analysis performed for ASI, ASI's independent actuaries determined that
the ASI fund could withstand losses sustained during adverse economic
conditions for up to 5 years, with or without insuring this large credit
union.

corporation, ASI is subject to state oversight and regulation in those
states where ASI insures credit unions. ASI was chartered in Ohio statute,
with the Ohio Division of Financial Institutions and the Ohio Department
of Insurance dually regulating it. ASI is licensed by the Ohio
Superintendent of Insurance and is subject to routine oversight by that
department and Ohio's Superintendent of Credit Unions.107 The Ohio
Division of Financial Institutions conducts annual assessments of ASI,
which evaluate ASI's underwriting and monitoring procedures, financial
soundness, and compliance with Ohio laws. Under Ohio law, its Department
of Insurance also is required to examine ASI at least once every 5 years.
The last Ohio Department of Insurance exam of ASI was completed in March
1999, which covered January 1995 through December 1997. When we met with
Ohio officials in June 2003, they told us that the Ohio Department of
Insurance planned to examine ASI in the third quarter of calendar year
2003. ASI is also required to submit annual audited financial statements,
including management's attestation, and quarterly unaudited financial
statements to Ohio insurance and credit union regulators.108 Ohio law also
requires ASI to provide copies of written communication with regulatory
significance to Ohio regulators, obtain the opinion of an actuary
attesting to the adequacy of loss reserves established, and apply annually
for a license to do business in Ohio. In our discussions with officials
from the Ohio Division of Financial Institutions and the Ohio Department
of Insurance, we found that, to date, ASI has complied with all
requirements and regulations, and no regulators have taken corrective
actions against ASI or limited ASI's ability to do business in Ohio.

Generally, state financial regulators have taken the primary lead for
monitoring ASI's actions, while state insurance regulators were not as
involved in overseeing ASI's operations. All states where ASI insures
credit unions have, at some point, formally certified ASI to conduct
business in that state.109 Ohio and Maryland have certified ASI in the
past year-as required by governing statutes in those states. Regarding the
other states in which ASI operates, while they have not formally
recertified ASI, Ohio's

107See Ohio Rev. Code Ann. Ch. 1761.

108While Ohio law requires ASI to submit annual audited financial
statements, Ohio law permits the superintendent of insurance to require
the submission of quarterly reports. The superintendent of insurance
imposes this requirement on ASI. See Ohio Rev. Code Ann. S:S: 1761.16 and
3901.42.

109The states are Alabama, California, Idaho, Illinois, Indiana, Maryland,
Nevada, and Ohio.

annual examination process of ASI involves regulators from most states.
State credit union regulators from Idaho, Illinois, Indiana, and Nevada
commonly participate in this assessment; according to ASI officials, their
acceptance of the final examination report infers that they approve of
ASI's continuing operation in their respective states. State credit union
regulators from California and Alabama, however, have not participated in
the annual on-site assessment of ASI. Regarding monitoring efforts by
state insurance regulators, according to ASI, the Ohio Department of
Insurance is the only state insurance department that imposes requirements
and insurance regulators from Idaho, Illinois, and Nevada only request
information.

Most state credit union regulators with whom we met told us they had
regular communication with ASI about the credit unions ASI insured. ASI
officials reported that they commonly conducted joint, on-site exams of
credit unions with state regulators. State credit union regulators imposed
safety and soundness standards and carried out examinations of
state-chartered credit unions in a way similar to how the federal
government oversees federally insured credit unions. According to state
regulators, state regulations, standards, and examinations apply to all
state-chartered credit unions, regardless of their insurance status
(whether federal, private, or noninsured). State credit union regulators
reported that they had adopted NCUA's examination program, and their
examiners had received training from NCUA. However, as previously
discussed, some state officials with whom we met indicated that they faced
challenges related to oversight of their credit unions; for example, some
states lacked examiner resources and had high examiner turnover.

Additionally, privately insured credit unions-as compared with federally
insured credit unions-are not subject to identical requirements and
regulations. For example, while federally insured, state-chartered credit
unions are subject to PCA-as discussed earlier, privately insured,
state-chartered credit unions are not subject to these federally mandated
supervisory actions. Although, as a matter of practice, many state
regulators reported that they have the authority to impose capital
requirements on privately insured credit unions and could take action when
a credit union's capital levels are not safe and sound. However, state
officials in California, Idaho, Illinois, Indiana, Ohio, and Nevada said
that their states required privately insured credit unions to maintain
specified reserve levels, which were codified in statute or regulation.
Additionally, Alabama requires credit unions seeking private insurance to
meet certain capital levels.

While some states had specific requirements for credit unions seeking to
purchase private share insurance, many states regulators reported that
they have the authority to "not approve" the conversion of credit unions
to private share insurance. Alabama, Illinois, and Ohio have written
guidelines for credit unions seeking to purchase private share insurance
and regulators reported that they have the authority to "not approve" a
credit union's purchase of private insurance. The other five states that
permitted private share insurance do not have written guidelines for
credit unions seeking to purchase private share insurance, but Idaho,
Indiana, and Nevada state regulators also noted that they have the
authority to "not approve" a credit union's purchase of private share
insurance.

Moreover, NCUA supervised the conversions of federally insured credit
unions to private share insurance. Specifically, NCUA has imposed
notification requirements on federally insured credit unions seeking to
convert to private share insurance and requires an affirmative vote of a
majority of the credit union members on the conversion from federal to
private share insurance. NCUA has required these credit unions to notify
their members, in a disclosure, that if the conversion were approved, the
federal government would not insure shares.110 We reviewed six recent
conversions to private share insurance, and found that, prior to NCUA's
termination of the credit union's federal share insurance, these credit
unions, including the large credit union that recently converted to ASI,
had generally complied with NCUA's notification requirements for
conversion.

Members of Many Privately Although actions taken by ASI and some state
regulators provide some Insured Credit Unions Are assurances that ASI is
operating in a safe and sound manner, ASI's Not Receiving Required
concentration risks and limited borrowing capacity raise concerns that

under stressful economic conditions it may not be able to fulfill
itsDisclosures about the Lack responsibilities to its membership. Congress
determined that it wasof Federal Share Insurance important for members of
privately insured credit unions to be informed

110Specifically, under the Federal Credit Union Act, if a federally
insured credit union terminates federal share insurance or converts to
nonfederal (private) insurance, the institution must give its members
"prompt and reasonable notice" that the institution has ceased to be
federally insured. 12 U.S.C. S: 1786(c). NCUA rules implement these
provisions by prescribing language to be used in (1) the notices of the
credit union's proposal to terminate federal share insurance or convert to
nonfederal (private) insurance, (2) an acknowledgement on the voting
ballot of the member's understanding that federal share insurance will
terminate, and (3) the notice of the termination or conversion. See 12
C.F.R. Part 708b (2003).

that their deposits in such institutions were not federally insured.
Specifically, among other things, section 43 of the Federal Deposit
Insurance Act requires depository institutions lacking federal deposit
insurance, which includes privately insured credit unions, to
conspicuously disclose to their membership that deposits at these
institutions are (1) not federally insured and (2) if the institution
fails, the federal government does not guarantee that depositors will get
back their money.111 These institutions are required to conspicuously
disclose this information on periodic statements of account, signature
cards, and passbooks, and on certificates of deposit, or instruments
evidencing a deposit (deposit slips). These institutions are also required
to conspicuously disclose that the institution is not federally insured at
places where deposits are normally received (lobbies) and in advertising
(brochures and newsletters).

The Federal Trade Commission (FTC) is responsible for enforcing compliance
with section 43.112 However, FTC has never taken action to enforce these
requirements, and has sought and obtained in its appropriations authority
a prohibition against spending appropriated funds to carry out these
provisions. We recently reported that because of a lack of federal
enforcement of this section, many privately insured credit unions did not
always make required disclosures.113 We conducted unannounced site visits
to 57 locations of privately insured credit unions (49 main and 8 branch
locations) in five states-Alabama, California, Illinois, Indiana, and Ohio
and found that 37 percent of the locations we visited did not
conspicuously post signage in the lobby of the credit union. During these
site visits, we also obtained other available credit union materials
(brochures, membership agreements, signature cards, deposit slips, and
newsletters) that did not include language to notify consumers that the
credit union was not federally insured-as required by section 43. Overall,
134 of the 227 pieces of materials we obtained from 57 credit union
locations-or 59 percent-did not include specified language. As part of our
review, we also reviewed 78 Web sites of privately insured credit unions
and found that many Web sites were not fully compliant with section 43
disclosure requirements. For example, 39 of the 78 sites

11112 U.S.C. S: 1831t (b).

11212 U.S.C. S: 1831t (g).

113U.S. General Accounting Office, Federal Deposit Insurance Act: FTC Best
Among Candidates to Enforce Consumer Protection Provisions, GAO-03-971
(Washington, D.C.: Aug. 20, 2003).

reviewed had not included language to notify consumers that the credit
union was not federally insured.

Our primary concern, resulting from the lack of enforcement of section 43
provisions, was the possibility that members of privately insured,
state-chartered credit unions might not be adequately informed that their
deposits are not federally insured and should their institution fail, the
federal government does not guarantee that they will get their money back.
The fact that many privately insured credit unions we visited did not
conspicuously disclose this information raised concerns that the
congressional interest in this regard was not being fully satisfied. In
our August 2003 report, we concluded that FTC was the best among
candidates to enforce and implement section 43 and provided suggestions on
how to provide additional flexibility to FTC to enforce section 43
disclosure requirements. The House Committee on Appropriations,
Subcommittee on Commerce, Justice, State, the Judiciary, and Related
Agencies, is currently considering adding language in FTC's 2004
appropriations bill that would require FTC to enforce and implement
section 43 disclosure provisions.

Conclusions	The financial condition of the credit union industry has
improved since 1991. Between 1992 and 2002, changes in the industry have
resulted in two distinct groups of credit unions-smaller credit unions
providing their members with basic banking services and larger credit
unions that seek to provide their members with a full range of financial
services similar to other depository institutions. These larger credit
unions control a larger percent of industry assets than they did in 1991.
This concentration of industry assets creates the need for greater risk
management on the part of credit union management and NCUA with respect to
monitoring and controlling risks to the federal share insurance fund.

Among the more significant changes that have occurred in the credit union
industry over the past two decades have been the weakening or blurring of
the common bond that traditionally existed between credit union members.
The movement toward geographic-based fields of membership, and other
expansions of the common-bond restrictions in conjunction with expanded
lines of financial services, have made credit unions more competitive with
banks. These changes have raised questions about the extent to which
credit unions are fulfilling their perceived historic mission of serving
individuals of modest means. However, no comprehensive data are available
to determine the income characteristics of those who receive credit unions
services, especially with respect to consumer loans and other

financial services. Available data, such as that provided by the SCF and
HMDA, provide some indication that credit unions serve low- and
moderate-income households but not to the same extent as banks. If credit
unions, as indicated by NCUA and the credit union industry, place a
special emphasis on serving low- and moderate-income households, more
extensive data would be needed to support this conclusion. These data
would need to include information on the distribution of consumer loans
because smaller credit unions are more likely to make consumer than
mortgage loans. Lack of data especially impairs NCUA's ability to
determine if credit unions that have adopted underserved areas are
reaching the households in the communities most in need of financial
services.

As the industry has changed and larger credit unions have become more like
banks in the services they have provided, NCUA has adopted a supervisory
and examination approach that more closely parallels that of the other
depository institution regulators. While it is too soon to determine
whether the risk-focused approach being implemented by NCUA will allow it
to more effectively monitor and control the risks being assumed by credit
unions, our work suggests that further opportunities exist for NCUA to
further leverage off the approaches and experiences of the other federal
depository institution regulators. For example, as NCUA is addressing
challenges in implementation of its risk-focused program, it has the
opportunity to use forums such as the FFIEC to learn how other depository
institution regulators dealt with similar challenges in implementing their
risk-focused programs. Also, NCUA might gain an evaluation of an
institution's internal controls, comparable to other depository
institution regulators, if credit unions were required, like banks and
thrifts, to provide management evaluations of internal controls and their
auditor's assessments of such evaluations. Finally, NCUA could gain better
oversight of third-party vendors if it had the same ability to examine the
activities of third-party vendors as do other depository institution
regulators.

As of December 2002, NCUSIF's financial condition appeared satisfactory
based on its fund-equity ratio and positive net income. However, it is not
clear whether or to what extent NCUSIF's recent decline in net income will
continue. Improvements in NCUA's processes for determining the overhead
transfer rate, pricing, and estimated losses could help to promote future
financial stability by providing more accurate information for financial
management. As currently determined by NCUA, the overhead transfer rate
may not have accurately reflected the actual time spent by NCUA staff on

insurance-related activities. Recent fluctuations are the result of
adjustments being made because of surveys that had not been conducted
regularly or over sufficient periods of time. In addition, NCUSIF's
pricing for federal share insurance coverage does not reflect the risk
that an individual credit union poses to the fund. Moreover, the process
used by NCUA to estimated losses to the insurance fund-the fund's most
significant liability and management estimate-has been based on overly
broad historical analysis. The risk-based pricing structure that is the
norm across the insurance industry and, for loss estimates, the more
detailed, risk-based historical analysis used by FDIC in insuring banks
and thrifts may provide useful lessons for NCUA in improving its
management of insurance for credit unions.

While systemic risks that might be created by private share insurance
appear to have decreased since 1990, the recent conversion of a large
credit union from federal to private share insurance has introduced new
concerns. Because the remaining private insurer's (ASI) insured shares are
overly concentrated in one large credit union and in certain states, and
because it does not have the backing of any governmental entity and it has
limited borrowing capacity, ASI may have a limited ability to absorb
catastrophic losses. This raises questions about the ability of ASI, under
severe economic conditions, to fulfill its obligations if its largest
credit unions were to fail. Given this risk, we believe it is important
that the members of privately insured credit unions are made aware that
their shares are not federally insured. As we previously reported, since
no federal entity currently enforces compliance with federal disclosure
requirements for privately insured credit unions, and with the high level
of noncompliance that we found in on-site visits to privately insured
credit unions, we believe that members of privately insured credit unions
might not be adequately informed that their shares are not federally
insured. As a result, we have previously recommended that Congress
consider providing additional flexibility to FTC to ensure compliance with
the federal disclosure requirements.114

Recommendations for 	To promote NCUA's ability to meet its goal of
assisting credit unions in safely providing financial services to all
segments of society, to enable

Executive Action more consistent federal oversight of financial
institutions, and to enhance

114 GAO-03-971.

share insurance management (for example, improving allocation costs,
providing insurance according to risk, and improving the loss estimation
process), we recommend that the Chairman of the National Credit Union
Administration

o 	use tangible indicators, other than "potential membership," to
determine whether credit unions have provided greater access to credit
union services in underserved areas;

o 	consult with other regulators through FFIEC more consistently about
risk-focused programs to learn how these regulators have dealt with past
challenges (for example, training of information technology specialists);

o 	continuously improve the process for and documentation of the overhead
transfer rate by consistently calculating and applying those rates,
updating the rates annually, and completing the survey with full
representation;

o 	evaluate options for implementing risk-based insurance pricing. In its
evaluation, the NCUA Chairman should consider the potential impact of
risk-based insurance pricing to the ability of credit unions to provide
services to various constituencies; and

o 	evaluate options for stratifying the industry by risk profile and
applying probable failure rates and loss rates, based in part on
historical data, for each risk profile category when estimating future
losses from institutions.

Matters for Congressional Consideration

Should Congress be concerned that federally insured credit unions,
especially those serving geographical areas, are not adequately serving
low-and moderate-income households, Congress may wish to consider
requiring NCUA to obtain data on the proportion of mortgage and consumer
loans provided to low-and moderate-income households within each federally
insured credit union's field of membership and obtain descriptions of
services specifically targeted to low- and moderate-income households.

To ensure the safety and soundness of the credit union industry, Congress
may wish to consider making credit unions with assets of $500 million or
more subject to the FDICIA requirement that management and external

auditors report on the internal control structure and procedures for
financial reporting, as well as compliance with designated safety and
soundness laws.

To improve oversight of third-party vendors, Congress may wish to consider
granting NCUA legislative authority to examine third-party vendors that
provide services to credit unions and are not examined through FFIEC.

Agency Comments and Our Evaluation

We requested comments on a draft of this report from the Chairman of the
National Credit Union Administration and the President and Chief Executive
Officer of American Share Insurance. We received written comments from
NCUA and ASI that are summarized below and reprinted in appendixes XI and
XII respectively. In addition, we received technical comments from NCUA
and ASI that we incorporated into the report as appropriate.

NCUA concurred with most of the report's assessment regarding the
challenges facing NCUA and credit unions since 1991. For example, NCUA
concurred with the report's assessment that overall the financial health
and stability of the credit union industry has improved since 1991. NCUA
also agreed with our recommendation to consult with other regulators
through FFIEC more consistently to leverage the knowledge and experience
the other regulators have gained in administering risk-focused programs.
NCUA stated that it plans to continue its coordination with its FFIEC
counterparts as it makes ongoing improvements to its approach to
supervising federally insured credit unions.

NCUA also concurred with our matter for congressional consideration that
credit unions with assets of $500 million or more should provide annual
management reports assessing the effectiveness of their internal controls
over financial reporting and their external auditor's attestation to
management's assertions. NCUA stated that it is providing guidance for
credit unions on the principles of the Sarbanes-Oxley Act that will, among
other things, strongly encourage large credit unions to voluntarily
provide this reporting on internal controls. However, NCUA believed that
legislation was not necessary because NCUA has the authority to implement
regulations requiring credit unions to provide these reports should it
become necessary. While we acknowledge NCUA's authority to issue
regulations on this issue, we note that regulations can be changed
unilaterally by the agency, whereas legislation is binding unless changed
by

Congress. Our intent in developing this matter for congressional
consideration was to ensure parity between credit unions, banks, and
thrifts with regard to internal control reporting requirements; therefore,
we have left this as a matter for congressional consideration in our
report.

NCUA also indicated that it did not oppose our recommendation that it be
given statutory authority to examine third-party vendors that provide
services to credit unions and are not examined through FFIEC, provided
that appropriate discretion was extended to the agency in the allocation
of agency resources and evaluation of risk parameters in using this
authority. NCUA stated that given that many of these third-party vendors
service numerous credit unions, a failure of a vendor poses systemic risk
issues. However, NCUA suggested that it be changed to a matter for
congressional consideration because it was a statutory issue rather than
one involving the use of existing NCUA regulatory authority. We agreed
with NCUA's assessment and have modified the report accordingly.

