1992 Bank Resolutions: FDIC Chose Methods Determined Least Costly, but
Needs to Improve Process (Chapter Report, 05/10/94, GAO/GGD-94-107).

The least-costly resolution provisions of the Federal Deposit Insurance
Corporation (FDIC) Improvement Act of 1991 require FDIC to resolve
failed banks with the least expensive method. Generally, FDIC complied
with the law in calculating the costs of resolution alternatives and
documenting cost evaluations during 1992.  FDIC's procedures to estimate
the net realizable value of failed bank's assets and to document cost
evaluations improved in the latter half of 1992 as FDIC gained
experience in carrying out new ways to estimate the net realizable value
of the assets of failed banks.  Some factors in the cost estimates were
more uncertain than others, such as estimates of insured and uninsured
deposits, future losses from loss-sharing agreements, and the future
market value of bridge banks.  GAO's analysis of 22 sample resolutions
indicated that FDIC consistently chose the resolution alternative that
it deemed to be the least costly compared to the other approaches
considered.  In 18 of the 22 cases, however, FDIC did not document its
rationale for the marketing strategy it selected.  Although technically
not required by law, such documentation would give the fullest effect of
FDIC's statutory mandate to resolve failed banks in the least costly
way.

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

 REPORTNUM:  GGD-94-107
     TITLE:  1992 Bank Resolutions: FDIC Chose Methods Determined Least 
             Costly, but Needs to Improve Process
      DATE:  05/10/94
   SUBJECT:  Bank failures
             Financial management
             Insured commercial banks
             Financial institutions
             Cost effectiveness analysis
             Bank management
             Financial statement audits
             Documentation
             Banking law
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.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Committees

May 1994

1992 BANK RESOLUTIONS - FDIC CHOSE
METHODS DETERMINED LEAST COSTLY,
BUT NEEDS TO IMPROVE PROCESS

GAO/GGD-94-107

1992 Bank Resolutions


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

  AVR - Asset Valuation Review
  BIF - Bank Insurance Fund
  DOL - Division of Liquidation
  DOR - Division of Resolutions
  DOS - Division of Supervision
  FDI - Federal Deposit Insurance
  FDIC - Federal Deposit Insurance Corporation
  FDICIA - Federal Deposit Insurance Corporation Improvement
     Actof1991
  FIRREA - Financial Institutions Reform, Recovery, and Enforcement
     Act of1989
  IDT - insured deposit transfer
  OCC - Office of the Comptroller of the Currency
  OIG - Office of Inspector General
  P&A - purchase and assumption
  RTC - Resolution Trust Corporation
  TAPA - Total Asset Purchase and Assumption

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


B-256292

May 10, 1994

The Honorable Donald W.  Riegle, Jr.
Chairman
The Honorable Alfonse M.  D'Amato
Ranking Minority Member
Committee on Banking, Housing, and
 Urban Affairs
United States Senate

The Honorable Henry B.  Gonzalez
Chairman
The Honorable Jim Leach
Ranking Minority Member
Committee on Banking, Finance and
 Urban Affairs
House of Representatives

This report presents the results of our review of the Federal Deposit
Insurance Corporation's (FDIC) compliance with Section 13(c)(4) of
the Federal Deposit Insurance Act, as amended by The Federal Deposit
Insurance Corporation Improvement Act of 1991.  This section of the
act requires FDIC to calculate and document its evaluation of the
costs of all possible methods for resolving a troubled depository
institution and to choose the resolution method that entails the
least possible cost to the deposit insurance fund.  These statutory
provisions establish the basic requirements that FDIC must meet in
making its least-cost determination. 

The act requires that we annually audit FDIC's compliance with the
least-cost provisions.  This report is intended to provide
information, analysis, and recommendations to improve the FDIC
resolution process. 

We are providing copies of this report to the Acting Chairman of the
FDIC and other interested parties. 

This report was prepared under the direction of Mark J.  Gillen,
Assistant Director, Financial Institutions and Markets Issues.  Other
major contributors are listed in appendix VII.  If there are any
questions about this report, please contact me on (202) 512-8678. 

