Inspectors General: Mandated Studies to Review Costly Bank and Thrift
Failures (Letter Report, 11/07/96, GAO/GGD-97-4).
Pursuant to a legislative requirement, GAO reviewed the compliance of
the Inspectors General (IG) offices of the Board of Governors of the
Federal Reserve System, Federal Deposit Insurance Corporation (FDIC),
and the Department of the Treasury with section 38(k) of the Federal
Deposit Insurance Act of 1991, focusing on: (1) the findings of the
material loss review (MLR) reports initiated between July 1, 1994 and
June 30, 1995; (2) recommendations to improve bank supervisory
practices; and (3) the economy and efficiency of the MLR process.
GAO found that: (1) the Federal Reserve, FDIC, and Treasury IG issued a
total of four MLR reports on banks that failed or whose losses were
recognized during the second year of the MLR mandate; (2) these reports
noted that the four banks failed for similar reasons, including rapid
growth, excessive loan concentrations in the commercial real estate
industry, poor internal controls, and violations of laws and
regulations; (3) in three of the four cases, bank regulators did not
take aggressive enforcement actions to correct identified safety and
soundness deficiencies or to ensure that troubled banks complied with
existing enforcement actions; (4) the relatively small number of reports
issued in the first 2 years of the mandate did not provide a sufficient
basis to reach overall conclusions about the quality of bank supervisory
practices; and (5) there are reasons to question the cost-effectiveness
of the MLR process, including the fact that certain MLR requirements are
relatively inflexible and divert IG staff and resources from broader
reviews of the quality of bank supervision.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: GGD-97-4
TITLE: Inspectors General: Mandated Studies to Review Costly Bank
and Thrift Failures
DATE: 11/07/96
SUBJECT: Bank examination
Reporting requirements
Losses
Insured commercial banks
Savings and loan associations
Bank loans
Auditing procedures
Internal controls
Bank management
Bank failures
IDENTIFIER: Bank Insurance Fund
BIF
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO report. Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved. Major **
** divisions and subdivisions of the text, such as Chapters, **
** Sections, and Appendixes, are identified by double and **
** single lines. The numbers on the right end of these lines **
** indicate the position of each of the subsections in the **
** document outline. These numbers do NOT correspond with the **
** page numbers of the printed product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
Cover
================================================================ COVER
Report to Congressional Committees
November 1996
INSPECTORS GENERAL - MANDATED
STUDIES TO REVIEW COSTLY BANK AND
THRIFT FAILURES
GAO/GGD-97-4
Mandated Reviews of Costly Bank and Thrift Failures
(233498)
Abbreviations
=============================================================== ABBREV
BIF - Bank Insurance Fund
CFO - Chief Financial Officers Act of 1990
FDIA - Federal Deposit Insurance Act
FDIC - Federal Deposit Insurance Corporation
IG - Inspector General
MLR - material loss review
MNB - Mechanics National Bank
OCC - Office of the Comptroller of the Currency
OTS - Office of Thrift Supervision
SBA - Small Business Administration
Letter
=============================================================== LETTER
B-272445
November 7, 1996
The Honorable Alfonse D'Amato
Chairman
The Honorable Paul Sarbanes
Ranking Minority Member
Committee on Banking, Housing,
and Urban Affairs
United States Senate
The Honorable James A. Leach
Chairman
The Honorable Henry B. Gonzalez
Ranking Minority Member
Committee on Banking and
Financial Services
House of Representatives
This report presents the results of our second review of the
compliance of the Inspectors General (IG) offices for the Board of
Governors of the Federal Reserve System, Federal Deposit Insurance
Corporation (FDIC), and the Department of the Treasury with section
38(k) of the Federal Deposit Insurance Act (FDIA) as amended in
1991.\1 This section of FDIA requires the IGs to issue reports on
depository institutions--banks or thrifts\2 --whose failures result
in "material losses" (i.e., basically losses that exceed $25 million)
to deposit insurance funds. The section directs the IGs to determine
why a bank's or thrift's problems resulted in a material loss to an
insurance fund and to make recommendations for preventing such losses
in the future. Finally, until amended in October 1996, the section
required us to annually review reports issued by the IGs, verify the
accuracy of one or more of these reports, and make recommendations as
needed to improve the supervision of depository institutions.\3
In carrying out our responsibilities under section 38(k), our
objectives were to (1) summarize the findings of the MLR reports
initiated during the second year of the mandate--July 1, 1994, to
June 30, 1995--including the verification of the accuracy of one of
these reports; (2) make recommendations, if necessary, to improve
bank supervisory practices; and (3) assess the economy and efficiency
of the current MLR process.
--------------------
\1 12 USC Section 1831o(k).
\2 Thrifts are federally insured financial institutions that have
traditionally used customer deposits to finance home mortgages.