NCUA concurred with the report's recommendation to make improvements to
the process for determining the overhead transfer rate and indicated that
management is in the process of improving the methodology for calculating
this rate. NCUA also concurred in part with our report's conclusion that
the NCUSIF loss reserve methodology warrants study, in order to further
refine NCUSIF's estimates. Regarding our recommendation that NCUA study
options for improving its estimates of future insurance losses, NCUA
stated that it is awaiting the receipt of recommendations that FDIC
received on revising its insurance process, and NCUA will review the
details of the revised FDIC process and how to integrate those practices
within NCUA's system.

In its response, NCUA proposed an alternative to risk-based insurance
pricing by using the adoption of a PCA approach where required net worth
levels would be tied to an institution's risk profile. While NCUA's
proposal may be one option to consider, we continue to recommend that NCUA
evaluate and study various options for achieving a risk-based pricing of
insurance to fairly distribute risk, provide incentives for member credit
unions to control their risk, and focus regulators on higher-risk credit
unions. While it is possible that the option suggested by NCUA would
achieve the objectives, we believe that NCUA should study the costs,
benefits, and risks associated with various options in order to determine
the most effective and cost-beneficial means of achieving a risk-based
system of insurance.

NCUA disagreed with our recommendation that it should use indicators,
other than "potential membership," to determine whether credit unions have
provided greater access to credit union services in underserved areas.
NCUA officials stated that they believe that their data indicated that
credit unions have reached out to underserved communities; implementation
of this recommendation could result in significant and unnecessary data
collection; and Congress has not imposed CRA-like requirements on credit
unions in the past. We agree that federally chartered credit unions have
added underserved areas in record numbers, increasing the numbers of
potential members in these areas, and that membership growth in credit
unions with underserved areas has been greater than for credit unions
overall. However, this information does not indicate whether underserved
individuals or households have received greater access to services (for
example, by using check-cashing services, opening no-fee checking
accounts, or receiving loans) as a result of these field of membership
expansions. Further, while we agree that documenting service to the
underserved would result in additional administrative requirements, the
magnitude and scale of this effort does not necessarily require imposition
of CRA as implemented for banks and thrifts, and could result in
information benefitting future credit union expansion efforts. At a
minimum, it would be useful to know whether membership growth in credit
unions that have added underserved areas has come from the underserved
areas themselves and the extent to which those census tracts within these
areas have been identified as low- or moderate-income. This type of
information, collected uniformly by a federal agency like NCUA, could
serve as first step towards documenting the extent to which credit unions
have reached for members outside of their traditional membership base.
Finally, without this information, it will be difficult for NCUA or others
that are interested to determine whether credit unions have extended
services of any kind to underserved individuals as authorized in CUMAA.

Finally, NCUA also concurred with the report's identification of possible
systemic risk that could be associated with private share insurance that
lacks the full faith and credit backing of a state or the federal
government. NCUA believed that the asset concentration, limited borrowing
capacity, and the lack of any reinsurance of the private insurer present
unique challenges for the eight state supervisory authorities where
private insurance exists today.

In commenting on the private share insurance section of a draft of this
report, ASI stated that we failed to adequately assess the private share

insurance industry. In summary, as discussed below, ASI raised objections
to the report statements that ASI's risks are concentrated in a few large
credit unions and a few states; ASI has limited ability to absorb large
losses because it does not have the backing of any governmental agency;
and ASI's lines of credit are limited in the aggregate as to amount and
available collateral. In response, we considered ASI's positions and
materials provided, including ASI's actuarial assumptions and ASI's past
performance, and believe our report addresses these issues correctly as
originally presented.

First, in regard to ASI's concentration risks, ASI stated that the
inclusion of a single large, high-quality credit union provided financial
resources that improved, not diminished, the financial integrity of ASI.
Our report acknowledges this fact. However, our report also notes that
this credit union made up about 25 percent of ASI's total insured shares,
and that ASI's five largest credit unions represent nearly 40 percent of
ASI's insured shares, as of December 2002. While not disputing that the
large credit union would improve ASI's current financial position, we
continue to believe that this level of concentration in a few credit
unions, under adverse economic conditions, could expose ASI to a
potentially high level of losses. ASI also stated that ASI's coverage and
the geographic distribution of ASI's insured credit unions is a matter of
state law. The report points out this fact, and we acknowledge that it
limits ASI's ability to diversify its risks. However, the fact remains
that ASI's risks are currently concentrated in eight states.

Second, in response to our report's assessment of ASI's limited ability to
absorb catastrophic losses, ASI noted "its sound private deposit insurance
program builds on a solid foundation of careful underwriting, continuous
risk management and the financial backing of its mutual member credit
unions, capable of absorbing large (catastrophic) losses." In addition,
ASI noted that over its 29-year history, it has paid over 110 claims on
failed credit unions, and that no member of an ASI-insured credit union
has ever lost money. ASI also noted that it could assess its member credit
unions up to 3 percent of their total assets in order to obtain more
capital. We acknowledge these facts in this report; however, our point
remains that ASI has limited borrowing capacity and, under stressful
economic conditions, may have difficulty securing funds from others to
meet its obligations. ASI also objected to the report's comparison of
private share insurance to the federal insurance program. As the last
remaining private share insurer, ASI has no peer on which to base a
comparison and the only alternative to private share insurance for credit
unions is NCUSIF.

Third, ASI commented that the draft report incorrectly views the company's
lines of credit as a source of capital. ASI noted that their lines of
credit are solely in place to provide emergency liquidity. We do not
disagree with ASI's statement. When incorporating ASI's previously
received technical comments, we clarified in the report that losses on
member deposits, in excess of available cash, investments, and other
assets of ASI-insured institutions, would only be covered up to ASI's
available resources and its secured lines of credit, which serve as a
back-up source of funds. Further, the report notes that ASI's lines of
credit required collaterization between 80 and 115 percent of current
market value of the U.S. government or agency securities ASI holds. As a
result, ASI's borrowing capacity is essentially limited to the securities
it holds and therefore, in a time of stressful economic conditions, ASI
may have difficulty maintaining its own liquidity if its insured credit
unions were failing and unable to meet the withdrawal requests of
depositors.

Lastly, ASI supported our previous conclusion that FTC is the appropriate
agency for monitoring and defining private share insurance consumer
disclosure requirements and believed that privately insured credit unions
would benefit from FTC's enforcement of such provisions. In our concluding
discussions with ASI officials, they emphasized that they were undertaking
efforts to educate their member credit unions on the required consumer
disclosures and taking steps, in conjunction with state credit union
leagues, to ensure compliance.

As agreed with your offices, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days from
the report date. At that time, we will send copies of this report to the
Chairman of the Senate Committee on Banking, Housing, and Urban Affairs,
the Chairman and Ranking Minority Member of the House Committee on
Financial Services, and other congressional committees. We also will send
copies to the National Credit Union Administration and American Share
Insurance and make copies available to others upon request. In addition,
the report will be available at no charge on the GAO Web site at
http://www.gao.gov.

This report was prepared under the direction of Debra R. Johnson and Harry
Medina, Assistant Directors. If you or your staff have any questions
regarding this report, please contact the Assistant Directors or me at
(202) 512-8678. Key contributors are acknowledged in appendix XIII.

Sincerely yours,

Richard J. Hillman Director, Financial Markets and Community Investment

Appendix I

                       Objectives, Scope, and Methodology

Our report objectives were evaluate (1) the financial condition of the
credit union industry; (2) the extent to which credit unions "make more
available to people of small means credit for provident purposes;"1 (3)
the impact, if any, of the Credit Union Membership Access Act of 1998
(CUMAA) on the credit union industry with respect to membership
provisions; (4) how the National Credit Union Administration's (NCUA)
examination and supervision processes have changed in response to changes
in the industry; (5) the financial condition of the National Credit Union
Share Insurance Fund (NCUSIF); and (6) issues concerning the use of
private share (deposit) insurance.

Financial Condition of Industry

To assess the financial condition of the credit union industry, we
obtained and analyzed annual call report financial data (Form 5300) and
regulatory ratings (CAMEL scores) for all federally insured credit unions
from 1992 to 2002.2 NCUA requires federally insured credit unions to
submit a quarterly call report, which contains information on the
financial condition and operations of the institution. Using the call
reports, we calculated descriptive statistics and key financial ratios and
determined trends in financial performance. NCUA provided us with a copy
of the electronic Form 5300 database for our analysis. The database
contained year-end information for December 1992-December 2002. We
reviewed NCUA established procedures for verifying the accuracy of the
Form 5300 database and found that the data that forms this database are
verified on an annual basis, either during each credit union's
examination, or through off-site supervision. We determined that the data
were sufficiently reliable for the purposes of this report. In addition we
received a database of regulatory ratings (CAMEL) from NCUA for 1992-2002,
on which we (1) reviewed the data by performing electronic testing of
required data elements, (2) reviewed existing information about the data
and the system that produced them, and (3) interviewed agency officials
knowledgeable

1While credit union legislation (see the Federal Credit Union Act at 12
U.S.C. S: 1751) uses "small means" and the credit union industry has not
defined the term, in this report, we used "low- and moderate-income," as
defined by banking regulators, to describe the type of people who credit
unions might serve.

2As do banking regulators, NCUA and state regulators use the "CAMEL"
rating system, a composite score, to help evaluate the safety and
soundness of institutions. CAMEL scores rate capital adequacy (C), asset
quality (A), management (M), earnings (E), liquidity (L), and overall
condition.

Appendix I
Objectives, Scope, and Methodology

about the data. We determined that the data were sufficiently reliable for
the purposes of this report.

In addition to using call report data for credit unions, we also used data
collected by the Federal Financial Institutions Examination Council
(FFIEC) and Office of Thrift Supervision (OTS) to compare the financial
condition of and services offered by credit unions with those of other
depository institutions insured by the Federal Deposit Insurance
Corporation (FDIC).3 We used call report (reporting forms FFIEC 031 and
FFIEC 041 for banks and OTS Form 1313 for thrifts) data obtained from
FDIC's Statistics on Depository Institutions Web site, which contains
consolidated bank and thrift data stored on FDIC's Research Information
System database.4 To assess the reliability of these data, we randomly
cross-checked selected data obtained from this Web site with selected
individual call reports and compared our calculations with aggregate
figures provided by FDIC. Given the context of the analyses, we determined
that these data were sufficiently reliable for the purposes of our report.
For broad, industrywide comparisons with banks involving industry
concentration and capital ratios, we used total assets and equity capital
data for all FDIC-insured institutions, excluding insured branches of
foreign-chartered banks. In order to determine bank and thrift
institutions for our more detailed review, we constructed five peer groups
in terms of institution size as measured by total assets, reported as of
December 31, 2002. See table 3 for the definitions we used to create peer
groups.

3FDIC is responsible for overseeing insured financial institution
adherence to FFIEC's reporting requirements, including the observance of
all bank regulatory agency rules and regulations, accounting principles,
and pronouncements adopted by the Financial Accounting Standards Board and
all other matters relating to call report submission. Call reports are
required by statute and collected by FDIC under the provision of Section
1817(a)(1) of the Federal Deposit Insurance Act. FDIC collects, corrects,
updates, and stores call report data submitted to it by all insured
national and state nonmember commercial banks and state-chartered savings
banks on a quarterly basis. Throughout the report, we use the terms,
"banks," "banks and thrifts," and "FDIC-insured institutions"
interchangeably.

4As of August 31, 2003, the address for this Web site was
www3.fdic.gov/sdi/.

                                   Appendix I
                       Objectives, Scope, and Methodology

Table 3: Peer Group Definitions

Group Asset size of institution

                     I Total assets of $100 million or less

II	Total assets greater than $100 million, but less than or equal to $250
million

III	Total assets greater than $250 million, but less than or equal to $500
million

IV Total assets greater than $500 million, but less than or equal to $1 billion

V	Total assets greater than $1 billion, but less than or equal to the
asset size, rounded up to the nearest billion dollars, of the largest
credit union (for example, $16 billion for 2001 and $18 billion for 2002)

Source: GAO.

We specified the maximum total assets of $18 billion by rounding up the
total assets of the largest credit union in our database as of December
31, 2002, to the nearest billion dollars. We also classified bank and
thrift institutions as to whether they emphasized credit card or mortgage
loans; this was done by determining if a given bank had (1) a total loans
to total assets ratio of at least 0.5 and (2) either a credit card loans
to total loans ratio of at least 0.5 or a mortgage loans to total loans
ratio of at least 0.5. The call report data that we used for our financial
condition and services analyses consisted of information on total assets
and total loans, as well as more specific loan holdings data (for example,
consumer loans and real estate loans). We also obtained additional data to
calculate bank capital ratios and return on average assets, including
equity capital, net income, and average assets.

Service to People with Low and Moderate Incomes

To evaluate the extent to which credit unions serve people with low and
moderate incomes, we analyzed existing data on the income levels of credit
union members, reviewed available literature, and interviewed regulatory
and industry officials. We analyzed 2001 Home Mortgage Disclosure Act
(HMDA) data, the Federal Reserve's 2001 Survey of Consumer Finances (SCF),
NCUA program literature, and statistical reports of industry trade and
consumer groups. To present our findings, we relied on the combined
message of all these studies and data sources because we found no single
source that contained data on the incomes of credit union and other
depository institution consumers. To compare the income characteristics of
households and neighborhoods that obtain mortgages from credit unions and
banks, we used four income categories--low, moderate, middle, and

Appendix I
Objectives, Scope, and Methodology

upper-used by financial regulators as part of the Community Reinvestment
Act (CRA) exams.5 See table 4 for definitions.

                    Table 4: Definition of Income Categories

Categories Definitions

Low income	For an individual income, when income is less than 50 percent
of the metropolitan statistical area's (MSA) median family income, and for
a geographic area, when the median family income is less than 50 percent

Moderate income	For an individual income, when income is at least 50
percent and less than 80 percent of the MSA's median family income, and
for a geographic area, when the median family income is at least 50
percent and less than 80 percent

Middle income	For an individual income, when income is at least 80 percent
and less than 120 percent of the MSA's median family income, and for a
geographic area, when the median family income is at least 80 percent and
less than 120 percent

Upper income	For an individual income, when income is at least 120 percent
or more of the MSA's median family income, and for a geographic area, when
the median family income is 120 percent or more

Source: 12 C.F.R. 228.12 (n).

We analyzed loan application records (LAR) from the HMDA database to
compare the proportion of mortgage loans made by credit unions and peer
group banks with households and communities with various income levels. We
used 2001 HMDA data, the most recent data set available from the Federal
Reserve Bank at the time of our review. For the purposes of comparing
credit union lending with that of banks, we included only those banks with
assets of $16 billion or less on December 31, 2001, which was the size of
the largest credit union in 2001, rounded up to the nearest billion. In
addition, we excluded lending institutions that only made mortgages. Our
HMDA analysis included records from 4,195 peer group banks. We obtained
the asset size and total membership for credit unions reporting to HMDA
from NCUA's 2001 call report database and obtained the asset size of other
lenders (to identify the peer group banks) from the

5In passing the CRA, Congress required federal financial supervisory
agencies, except NCUA, to assess an institution's record of helping to
meet the credit needs of the local communities in which the institution is
chartered.

Appendix I
Objectives, Scope, and Methodology

HMDA Lender File, which contains data on the characteristics of
institutions reporting to HMDA, supplied to us by the Federal Reserve.

Our HMDA analysis did not include all credit unions and banks because only
institutions that meet HMDA's reporting criteria, such as having a certain
amount of assets, must report their mortgage loans to HMDA. For example,
in 2001, depository institutions with more than $31 million in assets as
of December 31, 2000, were required to report loans to HMDA. Largely
because of this criterion, most credit unions-86 percent-were not required
to report mortgage loans to HMDA and, thus, were excluded from our
analysis. However, we believe our analysis is still of value because, in
2001, reporting credit unions held about 70 percent of credit union assets
and included 62 percent of credit unions' members.

For our analysis, we only analyzed LARs for originated loans for the
purchase of one-to-four family homes that served as the purchaser's
primary dwelling. Our analysis included about 71,000 loans reported by
credit unions and about 807,000 loans reported by peer group banks. We
determined that the data were sufficiently reliable for the purposes of
this report by performing electronic testing of the required data
elements, reviewing existing information about the data and the system
that produced them, and interviewing agency officials knowledgeable about
the data. We did not independently verify the accuracy of the contents of
the LARs reported to the HMDA database or the accompanying lender file.

After selecting the records, we determined what proportion of credit union
and bank loans were made to purchasers with low, moderate, middle, and
upper incomes. To do so, we categorized the purchaser's gross annual
income, as identified on the LAR, into one of four income categories based
on the median family income of the MSA in which the purchased home was
located. We did this by matching the Metropolitan Statistical Area (MSA)
on the HMDA LAR with the appropriate Department of Housing and Urban
Development (HUD)-estimated 2001 median family income. We used SAS version
8.02 version, which is a computer-based data analysis and reporting
software application, to perform all of these analyses. We did not analyze
about 16 percent of the credit union and bank LARs because they did not
contain a MSA. While it is possible that this information was simply not
recorded, lenders must only report MSAs for properties located in MSAs
where their institution has a home or branch office.

In addition, we determined what proportion of credit union and bank loans
were made for the purchase of properties in census tracts by the median

Appendix I
Objectives, Scope, and Methodology

family income of the census tract. The Federal Reserve Board, in
categorizing each census tract level, used the four income categories used
by the financial regulators (low, moderate, middle, and upper) and used
definitions corresponding to the ones identified in table 4. Because the
median income of each census tract is labeled within HMDA, we did not have
to determine the income category ourselves. We did not analyze about 16
percent of the credit union and bank LARs because they did not contain a
census tract. While it is possible that this information was simply not
recorded, lenders are not responsible for identifying census tract
information if the property is located in a county with less than 30,000
people or if the property was located in an area that did not have census
tracts for the 1990 census.

Finally, we analyzed the race and ethnicity data in HMDA to compare the
lending records of credit unions and banks whose loans met our criteria.
As noted in appendix VI, about 15 percent of records for credit unions
lacked race and ethnicity data and 6 percent of records for banks. While
it is possible that this information was simply not recorded, applicants
filing loan applications by mail or by telephone are not obligated to
provide this information.

We also analyzed the Federal Reserve's 2001 SCF, a triennial survey of
U.S. households sponsored by the Board of Governors of the Federal Reserve
System with the cooperation of Treasury, and reviewed secondary sources to
identify the characteristics of credit union members. We analyzed the SCF
because it is a respected source of publicly available data on financial
institution and consumer demographics that is nationally representative
and because it was the only comprehensive source of publicly available
data with information on financial institutions and consumer demographics
that we could identify. We analyzed the SCF to develop statistics on the
income, race, age, and education of credit union members and bank
customers. Because some customers use both credit unions and banks, we
performed our income analysis based on the assumption that households can
be divided into four user categories-those who use credit unions only,
those who primarily use credit unions, those who use banks only, and those
who primarily use banks. Dr. Jinkook Lee of Ohio State University
developed these categories. In addition, to identify existing research on
credit union research, we asked officials at NCUA and industry groups (for
example, the Credit Union National Association (CUNA) to identify relevant
studies and performed a literature search.