James L.  Bothwell
Director, Financial Institutions
 and Markets Issues


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


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

From 1986 through 1993, resolutions of failed banks cost the deposit
insurance fund about $31.93 billion.  In an effort to stem insurance
fund losses and to foster depositor discipline, Congress passed the
least-cost resolution provisions contained in the Federal Deposit
Insurance Corporation Improvement Act of 1991 (FDICIA).\1 These
provisions were effective immediately upon FDICIA's enactment on
December 19, 1991. 

The least-cost resolution provisions of FDICIA generally require the
Federal Deposit Insurance Corporation (FDIC) to resolve a failed bank
in the least costly of all possible methods.  The statute contains
specific rules for calculating the costs of resolution alternatives
and documenting the agency's evaluations of those costs.  On the
basis of this evaluation, the agency is required to select the least
costly alternative.  Finally, the statute requires GAO to annually
report to Congress on FDIC's compliance with FDICIA's least-cost
resolution provisions. 

In accordance with this requirement, GAO assessed FDIC's compliance
with the least-cost provisions, specifically with requirements for
calculating costs and documenting the evaluation of the costs of
resolution methods.  In addition, GAO reviewed the marketing aspect
of FDIC's resolution process. 


--------------------
\1 Section 13(c)(4) of the Federal Deposit Insurance Act, as amended
by FDICIA, P.L.  102-242, 105 Stat.  2236 (1992), effective December
19, 1991. 


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

One of the first steps FDIC is to take toward resolving a failed bank
is to estimate the cost of liquidation--basically, the amount of
insured deposits paid out minus the net amount recovered through
asset disposition activities (or net realizable value).  FDIC is to
compare the estimated cost of liquidation with the estimated costs of
other resolution methods. 

The least-cost provisions' cost-calculation requirements primarily
apply to the adjustments FDIC makes to the book value of a failed
bank's assets in estimating the net realizable value of those assets. 
The least-cost documentation provision requires FDIC to document its
evaluation of the costs of the resolution alternatives considered,
including the assumptions on which the evaluation is based. 

Historically, FDIC has resolved failed banks using three basic
methods.  These include (1) directly paying depositors the insured
amount of their deposits and disposing of the failed bank's assets
(deposit payoff and asset liquidation); (2) selling only the bank's
insured deposits and certain other liabilities, with some of its
assets, to an acquirer (insured deposit transfer); and (3) selling
all of the failed bank's deposits, certain other liabilities, and
some or all of its assets to an acquirer (purchase and assumption).\2
Within this third category, many variations exist based on specific
assets that are offered for sale.  For example, some purchase and
assumption resolutions have also included loss-sharing agreements--an
arrangement that allows the acquirer of the failed bank to share the
losses on certain assets with FDIC.\3

The resolution alternatives that FDIC considers in the agency's
least-cost evaluations result largely from FDIC's efforts to market
the failed bank.  On a case-by-case basis, FDIC develops a marketing
strategy for how to offer a failed bank to potential acquirers.  The
marketing strategy is shaped, according to FDIC policy, by the unique
characteristics of the institution and marketing conditions at the
time the strategy is developed.  Typically, the marketing strategy
includes any one approach or a combination of approaches to selling
the failed bank's assets.  In most cases, FDIC's marketing strategy
includes a potential bid framework for both purchase and assumption
and insured deposit transfer transactions.  Each bid received is an
individual resolution alternative that FDIC considers and evaluates
in the agency's least-cost determination.  According to FDIC policy,
the agency considers bids that conform to the identified marketing
strategy (conforming bids) as well as bids that do not conform to the
marketing strategy (nonconforming bids). 

To assess FDIC's compliance with FDICIA's least-cost rule, GAO
reviewed a judgmental sample of 22 bank resolutions as well as FDIC's
resolution procedures and interviewed FDIC officials. 