\3 This requirement was amended by the General Accounting Office
Management Reform Act of 1996, which the President signed on October
19, 1996. The amendment requires the Comptroller General of the
General Accounting Office to review material loss review (MLR)
reports, as the Comptroller General determines to be appropriate, and
recommend improvements in the supervision of depository institutions.
The amendment discontinues the requirement that GAO review MLR
reports on an annual basis. This report was prepared under section
38(k) of FDIA prior to the enactment of this amendment.
SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :1
To meet our first objective, we reviewed reports on two banks that
failed during the second year of the MLR mandate and two banks that
failed during the first year of the mandate but whose losses were not
recognized until the second year. In these latter two cases, FDIC's
initial estimates of the banks' losses were below $25 million, but
the estimates were revised above $25 million at a later date
requiring the IGs to initiate MLRs. We chose to verify the accuracy
of an MLR report issued by the Treasury IG because we verified
reports issued by the FDIC and Federal Reserve IGs in our first
report on the MLR process.\4 To verify the accuracy of this MLR
report, we reviewed the supporting workpapers, interviewed Treasury
IG officials, and met with examiners involved in the failure.
To meet our second objective, we analyzed the MLR reports issued to
date to determine whether an adequate base of evidence had been
established to make recommendations to the regulators on improving
supervisory practices. We asked IG officials to provide us with
their views on the MLR process and recommendations for improving its
cost-effectiveness. To meet our third objective, we assessed the
cost-effectiveness of the current MLR process by reviewing our
previous report which discussed this issue.
The IGs for FDIC, the Federal Reserve, and Treasury provided written
comments on a draft of this report, which are discussed on page 15
and reprinted in appendixes I, II, and III. We did our work in
Washington, D.C., and San Francisco between June and August 1996 and
in conformance with generally accepted government auditing standards.
--------------------
\4 Inspectors General: Mandated Studies to Review Costly Bank and
Thrift Failures (GAO/GGD-95-126, July 31, 1995).
BACKGROUND
------------------------------------------------------------ Letter :2
The FDIC, the Federal Reserve, the Office of the Comptroller of the
Currency (OCC), and the Office of Thrift Supervision (OTS)\5 share
responsibility for regulating banks and thrifts in the United States.
FDIC regulates federally insured state-chartered banks that are not
members of the Federal Reserve System while the Federal Reserve
regulates state-chartered banks that are members of the system.
State regulatory agencies are also responsible for overseeing the
operations of state-chartered institutions, in coordination with FDIC
or the Federal Reserve. OCC regulates nationally chartered banks,
while OTS regulates thrifts.\6 The regulators use a number of devices
to carry out their oversight responsibilities, such as conducting
periodic examinations and issuing enforcement actions against unsafe
and unsound banking practices.
Congress amended FDIA in 1991 after the failures of about 1,000 banks
between 1986 and 1990 resulted in about $14.4 billion in losses to
the Bank Insurance Fund (BIF) and threatened its solvency.\7 The
amendments were designed largely to strengthen bank supervision and
to help avoid a taxpayer bailout of the BIF similar to the
approximately $124.6 billion\8 in direct taxpayer funds that Congress
provided to protect the depositors of thrifts that failed during the
1980s and early 1990s. Among other provisions, the amendments
authorize the banking regulators to take specified enforcement
actions when they identify unsafe or unsound practices or conditions.
For example, the regulators can close banks whose capital levels fall
below predetermined levels. Congress also added section 38(k) to
FDIA to (1) ensure that the regulators learn from any weaknesses in
the supervision of banks whose failures cause material losses and (2)
make improvements as needed in the supervision of depository
institutions.
The IGs for the Federal Reserve, FDIC, and Treasury--the latter being
responsible for auditing OCC and OTS--are responsible for identifying
fraud, waste, and abuse and recommending improvements in agency
operations. Each IG oversees a staff of auditors and investigators
to assist in carrying out its mission. The staff engage in a range
of activities, including criminal investigations, financial audits,
and audits of the economy and efficiency of agency operations.
Section 38(k) of FDIA requires the IGs to initiate MLRs when the
estimated loss of a bank or thrift failure exceeds either $25 million
or a specified percentage of the institution's assets (see table 1)
when that percentage represents an amount greater than $25 million.
An MLR report must be completed within 6 months of the date that it
becomes apparent that the loss on a bank or thrift failure will meet
the criteria established by section 38(k).\9
Table 1
Percentage Loss of a Failed
Institution's Assets That Requires the
Initiation of an MLR\a
Time period\b Percentage
---------------------------------------- ----------------------------
1993 -1994 7
1994 -1995 5
1995 -1996 4
1996 -1997 3
After June 30, 1997 2
----------------------------------------------------------------------
\a The estimated loss must exceed $25 million.
\b Each time period is from July 1 to June 30 of the following year.
Source: Section 38(k) of the FDIA Act.