                                   Appendix I
                       Objectives, Scope, and Methodology

Impact of CUMAA	To study the impact of CUMAA on credit union field of
membership regulations, we reviewed and analyzed CUMAA and compared its
provisions with NCUA interpretive rulings and policy statements (IRPS) in
effect before and after CUMAA. In addition, we interviewed NCUA officials
and industry representatives to obtain their viewpoints on how NCUA
interpreted CUMAA's membership provisions. To obtain information about
state field of membership regulations in general and how many
state-chartered credit unions serve geographical areas, we surveyed
regulators in the 50 states and received responses from the 46 that
actively charter credit unions. This allowed us to compare the number of
federally chartered and state-chartered credit unions serving geographical
areas. Finally, we obtained historical trend data from NCUA on the charter
types of federally chartered credit unions, "potential" (that is, people
within a credit union's field of membership but not members of the credit
union) and actual membership, and service to underserved areas.

Regulatory Oversight	To evaluate how NCUA's supervision and examination of
credit unions has evolved in response to changes in the industry since
1991, we identified changes in the types of products, services, and
activities in which credit unions engage as well as key changes to NCUA
regulations. We also identified changes to NCUA's examination and
supervision approach, and evaluated oversight procedures of federally
insured, state-chartered credit unions. Finally, we studied NCUA's
implementation of prompt corrective action (PCA).

To identify changes in the types of products, services, and activities in
which credit unions engage, we analyzed 1992-2002 Form 5300 call report
data and conducted structured interviews with NCUA examiners, state
supervisory officials, and officials from seven large credit unions. To
identify key regulatory changes, we (1) reviewed the Federal Credit Union
Act and amendments made by Congress since 1991; (2) interviewed NCUA
officials, including NCUA's General Counsel and officials from NCUA's
Division of Examination and Insurance, NCUA and state examiners, and
officials from seven large credit unions; (3) reviewed NCUA legal opinions
and letters to credit unions; and (4) reviewed final rules published in
the Federal Register.

To identify changes to NCUA's examination and supervision approach, we
reviewed NCUA's examiner guide for key elements of the risk-focused
examination approach and compared current exam documentation

Appendix I
Objectives, Scope, and Methodology

requirements with previous requirements. We conducted structured
interviews with six of NCUA's regional directors, 23 NCUA examiners
covering all NCUA regions, and 13 state supervisory officials from
Alabama, California, Idaho, Illinois, Indiana, Maryland, Massachusetts,
Michigan, Nevada, Ohio, Texas, Washington, and Wisconsin. These states
contained 51 percent of the total number of federally insured,
state-chartered credit unions and 58 percent of the total assets of
federally insured, state-chartered credit unions as of December 31, 2002.
In addition, we interviewed officials from seven large credit unions;
selecting at least one credit union from NCUA's six regions. To obtain
information on the experiences of other depository institution regulators
with the risk-focused examination and supervision approach, we interviewed
officials from the FDIC, OTS, Office of the Comptroller of the Currency,
and the Federal Reserve Bank. Finally, to obtain information on other NCUA
initiatives intended to compliment the risk-focused program, we reviewed
NCUA documents on the large credit union pilot program, and the subject
matter examiner program.

To evaluate oversight procedures of federally insured, state-chartered
credit unions, we obtained information about the oversight procedures
during our structured interviews with the 13 states supervisory officials
and NCUA examiners. We also reviewed NCUA's examiner guide and memorandum
of understanding between NCUA and states describing NCUA's procedures for
conducting joint examinations of federally insured, state-chartered credit
unions with state regulators.

Finally, to study NCUA's implementation of PCA, we reviewed CUMAA, NCUA
rules and regulations pertaining to PCA, and NCUA's examiner guide. We
also analyzed data from NCUA on the number of credit unions subject to PCA
as of December 31, 2002. We interviewed agency officials knowledgeable
about this data and found that NCUA headquarters, as well as the region,
conducted reasonableness checks against the Form 5300 database, which
contains the net-worth ratio used for PCA. When data outliers were found,
examiners were required to review the data for accuracy and make any
necessary corrections. We determined that the data were sufficiently
reliable for the purposes of this report. In addition, we interviewed NCUA
officials and examiners, state supervisory officials, credit union
officials, and officials of other federal financial regulatory agencies to
obtain their perspectives on PCA.

                                   Appendix I
                       Objectives, Scope, and Methodology

Status of NCUSIF	To evaluate the financial condition of NCUSIF, we
obtained key financial data about the fund from NCUA's annual audited
financial statements for 1991-2002. For 2002, we compared NCUSIF's key
performance measure, which is the ratio of fund equity to insured shares
(deposits), to key performance measures of the Bank Insurance Fund,
Savings Association Insurance Fund, and American Share Insurance, the
remaining private insurer. We also reviewed NCUSIF's estimated loss and
overhead administrative expenses transfer process and applicable internal
controls. We reviewed other relevant industry studies on deposit-insurance
pricing and loan-loss allowance. In addition, we interviewed NCUA
officials, industry trade groups, and officials of other federal financial
regulatory agencies to obtain their perspectives on the funding of NCUSIF,
the overhead transfer rate, and the loan-loss allowance.

Private Share Insurance	To better understand the issues around share
(deposit) insurance, we reviewed and analyzed relevant studies on federal
and private insurers for both credit unions and other depository
institutions.6 In addition, we interviewed officials at NCUA, the
Department of the Treasury, and FDIC to obtain perspectives specific to
private share insurance. We also obtained views from credit union industry
groups including the National Association of Federal Credit Unions,
National Association of State Credit Union Supervisors, and CUNA.

To determine the extent to which private share insurance is permitted and
utilized by state-chartered credit unions, we conducted a survey of state
credit union regulators in all 50 states. Our survey had a 100-percent
response rate. In addition to the survey, we obtained and analyzed
financial and membership data of privately insured credit unions from a
variety of sources (NCUA, Credit Union Insurance Corporation, CUNA, and
ASI-the only remaining provider of primary share insurance). We found this
universe difficult to confirm because in our discussions with state
regulators, NCUA and ASI officials, and our review of state laws, we
identified other states that could permit credit unions to purchase
private share insurance.

6The scope of our work was limited to primary share insurance, which is
generally mandatory for all credit unions (whereas excess share insurance
is optional coverage above primary share insurance).

Appendix I
Objectives, Scope, and Methodology

To determine the regulatory differences between privately insured credit
unions and federally insured, state-chartered credit unions, we identified
and analyzed statutes and regulations related to share insurance at the
state and federal levels.7 In addition, we interviewed officials at NCUA
and conducted interviews with officials at the state credit union
regulatory agencies from Alabama, California, Idaho, Indiana, Illinois,
Maryland, Nevada, New Hampshire, and Ohio. Finally, we analyzed NCUA's
application of its conversion policies and looked at the cases of six
credit unions that terminated their federal share insurance and converted
to private share insurance in 2002 and 2003.

To identify factors influencing a credit union's decision to obtain
private or federal share insurance, we conducted structured interviews
with officials of both federally insured and privately insured credit
unions. Specifically, we interviewed management at 29 credit unions that,
since 1990, had converted from federal to private share insurance and
management at 26 credit unions that had converted from private to federal
share insurance. We did not interview credit union management in states
that did not permit private insurance.

To determine the extent to which privately insured credit unions met
federal disclosure requirements, we identified and analyzed federal
consumer disclosure provisions in section 43 of the Federal Deposit
Insurance Act, as amended, and conducted unannounced site visits to 57
privately insured credit unions (49 main and 8 branch locations) in
Alabama, California, Illinois, Indiana, and Ohio.8 The credit union
locations were selected based on a convenience sample using state and city
location coupled with random selection of main or branch locations within
each city. About 90 percent of the locations we visited were the main
institution rather than a branch institution. This decision was based on
the assumption that if the main locations were not in compliance, then the
branch locations would probably not be in compliance either. Although
neither these site visits, nor the findings they produced, render a
statistically valid sample of all possible main and branch locations of
privately insured credit unions necessary in order to determine the
"extent" of compliance, we believe that what we found is robust enough,
both in the

7We limited our analysis to those states with privately insured credit
unions-Alabama, California, Idaho, Indiana, Illinois, Maryland, Nevada,
and Ohio.

812 U.S.C. S: 1831t.

Appendix I
Objectives, Scope, and Methodology

aggregate and within each state, to raise concern about lack of disclosure
in privately insured credit unions. During each site visit, using a
systematic check sheet, we noted whether or not the credit union had
conspicuously displayed the fact that the institution was not federally
insured (on signs or stickers, for example).

In addition, from these same 57 sites visited, we collected a total of 227
credit union documents that we analyzed for disclosure compliance. While
section 43 requires depository institutions lacking federal deposit
insurance to disclose they are not federally insured in personal
documents, such as periodic statements, we did not collect them. We also
conducted an analysis of the Web sites of 78 privately insured credit
unions, in all eight states where credit unions are privately insured, to
determine whether disclosures required by section 43 were included. To
identify these Web sites, we conducted a Web search. We attempted to
locate Web sites for all 212 privately insured credit unions; however, we
were able to identify only 78 Web sites. We analyzed all Web sites
identified. Finally, we interviewed FTC staff to understand their role in
enforcement of requirements of section 43 for depository institutions
lacking federal deposit insurance.

To understand how private share insurers operate, we conducted interviews
with officials at three private share insurers for credit unions- ASI
(Ohio), Credit Union Insurance Corporation (Maryland), and Massachusetts
Credit Union Share Insurance Corporation (Massachusetts). Because ASI was
the only fully operating provider of private primary share insurance, ASI
was the focus of our review.9 We obtained documents related to ASI
operations such as financial statements and annual audits and analyzed
them for the auditor's opinion noting adherence with accounting principles
generally accepted in the United States. Additionally, to understand the
state regulatory framework for this remaining private share insurer, we
interviewed officials at the Ohio Department of Insurance.

9As of December 2002, we identified two entities that provide private
primary share insurance to credit unions in the 50 states and the District
of Columbia-ASI and Credit Union Insurance Corporation (CUIC). However,
CUIC in Maryland was in the process of dissolution and, therefore, we did
not include it in our analysis. During our review, we learned that
Massachusetts Credit Union Share Insurance Corporation only provides
excess deposit insurance, and therefore we did not include it in our
analysis.

Appendix II

Status of Recommendations from GAO's 1991 Report

We made 52 recommendations to Congress and the National Credit Union
Administration (NCUA) in our 1991 report on the credit union industry and
NCUA1 Of these, 28 were made to Congress, of which 8 were implemented or
partially implemented as of September 2003. We made 24 recommendations to
NCUA, and 19 were implemented as of September 2003. In addition, we issued
one matter for congressional consideration. Congress partially addressed
this matter.

Our recommendations spanned the range of issues addressed in our 1991
report, including

o 	the condition of the credit union industry and the National Credit
Union Share Insurance Fund (NCUSIF),

o  credit union law and regulation,

o  supervision of credit unions,

o  NCUA's management of failed credit unions,

o  corporate credit unions,

o  share insurance issues,

o  structural changes in NCUA, and

o  the evolution of credit unions' role in the financial marketplace.

NCUA implemented most of our recommendations to the agency. The key
changes implemented by NCUA affected (1) corporate credit unions, (2)
reporting requirements for credit unions, and (3) supervision of
state-chartered credit unions. With respect to corporate credit unions,
NCUA implemented various recommendations that established minimum capital
requirements, limited investment powers of state-chartered corporate
credit unions, increased detail and frequency of reporting requirements,
and established a new unit in NCUA that is responsible for oversight,
examination, and enforcement of corporate credit unions. We expect to
review corporate credit unions following this study and to report in
greater

1U.S. General Accounting Office, Credit Unions: Reforms for Ensuring
Future Soundness, GAO/GGD-91-85 (Washington, D.C.: July 10, 1991).

Appendix II
Status of Recommendations from GAO's 1991
Report

depth on issues affecting corporate credit unions. In the area of
reporting requirements, NCUA implemented a requirement in 1993 that all
federally insured credit unions with assets greater than $50 million file
financial and statistical reports (call reports) on a quarterly basis and
as of July 1, 2002, required all federally insured credit unions to file
quarterly call reports. Finally, NCUA affirmed its supervision of
state-chartered and federally insured credit unions by establishing
examination goals, as well as conducting examinations, at almost 16
percent of all state-chartered and federally insured credit unions in
2002.

NCUA told us that it chose not to implement five of our recommendations
because it either disagreed with the recommendations (see recommendation
24 in table 5), or believed it had already addressed the recommendations
(see recommendations 9, 11, 16, 17 in table 5). For example, NCUA
disagreed with our recommendation to separate its supervision and
insurance functions (see recommendation 24) and believed it was
unnecessary for credit unions to submit copies of their supervisory
committee audit reports to NCUA, as NCUA examiners routinely review the
reports as part of the examination process (see recommendation 9).

Congress implemented or partially implemented 8 of the 28 recommendations
we made, which (1) established minimum capital levels for credit unions,
(2) tightened commercial lending, and (3) established annual audit
requirements for credit unions with assets greater than $500 million. As
discussed in table 5, among those not implemented are recommendations
dealing with NCUA's Central Liquidity Facility (CLF) (see recommendations
49-52) and the structure of NCUA (see recommendations 43-48).2

See table 5 for our recommendations to NCUA and Congress and their status
as of August 31, 2003.

2CLF was created in 1978 to improve the general financial stability of
credit unions by serving as a liquidity lender to credit unions
experiencing unusual or unexpected liquidity shortfalls. The NCUA board
oversees the CLF.

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

Table 5: Status of GAO Recommendations to NCUA and Congress, as of August
31, 2003

                Issue GAO Recommendation to NCUA Status Comments

    Condition of Require credit unions Implemented Implemented in the March   
                 with assets greater               31, 1993,                  
credit unions  than $50 million to              quarterly call reports for 
                  file financial and               federally insured          
      and NCUSIF  statistical reports               credit unions with assets 
                      quarterly.                             greater than $50 
                                                   million. Effective July 1, 
                                                   2002, NCUA                 
                                                   expanded rule to cover all 
                                                            federally insured 
                                                         credit unions.       

Expand the information required from credit unions with assets greater
than $50 million on the financial and statistical reports in the areas of
asset quality, interest rate sensitivity, management, and common bond.

Implemented	According to NCUA, it established a reporting system for
common bond data in January 2002. The system monitors the approvals of
field of membership and is called Generated Efficient National Information
System for Insurances Services. Also, NCUA investment rules require credit
unions that make certain investments to perform shock tests on interest
rate sensitivity. According to NCUA, it performs shock tests of credit
unions using call report data and expects examiners to make contact with
credit unions if potential problems are identified.

    Law and    Assess its real estate   Implemented In June 1991, NCUA issued 
                   regulation and                               comprehensive 
regulation strengthen it to help                 guidelines and since then 
              ensure the sound                             issued a series of 
              underwriting of loans and              letters to credit unions 
                  their suitability for                to address this issue. 
                sale in the secondary               
                       market.                      
                Restrict the exclusions             A final rule addressing   
                    from its commercial Implemented all of our concerns       
              lending limit established             and recommendations went  
              in 1987 to help                       into effect in            
              ensure that credit unions                January 1992. The rule 
              are not used as                          established a limit on 
              vehicles underwriting                 the amount of loans that  
              large commercial                      may be made to            
                      ventures.                           one borrower to the 
                                                     greater of 15 percent of 
                                                      reserves or $75,000.    

Supervision Clarify the purposes,     Implemented  According to NCUA, the  
               unique values, and                           Office of         
               requirements for use of                        Examination and 
               each of its off-site                   Insurance completed the 
                       monitoring tools.             requirements for the use 
               Determine the appropriate                   of off-site        
               recipients of the tools               monitoring tools, such   
               and distribute them                   as the use of risk       
               accordingly, within each              reports, in fiscal year  
                        region.                      1995. Since then,        
                                                     NCUA has adopted         
                                                     additional off-site      
                                                       monitoring tools, such 
                                                          as the consolidated 
                                                     balance sheet and scope  
                                                     workbook.                

6    Require documentation at the Implemented NCUA requires this review as 
                            regional                              part of the 
          office level of examiners'             examination process and      
                      reviews of all             requires                     
       credit union call reports.                documentation of the review  
                                                 in the                       
                                                     examination report.      

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

                Issue GAO Recommendation to NCUA Status Comments

Invoke its statutory authority to refuse to accept state supervisors'
examinations when a state regulatory authority lacks adequate independence
from the credit union industry. Examine all NCUSIF-insured credit unions
in such states.

Implemented	According to NCUA, its examiner guide addresses oversight of
federally insured, state-chartered credit unions, including processes to
make an independent assessment of these credit unions. NCUA affirms it is
empowered by the Federal Credit Union Act to examine any federally insured
credit union, including those where questions are raised regarding the
independence of the state from the industry. NCUA claims that use of this
authority is evidenced by having conducted exams at 15.6 percent of all
federally insured, state-chartered credit unions in 2002.

Establish a policy goal for examination Implemented NCUA affirms that its
regions have

frequency of state-chartered credit unions.	established goals that include
monitoring the examination cycles and supervision efforts of each state.
State examinations not conducted within 18 months are tracked and
agreements are made and followed to bring the state into compliance.

Require all credit unions to submit copies of Not implemented This
recommendation pertains to federally

their supervisory committee audit reports to insured credit unions with
less than $500

NCUA upon completion.	million in assets. NCUA believes that the 1991
recommendation is unnecessary. NCUA claims it reviews the supervisory
committee audits as a required step during the risk-focused examination
process.

Conduct an Inspector General review Implemented The Inspector General
completed quality
focusing on NCUA's handling of problem assurance reviews of each NCUA
region as
credit unions since mid-1990, specifically its of July 1994.
use of enforcement powers, and submit a
report to the NCUA board.

NCUA's Require that waivers and special charges management of be
authorized by the Director of the Office failed credit of Examination and
Insurance, the General unions Counsel, and the regional director.

Not implemented	Under prompt corrective action, NCUA is required to take
various mandatory supervisory actions against credit unions depending on
their net worth ratio, including requiring earnings transfers for credit
unions that are less than well capitalized-7 percent net worth ratio or
less. NCUA has established guidelines under which regional directors can
grant earnings retention waivers as well as charges to the reserve. NCUA
claims that its regional offices track approval of waivers and charges.

12	Develop policy guidance concerning the Implemented NCUA maintains rules
regarding waivers and use of these provisions and monitor their special
charges in Section 702 of its rules use. and regulations.