--------------------
\2 In rare circumstances, unique resolution approaches are taken. 
For example, the recent resolution of CrossLand Savings Bank used a
public stock offering to multiple investors.  See Failed Bank:  FDIC
Sale of CrossLand Conservatorship Satisfied Least-Cost Test
(GAO/GGD-94-109, Apr.  20, 1994). 

\3 FDIC has also temporarily controlled and operated failed banks as
"bridge banks" until acceptable acquirers are selected and approved. 


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

Generally, FDIC complied with the provisions on calculating the costs
of resolution alternatives and documenting cost evaluations during
1992.  FDIC's procedures to estimate the net realizable value of a
failed bank's assets and to document cost evaluations improved in the
latter half of 1992 as FDIC implemented and gained experience in
using new procedures to estimate the net realizable value of the
assets of failed banks.  For a variety of reasons, some factors in
the cost estimates were more uncertain than others.  These factors
included estimates of insured and uninsured deposits, future losses
from loss-sharing agreements, and the future market value of bridge
banks. 

GAO's analyses of 22 sample resolutions indicated that FDIC
consistently chose the resolution alternative FDIC determined to be
the least costly compared to other alternatives considered.  In 18 of
the 22 cases, FDIC did not document its rationale for the marketing
strategy it selected.  FDIC's marketing decisions should be
thoroughly documented and submitted to the FDIC Board of Directors
for its consideration in making the least costly resolution decision. 
Although technically not required by FDICIA, such documentation would
give the fullest effect of FDIC's statutory mandate to resolve failed
banks in the least costly method. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4


      FDIC GENERALLY COMPLIED WITH
      COST CALCULATION AND
      DOCUMENTATION REQUIREMENTS
-------------------------------------------------------- Chapter 0:4.1

FDIC's methods to estimate the net realizable value of failed bank
assets improved in the latter half of 1992.  FDIC used three methods
in 1992, all of them consistent with FDICIA calculation requirements. 
In the first half of 1992, FDIC generally used its pre-FDICIA method
of reviewing on-site samples of assets and estimating the effect of
current market conditions and risk factors on asset values by asset
categories.  By mid-1992, FDIC was using an improved on-site review
of assets based on more narrowly defined categories, which enabled
greater specificity and reliability in valuations.  This improved
technique led to clearer documentation of the bases for underlying
assumptions.  In certain circumstances where an on-site review of the
failing bank's assets was not possible, FDIC used a third valuation
method based on off-site statistical projections.  FDIC officials
stated that these projections were not as precise as the valuations
based on on-site reviews. 

Some resolution cost factors involved greater uncertainty than
others.  Most uncertain were estimates of uninsured deposits and
losses to FDIC under loss-sharing and other asset disposition
strategies.  Typically, FDIC lacked precise information on uninsured
deposit amounts because of the (1) complexity of deposit insurance
coverage rules, (2) inadequacy of banks' reported financial
information on deposit accounts, and (3) potential for deposits to be
withdrawn at or near the time of bank closure.  Estimates of losses
under the relatively new loss-sharing agreements were uncertain
because FDIC lacked historical data for accurate projections.  FDIC
realizes the need for such historical data and has initiated efforts
to develop such information.  Furthermore, estimates of losses under
loss-sharing agreements were also uncertain because FDIC could not
precisely predict future market conditions, including possible
changes in asset values, interest rates, or economic conditions. 

In all 22 bank resolutions that GAO reviewed, FDIC chose the
resolution alternative that the agency determined was the least
costly of all resolution alternatives it considered and evaluated. 
In all 22 cases, FDIC generally evaluated all bids
received--including both conforming and nonconforming bids.  The 22
cases included a total of 206 bids, 34 (about 17 percent of the 206
bids received) of which were nonconforming bids.  Of the
nonconforming bids, 20 (about 10 percent of the 206 bids received)
varied in substantial ways from the conforming bids.  In 4 of the 22
resolutions, FDIC determined that the nonconforming bid was the least
costly resolution alternative. 