In July 1995, we issued our first report on the MLR process, as
required by section 38(k). It assessed the plans and procedures that
the IGs initiated to comply with the MLR mandate as well as the two
reports that were issued in the first year--July 1, 1993, to June 30,
1994. We found that the IGs had effectively positioned themselves to
meet the requirement by developing audit guidelines, providing
additional training, and hiring staff with appropriate banking
expertise. We also found that the two reports issued the first year,
one by the FDIC IG and one by the Federal Reserve IG, were generally
accurate in describing the causes of the banks' failures and the
quality of their supervision. We recommended that the FDIC IG take
further steps to assess the adequacy of regulatory enforcement
actions, which the IG agreed to implement. We also found that the
MLR process as currently structured may not be the most
cost-effective means of achieving improved bank supervision because,
among other reasons, the limited number of reports issued did not
provide a comprehensive basis to recommend overall changes in bank
regulatory practices.
--------------------
\5 OCC and OTS are both part of the Department of the Treasury.
\6 State regulatory authorities also share responsibility with OTS
for regulating and supervising state-chartered thrifts.
\7 BIF had a deficit balance of $7 billion at the end of 1991 but has
since recovered substantially and had a balance exceeding $25 billion
in 1995.
\8 In addition to $124.6 billion in direct taxpayer costs, which
included funds provided to the Resolution Trust Corporation, there
was also $7.5 billion in indirect taxpayer costs, such as tax
benefits provided to the acquirers of troubled thrifts. See
Financial Audit: Resolution Trust Corporation's 1995 and 1994
Financial Statements (GAO/AIMD-96-123, July 2, 1996).
\9 The section defines a loss as incurred when (1) FDIC is appointed
receiver of the institution and it becomes apparent that the present
value of a deposit insurance fund's outlays with respect to the
institution will exceed the present value of the receivership
dividends or other payments on claims held by FDIC or (2) if FDIC
provides assistance to the institution while an ongoing concern and
it is not substantially certain that the assistance will be repaid
within 24 months after the date on which the assistance was initiated
or the institution ceases to repay the assistance in accordance with
its terms.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :3
The Federal Reserve, FDIC, and Treasury IGs issued a total of four
MLR reports on banks that failed or whose losses were recognized
during the second year of the MLR mandate. Our review of these
reports found that the four banks failed for similar reasons
including: rapid growth, excessive loan concentrations in the
commercial real estate industry, poor internal controls, and
violations of laws and regulations. The reports also identified
certain weaknesses in the bank regulators' oversight of these
institutions; for example, in three of the four cases bank regulators
either did not take sufficiently aggressive enforcement actions to
correct identified safety and soundness deficiencies or to ensure
that troubled banks complied with existing enforcement actions. Our
review of the Treasury IG's report on the failure of Mechanics
National Bank (MNB) of Paramount, California, found that it
accurately described the causes of the bank's failure and the
weaknesses in regulatory oversight of the bank.
We are not making any general recommendations to the bank regulators
because the relatively small number of reports issued in the first 2
years of the mandate--six--does not provide a sufficient basis to
reach overall conclusions about the quality of their supervisory
practices. In our view, the limited basis that these reports provide
for making recommendations about overall bank supervision raises
questions about the cost-effectiveness of the MLR process as
currently structured. There are other reasons to question the
cost-effectiveness of the current process, including the fact that
certain MLR requirements are relatively inflexible and divert IG
staff and resources from broader reviews of the quality of bank
supervision. In our previous report, we identified options that
could improve the efficiency and/or effectiveness of the current MLR
process which we believe continue to merit congressional
consideration.
IG REPORTS FOUND THAT BANKS
FAILED FOR SIMILAR REASONS
------------------------------------------------------------ Letter :4
The four MLR reports that we reviewed (see table 2) found that the
banks failed for similar reasons. In particular, the reports found
that the banks, primarily during the 1980s, engaged in unsafe and
unsound practices, such as rapid growth, significant concentrations
in real estate loans, inadequate lending practices, and violations of
laws and regulations. When the real estate industry suffered a
substantial downturn in the early 1990s, the banks suffered
substantial losses which ultimately caused their failures. The
reports also identified various shortcomings in bank regulatory
practices; for example, the reports on MNB and The Bank of Hartford
cited OCC and FDIC for not ensuring that the banks complied with
existing enforcement actions, and the Pioneer Bank report cited the
Federal Reserve for not taking more aggressive enforcement actions to
correct identified deficiencies. However, only the FDIC IG issued
recommendations in its two reports to improve bank supervision. The
Treasury and Federal Reserve IGs did not issue specific
recommendations because, among other reasons, they did not believe
that one MLR report provided an adequate basis for such
recommendations.