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

                Issue GAO Recommendation to NCUA Status Comments

Adhere to the criteria for assisting credit unions.

Implemented	NCUA claims that the implementation of prompt corrective
action in February 2000 greatly changed its ability to assist credit
unions. To address the issue of assistance to credit unions, NCUA affirms
that the board approved a Special Assistance Program in February 2001, and
that it maintains a Special Assistance Manual regarding the documentation
and quality of requests for assistance. Finally, NCUA claims it has
implemented an approval process for different levels of assistance to
credit unions.

                         Establish minimum                Section 704 of NCUA 
                  capital requirements for             regulations requires a 
        Corporate  corporate credit unions                  minimum 4 percent 
    credit unions  and U.S. Central Credit Implemented      capital ratio for 
                   Union, taking all risks                 retail, as well as 
                            into account.a               wholesale, corporate 
                                                               credit unions, 
                  In the interim,                       such as U.S. Central  
                  establish a minimum                      Credit Union.      
                  level                                
                  based on assets, and set             
                  a time frame for                     
                     achieving this level.             
                    This could be achieved             
                   by increasing reserving             
                          requirements and             
                  using subordinated debt              
                   arrangements, such as               
                  membership capital share             
                         deposits.                     

      Restrict the investment Implemented NCUA's corporate credit union rules 
             powers of state-                                        apply to 
chartered corporate credit                 all federally insured corporate 
                unions to the                                  credit unions. 
    limits imposed on federal             NCUA requires all nonfederally      
             corporate credit             insured                             
            unions.                       corporate credit unions to adhere   
                                          to the                              
                                          same rules as a condition of        
                                          receiving                           
                                          shares or deposits from federally   
                                          insured                             
                                                    credit unions.            

Limit the investments of corporate credit Not implemented NCUA believes it
is more appropriate to
unions and U.S. Central Credit Union in a establish concentration limits
on capital
single obligor to 1 percent of the investor's rather than assets and
established a
total assets. Exceptions should include regulation limiting aggregate
investments in
obligations of the U.S. Government, any single obligor to the greater of
50
repurchase agreements that equal up to 2 percent of capital or $5 million.
percent of assets, and all investments by
corporate credit unions in U.S. Central
Credit Union.

17        Limit loans to one Not implemented NCUA believes it is more      
          borrower by corporate                 appropriate to set            
         credit unions and U.S.                 limits based on capital       
           Central Credit Union                 instead of assets. In         
            to 1 percent of the                  October 1997, the loan limit 
          lender's assets. NCUA                                was 10 percent 
        should be authorized to                       of capital-an amount we 
                make exceptions                              determined could 
      on a loan-by-loan basis.                  exceed 1 percent of assets.   
                                                As of January                 
                                                 2003, NCUA rules capped the  
                                                           maximum            
                                                aggregate loan amount to any  
                                                one member                    
                                                 to 50 percent of capital for 
                                                             unsecured loans, 
                                                   and 100 percent of capital 
                                                           for secured loans, 
                                                with exceptions. We view this 
                                                               as a departure 
                                                from the 1991 recommendation. 

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

Issue GAO Recommendation to NCUA Status Comments

      Obtain more complete and Implemented According to NCUA, it requires     
                        timely             corporate                          
information about corporate                credit unions to submit monthly 
                     financial                                   call reports 
           operations.                     to NCUA as well as information to  
                                           examiners. Also, NCUA affirms that 
                                                                   it revises 
                                                   the corporate call reports 
                                                           annually to ensure 
                                           proper supervision of corporate    
                                           credit                             
                                                        unions.               

Establish a unit at NCUA headquarters that Implemented According to NCUA,
the NCUA board would be responsible for corporate separated corporate
credit union supervisory oversight, examination, and enforcement
responsibility from the Office of Examination actions. and Insurance and
created the Office of

Corporate Credit Unions in August 1994.

Review the CAMEL rating system Implemented In January 1999, NCUA           
                              for             implemented a                   
       corporate credit unions to             system for evaluating the risk  
                       reduce the             associated                      
inconsistencies and focus more                with corporate credit unions 
                       clearly on                           that is different 
     the component being rated.               from the CAMEL ratings used for 
                                                                 other credit 
                                                 unions. The system, known as 
                                                                the Corporate 
                                              Risk Information System, has 12 
                                                                    component 
                                              ratings regarding financial     
                                              risk and risk                   
                                                        management.           

     Share   Place NCUSIF's fiscal Implemented In November 1993, the NCUA     
                year on a calendar             Board of                       
insurance                                   Directors approved the change  
                     year.                     to a fiscal                    
                                                   year based on the calendar 
                                                               year (January- 
                                                      December), which became 
                                                            effective January 
                                                          1, 1995.            

Reduce the time lag in adjusting NCUSIF's Implemented According to NCUA,
establishing a fiscal

financing.	year based on the calendar year for NCUSIF reduced time lags in
collection of assessments from 7 to 3 months.

Require credit unions to exclude their 1 Implemented Action taken by
Congress addressed our
percent deposit in NCUSIF from both sides concern. Minimum net worth
ratios
of their balance sheet when assessing established in the 1998 Credit Union
capital adequacy. Then, that amount would Membership Access Act (CUMAA),
which is
not be counted as credit union capital. 7 percent for well-capitalized
credit unions,

compensated for the NCUSIF deposit (1 percent of assets) that credit
unions account for on their balance sheet. The minimum capital ratio for
banks insured by FDIC is 6 percent.

24	NCUA Immediately establish separate supervision Not implemented NCUA
disagrees with this recommendation. structural and insurance offices that
report directly to changes the board.

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

Issue GAO Recommendation to Congress Status Comments

Condition of Hold annual oversight hearings at which the Not implemented
As of September 1994, the Senate did not credit unions NCUA board
testifies on the condition of hold hearings, but the House Banking and
NCUSIF credit unions and NCUSIF and assesses Committee had. NCUA has no
objections to

risk areas and reports on NCUA's this recommendation. responses.

    Law and    Amend Federal Credit Union Implemented  Implemented as part of 
                            Act (FCUA) to                   prompt corrective 
regulation   require NCUA to establish             action in CUMAA (August 
                          minimum capital             1998) and               
              levels for credit unions no             promulgated as NCUA     
              less stringent                          regulation in           
              than those applicable to                    February 2000.      
              other insured                           
              depository institutions,                
              providing for an                        
                 appropriate phase-in                 
                        period.                       

Amend the FCUA to limit the amount that credit unions can loan or invest
in a single obligor, other than investments in direct or guaranteed
obligations of the U.S. Government or in the credit union's corporate
credit union, to not more than 1 percent of the credit union's total
assets. Limits permitted in 1991 with respect to credit union service
organizations should continue, and exposures of not more than 2 percent of
assets should be provided for in repurchase agreement transactions.
Authorize NCUA to set a higher limit for secured consumer loans made by
small credit unions and for overnight funds deposited with correspondent
institutions.

Not implemented	NCUA's position has changed since 1994, when it believed a
5 percent of assets limitation on exposure to single obligors would be
satisfactory. According to NCUA, the 5-percent limitation is too
restrictive for some credit unions, especially for smaller credit unions.
According to NCUA, its current regulations for credit unions do not
provide specific limits, but provides flexibility to well-run and managed
credit unions. NCUA believes that setting obligor limitations is better
handled through the agency's regulation process because it permits prompt
changes, is considerate of the fluid financial environment, and maintains
emphasis on overall risk.

       Amend the FCUA to require Implemented  Implemented as part of CUMAA in 
                         NCUA to                                     1998 and 
          tighten the commercial             promulgated as NCUA regulation   
              lending regulation             in May                           
and include an overall limit.                   1999. NCUA established the 
                                                              aggregate limit 
                                             on a credit union's outstanding  
                                             member                           
                                              business loans to the lesser of 
                                                               1.75 times the 
                                                  credit unions' net worth or 
                                                             12.25 percent of 
                                                      total assets.           

Amend the FCUA to modify borrowing authority and specify that credit
unions may not borrow for the purpose of growth, unless prior approval of
NCUA is obtained.

Not implemented	NCUA believes that this recommendation is not necessary
because Congress indirectly addressed this issue through PCA provisions in
CUMAA in 1998. According to NCUA, if a credit union is undercapitalized
under PCA, then growth can be restricted. Also according to NCUA, PCA
requirements indirectly influence borrowing because borrowing could impact
net worth classification.

For clarification, we intended this recommendation to apply to all credit
unions, not just those under PCA.

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

Issue GAO Recommendation to Congress Status Comments

Amend the FCUA to require credit unions to Implemented Implemented as part
of comprehensive
adequately disclose that dividends on banking reforms in 1991. NCUA issued
a
shares and other accounts cannot be regulation under the Truth in Savings
Act.
guaranteed in advance but are dependent
on earnings.

Amend the FCUA to require all insured Not implemented NCUA is opposed to
this recommendation

credit unions to obtain NCUA permission and believes that current
regulations are

before opening a new branch.	appropriate. NCUA's regulations require
federally insured credit unions with over $1 million in assets to obtain
NCUA approval to invest in fixed assets, including branch offices, if the
aggregate of all such investments exceeds 5 percent of shares and retained
earnings. Credit unions eligible under NCUA's Regulatory Flexibility
Program are exempt from this requirement.

Amend the FCUA to require credit unions Partially Implemented as part of
CUMAA in 1998 and above a minimum size to obtain annual implemented
promulgated as NCUA regulation in July independent certified public
accountant 1999. Credit unions with assets greater than audits and to make
annual management $500 million are required to obtain annual reports on
internal controls and compliance independent certified public accountant
with laws and regulations. audits. However, no requirement has been

made requiring annual management reports on internal controls and
compliance with laws and regulations.

Amend the FCUA to authorize and require Not implemented NCUA agrees with
this recommendation.
NCUA to compel a federally insured, state-
chartered union to follow the federal
regulations in any area in which the credit
union's powers go beyond those permitted
federally chartered credit unions and are
considered to constitute a safety and
soundness risk.

NCUA's Amend FCUA to authorize NCUA to provide Not implemented According
to NCUA, it maintains a policy of

management of assistance in resolving a failing credit union assisting
failing credit unions at the least

failures only when it is less costly than   cost. Also, NCUA believes that 
                     liquidation or                                changes to 
             essential to provide adequate   the FCUA are unnecessary because 
                       depository                                        NCUA 
               services in the community.    has enough flexibility to assist 
                                                               failing credit 
                                                  unions when the benefits of 
                                                               preserving the 
                                             credit union outweigh the cost.  

35  Require NCUA to maintain Not implemented    According to NCUA, its     
                  documentation                         policies and          
      supporting its resolution                    practices emphasize the    
             decisions,                                 importance of         
      including the statistical                 maintaining documentation of  
            and economic                                 resolutions          
          assumptions made.                        and that decisions are     
                                                        supported. In         
                                                addition, and according to    
                                                NCUA, it actively             
                                                  updates expectations and    
                                                        processes for         
                                                retrieving and maintaining    
                                                data through the              
                                                   revision of the Examiner's 
                                                            Guide, Accounting 
                                                  Manual, Directives, Special 
                                                              Actions Manual, 
                                                   and Guidance to Credit     
                                                           Unions.            

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

              Issue GAO Recommendation to Congress Status Comments

                 Amend the FCUA to                        While not expressly 
     Corporate   confine insured      Not implemented   implemented, NCUA has 
                 credit                               
credit unions union investments in                  taken some action in   
                   corporate credit                       this area. NCUA     
                 unions and U.S.                        regulations require   
                 Central Credit Union                  nonfederally insured   
                 to                                   
                      those that have                 corporate credit unions 
                     obtained deposit                   to agree to adhere to 
                            insurance                 
                                                      its corporate credit    
                     from NCUSIF.                     union rule and to       
                                                      submit                  
                                                      to NCUA examinations as 
                                                          a condition of      
                                                      receiving shares or     
                                                      deposits from federally 
                                                      insured credit unions.  
                                                      According to NCUA,      
                                                      there is only one       
                                                      corporate credit union  
                                                      that                    
                                                         is not federally     
                                                             insured.         

Require NCUA to establish a program to promptly increase the capital of
corporate credit unions and establish minimum capital standards.

Implemented	NCUA's regulations require corporate credit unions to maintain
a minimum capital ratio of 4 percent. In addition, NCUA may issue a
capital directive to corporate credit unions to achieve adequate
capitalization within a specified time frame by taking any action deemed
necessary, including increasing the amount of capital to specific levels.
NCUA's corporate credit union rule also imposes an earnings retention
requirement of either 10 or 15 basis points per annum if a corporate
credit union's retained earnings ratio falls below 2 percent.

Share insurance Require credit unions to expense the 1 percent deposit in
NCUSIF over a reasonable period of time-to be determined by NCUA. At the
same time, emphasize that the assets represented by a failed credit
union's insurance deposit should be available first to NCUSIF. This action
should be coordinated with and consistent with any legislation to
recapitalize the Bank Insurance Fund in order to avoid placing credit
unions at a competitive disadvantage.

Implemented	We determined that Congress' passage of CUMAA, which set net
worth levels for credit unions 1 percent higher to compensate for NCUSIF's
accounting of the deposit as an asset, addressed our concerns about the
double counting of capital at NCUSIF and credit unions. We determined that
the recommendation regarding NCUSIF's access to the assets of a failed
credit union has not been implemented, but we determined that this
recommendation is implemented because our greatest concern was addressed
regarding the double counting of capital between NCUSIF and credit unions.

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

              Issue GAO Recommendation to Congress Status Comments

Amend the FCUA to establish an available assets ratio for NCUSIF.

Implemented	In passing CUMAA in August 1998, Congress amended the FCUA to
establish a minimum 1.0 percent available assets ratio for NCUSIF. In
addition, the NCUA board is to make a distribution to insured credit
unions after each calendar year if, at the end of the calendar year: the
NCUSIF's available assets ratio exceeds 1.0 percent, any loans from the
federal government as well as interest on those loans have been repaid,
and NCUSIF's equity ratio exceeds the normal operating level.

Amend the FCUA to authorize NCUA to raise the basic NCUSIF equity ratio,
available assets ratio, and premiums, and delete NCUSIF ability to set a
normal operating level below the statutory minimum.

Implemented	Under CUMAA, Congress authorized NCUA to assess a premium
charge on insured credit unions if NCUSIF's equity ratio was less than 1.3
percent and the premium charge would not exceed the amount necessary to
restore the equity ratio to 1.3 percent. Congress also defined NCUSIF's
normal operating level as an equity ratio to be specified by the NCUA
board between 1.2 and 1.5 percent. However, Congress set the available
assets ratio at 1.0 percent with no authority given to NCUA to change it.

Amend the FCUA to provide for additional Not implemented NCUA believes
that borrowing authority is NCUA borrowing from Treasury on behalf of
appropriate so long as the CLF and NCUSIF NCUSIF. continue to have
borrowing authority.

Amend the FCUA to place NCUSIF in a Not implemented NCUA sees no
compelling reason to make
position second to general creditors but this change.
rank this position ahead of uninsured
shares.

      NCUA           Amend the FCUA to Not implemented NCUA believes there is 
                 require that NCUA, in                      no need for       
structural consultation with                        legislative change, as 
              Congress and the credit                    PCA provisions in    
              union industry, to                                CUMAA address 
    changes   identify specific unsafe                    declining net worth 
                                                                    levels in 
                 and unsound practices                     credit unions.     
                   and conditions that                 
              merit enforcement                        
              action, identify the                     
               appropriate corrective                  
                    action, and                        
                  promulgate these                     
                  requirements by                      
                    regulation.                        

44   Amend the FCUA to require NCUA to take Not implemented Same as above. 
         appropriate enforcement action when                   
          unsafe and unsound conditions or                     
        practices, as specified in law or NCUA                 
            regulations, are identified.                       

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

Issue GAO Recommendation to Congress Status Comments

Amend the FCUA to provide for a five-Not implemented NCUA is opposed to
this recommendation.
member NCUA board, with two members
ex officio, (the Chairman of the Federal
Reserve Board and the Secretary of the
Treasury). Authorize the two ex officio
members to delegate their authority to
another member of the Federal Reserve
Board or to another official of the
Department of the Treasury who is
appointed by the President with the advice
and consent of the Senate.

Consider placing credit union's examination Not implemented NCUA opposes
this recommendation
and supervision functions under a single because it believes the change
would affect
federal regulator once such an entity is the identity of credit unions,
limit the financial
operating effectively, if there is broad reform choices for consumers,
create competing
of the depository institution regulatory and conflicting priorities for
the single
structure. The insurance function could regulator, and stifle the
financial
then be placed under FDIC or under a marketplace.
separate entity.

Remove the power of federally chartered Not implemented NCUA has no
objection to this credit unions to borrow from Farm Credit recommendation.
Banks, as provided for in FCUA.

Amend the Community Development Credit Not implemented NCUA opposes this
recommendation

Union Revolving Fund Transfer Act to because such a change would create

designate an entity other than NCUA as additional bureaucratic
requirements for

administrator of the revolving fund.	small financial institutions.
According to NCUA, the agency does not receive appropriations for
administering the program and funds the program through the operating and
overhead transfer fees collected from both federally chartered and
federally insured credit unions.

Dissolve the CLF, as established by Title III Not implemented NCUA opposes
this recommendation. of the FCUA.

If CLF continues to operate, sharply reduce Not implemented NCUA opposes
this recommendation and
CLF borrowing authority from the current believes that restricting CLF's
capacity could
level of 12 times subscribed capital and undermine its purpose.
surplus.

51     If CLF continues to Not implemented NCUA believes that the rates of 
         operate, require the                 CLF loans                       
      terms and conditions of                 are prudent. According to NCUA, 
           CLF loans to be no                 rates on                        
      more liberal than those                  CLF loans to credit unions are 
            made by the                                          based on the 
         Federal Reserve.                        Federal Financing Bank (FFB) 
                                                               fixed rate, as 
                                                the CLF borrows from the FFB. 
                                                                 Furthermore, 
                                                 according to NCUA, FFB rates 
                                                               are related to 
                                                   U.S. Treasury rates.       

                                  Appendix II
                   Status of Recommendations from GAO's 1991
                                     Report

                         (Continued From Previous Page)

Issue GAO Recommendation to Congress Status Comments

If CLF continues to operate, prohibit CLF Not implemented According to
NCUA, CLF and NCUSIF are
loans or guarantees of any kind to NCUSIF, distinct entities and CLF does
not extend
and, in the event the NCUA board certifies loans or guarantees to NCUSIF.
that CLF does not have sufficient funds to
meet liquidity needs of credit unions,
authorize the Department of the Treasury to
lend to NCUSIF, rather than to CLF, in order
to meet such needs.