      IMPROVED DOCUMENTATION OF
      MARKETING STRATEGIES IS
      NEEDED
-------------------------------------------------------- Chapter 0:4.2

GAO recognizes that, as a practical matter, FDIC must make judgments
regarding how best to offer an institution for sale.  In developing a
marketing strategy, FDIC has considerable discretion to construct a
strategy by selecting from among a large number of variations within
the basic resolution methods in any given resolution.  At the same
time, the process by which FDIC selects its marketing strategy can
affect the range of alternatives that are later considered in
least-cost calculations once bids are received. 

FDIC guidelines for developing marketing strategies were generally
reasonable and designed to ensure that all resolution possibilities
were to be considered in the development of such marketing
strategies.  However, the guidelines provide greater assurance of
this for failed banks with assets under $1 billion than for those
with assets greater than $1 billion.  Under the guidelines, FDIC
regional staff are to develop marketing strategies for banks having
assets under $1 billion, and FDIC headquarters staff are to formally
review these strategies.  The guidelines also require FDIC
headquarters staff to develop marketing strategies for banks having
assets over $1 billion, but the guidelines do not provide for a
formal review of the strategies developed. 

GAO was unable to determine from available documentation how FDIC
arrived at the marketing strategy for most of the 22 resolutions
reviewed.  In all 7 of the resolutions involving failed banks with
assets over $1 billion and in 11 of the 15 resolutions involving
smaller banks, GAO found that documentation of the rationale for the
marketing strategy used to solicit bids was inadequate for a
determination of FDIC's reasons for (1) selecting or rejecting
various resolution methods, (2) offering or not offering loss-sharing
agreements on certain assets, or (3) not offering poor quality assets
for sale.  Because these marketing decisions determined how failed
banks were presented to potential acquirers, they may have affected
the ranges of alternatives considered by both FDIC and potential
acquirers. 

To give the fullest effect to FDIC's statutory mandate to choose the
least costly method, GAO believes that marketing decisions should be
thoroughly documented and reviewed.  In particular, the record in
each resolution should address those methods that are potentially
available and explain FDIC's rationale for selecting some and
rejecting others.  Further, the record should reflect a formal review
of the marketing strategy developed to ensure all resolution
possibilities have been considered.  GAO believes that thoroughly
documenting the record in each case would both enhance the quality of
FDIC's decisionmaking and provide greater assurance to Congress and
the public that resolution costs are being minimized. 


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

GAO makes recommendations designed to improve the marketing aspect of
FDIC's resolution process.  It recommends that the Acting Chairman of
the FDIC require the Division of Resolutions to (1) document the
rationale for its marketing strategies for resolving all failing or
failed banks and (2) submit the documented record of the marketing
strategies to the FDIC Board for its consideration in making the
least-cost resolution decisions. 


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

FDIC provided written comments on a draft of this report.  These
comments are presented and evaluated in chapter 3 and reprinted in
appendix VI.  FDIC concurred with GAO's findings and recommendations. 
FDIC said that it is updating its internal procedures to require
documentation of the rationale for its marketing strategies and
submission of the documented marketing strategies to the FDIC Board. 
GAO believes that these procedures, if effectively implemented,
should provide FDIC better assurance that its statutory mandate to
choose the least costly resolution method is being met. 


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

Resolving failed banks is a primary responsibility of the Federal
Deposit Insurance Corporation (FDIC).  Until 1991, FDIC could select
any resolution alternative for a failed bank as long as that method
was less costly to FDIC's insurance fund than liquidation, including
the payment of insured accounts.  However, the Federal Deposit
Insurance Corporation Improvement Act of 1991 (FDICIA) generally
requires FDIC to select the resolution alternative it estimates to be
the least costly to its insurance fund.\1 The statute also requires
FDIC to follow specific rules for calculating the costs of resolution
alternatives and documenting the agency's evaluations of those costs,
including the underlying assumptions of those evaluations.  Finally,
the statute requires us to report to Congress annually on FDIC's
compliance with these requirements. 