Table 2
Reports Initiated During the Second Year
of the MLR Mandate
($ in millions)
Estimated
Bank loss to BIF IG office
---------------------------------------- ------------- -------------
Mechanics National Bank, Paramount, $ 36.6 Treasury
California
The Bank of San Pedro, San Pedro, 28.8 FDIC
California
The Bank of Hartford, Hartford, 31.3 FDIC
Connecticut
Pioneer Bank, Fullerton, California 27.3 Federal
Reserve
----------------------------------------------------------------------
Source: IG MLR reports on these bank failures.
MECHANICS NATIONAL BANK
---------------------------------------------------------- Letter :4.1
OCC closed MNB on April 1, 1994, during the first year of the MLR
mandate, but FDIC did not estimate that its total loss on the MNB
failure would exceed $25 million until April 18, 1995, which was
during the second year of the mandate. Consequently, the Treasury IG
initiated its MLR as of April 18, 1995, and issued the report on
September 29, 1995. In our view, the report accurately described the
causes of MNB's failure and identified shortcomings in OCC's
oversight of the institution.
The report concluded that MNB engaged in a variety of unsafe and
unsound banking practices between 1988 and 1994 that contributed
substantially to its failure. For example, the report found that the
bank nearly doubled in size between 1988 and 1991 and made
significant loan commitments to the commercial real estate sector.
Further, the bank established a substantial inventory of loans
guaranteed by the Small Business Administration (SBA); many of these
loans were for financing gasoline service stations and highly
risky.\10 In addition to credit concentrations in the commercial real
estate and service station sectors, MNB suffered from inadequate
underwriting practices, weak appraisal procedures, poor loan
documentation, suspected fraudulent activities in the SBA
portfolio,\11 and insider abuses. When the California commercial
real estate sector suffered a substantial downturn in the early
1990s, many MNB loans defaulted and the bank had inadequate capital
to prevent its eventual insolvency and closure in April 1994.
The Treasury IG also identified various deficiencies in OCC's
oversight of MNB. For example, the IG found that the limited safety
and soundness exams that OCC conducted of MNB in 1989 and 1990 were
insufficient to detect the bank's weak underwriting and loan
administration practices. Therefore, many of the bad loans that
resulted in MNB's eventual failure were already on its books by the
time OCC conducted the full-scope examination in 1991 that first
identified the severity of the bank's problems. The report also
concluded that OCC's enforcement actions proved ineffective in
getting MNB management to correct identified unsafe and unsound
practices. Specifically, MNB continued to make a substantial number
of poorly underwritten commercial real estate loans even after OCC
issued a cease and desist order against the bank in 1991 which
required substantial improvements in the bank's lending operations.
The report faulted OCC for not imposing civil money penalties on bank
officials for continued violations of the cease and desist order.
However, the IG made no specific recommendations in the report to
improve OCC enforcement practices because the IG believed that (1)
the passage of laws, such as the 1991 amendments to FDIA, and rules
and procedures should address many of the supervisory weaknesses
identified in the MNB report; and (2) one report did not provide a
sufficient base of evidence for making recommendations.
Our analysis of the MNB report found that it accurately described the
causes of the bank's failure and the shortcomings in OCC's oversight
activities. This assessment is based on our review of the report's
supporting documentation and on discussions with IG staff and OCC
officials. We also note that OCC's Senior Deputy Comptroller for
Administration agreed with the IG's findings in OCC's official
written response to the report.
--------------------
\10 Service station loans are risky because the collateral values for
such loans can be significantly reduced by hazardous waste
contamination at the site of the property. The bank may also be
liable for the costs of cleaning up the environmental pollution on
foreclosed properties.
\11 SBA guarantees between 70 to 90 percent of each SBA loan with the
bank liable for the remaining percent. MNB's SBA loan portfolio
primarily consisted of the nonguaranteed portion of SBA loans because
the bank typically sold the guaranteed portion to the secondary
market. In addition, MNB could be liable for the guaranteed portion
of the SBA loans that were sold if they were approved based on
misrepresentation or fraud.
THE BANK OF SAN PEDRO
---------------------------------------------------------- Letter :4.2
The State of California closed The Bank of San Pedro on July 15,
1994, and the FDIC IG issued its MLR report on December 21, 1994.
The report cited numerous causes of the bank's failure, including
excessive concentrations in a real estate development subsidiary,
poor underwriting practices, inadequate loss reserves and capital,
failure to comply with regulatory enforcement actions, and repeated
violations of laws and regulations. As was the case with MNB, the
substantial downturn in the California real estate economy, coupled
with the bank's poor management practices, caused significant losses
to San Pedro's loan portfolio that ultimately caused the bank to
fail.