                     Matter for congressional consideration

Credit unions' role in the financial marketplace

If credit unions are to remain distinct from Partially
other depository institutions because, in implemented
part, of their common-bond membership
requirement, and if this requirement is
intended to further the safe and sound
operation of credit unions, consider stating
this general intent in legislation and
establish guidelines on the limits of
occupational, associational, and community
common bonds as well as the purpose and
limits of multiple group charters. These
guidelines should apply to all federally
insured credit unions.

In passing CUMAA in August 1998, Congress established membership limits
for federally chartered credit unions with respect to common-bond and
community-chartered credit unions. Furthermore, Congress established
numerical limitations for groups to be eligible for inclusion in multiple
common-bond credit unions and established geographical guidelines for
community credit unions.

However, the legislation only applied to federally chartered credit
unions. It did not apply to federally insured,state-chartered credit
unions, which held 46 percent of total industry assets as of December 31,
2002. Therefore, this recommendation is partially implemented.

Sources: GAO; NCUA; Department of Treasury; Federal Register; CUMAA.

aU.S. Central Credit Union, founded in 1974, solely assists corporate
credit unions with financial services, including investment, liquidity,
and cash management products and services; risk management and analytic
capabilities; settlement, funds transfer and payment services; and
safekeeping and custody services. It is owned and directed by its member
corporate credit unions.

Appendix III

Financial Condition of Federally Insured Credit Unions

As we reported earlier, the financial condition of federally insured
credit unions-the industry-has improved since 1991, based on various
measures such as capital ratios, assets, and regulatory ratings. This
appendix provides greater detail on these measures. We used annual call
reports from December 31, 1992, to December 31, 2002, as well as a
database of regulatory ratings from the National Credit Union
Administration (NCUA) for the same time period. In addition, we used
consolidated data based on annual call reports for banks and thrifts in
order to compare them with credit unions.

Industry Capital Ratios The capital of federally insured credit unions as
a percentage of total

Have Increased over Time	industry assets-the capital ratio-grew from 8.10
to 10.86 percent from December 31, 1992, to December 31, 2002 (see fig.
18). Over this period, larger credit unions had consistently higher
capital ratios than smaller credit unions.

      Appendix III Financial Condition of Federally Insured Credit Unions

    Figure 18: Capital Ratios in Federally Insured Credit Unions, 1992-2002

                               Capital ratios 16

                                       14

                                       12

                                      10 8

                                       6

                                       4

                                       2

            0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Small

                                     Medium

Large

                               All credit unions

                           Source: Call report data.

Note: In this figure, small credit unions are defined as those with less
than $10 million in assets; medium credit unions are those with assets
ranging from $10 million to less than $50 million in assets; and large
credit unions are those with $50 million or more in assets. The capital
ratio of a given size category is calculated as the total equity of all
credit unions in that size category divided by the total assets of all
credit unions in that size category.

Growth of the Industry	The credit union industry grew dramatically since
December 31, 1992, as measured by assets and the value of shares (see
table 6). From December 31, 1992, to December 31, 2002, assets in
federally insured credit unions increased from $258 billion to $557
billion, or 116 percent, while shares increased from $233 billion to $484
billion, or 108 percent. From December 31, 1992, to December 31, 2000, the
annual percentage growth rates of assets and shares generally fluctuated
from around 3 percent to around 7 percent, with a significant rise in 1998
to over 10 percent. In the last 2 years (2001-2002), however, the annual
percentage growth in assets and shares again rose sharply. According to
NCUA officials, the more recent growth in assets and shares reflected a
"flight to safety" on the part of consumers

Appendix III Financial Condition of Federally Insured Credit Unions

seeking low-risk investments in reaction to the generally depressed
condition of the securities market.

 Table 6: Federally Insured Credit Union Growth in Assets and Shares, 1992-2002

Dollars in billions

Assets Shares

                                      Percentage                   Percentage 
          December 31  Dollar value     growth     Dollar value        growth 
                 1992        $258.37                     $233.01 
                 1993         277.13         7.26         246.96         5.99 
                 1994         289.45         4.45         255.02         3.26 
                 1995         306.64         5.94         270.14         5.93 
                 1996         326.89         6.60         286.71         6.13 
                 1997         351.17         7.43         307.18         7.14 
                 1998         388.70        10.69         340.00        10.68 
                 1999         411.42         5.84         356.92         4.98 
                 2000         438.22         6.51         379.24         6.25 
                 2001         501.54        14.45         437.13        15.27 
                 2002         557.07        11.07         484.19        10.77 

Source: Call report data.

As noted earlier, the industry has consolidated and become slightly more
concentrated. As of December 31, 1992, there were 12,595 credit unions,
but by December 31, 2002, that number had declined to 9,688 (see table 7).
The number of credit unions with less than $10 million in assets declined
during this period, while the number of credit unions with more than $30
million in assets grew. Those credit unions with over $100 million in
assets had around 52 percent of total industry assets as of December 31,
1992, but by December 31, 2002, credit unions of this size had around 75
percent of total industry assets. The 50 largest credit unions held 18
percent of industry assets in 1992, but by 2002 the 50 largest credit
unions held 23 percent of industry assets.

      Appendix III Financial Condition of Federally Insured Credit Unions

      Table 7: Distribution of Credit Unions by Asset Size, 1992 and 2002

Asset size (dollars in millions)

December Less than .5 to less 2 to less 10 to less 30 to less 50 to less
100 .5 than 2 than 10 than 30 than 50 than 100 or more Total

      Number of                                                       
    credit unions   1,696   2,818   4,304   2,121   625    519   512   12,595 
     Percent of                                                       
    credit unions   13.47   22.37   34.17   16.84   4.96  4.12  4.07  

Total assets
(dollars in
millions) $433,203 $3,243,850 $21,230,518 $37,355,589 $24,331,358
$36,133,301 $135,637,393 $258,365,211

Percent of
total assets 0.17 1.26 8.22 14.46 9.42 13.99 52.50

      Number of                                                        
    credit unions   620    1,327   3,022   2,121   801    751   1,046   9,688 
     Percent of                                                        
    credit unions   6.40   13.70   31.19   21.89   8.27   7.75  10.80  

Total assets
(dollars in
millions) $165,054 $1,543,306 $16,181,104 $37,913,707 $31,135,123
$52,762,245 $417,374,026 $557,074,565

Percent of
total assets 0.03 0.28 2.9 6.81 5.59 9.47 74.92

Source: Call report data.

As industry assets have increased, the composition of these assets has
changed. Total loans as a percentage of total assets increased from 54
percent as of December 31, 1992, to 62 percent as of December 31, 2002
(see table 8). While consumer loans, which broadly consist of unsecured
credit card loans, new and used vehicle loans, and certain other loans to
members, remained the largest category of credit union loans, the most
significant growth in credit union loan portfolios was in real estate
loans. These loans grew from 19 percent of total assets as of December 31,
1992, to 26 percent of total assets as of December 31, 2002.

      Appendix III Financial Condition of Federally Insured Credit Unions

  Table 8: Asset Composition of Credit Unions as a Percentage of Total Assets,
                                   1992-2002

                               Figures in percent

            Dec. Dec.  Dec.  Dec.  Dec.  Dec.  Dec.  Dec.  Dec.  Dec.31,  Dec. 
             31,  31,  31,   31,   31,   31,   31,   31,   31,             31, 
            1992  1993  1994  1995  1996  1997  1998  1999  2000    2001  2002 
    Cash    2.42  2.27  2.18  2.32  2.22  2.25  2.28  6.39  7.64   10.09  9.62 
  Consumer                                                               
loans   29.77 31.35 35.83 37.70 39.01 38.57 35.59 36.31 37.61   33.91 31.47 
    Real                                                                 
estate                                                                
loans   19.05 18.67 19.96 20.14 21.63 22.92 23.35 25.28 26.61   26.27 26.41 
Other    5.19  4.94  4.97  4.82  4.76  4.66  4.30  4.43  4.55    4.11  3.63 
loans                                                                 
Total                                                                       
loans   54.01 54.96 60.76 62.66 65.40 66.15 63.24 66.02 68.77   64.29 61.51

U.S.
government
and agency
securities 16.26 18.05 18.33 16.41 15.63 14.52 13.67 13.19 11.99 12.32
13.89

Investments in
corporate credit
unions 13.33 11.41 8.28 8.07 6.97 7.41 9.29 5.15 3.36 3.85 4.77

Bank and thrift
deposits 0.00 0.00 5.52 5.43 4.87 4.71 5.46 3.78 2.85 3.71 4.18

Other
investments 1.05 1.13 1.08 1.02 0.85 0.92 1.29 1.44 1.35 1.47 1.56

Fixed and other
assets 12.93 12.17 3.85 4.09 4.06 4.04 4.77 4.03 4.04 4.27 4.47

Total assets ($
in billions) $258.37 $277.13 $289.45 $306.64 $326.89 $351.17 $388.70
$411.42 $438.22 $501.54 $557.07

Source: Call report data.

      Appendix III Financial Condition of Federally Insured Credit Unions

Despite the growth in credit union real estate loans, credit unions had a
lower percentage of real estate loans to total assets (26 percent) than
their peer group banks and thrifts, which had 37 percent of real estate
loans to total assets (see table 9). Credit unions had a significantly
higher percentage of consumer loans to total assets (31 percent) compared
with their peer group banks and thrifts (8 percent). These banks and
thrifts, however, had a significantly higher percentage of agricultural
and commercial loans to total assets (12 percent) compared with credit
unions (slightly more than 1 percent).

 Table 9: Comparison of the Loan Portfolios of Federally Insured Credit Unions
                 with Peer Group Banks and Thrifts, as of 2002

Credit unions Banks

          Loan types            Dollar value Percent     Dollar value Percent 
        Consumer loans      $175,300,187,240  31.47  $189,841,654,000 
      Real estate loans      147,131,474,868  26.41   944,031,005,000   37.44 
       Agricultural and                                               
          commercial                                                  
            loans              6,644,982,024  1.19    303,205,739,000   12.03 
         Other loans          13,571,878,174  2.44     65,472,408,000 

Source: Call report data.

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions and banks and thrifts filing call reports. Insured
U.S. branches of foreign-chartered banks, banks with more than $18 billion
in assets, and banks we determined had emphases in credit card or mortgage
loans are excluded.

Credit Union Profits Have Been Relatively Stable in Recent Years

The profitability of credit unions, as measured by the return on average
assets, has been relatively stable in recent years. According to this
measure, credit union profitability was higher in the early to mid-1990s
than in the late 1990s and early 2000s. While declining from 1993 through
1999, the return on average assets has since stabilized. It has generally
hovered around 1 percent, which, by historical banking standards, is a
performance benchmark, and it was reported at 1.07 as of December 31, 2002
(see fig. 19). Profits are an especially important source of capital for
credit unions because they are mutually owned institutions that cannot
sell equity to raise capital.

      Appendix III Financial Condition of Federally Insured Credit Unions

     Figure 19: Profitability of Federally Insured Credit Unions, 1992-2002

                                 Percentage 1.6

                                      1.4

                                      1.2

                                      1.0

                                      0.8

                                      0.6

                                      0.4

                                      0.2

           0.0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

                           Source: Call report data.

Notes: Profitability is measured by the return on average assets, in which
average assets are the simple average of total assets as of the current
period and prior yearend. The return on average assets was not available
for 1992 since we did not have 1991 total assets data.

Credit Unions' Regulatory The number of credit unions with a CAMEL rating
of 1 (strong) increased Ratings Have Improved from 1,082 (9 percent) in
1992 to 2,186 (23 percent) in 2002 (see fig. 20). Since December 1992
During the same time period, institutions classified as problem credit

unions-those with CAMEL ratings of 4 (poor) or 5 (unsatisfactory)-

decreased from 578 (5 percent) in 1992 to 211 (2 percent) in 2002.

Appendix III Financial Condition of Federally Insured Credit Unions

Source: NCUA.

Appendix IV

Comparison of Bank and Credit Union Distribution of Assets

Figures 21, 22, and 23 illustrate the marked size disparity between credit
unions and institutions insured by the Federal Deposit Insurance
Corporation (FDIC), with figure 21 highlighting how small most credit
unions are.1 At the end of 2002, the largest credit union had less than
$18 billion in assets, while the largest bank, with over $600 billion in
assets, was larger than the entire credit union industry.

Figure 21: Total Assets of All Credit Unions and All Banks, as of 2002
10,000

8,000

6,000

4,000

2,000

0 $0.1 $0.25 $0.5 $1 $5 $10 $15 $20 $50 $100 $500 $750 Total assets (in
billions)

Source: Call report data.

Credit unions

Banks

Note: Data are as of December 31, 2002, and include all federally insured
credit unions and banks and thrifts filing call reports. Insured U.S.
branches of foreign-chartered institutions are excluded. This figure
depicts the number of institutions in a particular asset size category.
Each category represents a range-for example, the first category includes
all institutions with assets of $100 million or less, while the second
category includes all institutions with assets greater than $100 million
and less than or equal to $250 million, up to the last category, which
includes all institutions with assets greater than $500 million and less
than or equal to $750 billion.

1Throughout the report, we refer to institutions insured by the FDIC
interchangeably as "banks," "banks and thrifts," and "FDIC-insured
institutions."

                                  Appendix IV
                      Comparison of Bank and Credit Union
                             Distribution of Assets

                             Number of institutions

3,500

3,000

2,500

2,000

1,500

1,000 $5,000 $10,000

$15,000 $20,000 $25,000 $30,000 $35,000 $40,000 $45,000 $50,000 $55,000 $60,000
 $65,000 $70,000 $75,000 $80,000 $85,000 $90,000 $95,000 $100,000 Total assets
                                 (in thousands)

Source: Call report data.

Credit unions

Banks

Note: Data are as of December 31, 2002, and include all federally insured
credit unions and banks and thrifts filing call reports. Insured U.S.
branches of foreign-chartered institutions are excluded. This figure
depicts the number of institutions in a particular asset size category.
Each category represents a range-for example, the first category includes
all institutions with assets of $5 million or less, while the second
category includes all institutions with assets greater than $5 million and
less than or equal to $10 million, up to the last category, which includes
all institutions with assets greater than $95 million and less than or
equal to $100 million.

                                  Appendix IV
                      Comparison of Bank and Credit Union
                             Distribution of Assets

350

300

250

200

150

100

50

0 $250,000 $500,000

$750,000 $1,000,000 $1,250,000 $1,500,000 $1,750,000 $2,000,000 $2,250,000
  $2,500,000 $2,750,000 $3,000,000 $3,250,000 $3,500,000 $3,750,000 $4,000,000
                  $4,250,000 $4,500,000 $4,750,000 $5,000,000

Total assets

                           Source: Call report data.

Note: Data are as of December 31, 2002, and include all federally insured
credit unions filing call reports. This figure depicts the number of
institutions in a particular asset size category. Each category represents
a range-for example, the first category includes all institutions with
assets of $250,000 or less, while the second category includes all
institutions with assets greater than $250,000 and less than or equal to
$500,000, up to the last category, which includes all institutions with
assets greater than $4.75 million and less than or equal to $5 million.

Given the disproportionate size of the banking industry relative to the
credit union industry, peer groups were defined to mitigate the effects of
this discrepancy. Therefore, for our more detailed reviews, we constructed
five peer groups in terms of institution size as measured by total assets,
reported as of December 31, 2002. We further refined the sample of
FDIC-insured institutions to exclude those banks and thrifts we determined
had emphases in credit card or mortgage loans. The largest bank included
in our analyses had total assets of nearly $18 billion in 2002. See
appendix I for details.

Figures 24, 25, 26, and 27 illustrate that differences in services (as
measured by the number of institutions holding various consumer, mortgage,
and business loans) between credit unions and peer group banks

Appendix IV
Comparison of Bank and Credit Union
Distribution of Assets

are manifested in terms of institution size. Overall, the credit union
industry in aggregate did not appear to be that similar to the banking
industry (as captured by our sample of peer group banks) in terms of
services; however, when broken out by size, the larger credit unions
(those with more than $100 million in assets, or credit unions in Groups
II, III, IV, and V) appeared to be offering very similar services to peer
banks. Moreover, as nearly 90 percent of all credit unions had less than
$100 million in assets as of December 31, 2002, the results depicted in
Figure 24 are influenced more heavily by these institutions.

Figure 24: Percentage of All Credit Unions and All Banks Holding Various
Loans, as of 2002

Percentage of institutions

100 98 10098 98

80

60

40

20

0

First Junior mortgage Credit Other Agricultural mortgage and home card
consumer and business loans equity loans loans loans loans

Credit unions

Banks

Source: Call report data.

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions and banks and thrifts filing call reports. Insured
U.S. branches of foreign-chartered institutions and banks we determined
had emphases in credit card or mortgage loans are excluded. Bank data on
mortgages exclude thrifts. Credit union data on other consumer loans may
include member business and agricultural loans.

Appendix IV
Comparison of Bank and Credit Union
Distribution of Assets

Less Holding Various Loans, as of 2002 Percentage of institutions

                                      100

100

80

60

40

20

0

First Junior mortgage Credit Other Agricultural mortgage and home card
consumer and business loans equity loans loans loans loans

Credit unions

Banks

Source: Call report data.

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions and banks and thrifts filing call reports. Insured
U.S. branches of foreign-chartered institutions and banks we determined
had emphases in credit card or mortgage loans are excluded. Bank data on
mortgages exclude thrifts. Credit union data on other consumer loans may
include member business and agricultural loans.

Appendix IV
Comparison of Bank and Credit Union
Distribution of Assets

Percentage of institutions

100 100 100

100

80

60

40

20

0

First Junior mortgage Credit Other Agricultural mortgage and home card
consumer and business loans equity loans loans loans loans

Credit unions

Banks

Source: Call report data.

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions and banks and thrifts filing call reports. Insured
U.S. branches of foreign-chartered institutions and banks we determined
had emphases in credit card or mortgage loans are excluded. Bank data on
mortgages exclude thrifts. Credit union data on other consumer loans may
include member business and agricultural loans.

                                  Appendix IV
                      Comparison of Bank and Credit Union
                             Distribution of Assets

Source: Call report data.

Credit unions

Banks

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions and banks and thrifts filing call reports. Insured
U.S. branches of foreign-chartered institutions and banks we determined
had emphases in credit card or mortgage loans are excluded. Bank data on
mortgages exclude thrifts. Credit union data on other consumer loans may
include member business and agricultural loans. Group I credit unions had
assets of $100 million or less; Group II credit unions had assets greater
than $100 million and less than or equal to $250 million; Group III credit
unions had assets greater than $250 million and less than or equal to $500
million; Group IV credit unions had assets greater than $500 million and
less than or equal to $1 billion; and Group V credit unions had assets
greater than $1 billion and less than or equal to $18 billion, which is
the asset size, rounded up to the nearest billion dollars, of the largest
credit union as of December 31, 2002.