--------------------
\1 Section 13(c)(4) of the Federal Deposit Insurance Act, as amended
by FDICIA, P.L.  102-242, 105 Stat.  2236 (1991), effective December
19, 1991. 


   BANK RESOLUTIONS ACCELERATED
   BEGINNING IN THE MID-1980S
---------------------------------------------------------- Chapter 1:1

From its formation on January 1, l934, through 1942, FDIC handled an
annual average of 47 failed banks.  From 1943 through 1985, FDIC
averaged only 11 failed banks a year.  However, from 1986 through
1993, FDIC handled an average of 155 annual failures.  During this
8-year period, failed bank assets totaled over $226 billion, and
resolutions cost FDIC's insurance fund an estimated $31.93 billion. 
As shown in table 1.1, the number of failed bank resolutions had
declined from its 1988 peak of 221 institutions to 42 institutions
during 1993. 



                          Table 1.1
           
           FDIC Resolved and Assisted Banks, 1986-
                             1993

                    (Dollars in billions)

Year                       Number of banks      Total assets
------------------------  ----------------  ----------------
1986                                   145             $7.63
1987                                   203              9.23
1988                                   221             52.62
1989                                   207             29.40
1990                                   169             15.74
1991                                   127             63.40
1992                                   122             44.23
1993                                    42              4.06
============================================================
Total                                1,236           $226.31
------------------------------------------------------------
Source:  FDIC Failed Bank Cost Analysis 1986-1992 and DOR 1993
statistics. 

The significant increase in bank resolution activity since the
mid-1980s led FDIC to reorganize its operating divisions--Division of
Supervision (DOS) and Division of Liquidation (DOL).\2 Before the
reorganization, DOS was responsible for resolving failed banks with
assets in excess of $100 million.  The primary responsibility of DOS
was to examine state-chartered institutions that were not members of
the Federal Reserve System.  DOL resolved small failed institutions
(those having less than $100 million in assets).  The primary
responsibility of DOL was to sell failed institution assets remaining
and assumed by FDIC after resolution. 

In March 1991, FDIC established the Division of Resolutions (DOR) to
plan for and handle bank resolutions.  By centralizing much of the
work related to bank failures that had been split between DOS and
DOL, FDIC expected its new division, DOR, to yield increased
efficiency, reduced costs, consistent decisionmaking, and more
in-depth resolution expertise.  DOR was staffed primarily by
transferees from the Resolution Trust Corporation (RTC), DOS, and
DOL; thereby enabling DOR to assume its responsibilities with a
historical perspective of how FDIC resolved failing institutions.  In
addition, some senior management positions were filled through
external hires from the financial services industry and from other
federal agencies. 


--------------------
\2 In October 1993, DOL was reorganized and renamed the Division of
Depositor and Asset Services.  DOS currently operates under its
original name. 


   FDIC'S EVOLVING RESOLUTION
   METHODS
---------------------------------------------------------- Chapter 1:2

Initially, FDIC had two methods for handling bank failures:  deposit
payoffs or deposit assumptions.  When deposits were paid off, FDIC
made payments directly to the depositors up to the insurance limit
and subsequently sold the failed bank's assets.  The cost of the
payoff essentially was the difference between the amount disbursed to
insured depositors minus the net funds received from asset sales.  In
deposit assumption transactions, an acquiring bank assumed the
insured, and generally the uninsured, deposits of the failed bank. 

By the late-1960s, FDIC began to actively pursue purchase and
assumption (P&A) transactions.  In P&A transactions, the acquiring
bank agreed not only to assume deposits and pay a premium, but also
to purchase at least some of the failed bank's assets.  FDIC handled
most bank failures this way until the late-1980s. 