The report found that FDIC was in compliance with applicable laws and
regulations and properly identified the causes of San Pedro's
failure. However, the IG did recommend that FDIC evaluate the need
for developing regulations to control the use of money desks by
troubled banks.\12 Even though San Pedro was undercapitalized,
management used the bank's money desk to raise volatile deposits
nationwide by offering above-market interest rates.\13 The use of
volatile deposits by irreversibly troubled institutions can delay
their failure and potentially increase insurance fund losses,
according to the report. The report also reiterated the importance
of recommendations that the IG made in a previous MLR report covering
such areas as controlling loan concentrations and excessive overhead
expenses.\14
--------------------
\12 A money desk is a mechanism by which orders are executed for bank
customers, correspondent banks, or a bank's own account. Banks can
use money desks to attract deposits from outside of their
geographical area.
\13 Because these deposits were highly sensitive to interest rates,
they were considered volatile by the regulators.
\14 FDIC Inspector General, The Failure of the Bank of San Diego
(April 29, 1994).
THE BANK OF HARTFORD
---------------------------------------------------------- Letter :4.3
The State of Connecticut closed the Bank of Hartford on June 10,
1994, but FDIC did not estimate that the associated losses would
exceed $25 million until the spring of 1995. The FDIC IG initiated
its MLR as of June 6, 1995, and the report was issued on December 1,
1995.
According to the MLR report, The Bank of Hartford failed as a result
of several unsafe and unsound practices. In particular, the bank
committed a significant percentage of its loan portfolio to the
multifamily and commercial real estate sectors; such loans
represented 47 percent of the bank's total loan portfolio in 1991,
which exposed the bank to significant losses when the Connecticut
real estate sector declined. The bank also failed to adequately
assess the ability of loan guarantors to fulfill their commitments
and engaged in poor loan administration practices, such as the
failure to obtain updated financial information on large borrowers.
Further, within 12 months prior to its failure, the bank purchased
$27 million in derivative financial instruments without performing
the necessary analysis as required by FDIC policy guidelines. FDIC
experienced a loss of $5.8 million on the sale of the derivative
instruments after the Bank of Hartford was closed.
The report identified weaknesses in FDIC's oversight of the bank and
recommended improvements. For example, the report found that FDIC
and state examiners did not identify the bank's credit concentrations
in multifamily loans until 1991 even though they were apparent in
1987. By the time FDIC entered into a cease and desist order with
the bank in July 1992 that required the bank to reduce its loan
concentrations, more than 33 percent of its commercial real estate
loans were considered in danger of default. Moreover, the report
found that the bank's purchase of $27 million in derivative
instruments without a proper risk analysis was contrary to a cease
and desist order against the bank and FDIC investment guidelines.
However, the report pointed out that FDIC increased its oversight
activities when the derivatives purchases were identified. Among
other suggestions, the IG recommended that all undercapitalized banks
be required to obtain written approval from FDIC prior to purchasing
derivative instruments.
PIONEER BANK
---------------------------------------------------------- Letter :4.4
The State of California closed Pioneer Bank on July 8, 1994, and the
Federal Reserve IG issued its MLR report on January 6, 1995. Like
the other three banks discussed above, Pioneer failed primarily as
the result of its rapid growth and substantial commitment to the
commercial real estate sector. Commercial real estate loans
accounted for 44 percent of Pioneer's total loans in 1991 and exposed
the bank to substantial losses when the economy deteriorated. In
addition, Pioneer suffered from weak liquidity and poor appraisal and
underwriting practices.
The report found that the Federal Reserve identified Pioneer's
deficiencies during on-site examinations and tried to get bank
officials to correct safety and soundness violations through
enforcement actions. Further, the report concluded that Federal
Reserve's actions, both formal and informal, were within the range of
acceptable actions for a bank with Pioneer's problems. However, the
report also identified certain weaknesses in the Federal Reserve's
supervisory practices. For example, examiners did not always detect
appraisal violations, or adequately address underwriting and credit
administration weaknesses. Also, enforcement actions taken by the
Federal Reserve were not always sufficiently aggressive given the
severity of the safety and soundness deficiencies identified. The IG
did not make any recommendations in the report since it only assessed
one bank failure. However, the IG did suggest that the Federal
Reserve take the report's findings on appraisal violations,
underwriting weaknesses, and supervisory actions under advisement.
MLR PROCESS APPEARS TO HAVE
LIMITED COST-EFFECTIVENESS
------------------------------------------------------------ Letter :5
We believe there are reasons to question whether the time and
resources that the IGs commit to the MLR process, as currently
structured, are cost-effective. Although the six MLR reports
initiated during the first 2 years of the mandate identified
important information about individual bank failures, they do not
provide us with an adequate base of evidence to recommend
cross-regulator improvements in bank supervision. IG officials have
also found it difficult to justify recommendations based on a limited
sample of reports, and have said that certain MLR requirements,
particularly the $25 million threshold for initiating an MLR, are
relatively inflexible, resulting in administrative burdens and
expenditures. Even if there were a substantial increase in costly
bank failures requiring MLR reports, we believe it is questionable
whether the labor-intensive demands and short reporting deadlines of
the MLR process would make the potential benefits worth the costs.