Appendix V

                        Credit Union Services, 1992-2002

In the absence of detailed time series data on the provision of services
by credit unions, we used holdings of various loans, including mortgage
and consumer loans, as well as other variables, as rough measures of
credit union services over time. We also separated credit unions by asset
size to illustrate any differences in provision of services by this
criterion. For illustrative purposes, we compared the smallest credit
unions (those with assets of $100 million or less) with the largest credit
unions (those with more than $1 billion in assets).

The percentage of all credit unions holding first mortgage loans has
increased every year since 1992 (see fig. 28). However, nearly twice as
many credit unions hold new and used vehicle loans as first mortgage
loans.

                                   Appendix V
                        Credit Union Services, 1992-2002

Source: Call report data.

Note: Data are as of December 31 and are based on all federally insured
credit unions filing call reports.

Appendix V
Credit Union Services, 1992-2002

Calculating the percentage of loan amounts held to total assets can reveal
the relative importance of each type of loan to credit unions. Figure 29
shows that first mortgage loans have increased in importance, surpassing
each of the other loan holdings.

                                   Appendix V
                        Credit Union Services, 1992-2002

Source: Call report data.

Note: Data are as of December 31 and are based on all federally insured
credit unions filing call reports.

Appendix V
Credit Union Services, 1992-2002

Although nearly all credit unions have offered regular shares (savings
accounts), over the years, the percentage of those offering share drafts
(checking accounts) and money market shares has increased, as illustrated
in figure 30.

Appendix V
Credit Union Services, 1992-2002

Source: Call report data.

Note: Data are as of December 31 and are based on all federally insured
credit unions filing call reports. Regular shares are savings accounts and
share drafts are checking accounts.

Appendix V
Credit Union Services, 1992-2002

The number of employees could have an effect on the provision of services
as well. Figure 31 shows that industry consolidation has not adversely
affected employment. Even though the industry shrank in terms of the
number of institutions from 12,595 in 1992 to 9,688 in 2002, a decline of
23 percent, the number of full-time employees went from 119,480 in 1992 to
180,401 in 2002, an increase of 51 percent.

Figure 31: Credit Union Employees and Number of Credit Unions, 1992-2002
Number of credit unions Full-time employees

15,000 200,000

175,000

12,000 150,000

9,000 125,000

100,000

6,000 75,000

50,000 3,000

25,000

00 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Number of credit unions

Full-time employees

Source: Call report data.

Note: Data are as of December 31 and are based on all federally insured
credit unions filing call reports.

The differences between the smallest credit unions (those with $100
million or less in assets) and the largest credit unions (those with more
than $1 billion in assets) are also apparent in the types of loans held
and their relative importance for each group over time (see figs. 32 and
33). Nearly all of the smallest credit unions have emphasized new and used
vehicle loans, but typically less than one-half of these credit unions
have held other loan types. As of December 31, 2002, used vehicle loans
were the relatively most important loan holding for the smallest credit
unions, surpassing new vehicle loans. Almost all of the largest credit
unions have

Appendix V
Credit Union Services, 1992-2002

held most types of loans over the past decade, with the exception of
member business loans-but the percentage of the largest credit unions
holding these has been steadily growing and, as of December 31, 2002,
roughly three out of four of these credit unions held them. First mortgage
loans have consistently been the most important loan holding of the
largest credit unions, and they now represent nearly one-quarter of the
asset mix of these credit unions.

                                   Appendix V
                        Credit Union Services, 1992-2002

Source: Call report data.

Note: Data are as of December 31 and are based on all federally insured
credit unions filing call reports. The smallest credit unions (Group I)
are those with $100 million or less in assets while the largest credit
unions (Group V) are those with more than $1 billion in assets.

                                   Appendix V
                        Credit Union Services, 1992-2002

Source: Call report data.

Note: Data are as of December 31 and are based on all federally insured
credit unions filing call reports. The smallest credit unions (Group I)
are those with $100 million or less in assets while the largest credit
unions (Group V) are those with more than $1 billion in assets.

Appendix V
Credit Union Services, 1992-2002

As of December 31, 2002, we observed a gap in services offered by smaller
credit unions and larger credit unions (see fig. 34). While larger credit
unions-those with assets of more than $100 million-accounted for just over
10 percent of all credit unions, they offered more services than smaller
credit unions. For example, nearly all of the larger credit unions held
mortgage loans and credit card loans, while only around one-half of the
smaller credit unions held these loans.

Figure 34: Differences among Services Offered by Smaller and Larger Credit

Unions, as of 2002 Percentage of institutions
100 98.2 98.7 99.9 100

80

60

40

20

0 First Junior mortgage Credit Other Agricultural mortgage and home card
consumer and business loans equity loans loans loans loans

Smaller credit unions

Larger credit unions

Source: Call report data.

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions filing call reports. In this figure, larger credit
unions are those with more than $100 million in assets while smaller
credit unions are those with $100 million or less in assets.

The discrepancy in the services offered by smaller and larger credit
unions is more accurately illustrated through an analysis of more recently
collected data on more sophisticated product and service offerings, such
as the availability of automatic teller machines (ATM) and electronic
banking (see fig. 35). While less than half of the smallest credit unions
offered ATMs

                                   Appendix V
                        Credit Union Services, 1992-2002

and one-third offered financial services through the Internet, nearly all
larger credit unions offered these services.

Source: Call report data.

Note: Data are as of December 31, 2002, and are based on all federally
insured credit unions filing call reports. Group I credit unions had
assets of $100 million or less; Group II credit unions had assets greater
than $100 million and less than or equal to $250 million; Group III credit
unions had assets greater than $250 million and less than or equal to $500
million; Group IV credit unions had assets greater than $500 million and
less than or equal to $1 billion; and Group V credit unions had assets
greater than $1 billion and less than or equal to $18 billion, which is
the asset size, rounded up to the nearest billion dollars, of the largest
credit union as of December 31, 2002.

Appendix VI

Characteristics of Credit Union and Bank Users

This appendix provides additional information on the characteristics-age,
education, and race/ethnicity-of households that use banks and credit
unions. For figures 36, 37, and 38, we analyzed data from the Federal
Reserve's 2001 Survey of Consumer Finances (SCF). The categories we used
to describe these households-credit union users and bank users- included
those who only and primarily used each of these institutions. To
supplement our analyses of households by race, we also analyzed 2001 loan
application records from the Home Mortgage Disclosure Act database (HMDA)
(see fig. 39). As we did with our analysis of HMDA income data, we only
analyzed records for home purchase loans actually made for the purchase of
one-to-four family homes.

30

25

20

15

10

5

0

Households only and Households only and primarily using credit unions
primarily using banks

Less than high school High school graduate Some college College/graduate
degree Source: 2001 SCF.

Appendix VI
Characteristics of Credit Union and Bank
Users

50

46

40

30

20

10

0 Households only and Households only and primarily using credit unions
primarily using banks

Ages 18-33
Ages 34-49
Ages 50-65
Age 66+
Source: 2001 SCF. Note: Percentages do not add to 100 percent due to
rounding.

Appendix VI
Characteristics of Credit Union and Bank
Users

79

80

70

60

50

40

30

20

10

0

Households only and Households only and primarily using credit unions
primarily using banks

Other Hispanic Black White Source: 2001 SCF. Note: Percentages do not add
to 100 percent due to rounding.

Appendix VI
Characteristics of Credit Union and Bank
Users

80 77

70

60

50

40

30

20

10

0 Other Black Hispanic White Not provided Race

Loans made by credit unions

Loans made by peer group banks

Source: 2001 HMDA database.

Notes: The "other" category includes data reported for American Indians,
Alaskan natives, Asian or Pacific islanders, and those from the HMDA
"other" category. We collapsed these categories to create groups similar
to the ones used by the SCF. However, in our HMDA analysis, we only
included mortgages made by peer group banks (banks with less than $16
billion in assets) whereas the SCF did not exclude households using banks
with more than $16 billion in assets.

Fifteen percent of the HMDA data reported by credit unions and 6 percent
of the HMDA data reported by banks lacked race and ethnicity data. As
such, the data in this figure may not represent the exact proportion of
mortgage loans by race. We also found that the proportion of loans without
data varied by the asset size of institutions. For example, race data were
missing for 23 percent of credit unions with assets of more than $500
million compared with about 3 percent for credit unions with less than $50
million in assets. Similarly, race data were missing for about 8 percent
of peer group banks with more than $500 million in assets compared with
about 4 percent of banks with less than $50 million in assets. However,
since these larger institutions made most of the loans, missing data from
these institutions account for more than 80 percent of all the missing
data.

Appendix VII

Key Changes in NCUA Rules and Regulations, 1992-2003

Since 1992, changes to the National Credit Union Administration's (NCUA)
rules and regulations governing credit unions generally expanded the
powers of credit unions to offer products and services, and broadened the
activities in which they could engage. With the exception of member
business lending, which NCUA constrained during the 1990s, federally
chartered credit unions gained authority to, among other things, (1)
invest in a wider variety of financial instruments, (2) offer services
through the Internet, and (3) profit from referring members to products,
such as insurance and investments, sold by third parties. Also, NCUA
increased the number of activities in which credit union service
organizations (CUSO) could engage, including student loan and business
loan origination. In September 2003, NCUA expanded credit union powers in
member business lending to permit well-capitalized credit unions to make
unsecured member business loans within certain limits, among other things.
See table 10 for a timeline of key changes to NCUA rules and regulations.

    Table 10: Timeline of Key Changes to NCUA Rules and Regulations, January
                              1992-September 2003

                           Effective date Key change

January 1992	NCUA limited member business loans in response to losses to
credit unions, their members, and the National Credit Union Share
Insurance Fund. NCUA established loan security requirements, limits on
loans to one borrower, and an aggregate portfolio cap on construction and
development loans at 15 percent of reserves for federally insured credit
unions.

September 1996	NCUA allowed credit unions serving predominantly low-income
members to raise secondary capital from foundations and other
philanthropic-minded institutional investors, to help credit unions make

a

more loans, and improve services to low-income members. NCUA required
credit unions to establish certain uninsured or other form of nonshare
accounts for secondary capital.

January 1998	NCUA codified additional powers of federally chartered credit
unions to act as trustees and custodians of Roth Individual Retirement
Accounts (IRA) and Education IRAs, which is in addition to those trustee
and custodian services they had been authorized to provide for other kinds
of pension and retirement plans for approximately the previous 23 years.

NCUA changed its investment rule to focus on risk management (previous
focus was on specific financial instruments for federal credit unions).
NCUA established new requirements for assessing and managing risk
associated with federally chartered credit union investment activities.

April 1998	NCUA codified additional preapproved CUSO activities to include
student loan origination, disaster recovery services, additional checking
and currency services, and electronic income tax filing services, among
others.

August 1998	Credit Union Membership Access Act (CUMAA) became law. CUMAA
provisions cap the aggregate portfolio amount of member business loans for
federally insured credit unions, with exceptions.

March 2000	NCUA allowed federally chartered credit unions in specified
locations outside the United States to offer trustee or custodian services
for IRAs.

                                  Appendix VII
                   Key Changes in NCUA Rules and Regulations,
                                   1992-2003

(Continued From Previous Page)

                           Effective date Key change

August 2001	NCUA issued legal opinion that permitted a federally chartered
credit union employee to be a shared employee with a third party and,
while acting in the capacity of an employee of the third party, to sell
nondeposit investment products and provide investment advice. NCUA
continued to restrict federally chartered credit union employees, acting
as an employee of the credit union, from selling nondeposit investment
products or providing investment advice.

September 2001	NCUA's Incidental Powers Regulation became effective. This
rule codified a broad range of activities, products, and services that
federally chartered credit unions could offer directly to members, and
which NCUA had previously recognized in legal opinions or had recognized
in other regulations. One change, which permits federally chartered credit
unions to earn income directly from finder activities (the referral of
members to outside vendors, such as investment and insurance brokers), had
the effect of making it unnecessary to use a CUSO in third-party
networking arrangements in order to receive income. Key powers codified in
the regulation include: electronic financial services, finder activities,
loan-related products, such as debt suspension agreements, and trustee
services.b There is overlap of the activities in which federally chartered
credit unions and CUSOs may engage (for example, consumer mortgage
origination), but there are also activities only permissible for CUSOs
(for example, general trust services and travel agency services).

February 2002	NCUA issued a legal opinion on how federally chartered
credit unions can provide nonmembers, such as agricultural workers with
familial ties to foreign countries, with wire transfer services. While
expressly restricting unlimited services to nonmembers, NCUA permitted
federally chartered credit unions to (1) establish nondividend-bearing
accounts for people within its field of membership, (2) provide wire
transfer services as a promotional activity on a limited basis, and (3)
provide services as a charitable activity, so long as the recipients of
the charitable services were within the credit union's field of
membership.

March 2002	NCUA's Regulatory Flexibility Program became effective. NCUA
relieved eligible federally and state-chartered credit unions from certain
NCUA regulations relating to permissible investments and investment
management requirements, limits on share deposits from public entities and
nonmembers, approval processes for charitable contributions, and limits on
ownership of fixed assets.

July 2003	NCUA expanded investment powers of certain federally chartered
credit unions to allow them to purchase financial instruments that were
previously prohibited, including commercial mortgage-

                                       c

related securities and equity options.

NCUA permitted federally insured credit unions to open branches in foreign
countries, with conditions.

September 2003 NCUA amended its CUSO rule to permit CUSOs to originate business
                                     loans.

NCUA amended its member business loan rule to allow eligible federally
insured credit unions to make unsecured member business loans, with
limits, and to permit the exclusion of purchased nonmember loans and
nonmember participation interests from the aggregate business loan limit,
among other things.

Sources: GAO, NCUA, Federal Register.

Note:

aSecondary capital can take the form of investments into an institution by
nonmembers, such as foundations, corporations, and other financial
institutions. The investments are subordinated to all other credit union
debt, and are used to absorb losses.

bDebt suspension agreements are contracts between a lender and a borrower
where the lender agrees to suspend scheduled installment payments for an
agreed period in the event the borrower experiences financial hardship.

cEquity options are limited to those that would be purchased for the sole
purpose of offering dividends based on the performance of an equity index.

Appendix VIII

                    NCUA's Budget Process and Industry Role

The National Credit Union Administration (NCUA) changed its budget process
in 2001 to allow outside parties, including credit unions and trade
organizations, to submit comments on the budget. While outside parties can
submit their budget suggestions and concerns at any time, NCUA has a
formal budget briefing where these parties can officially submit their
comments. This briefing takes place at the latter stage of NCUA's budget
process. The changes NCUA has made to its budget process come during a
period in which NCUA has been reducing the growth in its budgets.

NCUA has two main sources of funding for its operating costs. According to
NCUA, 62 percent of the funds for operating costs in their 2002 budget
came from the National Credit Union Share Insurance Fund (NCUSIF),
administered by NCUA. NCUSIF is principally financed from earnings
(income) on investments purchased using the deposits of federally insured
credit unions. Funds are transferred from the insurance fund through a
monthly accounting procedure known as the overhead transfer to cover costs
associated with ensuring that insured deposits are safe and sound. The
remaining 38 percent of NCUA's funds for its operating costs came
primarily from operating fees assessed on federally chartered credit
unions, for which NCUA has oversight responsibility.

NCUA Budget Process Now Includes Step for Outside Parties to Submit
Comments

NCUA budgets on a calendar-year basis, and its board sets the policies and
overall direction for the budget. In July and August prior to the next
budget year, the NCUA regional offices submit their workload and program
needs. NCUA's examination and insurance officials in headquarters assess
the information and formulate proposed program hours, which along with
historical actual expenditures are the basis for the proposed budget. In
September and October, the Chief Financial Officer (CFO) reviews and
analyzes the figures, conducts briefings with office directors, and makes
adjustments. In November, NCUA holds a public briefing where interested
parties, including credit unions and trade associations, have the
opportunity to comment. Later in November, the CFO briefs the board prior
to final budget adjustments. Additionally, in July of the budget year,
there is a midyear budget review to determine if any adjustments need to
be made to the budget. According to NCUA officials, NCUA also conducts a
variance analysis on the budget on a monthly basis and a more
comprehensive review at the end of the year.

According to NCUA, credit unions and other stakeholders can submit their
budget suggestions and concerns at any time. Normally, suggestions come
between August and November while NCUA is working on the budget. For

             Appendix VIII NCUA's Budget Process and Industry Role

the public budget hearing, credit unions can address the board for 5
minutes or submit a written document.

Recent budget concerns by credit unions have centered on lessening the
costs to credit unions for NCUA oversight. Credit unions have raised
specific concerns about the number of NCUA staff or full-time equivalents,
the salaries of NCUA staff, and the overhead transfer rate from the
insurance fund. According to NCUA data, its average full-time equivalent
cost is less than that of the Federal Deposit Insurance Corporation (FDIC)
and the Office of the Comptroller of the Currency (OCC) and equal to that
of the Office of Thrift Supervision (OTS). Nevertheless, NCUA has
responded to concerns over its salary levels by deciding to undertake a
pay study.

NCUA Has Reduced Its Budget Growth in Recent Years

In recent years, NCUA has been successful in slowing its budget growth.
After 10-percent annual growth from 1998 to 2000, NCUA budget growth has
decreased to an average of about 3 percent in 2000-2003 (see fig. 40). The
NCUA board's budget priorities have been to streamline business processes,
increase efficiencies, control budget growth, and match resources to
mission requirements, while maintaining effective examination processes
and products. NCUA is seeking budget savings by adopting a risk-focused
examination approach, extending the examination cycle, adopting more
flexible rules and regulations, increasing efficiencies from technology
(such as videoconferencing), and consolidating two of their regions into
one.

Appendix VIII NCUA's Budget Process and Industry Role

Figure 40: NCUA Budget Levels, 1992-2004 Percentage change Dollars in
millions

12 150

10 120

8 90

6

60

4

30 2

00

-2

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
(projected)

Percentage change

Total dollars

Source: NCUA.

Note: The 2004 projected budget is expected to increase between 4.0 and
4.5 percent from the 2003 budget level.

NCUA's authorized full-time equivalent staff level decreased over 7
percent from 1,049 in 2000 to 971 in 2003 (see fig. 41). This level of
staff reductions has been partly in response to changes in the industry.
Since 1998, the number of federally insured credit unions has decreased
steadily by about 3 percent per year.

Appendix VIII NCUA's Budget Process and Industry Role

1,200

1,042 1,049 1,029

1,000

800

600

400

200

0
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Source: NCUA.