By 1987, FDIC's preferred resolution method evolved to a whole bank
P&A transaction in which most if not all of a failed bank's assets
and deposits, as well as other bank liabilities, were acquired by
another institution.  This method generally protected uninsured
depositors and minimized the assets requiring sale by FDIC.  Other
methods that were available to FDIC, and are still being used to
resolve failing or failed banks, included other types of P&A
transactions, an insured deposit transfer (IDT), open bank
assistance, and bridge bank arrangements.  In each of these methods,
except for IDT, the acquirer has the option of assuming all deposits
or only the insured deposits.  Following is a brief description of
these other types of resolution methods. 

  Other P&A transactions. 

  Clean bank P&A.  The acquirer purchases only high-quality assets,
     which are basically those loans that are performing and have a
     high likelihood of being repaid. 

  P&A transactions with asset pools.  The acquirer's purchases could
     include many variations of asset groupings that are referred to
     as asset pools. 

  P&A transactions with putback or loss-sharing agreements.  These
     agreements are used for asset groupings that both FDIC and
     acquirers believe may be difficult to get borrowers to repay,
     such as those loans in which the collateral value has fallen
     below the loan amount remaining due.  Putback agreements enable
     the acquirer to return certain assets to FDIC within a specified
     time period, while loss-sharing agreements enable the acquirer
     to recover from FDIC a stipulated percentage of losses incurred
     on certain purchased assets.  During 1992, several resolutions
     included loss-sharing agreements on certain assets, in which
     FDIC will reimburse acquirers 80 percent of net charge-offs, up
     to a pre-set limit, of those assets over a specified period of
     time.\3

  Insured deposit transfer.  The acquirer accepts the failed bank's
     insured deposits and makes them available to depositors.  The
     acquirer may also assume certain other liabilities and purchase
     some of the assets. 

  Open bank assistance.  FDIC can provide financial assistance to a
     troubled bank before it is declared insolvent or closed.  In
     such instances, FDIC uses its marketing techniques to actively
     seek an acquirer for the bank or otherwise ensures itself that
     the bank's current management can provide for the bank's
     viability. 

  Bridge bank.  FDIC can establish a bridge bank to take interim
     control of the operations of a troubled bank.  It can operate
     the bank temporarily (up to 2 years with the option for three
     1-year extensions) to preserve the franchise value--that is, its
     value as an ongoing entity--until a final resolution decision is
     made. 

FDIC's preference for whole bank transactions did not preclude
potential acquirers from making other types of bids.  However, FDIC's
bid evaluation process prior to FDICIA clearly focused on its
preference for whole bank bids.  The bid evaluation process was to
separate whole bank bids from all other types of bids received.\4
Staff were to then evaluate the whole bank bids, select the best one,
and determine whether it would be less costly to FDIC than a
liquidation.  If so, staff were to recommend its acceptance and stop
the bid evaluation process.  All other bids were to be returned
unopened to the potential acquirers.  If staff found there were no
acceptable whole bank bids, they were to evaluate other types of bids
based on a pre-established order, which gave preference to the
greater amount of assets to be purchased by the potential acquirer. 
Since FDIC did not evaluate all bids received, it did not have the
data to know the effect its preference for the whole bank resolution
method had on resolution costs and thus on its deposit insurance
fund. 

As the number of institutions needing resolution increased in the
late-1980s, so did their cost, and thus the losses, to FDIC's deposit
insurance fund.\5 Basically, the deposit insurance fund receives
revenue from deposit insurance premiums from FDIC-insured
institutions, interest earned on investments in U.S.  Department of
the Treasury obligations, and sales of FDIC-owned failed bank assets. 
The fund's proceeds are used to meet FDIC's obligation to provide
insurance coverage for insured deposits in a failed institution and
to provide financial assistance in resolving that institution.  In
1987, the fund's reserves reached $18.3 billion, the highest level
ever.  However, the upsurge in bank failures caused the fund to lose
more than $25 billion over the next 4 years, and by December 1991, it
was $7 billion in the red. 

Congress became increasingly concerned that the declining health of
the banking industry and mounting BIF losses would necessitate a
taxpayer-assisted bailout similar to that for the savings and loan
industry.  Congress also questioned whether BIF should continue to
pay uninsured depositors.  Paying uninsured depositors was typically
part of FDIC's preferred resolution strategy for whole bank
transactions. 