As stated in our previous report on the MLR process, we believe it
would make sense for Congress to provide the IGs with more
flexibility in initiating MLRs.
LIMITED NUMBER OF REPORTS
PROVIDE INADEQUATE BASIS FOR
MAKING RECOMMENDATIONS
---------------------------------------------------------- Letter :5.1
We stated in our first report on the MLR requirement that reports on
individual bank failures provide important information about the
causes of individual bank failures and the quality of the banks'
supervision. The reports also familiarize IG staff with the
examination process and have identified areas that require further
investigation. As examples, FDIC IG officials told us that the
report on the Bank of San Pedro served as the basis for doing a
broader audit on troubled banks' use of money desks, and Treasury IG
officials said that the MNB report served as the basis for initiating
a study on OCC officials' compliance with certain
conflict-of-interest disclosure requirements. Additionally, in March
1996, FDIC Division of Supervision officials in Washington, D.C.,
disseminated the findings of two IG MLR reports for the benefit of
bank examiners in regional offices across the country.
However, we do not believe that the six MLR reports on bank failures
issued during the first 2 years of the mandate provide a sufficient
basis for identifying trends in bank supervisory practices or making
recommendations that would apply equally to the Federal Reserve,
FDIC, OCC, and OTS. The limited basis for making overall
recommendations is demonstrated by the fact that four of the six
failed banks were located in California and only one of the reports
applies to OCC while none applies to OTS. IG officials have also
found it difficult to justify recommendations to the regulators based
on the MLR reports initiated during the first 2 years of the mandate.
For example, the Treasury and Federal Reserve IGs did not make
recommendations in the MNB and Pioneer Bank reports, respectively, at
least in part because the IG officials did not believe that the
reports provided a sufficient evidentiary base. We believe that such
recommendations, if necessary, should be based on larger samples of
failed bank cases located in different regions of the country or on
broader reviews of bank supervisory practices.
We recognize that the four MLR reports discussed above identify
certain commonalities (e.g., the banks failed due to real estate
concentrations and regulators did not take adequate enforcement
actions) which suggest potential areas for improvements in
supervisory practices. However, we are not convinced that four
reports provide an adequate basis for specific recommendations.
Further, the reports' major findings address weaknesses in bank
supervision during the 1980s, such as the reluctance of the
regulators to take strong enforcement actions to address repeated
safety and soundness violations, which have already been documented
and served as the basis for the regulatory reform amendments to FDIA
that were enacted in 1991.\15 For example, FDIA now directs that the
regulators require troubled banks whose capital has fallen to
preestablished levels to file plans that detail how they will improve
their financial condition and correct unsafe and unsound practices.
Therefore, to the extent that the reports are focused on historical
practices that have been addressed in whole or in part by legislation
or changes in supervisory practices, their applicability may be
limited.
Although a substantial increase in the number of bank failures
requiring MLRs would provide a better base for making
recommendations, we believe there are reasons to question whether the
benefits of issuing MLR reports under such circumstances would
justify the costs. The MLRs initiated to date have been
labor-intensive efforts that required substantial interaction between
IG staff and regulatory officials as well as several visits to bank
failure locations to review loan files and other records. The IG
offices are required to initiate these efforts to meet the 6-month
deadline established by section 38(k), but the presence of IG staff
at bank locations during a wave of bank failures as occurred during
the late 1980s and early 1990s may not be desirable. During such
periods, bank examiners, FDIC failed-bank resolution officials, and
IG staff may not have adequate time and resources to devote to the
preparation of in-depth case studies on the causes of individual bank
failures. Further, our past reports have found that poorly managed
banks tend to make weakly underwritten loans during strong economic
periods and that by the time banks begin to experience substantial
financial deterioration, the ability of bank managers and regulators
to avert failure may be substantially limited. Therefore, a series
of MLR reports issued in the midst of a bank crisis may have limited
short-term practical effects because the bad loans generating the
failures may have been on the banks' books for years. We believe the
four MLR reports discussed in this report illustrate this point
because they found that the banks engaged in shortsighted and
ultimately unsafe real estate lending policies, primarily during the
strong economic period of the 1980s, that were directly responsible
for their failures when the local economies weakened substantially in
the early 1990s.
--------------------
\15 See Deposit Insurance: A Strategy for Reform (GAO/GGD-91-26,
Mar. 4, 1991).