Appendix IX

NCUA's Implementation of Prompt Corrective Action

Section 301 of the Credit Union Membership Access Act (CUMAA) amended the
Federal Credit Union Act to require the National Credit Union
Administration (NCUA) to adopt a system of prompt corrective action (PCA)
for use on credit unions experiencing capitalization problems.1 The goal
of requiring PCA is to resolve the problems of insured credit unions with
the least possible long-term loss to the National Credit Union Share
Insurance Fund (NCUSIF). In that regard, NCUA was required to prescribe a
system of PCA consisting of three principal components: (1) a
comprehensive framework of mandatory supervisory actions and discretionary
supervisory actions, (2) an alternative system of PCA for "new" credit
unions, and (3) a risk-based net worth (RBNW) requirement for "complex"
credit unions.2 Furthermore, section 301 also required NCUA to report to
Congress on how PCA was implemented and how PCA for credit unions differs
from PCA for other depository institutions. NCUA submitted this report in
May 2000. In addition, NCUA submitted a further report to Congress that
described how NCUA carried out the RBNW requirements for credit unions and
how these requirements differed from RBNW requirements of other depository
institutions (see table 11).

1Pub. L. No. 105-219 (Aug. 7, 1998).

2CUMAA defines a "new" credit union as one that has been in operation for
less than 10 years and having less than $10 million in assets. 12 C.F.R.
S:702.2(h). NCUA defines a credit union as "complex" when its total assets
at the end of a quarter exceed $10 million and its RBNW calculation
exceeds 6 percent net worth. 12 C.F.R. S:702.103.

Appendix IX NCUA's Implementation of Prompt Corrective Action

Table 11: CUMAA Mandates and NCUA Actions on PCA Regulation Implementation

CUMAA CUMAA mandates to NCUA deadlines NCUA action dates PCA actions:

Issue PCA proposed rule May 1999 Issued May 1999

Issue the PCA final rule February 2000 Issued February 2000

Issue PCA report to Congress February 2000 Issued May 2000

              Implement PCA August 2000 Implemented August 2000 a

                           RBNW requirements actions:

Issue RBNW requirements February 1999 Issued October 1998
(Advance Notice of Proposed
Rulemaking)

c

Issue RBNW requirements Issued February 2000 proposed ruleb

Issue RBNW requirements final August 2000 Issued July 2000 rule

c

Issue RBNW requirements report Issued November 2000 to Congress b

Implement RBNW requirements January 2001 Implemented January 2001 final
rule

Sources: Federal Register 64, no. 95 (18 May 1999): 27090; Federal
Register 65, no. 34 (18 February 2000): 8560; Federal Register 63, no. 209
(29 October 1998): 57938; Federal Register 65, no. 34 (18 February 2000):
8597; Federal Register 65, no. 140 (20 July 2000): 44950; and NCUA reports
to Congress.

Note:

aThe PCA final rule applied to credit unions beginning in the fourth
quarter of 2000.

bCUMAA did not set any deadline for NCUA to issue the RBNW requirement
proposed rule and did not require NCUA to issue a RBNW report to Congress.

cNot mandated by CUMAA.

After NCUA implemented the initial PCA and RBNW regulations, it formed a
PCA Oversight Task Force to review at least a full year of PCA
implementation and recommend necessary modifications.3 The task force
reviewed the first six quarters of PCA implementation. It made several
recommendations to improve PCA, including revising definitions of terms
and clarifying implementation issues. In June 2002, NCUA issued a proposed
rule setting forth revisions and adjustments to improve and simplify PCA.
In November 2002, after incorporating public comments on the proposed
rule, NCUA issued the final PCA rule adopting the proposed

3NCUA established a PCA Oversight Task Force in February 2000. This task
force consisted of NCUA staff and state regulators. See Federal Register
65, no. 140 (20 July 2000): 44964.

         Appendix IX NCUA's Implementation of Prompt Corrective Action

revisions and adjustments.4 The final rule became effective on January 1,
2003.

PCA Incorporates a Comprehensive Framework of Mandatory and Discretionary
Supervisory Actions

The PCA rule consists of a comprehensive framework of mandatory and
discretionary supervisory actions for all federally insured credit unions
except "new" credit unions.5 The PCA system includes the following five
statutory categories and their associated net worth ratios:

o  well-capitalized-7.0 percent or greater net worth,

o  adequately capitalized-6.0 to 6.99 percent net worth,

o  undercapitalized-4.0 to 5.99 percent net worth,

o  significantly undercapitalized-2.0 to 3.99 percent net worth, and

o  critically undercapitalized-less than 2.0 percent net worth.

As noted earlier in the report, mandatory supervisory actions apply to
credit unions that are classified adequately capitalized or lower. The PCA
system also includes conditions triggering mandatory conservatorship and
liquidation.

CUMAA also authorized NCUA to develop a comprehensive series of
discretionary supervisory actions to complement the mandatory supervisory
actions. Some or all of these 14 discretionary supervisory actions can be
applied to credit unions that are classified undercapitalized or lower
(see table 12).

4The final PCA rule contains 17 revisions and adjustments. See Federal
Register 67, no. 230 (29 November 2002): 71078.

5NCUA issued staff instructions on discretionary supervisory actions in
April 2003, but has yet to impose a discretionary supervisory action
against any credit union.

         Appendix IX NCUA's Implementation of Prompt Corrective Action

 Table 12: Discretionary Supervisory Actions Discretionary supervisory actions
                          Statutory net worth category

Require NCUA prior approval for acquisitions, "Undercapitalized" and lower
branching, new lines of business

 Restrict transactions with and ownership of CUSOs "Undercapitalized" and lower

              Restrict dividends paid "Undercapitalized" and lower

          Prohibit or reduce asset growth "Undercapitalized" and lower

Alter, reduce, or terminate any activity by credit union
"Undercapitalized" and lower or its CUSO

            Prohibit nonmember deposits "Undercapitalized" and lower

 Other actions to further the purpose of part 702 "Undercapitalized" and lower

     Order new election of board of directors "Undercapitalized" and lower

  Dismiss directors or senior executive officers "Undercapitalized" and lower

    Employ qualified senior executive officers "Undercapitalized" and lower

Restrict senior executive officers' compensation and bonus

"Significantly Undercapitalized" and lower

     Require merger if grounds exist for conservatorship or "Significantly
                    Undercapitalized" liquidation and lower

Restrict payments on uninsured secondary capital "Critically
Undercapitalized"

Require NCUA prior approval for certain actions "Critically
Undercapitalized"

Source: Federal Register 64, no. 95 (18 May 1999): 27096-27098.

The discretionary supervisory actions are tailored to suit the distinctive
characteristics of credit unions.

An Alternative System for New Credit Unions

CUMAA required NCUA to develop an alternative PCA system for "new" credit
unions. In doing so, NCUA recognized that new credit unions (1) initially
have no net worth, (2) need reasonable time to accumulate net worth, and
(3) need incentives to become adequately capitalized by the time they are
no longer new. Accordingly, the PCA system for new credit unions has
relaxed net worth ratios, allows regulatory forbearance, and offers
incentives to build net worth. The PCA system for new credit unions
includes six net worth categories and their associated net worth ratios
(see table 13).

         Appendix IX NCUA's Implementation of Prompt Corrective Action

Table 13: Net Worth Category Classification for New Credit Unions

         New credit union net worth category Net worth ratio (Percent)

                        "Well-Capitalized" 7.0 or above

                      "Adequately Capitalized" 6.0 to 6.99

                      "Moderately Capitalized" 3.5 to 5.99

                      "Marginally Capitalized" 2.0 to 3.49

                      "Minimally Capitalized" 0.0 to 1.99

                          "Uncapitalized" Less than 0

           Source: Federal Register 64, no. 95 (18 May 1999): 27099.

Risk-based Net Worth CUMAA also required NCUA to formulate the definition
of a "complex" Requirement for "Complex" credit union according to the
risk level of its portfolios of assets and Credit Unions liabilities.
Well-capitalized and adequately capitalized credit unions

classified as complex are subject to an additional RBNW requirement to
compensate for material risks against which a 6.0 percent net worth ratio
may not provide adequate protection. (We describe the RBNW requirement in
more detail elsewhere in this appendix.)

NCUA Submitted Required PCA Report to Congress

CUMAA mandated that NCUA submit a report to Congress addressing PCA. The
report, dated May 22, 2000, explains how the new PCA rules account for the
cooperative character of credit unions and how the PCA rules differ from
the Federal Deposit Insurance Act's (FDIA) "discretionary safeguards" for
other depository institutions as well as the reasons for the differences.

The report discusses how the PCA rules account for credit unions'
cooperative character in three areas: their not-for-profit nature, their
inability to issue stock, and their board of directors consisting
primarily of volunteers.6 First, the final rule accounts for credit
unions' not-for-profit nature by permitting a less-than-well-capitalized
credit union to seek a reduction in the statutory earnings retention
requirement to allow the continued payment of dividends sufficient to
discourage an outflow of shares. In addition, a well-capitalized credit
union whose earnings are

6Credit unions cannot issue capital stock and, therefore, must rely on
retained earnings to build net worth.

Appendix IX NCUA's Implementation of Prompt Corrective Action

depleted may be permitted to pay dividends from its regular reserve
provided that such payment would not cause the credit union to fall below
the adequately capitalized level. Secondly, to account for the inability
of credit unions to issue capital stock, the final rule relies on the Net
Worth Restoration Plan, which must be submitted by credit unions
classified as undercapitalized or lower. Finally, to recognize that credit
unions' boards of directors consist primarily of volunteers, the rule
exempts credit unions that are near to being adequately capitalized from
the discretionary supervisory action authorizing NCUA to order a new
election of the board of directors.

NCUA reported that the final rule established discretionary supervisory
actions that are essentially comparable to section 38 of FDIA, which
specifies "discretionary safeguards" for other depository institutions.
The report notes that NCUA adopted discretionary supervisory actions that
are similar to all but two of FDIA's 14 discretionary safeguards.

NCUA did not adopt FDIA's safeguards requiring selling new shares of stock
and prior approval of capital distributions by a bank holding company.
NCUA's rationale for these exclusions was that, unlike banks, credit
unions cannot sell stock to raise capital and are not controlled by
holding companies.

NCUA departed from FDIA discretionary safeguards in fashioning three of
the discretionary supervisory actions: (1) dismissals of senior officers
or directors, (2) exemption of officers from discretionary supervisory
actions, and (3) ordering a new election of the boards of directors. NCUA
reported that the discretionary supervisory action for director dismissals
departs significantly from its FDIA counterpart. The FDIA safeguard
protects from dismissal of officials with office tenures of 180 days or
less, when an institution becomes undercapitalized. In contrast, NCUA
contends that such a "safe harbor" is unnecessary for credit unions.
Moreover, NCUA field experience supports the view that short-tenured
officers can be as responsible as others for rapidly declining net worth.

With regard to exempting officers from discretionary supervisory actions,
NCUA provides conditional relief to credit unions in contrast to the FDIA.
For example, the report notes that FDIA allows 11 discretionary safeguards
to be imposed on undercapitalized institutions. On the other hand, NCUA's
comparable discretionary supervisory actions can be imposed against
undercapitalized credit unions in the first tier of that category only
when they fail to comply with any of CUMAA's four mandatory supervisory

         Appendix IX NCUA's Implementation of Prompt Corrective Action

actions or fail to implement an approved Net Worth Restoration Plan.7
NCUA's rationale for granting relief from the relevant discretionary
supervisory actions is to avoid treating credit unions that are just short
of adequately capitalized as harshly as those that are almost
significantly undercapitalized.

NCUA's report states that it modified the discretionary supervisory action
ordering a new election of the board of directors. Specifically, NCUA
excludes undercapitalized credit unions from this requirement but applies
it to significantly undercapitalized and critically undercapitalized
credit unions. NCUA's exception was based on the belief that the safeguard
would undermine a defining characteristic of credit unions-membership
election of directors-and possibly discourage members from volunteering to
serve as directors. Moreover, NCUA noted that its discretionary
supervisory action does not compel a credit union to replace its board
with a NCUA-designated slate; it simply requires the membership to
reconsider its original choice of directors. Finally, the report states
that ordering a wholesale election of the board of directors may be an
overreaction when a credit union's net worth is within reach of becoming
adequately capitalized.

NCUA Submitted RBNW Report to Congress

NCUA submitted a report to Congress addressing its RBNW provisions on
November 3, 2000. In general, the report describes NCUA's comprehensive
approach to evaluating a credit union's individual risk exposure. It
explains the RBNW requirement that applies to complex credit unions. The
RBNW requirement takes into account whether credit unions classified as
adequately capitalized provide adequate protection against risks posed by
contingent liabilities, among other risks. According to the RBNW report,
NCUA's approach (1) targets credit unions that carry an above-average
level of exposure to material risk, (2) allows an alternative method to
calculate the amount of net worth needed to remain adequately capitalized
or well-capitalized, and (3) makes available a risk mitigation credit to
reflect quantitative evidence of risk mitigation.

NCUA reported that its final rule targets credit unions that have higher
material risk levels, thus warranting an extra measure of capital to
protect them and NCUSIF from losses. As noted previously, credit unions do
not

7The net worth ratio of credit unions in the undercapitalized category is
4.0-5.99 percent. The first tier of the undercapitalized net worth
category is 5.0-5.99 percent, and the second tier of that net worth
category is 4.0-4.99 percent.

Appendix IX NCUA's Implementation of Prompt Corrective Action

issue stocks that create shareholder equity. Without shareholder equity to
absorb losses, the RBNW requirement serves to mitigate most forms of risk
in a complex credit union's portfolio. Specifically, the RBNW measures the
risk level of on- and off-balance sheet items in the credit union's "risk
portfolios."8 The requirement applies only if a credit union's total
assets at the end of a quarter exceed $10 million, and its RBNW
requirement under the standard calculation exceeds 6 percent. The $10
million asset floor eliminates the burden on credit unions that are
unlikely to impose a material risk.9

NCUA uses two methods to determine whether a complex credit union meets
its RBNW requirement. Under the "standard calculation," each of eight risk
portfolios is multiplied by one or more corresponding risk weightings to
produce eight "standard components."10 The sum of the eight standard
components yields the RBNW requirement that the credit union's net worth
ratio must meet for it to remain either adequately capitalized or
well-capitalized. If the RBNW requirement is not met, the credit union
falls into the undercapitalized net worth category. NCUA allows a credit
union that does not meet its RBNW requirement under the standard
calculation to substitute for any of the three standard components, a
corresponding "alternative component" that may reduce the RBNW
requirement. The alternative components recognize finer increments of risk
in real estate loans, member business loans, and investments.

Finally, in reporting on the RBNW requirement, NCUA recognized that credit
unions, which failed under the standard calculation and with the
alternative components, nonetheless might individually be able to mitigate
material risk. In such instances, a risk mitigation credit is available to
credit unions that succeed in demonstrating mitigation of interest rate or

8The RBNW report notes that the "risk portfolios" of balance sheet assets
consist of long-term real estate loans, member business loans outstanding,
investments, low-risk assets, and average-risk assets. The "risk
portfolios" of off-balance sheet assets are loans sold with recourse and
unused member business loan commitments.

9According to the report, the principal banking industry trade association
advocated $10 million as an appropriate minimum asset "floor."

10Risk portfolios include real estate loans, member business loans (MBL)
outstanding, investments, low-risk assets, average-risk assets, loans sold
with recourse, unused MBL commitments, and allowances. See Federal
Register 65, no. 34 (18 February 2000): 8606.

Appendix IX NCUA's Implementation of Prompt Corrective Action

credit risk.11 If approved, a risk mitigation credit will reduce the RBNW
requirement a credit union must satisfy to remain classified as adequately
capitalized or above.

11According to NCUA data, as of May 2003, no credit union failed to meet
an RBNW requirement under the standard calculation and with the
alternative component, and so none has applied for a risk mitigation
credit to date.

Appendix X

                         Accounting for Share Insurance

The National Credit Union Share Insurance Fund (NCUSIF) capitalizes its
insurance fund differently than the Federal Deposit Insurance Corporation
(FDIC) capitalizes the Bank Insurance Fund (BIF) and the Savings
Association Insurance Fund (SAIF). For NCUSIF, a cash deposit in the fund
equal to 1 percent of insured shares, adjusted at least annually, must
remain on deposit with the fund for the period a credit union remains
federally insured. This deposit is treated as an asset on the credit
union's financial statements, and as part of equity on NCUSIF's financial
statements in an account entitled "Insured credit unions' accumulated
contributions." If a credit union leaves federal insurance, for example to
become privately insured, the deposit with NCUSIF is refunded. However, if
the National Credit Union Administration's (NCUA) board assesses
additional premiums in order to maintain the minimum required equity
ratio, the premiums are treated as an operating expense on the credit
unions' financial statements and would not be refunded. Since 2000, NCUA
has not made any distributions to contributing credit unions because the
fund did not exceed the NCUA board's specific operating level. And,
between 1990 and 2002, federally insured credit unions were assessed
premiums only in 1991 and 1992, when the fund's equity declined below the
mandated minimum 1evel ofnormal o 1.20 pp eerating l rcent of insured
shares.

However, unlike federally insured credit unions, federally insured banks
and thrifts operate exclusively under a premium-based insurance system.
This system requires banks and thrifts to remit a premium payment of a
specified percent of their balance of insured deposits twice a year to
FDIC to obtain federal deposit insurance. Each bank or thrift treats the
premium as an expense in its financial statements, while FDIC recognizes
the premium as income in its financial statements. If a bank or thrift
elects to not continue its federal deposit insurance, its premiums are,
unlike the NCUSIF insurance deposit, nonrefundable.

The Federal Deposit Insurance Corporation Improvement Act (FDICIA),
enacted in December 1991, contained some important provisions including
risk-based premiums for BIF and SAIF. FDIC developed and then implemented
the risk-based premium system on January 1, 1993. Under the system,
institutions were categorized according to a capital subgroup (1, 2,

1Federal Credit Union Act.

Appendix X Accounting for Share Insurance

or 3) and a supervisory subgroup (A, B, or C).2 This resulted in the
best-rated institutions being categorized as 1-A and the worst
institutions as 3-C. These categorizations result in a range of premium
costs, with the best-rated institutions paying the lowest premium and the
worst-rated institutions paying the highest premium.

In August 2000, FDIC issued a report that discussed the current deposit
insurance system, including the existence of two separate funds, an
insurance pricing system that may provide inappropriate incentives for
risk and growth, and issues of fairness and equitable insurance coverage,
and offered possible solutions. The report warned that this system might
require banks to fund insurance losses when they can least afford it.
Solutions offered in the report included (1) merging BIF and SAIF, (2)
improving the pricing of insurance premiums through a number of options,
and (3) setting a "soft" target for the reserve ratio, which would allow
the deposit insurance fund balances to grow during favorable economic
periods, thereby smoothing premium costs over a longer period of time. As
a result of FDIC's report, legislation is pending that may provide
additional reforms of the deposit insurance system, including pricing of
insurance.