--------------------
\3 For the loss-sharing agreements used in 1992 resolutions, FDIC
typically increased the loss-sharing level to 95 percent for losses
above the pre-set limit. 

\4 FDIC refers to this process, which established the agency's
priority for evaluating bids, as the "Robinson Resolution."

\5 The deposit insurance fund was named the Bank Insurance Fund (BIF)
with the enactment of the Financial Institutions Reform, Recovery,
and Enforcement Act of 1989.  BIF was established to provide deposit
insurance to all federally insured banks. 


   CONGRESS PASSED FDICIA
---------------------------------------------------------- Chapter 1:3

To address its concerns about the financial health of the deposit
insurance funds, Congress in December 1991 passed the least-cost
resolution provisions contained in FDICIA.  Essentially, the purpose
of FDICIA, as it relates to resolutions, is to (1) provide backup
funding for federal deposit insurance and (2) reduce losses when
depository institutions fail.\6 Because of the deteriorating
condition of BIF, Congress increased FDIC's line of credit with the
Treasury to $30 billion and required FDIC to prescribe a
recapitalization schedule for BIF. 

Congress sought to limit BIF losses, in part, by improving the
regulatory process.  In addition, FDICIA contains key provisions
pertaining to the resolution of failing or failed banks.  Most
significantly, Section 141 requires FDIC to choose the resolution
method least costly to BIF of all possible methods for meeting FDIC's
obligation to provide insurance coverage for insured deposits in
failed institutions.\7 To make this least-cost determination,\8 FDIC
must

  consider and evaluate all possible resolution alternatives by
     computing and comparing their costs on a present-value basis,
     using realistic discount rates;

  select the least costly alternative based on the evaluation;

  document the evaluation and the assumptions on which it is based,
     including any assumptions with regard to interest rates, asset
     recovery rates, asset holding costs, and contingent liabilities;
     and

  retain documentation for at least 5 years. 

FDICIA further required that we annually audit FDIC's compliance with
the least-cost provisions. 

In passing FDICIA, Congress expected to reduce the cost of bank
failures to BIF.  For instance, Congress intended for uninsured
depositors to share in resolution costs.  Thus, FDIC now can provide
coverage to uninsured depositors only in conjunction with a
least-cost resolution. 

In an earlier report on FDIC's resolution of CrossLand Savings
Bank,\9 we were critical of the quality and extent of its
documentation.  In that report, we established the following
criteria, which FDIC accepted. 

  Documentation should be clear, consistent, concise, and complete so
     that an outside observer can identify and understand the
     estimated cost of each alternative, including the assumptions
     and discount rates used. 

  Data sources for the cost evaluations should be clearly identified
     so that cost figures can be traced to their source(s). 

  Assumptions integral to the cost evaluations should be documented
     and supported.  In particular, each assumption should be (1)
     clearly identified and (2) supported by empirical data or, in
     the absence of such data, by judgment based on relevant
     experience.  This support should be explicitly described in the
     documentation and, where appropriate, the source(s) used in
     making the assumption should be identified. 

  If there is uncertainty about the validity of an assumption that
     materially affects the cost evaluation results, some effort to
     gauge that uncertainty should be made and documented by showing
     a range of possible outcomes. 

We also used these criteria in assessing the sampled resolutions for
this review to determine the adequacy of FDIC's documentation of its
cost evaluations and underlying assumptions. 


--------------------
\6 FDICIA applies to both FDIC and RTC relative to their
responsibilities for resolving failing or failed banks or thrifts,
respectively.  We reported separately on the RTC's compliance with
FDICIA's least-cost provisions.  See 1992 Thrift Resolutions:  RTC
Policies And Practices Did Not Fully Comply With Least-Cost
Provisions (GAO/GGD-94-110, 1994). 