CERTAIN MLR REQUIREMENTS ARE
RELATIVELY INFLEXIBLE AND
PLACE ADMINISTRATIVE BURDENS
ON IG OFFICES
---------------------------------------------------------- Letter :5.2
IG officials we contacted said that certain MLR requirements
specified in section 38(k) can place administrative burdens on their
offices. In particular, they said that FDIC sometimes underestimates
potential insurance fund losses when banks fail, so the IGs must
sometimes initiate MLRs months after failures have occurred when FDIC
revises its initial loss estimates above $25 million. The Treasury
and FDIC IGs did not initiate the MNB and Bank of Hartford MLRs until
about a year after their failures, because FDIC's initial loss
estimates were revised upward substantially.
Under a Statement of Understanding between the IG offices that
coordinates the MLR process, the IGs are to monitor certain bank
failures whose estimated losses are below $25 million for five full
quarters after the failure date to ensure that FDIC loss estimates do
not exceed the statutory threshold over time.\16 IG staff said that
monitoring certain failures, which may involve preliminary contacts
with regulatory officials and obtaining documents, whose estimated
losses could exceed $25 million, proves unnecessary if the estimates
stay below the statutory threshold. More important, IG officials may
experience difficulties obtaining information they believe is
necessary to do an MLR when FDIC substantially revises its initial
loss estimates upward with minimal warning months after a failure has
occurred.\17 For example, in early 1996, FDIC revised the estimated
loss on Guardian Bank in California, which failed in January 1995,
from $8 million to more than $25 million. This revised estimate
required the Federal Reserve IG to initiate an MLR since the bank was
regulated by the Federal Reserve. IG officials said they were not
fully prepared to initiate the MLR and had some difficulties
obtaining required bank records because they had already been shipped
to other storage locations throughout the country. In one instance,
IG officials said they had to visit an FDIC storage facility in
Chicago to review certain records. In addition, the IG officials
said that key regulatory personnel involved in overseeing Guardian
were no longer available or could not recall certain events due to
the passage of time.
We also pointed out in our previous report that the 6-month deadline
for MLRs may require IG officials to divert staff from ongoing, broad
reviews of bank regulation and other issues. For example, Treasury
IG officials said that 30 to 50 percent of their staff resources are
already committed to assessing executive agency financial systems as
required by the Chief Financial Officers Act of 1990 (CFO).
Therefore, MLR audits could place additional workload demands on the
Treasury IG staff that are not committed to CFO work, particularly
since the organization did not hire additional personnel to do MLRs.
Moreover, the FDIC IG office is undergoing a substantial downsizing,
which means that MLRs will place more stringent requirements on the
organization in the future. FDIC IG officials said the staff
responsible for MLRs and other bank supervision work declined from an
authorized level of 37 in 1994 to 25 in 1996 and may fall to 11 in
1997. The Federal Reserve IG staff available to do MLRs has remained
stable, but officials we contacted said they were not able to
initiate all of their planned audits on bank supervision in 1996 due
to the Guardian Bank MLR. As discussed earlier, we believe there is
reason to question whether the IGs' resource commitments to the MLR
process are justifiable, given their limited impact on improving bank
supervision.
--------------------
\16 The initial estimates on the cost of bank failures are based on
FDIC officials' best estimates at the time of failure of what the
recoveries will be on the sale of a particular bank's assets. They
may also be based on FDIC's historical recoveries on the sale of
failed bank assets. The initial estimates are revised over time as
FDIC actually sells assets and tabulates its recoveries.
\17 In March 1994, the FDIC IG issued a report on the FDIC's loss
estimation process entitled Audit of the Cost Estimate Process for
Failed Bank Resolutions. In September 1996, the FDIC IG issued a
related study entitled Follow-Up Audit: Cost Estimate Process for
Bank Resolutions.
PROVIDING IGS WITH MORE
FLEXIBILITY COULD IMPROVE
THE COST-EFFECTIVENESS OF
THE MLR PROCESS
---------------------------------------------------------- Letter :5.3
We suggested in our previous report that providing the IGs with more
discretion on the number and timing of MLRs to initiate each year
could improve the cost-effectiveness of the MLR process. For
example, more discretion would allow the IGs to focus their efforts
on broader reviews of the overall quality of bank supervision rather
than diverting staff to conduct mandatory MLRs within the 6-month
deadline. Additional discretion could also minimize the current
administrative burdens that have been generated by difficulties in
estimating the cost of bank failures; for example, the IGs could be
allowed to make decisions on whether to initiate an MLR based on
their professional judgment rather than a predetermined loss
estimate. We did not suggest repealing the MLR mandate entirely
because we believe the process holds the regulators accountable for
their supervisory practices. Further, IG officials told us that MLRs
provide important information that can be incorporated into broader
studies of bank supervision.
CONCLUSIONS
------------------------------------------------------------ Letter :6
The four MLR reports we reviewed found that the banks failed for
similar reasons. In particular, the banks grew at rapid rates and
committed an excessive percentage of their lending portfolios to the
commercial real estate sector, primarily during the 1980s. When the
real estate industry suffered substantial downturns in California and
Connecticut during the early 1990s, the banks' exposure to that
industry and poor lending practices caused substantial losses and
eventual insolvency. The reports also identified various weaknesses
in the regulators' oversight efforts, such as the failure to get
banks to comply with existing enforcement actions, and the FDIC IG
chose to make recommendations for improvement to FDIC while the
Treasury and Federal Reserve IGs did not. Our review of the MNB
report verified its major findings.