As did BIF and SAIF, American Share Insurance (ASI), the private primary
share insurer, adopted a form of risk-based insurance plan at the end of
2000. As does NCUSIF, ASI's member credit unions pay a deposit rather than
an annual premium assessment to purchase their insurance coverage. Prior
to December 31, 2000, all of ASI's insured credit unions were required to
maintain a deposit of 1.3 percent of each member's total insured share
amounts, compared with 1.0 percent that federally insured credit unions
maintain with NCUSIF. With its change to a risk-based system, ASI's
insurance coverage now requires a range-a minimum deposit of 1.0

2The capital subgroup is assigned on the basis of the institution's total
risk-based capital ratio, tier 1 risk-based capital ratio, and tier 1
leverage capital ratio. The institutions report this data quarterly to
FDIC on their Report of Income and Condition (call report). For instance,
according to FDIC Risk-Based Assessment System - Overview, Group 1
("Well-Capitalized") has a "Total Risk-Based Capital Ratio equal to or
greater than 10 percent, and Tier 1 Risk-Based Capital Ratio equal to or
greater than 6 percent, and Tier 1 Leverage Capital Ratio equal to or
greater than 5 percent." Each semiannual period, FDIC assigns the
supervisory subgroup based on various factors including results of the
most recent examination report, the amount of time since the last
examination, and statistical analysis of call report data. For example,
according to the FDIC's Risk-Based Assessment System-Overview, a subgroup
A institution is "financially sound institution with only a few minor
weaknesses and generally corresponds to the primary federal regulator's
composite rating of `1'or `2'."

Appendix X Accounting for Share Insurance

percent up to a maximum of 1.3 percent for each credit union depending on
the credit union's CAMEL rating.3

The FDIC study of risk-based pricing indicated that one of the negative
aspects of not pricing to risk is that new institutions and fast-growing
institutions are benefiting at the expense of their older and
slower-growing competitors. Rapid deposit growth lowers a fund's equity
ratio and increases the probability that additional failures will push a
fund's equity ratio below the minimum requirements, resulting in a rapid
increase in premiums for all institutions.

3Credit unions are rated on their condition by NCUA and state regulators
using a "CAMEL" system that evaluates their capital adequacy (C), asset
quality (A), management (M), earnings (E), liquidity (L), and their
overall condition.

Appendix XI

Comments from the National Credit Union Administration

Appendix XI
Comments from the National Credit Union
Administration

Appendix XI
Comments from the National Credit Union
Administration

Appendix XI
Comments from the National Credit Union
Administration

Appendix XI
Comments from the National Credit Union
Administration

Appendix XI
Comments from the National Credit Union
Administration

                                  Appendix XII

                     Comments from American Share Insurance

Appendix XII
Comments from American Share Insurance

Mr. Richard J. Hillman
October 14, 2003
Page 2

B. 	ASI has limited ability to absorb large (catastrophic) losses because
it does not have the backing of any government entity.

In its 29-year history, ASI has paid over 110 claims on failed credit
unions, and more importantly, no member of a privately insured credit
union has ever lost money in an ASI-insured account. Also, ASI's statutory
ability to reassess its member credit unions provides a significant amount
of committed equity for catastrophic losses. Further, the company employs
numerous programs to mitigate the risk of large losses and field examines
more than 60% of its insured risk annually. Therefore, a sound private
deposit insurance program, built upon a solid foundation of careful
underwriting, continuous risk management and the financial backing of its
mutual member credit unions, can absorb large (catastrophic) losses.

With regard to the government backing, the GAO fails to consider that ASI
is a private business, licensed at the state level; owned by the credit
unions it insures; and, managed by a board of directors elected by such
member credit unions. Private share insurance was never intended to have
any state or federal guarantees.

C. ASI's lines of credit are limited in the aggregate as to amount and
available collateral.

The Study Section erroneously views the company's lines of credit as a
source of capital, when they are solely in place to provide emergency
liquidity. Proportionately, ASI's committed lines of credit with third
parties, as a percentage of fund assets, are greater than that of the
federal share insurer. Comparisons throughout the Study Section are often
provided on an absolute basis, not a proportionate basis, which we believe
skews many of the results included in the Study Section.

D. 	Many privately insured credit unions have failed to make required
consumer disclosures about the absence of federal insurance of member
accounts as required under the Federal Deposit Insurance Corporation
Improvement Act of 1991 (FDICIA), and the Federal Trade Commission (FTC)
is the appropriate federal agency to enforce such compliance.

FDICIA was passed in December 1991, and not long thereafter, the FTC
sought and received an exemption from Congress from enforcing the consumer
disclosure provisions of FDICIA. We concur with the Study Section's
observations in this regard, and believe privately insured credit unions
would benefit from FTC's enforcement of such provisions.

Detailed comments supporting and supplementing our above comments are
attached as Exhibit A.

DENNIS R. ADAMS President/CEO

DRA/krb

Attachment

Appendix XII
Comments from American Share Insurance

EXHIBIT A

DETAILED COMMENTS ON THE GAO'S DRAFT STUDY OF PRIVATE SHARE INSURANCE

A Component of the GAO's Study Titled: Credit Unions: Financial Condition
Has Improved But Opportunities Exist

To Enhance Oversight and Share Insurance Management

                                 Submitted By:
                            American Share Insurance
                                October 14, 2003

A. ASI's risks are concentrated in a few large credit unions and in
certain states.

All businesses face some degree of concentration risk. For example, 55% of
all federally insured shares are on deposit at only 230 NCUSIF-insured
credit unions -- this represents less than 3% of all federally insured
credit unions nationally. Despite this natural phenomena, the GAO proceeds
to raise concern over ASI's risk distribution.

Geographic Risk

The Study Section states that compared to federally insured credit unions,
"...relatively few credit unions are privately insured." As of December
31, 2002, about 2% of all credit unions are privately insured. ASI is
currently authorized in nine states and insuring credit unions in eight
nationally, and is limited to insuring only state-chartered credit unions
in those states in which the company is authorized to do business. In its
current states of operation, the company insures 212 credit unions,
comprising $10.8 billion in insured shares. What the Study Section fails
to report is that these credit unions represent 19% of all 1,095
state-chartered credit unions within that limited market, and 13.67% of
the $80 billion in shares in those same 1,095 credit unions. Clearly,
private share insurance is more significant to those affected states than
the Study Section's 2% statistic infers.

The Study Section also reports that 45% of all shares insured by ASI are
in credit unions chartered in California, as compared to 14.7% for the
NCUSIF. These facts can be misleading given that ASI has a limited market,
and the NCUSIF operates in all 50 states. An entirely different, but more
comparable, result is achieved when one isolates the relative risk in
these eight states only. Under an assumption that both entities are
limited to doing business in just the eight ASI states, ASI's 45%
concentration in California looks significantly less daunting when
compared to 55% for the NCUSIF. This should offer evidence that when
placed on equal footing, the relative risk concentration variances are
reduced materially.

While eight states represent a limited market, they do not necessarily
represent a geographic concentration risk, as inferred by the Study
Section. We argue that the company's states of operation represent a
diverse cross-section of our nation, for example: East Coast - Maryland;
Midwest - Ohio, Indiana and Illinois; West Coast - California and Nevada;
Northwest - Idaho; and, Southeast - Alabama.

Statutory Factors

As a private company, ASI faces various admission obstacles when seeking
new markets. First, a state must have a state statute that allows for an
option in share insurance. According to the Study Section, a total of
approximately 20 state statutes currently allow for the share insurance
option for their state-chartered credit unions. Based on this data, ASI is
operating in about 40%-50% of the available markets. Furthermore, the
actual power to approve such coverage, when permitted by statute, is
generally resident with the specific state's credit union supervisory
authority. So, as a private company, to do business in any state requires
that three basic conditions exist: (1) credit union demand; (2) a
permissible statute; and, (3) regulatory acceptance of the option.

                                  Page 1 of 4

Appendix XII
Comments from American Share Insurance

Based on these legislative and regulatory barriers, we take exception to
the GAO constantly using the federal share insurer, the NCUSIF, as a
benchmark in evaluating a private company's geographic concentration risk.
Due to the agency's federal franchise, none of the above conditions need
be present for the NCUSIF to do business in a state.

Mitigating Concentration Risk

The business of insuring credit union member deposits is a business of
risk assumption. Accordingly, the type of risk one assumes drives the cost
of the program and the risk of ultimate loss to the fund. ASI has been
very selective in assuming the risk it underwrites, and does a thorough
job of monitoring and field examining its insured institutions on a
recurring basis as reported in the Study Section. In addition, the Study
Section reports that the company has denied insurance coverage to certain
credit unions representing inordinate risk to the fund, and conversely has
approved many that satisfy the company's Risk Eligibility Standards. Of
the 29 credit unions that have converted to private share insurance during
the past decade, all were at the time, and are now, safe and sound credit
unions, and all strictly complied with the federal requirements to convert
insurance. These were not problem credit unions fleeing federal
supervision. Included in these federal requirements is a mail ballot vote
of the credit union's entire membership.

Risk in a Few Large Credit Unions

The Study Section reports that ASI has one insured institution that
represents approximately 25% of its total insured shares, and that its
"Top Five" credit unions represent 40% of total insured shares. The first
statistic compares unfavorably to the NCUSIF's reported concentration risk
in a single institution of 3%, to which we take no exception. The risk of
a single institution, however, has been significantly misrepresented in
the Study Section. A large, well managed credit union contributes
significantly to the financial stability of a share insurance program.

When underwriting its current largest institution in 2002, ASI considered
several risk-mitigating factors, and, as with all applicant credit unions,
performed a careful analysis of the institution. First, the subject
institution received (and continues to receive) the highest rating
available for credit unions. Second, ASI's independent actuaries evaluated
the adequacy of ASI's capital prior to, and following, the underwriting of
this credit union, and determined that ASI would continue to have a
sufficiently high probability of sustaining runs even with this credit
union in its insurance fund. Lastly, the federal insurer and state
regulator both approved of the credit union's insurance conversion, but
only after the credit union took a full mail ballot vote of its almost
200,000 members and agreed to satisfy all the requirements of consumer
disclosure under FDICIA.

With regard to the risk concentrated in a few large credit unions, the
Study Section fails to report the concentration risk in what would be the
equivalent of the NCUSIF's "Top Five" federally insured credit unions.
Proportionately, this would equate to the NCUSIF's top 230 federally
insured credit unions. In terms of asset size, this group of 230 credit
unions represents 45% of the NCUSIF's total insured shares. Clearly, the
two funds compare on this statistic, when measured on a proportionate, not
absolute basis.

B. 	ASI has limited ability to absorb large (catastrophic) losses because
it does not have the backing of any government entity.

The credit union movement introduced share insurance on the state level
long before Title II of the Federal Credit Union Act was enacted in 1971,
providing the first federal deposit insurance for credit unions. However,
private share insurance didn't come of age until the mid 1970s, as states
began to realize the loss of sovereignty in a state charter under an
all-federal insurance setting.

                                  Page 2 of 4

Appendix XII
Comments from American Share Insurance

It was never envisioned that private share insurance would seek, or need,
any guarantee from a state or federal government to operate. In the
cooperative spirit of the credit union movement, private share insurance
was designed to be a credit union-owned and credit union-operated private
fund. Nor was it ever the intent of the framers of private share insurance
for it to operate without supervision, or financial capacity. Accordingly,
various state laws were proactively sought and passed to permit the
private share insurance option, subject to admission standards and
required approvals. Private share insurance was designed to provide credit
unions with a comparable - not identical -- alternative means for
protecting member share accounts. Accordingly, a government backing for
private share insurance was never anticipated, and to use the lack of such
a guarantee as a criticism of private share insurance does not take into
account its legislative intent, past performance or founding principles.

To our knowledge, no private insurance company, licensed by individual
states, has a guarantee from the federal government. Further, no private
insurance company in the U.S. would be able to meet the "deep pockets"
test of the federal or state governments inferred in the Study Section. As
evidence of this, the largest insurance company in the country reports
just under $32 billion in capital from all of its various insurance
product lines. This is barely 50% of the aggregate capital available to
the NCUSIF. (Note: This amount is the estimated sum of the NCUSIF's
balance sheet capital plus the off-balance sheet recapitalization
liability of its insured credit unions).

Credit union-only insurance funds have a stable history that does not
track with insurers of thrifts or a combination of thrifts and credit
unions. Funds that have insured only credit unions (like ASI and the
NCUSIF) have had very successful track records when it comes to loss and
risk management. In over 29 years, ASI's loss ratio has been significantly
below that of its federal counterpart, and ASI has never had a year with
an operating loss, nor has it ever had to seek any form of
recapitalization from its member credit unions to bolster the fund due to
losses.

The reality is that a sound deposit insurance program, built upon a solid
foundation of careful underwriting, continuous risk management and the
financial backing of its mutual member credit unions, can exist as long as
consideration is given to an actuarial analysis of the capital adequacy of
the program in terms of sufficiently high probabilities (over 90%) of
being able to withstand runs and multiple runs on the system. This is a
common analysis that is accepted in the insurance industry for various
kinds of low frequency, high-severity risk programs and is the foundation
that the ASI insurance program is built upon. Our actuarial analyses and
independent actuarial reports were provided to the GAO during its
investigation. Alternative share insurance can be comparable to the
NCUSIF, and still not have a government backing.

C. ASI's lines of credit are limited in the aggregate as to amount and
available collateral.

With regard to ASI's committed bank lines of credit, the Study Section
infers that ASI's ability to absorb losses is reduced since its lines of
credit are limited in the aggregate as to amount and available collateral.
We disagree with this inference. The company's lines of credit are
designed to be solely a liquidity facility. The committed lines ensure
liquidity of ASI's invested funds; i.e., they provide a mechanism for ASI
to quickly generate cash to meet liquidity needs, without having to
liquidate the portfolio. Resources available for funding losses are not
the same as resources available for providing liquidity. Lines of credit
are not intended to be a source for funding insurance losses. In fact,
banks would not provide a loan for such a purpose. ASI's assets and its
off-balance sheet sources of funding (i.e., the power to recapitalize the
fund by insured credit unions under the ASI's governing statute and
insurance policy) are its capital sources for funding losses, not the bank
lines of credit.

Proportionately, ASI's lines of credits are greater than that of the
NCUSIF. ASI's $90 million in committed lines of credit equates to
approximately 47% of the company's total assets. NCUSIF's $1.6 billion
maximum borrowing capacity ($100 million from the U.S. Treasury and $1.5
billion from the Central Liquidity Facility, as disclosed in the NCUSIF's
and CLF's audited financial statements for the year ended December 31,
2002), equates to approximately 28% of its total assets.

                                  Page 3 of 4

Appendix XII
Comments from American Share Insurance

ASI has other sources of liquidity when it liquidates a credit union --
that is the credit union's own liquid assets. Approximately 42% of ASI's
primary insured credit unions' total assets are comprised of cash and
investments - we believe this is significant. In addition, the non-liquid
assets (namely loans and fixed assets) of a failed institution can be
pledged as collateral for additional borrowings to generate short-term
liquidity until such loans and other assets can be collected and/or sold.
In essence, a failed credit union's total assets over time often generate
sufficient liquidity to pay shareholders. Any shortage (historically less
than 4% of total assets of the failed institution) is usually funded as a
loss by ASI's assets. This is the same principle under which NCUSIF
operates.

D. 	Many privately insured credit unions have failed to make required
consumer disclosures about the absence of federal insurance of member
accounts as required under the Federal Deposit Insurance Corporation
Improvement Act of 1991 (FDICIA), and the Federal Trade Commission (FTC)
is the appropriate federal agency to enforce such compliance.

The Study Section reference to the GAO's August 20, 2003 study titled:
Federal Deposit Insurance Act: FTC Best Among Candidates to Enforce
Consumer Protection Provisions (GAO-03-971) reiterates the GAO's earlier
concern that "...members of privately insured credit unions might not be
adequately informed that their deposits are not federally insured..."

Although the statement may be accurate, any implication that ASI and its
member credit unions are purposefully misleading consumers fails to
directly implicate the Federal Trade Commission (FTC) who, with the
concurrence of Congress, has totally disregarded its statutory
responsibility to regulate the disclosure requirements as defined by
Section 151 (g) of FDICIA, codified at 12 U.S.C. S: 1831 (t)(g).

We believe that the GAO's earlier study brought to light the problems that
arise when a federal law effectively lacks an enforcement agency, and we
support the GAO's previous conclusion that the FTC is the appropriate
agency for monitoring and defining private share insurance consumer
disclosure requirements.

This concludes ASI's detailed comments in response to the GAO's draft
report on its study of private share insurance in the credit union
movement -- a component of the GAO's broader study titled, Credit Unions:
Financial Condition Has Improved But Opportunities Exist to Enhance
Oversight and Share Insurance Management.

                                  Page 4 of 4

Appendix XIII

                     GAO Contacts and Staff Acknowledgments

GAO Contacts	Richard J. Hillman (202) 512-8678 Debra R. Johnson (202)
512-8678 Harry Medina (415) 904-2220

Staff Acknowledgments

(250097)

In additional to those named in the body of this report, the following
individuals made key contributions.

William Bates
Sonja Bensen
Anne Cangi
Theresa L. Chen
William Chatlos
Jeanette Franzel
Charla Gilbert
Paul Kinney
Jennifer Lai
May Lee
Kimberley McGatlin
Grant Mallie
Jose R. Pena
Donald Porteous
Mitch Rachlis
Emma Quach
Barbara Roesmann
Nicholas Satriano
Kathryn Supinski
Paul Thompson
Richard Vagnoni

GAO's Mission	The General Accounting Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost
is through the Internet. GAO's Web site (www.gao.gov) contains abstracts
and full-text files of current reports and testimony and an expanding
archive of older products. The Web site features a search engine to help
you locate documents using key words and phrases. You can print these
documents in their entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as "Today's Reports," on its
Web site daily. The list contains links to the full-text document files.
To have GAO e-mail this list to you every afternoon, go to www.gao.gov and
select "Subscribe to e-mail alerts" under the "Order GAO Products"
heading.

Order by Mail or Phone	The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C.
20548

To order by Phone: 	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, 	Contact: Web site: www.gao.gov/fraudnet/fraudnet.htm

Waste, and Abuse in E-mail: [email protected]

Federal Programs Automated answering system: (800) 424-5454 or (202)
512-7470

Public Affairs	Jeff Nelligan, Managing Director, [email protected] (202)
512-4800 U.S. General Accounting Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

                               Presorted Standard
                              Postage & Fees Paid
                                      GAO
                                Permit No. GI00

United States
General Accounting Office
Washington, D.C. 20548-0001

Official Business
Penalty for Private Use $300

Address Service Requested
*** End of document. ***