We are not making any recommendations for improving overall bank
supervision in this report because the limited number of MLR reports
produced so far does not provide an adequate basis for identifying
improvements. As currently structured, there are reasons to question
the cost-effectiveness of the MLR process, such as the limited basis
it provides for making recommendations to improve bank supervision
and the administrative burden and expenditures that the requirement
places on the IG offices. As we found in our previous report,
providing the IGs with more discretion in choosing the number and
timing of MLRs to initiate each year could improve the
cost-effectiveness of the MLR process while preserving the
congressional intent of section 38(k), which was to hold bank
regulators accountable for their actions.
MATTERS FOR CONGRESSIONAL
CONSIDERATION
------------------------------------------------------------ Letter :7
As we stated in our previous report, Congress may wish to consider
whether the current MLR requirement is a cost-effective means of
achieving improved bank supervision. If it determines that the
requirement is not cost-effective, we suggest that Congress consider
amending the requirement so that the IGs have more flexibility in
choosing the number and timing of MLRs to initiate each year.
AGENCY COMMENTS AND OUR
EVALUATION
------------------------------------------------------------ Letter :8
The IGs for the Federal Deposit Insurance Corporation, the Board of
Governors of the Federal Reserve System, and the Department of the
Treasury provided written comments on our draft report, which are
reprinted in appendixes I, II, and III, respectively. The IGs
generally agreed with our findings and conclusions as well as our
suggestion that Congress consider amending section 38(k) of FDIA so
that the IGs have more discretion in the number, timing, and scope of
MLRs to initiate each year. The IGs also provided comments that were
generally technical in nature and are incorporated in this report
where appropriate.
We are sending copies of this report to the Inspectors General for
the Federal Deposit Insurance Corporation, the Board of Governors of
the Federal Reserve System, the Department of the Treasury, and other
interested parties. We will make copies available to others upon
request.
This report was prepared under the direction of Susan S. Westin,
Assistant Director, Financial Institutions and Markets Issues. The
other major contributor was Wesley M. Phillips, Evaluator-in-Charge.
If you have any questions or comments about this report, please call
me on (202) 512-8678.
Thomas J. McCool
Associate Director, Financial Institutions
and Markets Issues
(See figure in printed edition.)Appendix I
COMMENTS FROM THE INSPECTOR
GENERAL OF THE FEDERAL DEPOSIT
INSURANCE CORPORATION
============================================================== Letter
(See figure in printed edition.)Appendix II
COMMENTS FROM THE INSPECTOR
GENERAL OF THE FEDERAL RESERVE
SYSTEM
============================================================== Letter
(See figure in printed edition.)Appendix III
COMMENTS FROM THE INSPECTOR
GENERAL OF THE DEPARTMENT OF THE
TREASURY
============================================================== Letter
(See figure in printed edition.)
The following are GAO's comments on the Treasury IG's comments dated
October 3, 1996.
GAO COMMENTS
------------------------------------------------------------ Letter :9
1. The IG stated that unsafe banking practices were the primary
cause of MNB's failure while a declining economy precipitated
eventual losses. We have clarified the language on pages 6-7 in
response to the IG's comment.
2. The IG said that an increasing MLR workload would not affect the
office's CFO work since such work is done by a separate unit. We
clarified the language on page 13 to make clear that MLRs could place
greater resource demands on IG staff not engaged in CFO work.
3. The IG stated that an underlying implication of our argument is
that the weak credit losses of the 1980s will also be the primary
cause of bank failures in the future. However, the IG said that this
implication is not necessarily the case and that banks are moving
into new areas, such as fee-based services, that involve risks
different from credit risk that could cause future bank failures.
We acknowledge that many banks are moving into new activities that
represent risks different from traditional credit risks. Such
activities must be carefully managed and monitored to ensure that
they do not pose unacceptable risks to the safety and soundness of
banking institutions. It is possible that the period between the
time these problems arise and losses are ultimately recognized could
be shorter.
4. The IG said that doing a few MLRs during strong economic periods
may be beneficial in providing important prospective information
about bank supervision to Congress, regulators, and the IGs as
compared with initiating a series of MLRs during a banking downturn.
We state in our report that MLRs have certain beneficial impacts on
bank supervision but raise questions about the cost-effectiveness of
the process as currently structured. By allowing the IGs greater
discretion in the number, timing, and scope of MLRs to initiate as we
have suggested, we believe the benefits of the MLR process will be
enhanced while costs will be reduced.
*** End of document. ***