\7 Section 13(c)(4)(G) of the Federal Deposit Insurance Act provides
for a systemic risk exception to the least-cost requirement if a
finding is made that compliance with the least-cost requirement would
have serious adverse effects on economic conditions and that a more
costly alternative would mitigate such adverse effects.  To date,
FDIC has not relied on this exception. 

\8 Section 13(c)(4) imposes specific requirements concerning the
treatment of certain costs.  For example, Section 13(c)(4)(B)(ii)
provides that if FDIC considers an alternative that would result in
forgone federal tax revenues, it must treat those revenues as if they
were revenues forgone by the deposit insurance fund.  Section
13(c)(4)(D) deals with the calculation of liquidation costs for
purposes of cost comparisons. 

\9 Failed Bank:  FDIC Documentation of CrossLand Savings, FSB,
Decision Was Inadequate (GAO/GGD-92-92, July 7, 1992). 


   FDIC'S CURRENT RESOLUTION
   PROCESS
---------------------------------------------------------- Chapter 1:4

DOR is responsible for resolving failed banks within FDIC.  A bank
fails when its chartering authority, the Office of the Comptroller of
the Currency (OCC) for national banks or a state authority for state
chartered banks, closes the institution.  At that time, FDIC is
usually appointed receiver.\10 The FDIC Board may also appoint FDIC
as receiver for any insured depository institution, after
consultation with the appropriate federal and state regulators, when
such action is necessary to reduce the risk of loss to the deposit
insurance fund. 

The marketing and resolution of failed banks is done through either a
regional or headquarters resolution process.  The processes are quite
similar.  Generally, DOR's regional offices handle institutions with
less than $1 billion in total assets, known as regional resolutions,
while larger institutions or major resolutions are done at DOR
headquarters in Washington, D.C. 

FDIC's formal resolution activities commence when it receives a
letter from the chartering authority advising FDIC of the imminent
failure.  DOS then is to provide DOR a failing bank case memorandum,
which summarizes the bank's problems, its ownership, and the results
of previous examinations.  DOS also is to give DOR a list of
institutions approved to bid on the bank because they meet certain
criteria such as asset size and minimum capital requirements.\11
Figure 1.1 shows FDIC's typical resolution process. 

   Figure 1.1:  FDIC's Typical
   Resolution Process

   (See figure in printed
   edition.)

\a In October 1993, DOL was reorganized and renamed the Division of
Depositor and Asset Services. 

Source:  GAO review of DOR procedures. 

DOR then is to prepare an information package on the institution,
which provides an in-depth description and accounting of the bank's
financial condition.  The purpose of the information package is to
inform potential buyers of the composition of assets to be acquired
and liabilities to be assumed.  DOR staff also are to use the
information package to begin developing a marketing strategy. 

In selecting a marketing strategy, DOR is to consider the alternative
resolution methods identified in its Resolutions Procedures Manual. 
Numerous approaches may be considered--including some very complex
combinations of resolution approaches.  For those resolutions done in
the field, regional officials are to prepare and submit to a DOR
headquarters official a failing bank recommendation, which is to
outline those resolution methods considered viable.  The headquarters
official is to review this recommendation and approve the resolution
methods to be offered to potential acquirers in DOR's marketing
strategy.  For those resolutions done at headquarters, DOR
headquarters officials are to determine the marketing strategy.  FDIC
sells no assets of a failed bank until the bank has been closed and
turned over to FDIC for resolution.  FDIC's marketing activities and
its analysis of resolution alternatives are discussed further in
appendix III. 

Concurrent with developing a marketing strategy, DOR staff are to
contract for or prepare a valuation of the failed bank's assets.  The
valuation is an estimate of the amount that FDIC would recover if it
were to sell the assets--the net realizable value.  The net
realizable value usually differs from the book value of the assets,
as recorded in the information package, because of reasons such as
costs associated with the length of time FDIC holds the assets prior
to sale; changes in interest rates since the loan was initially made;
and changes in the value of the underlying collateral, which may
affect the amount recovered by FDIC on the assets.  Detailed computer
models are to be used in valuing assets that employ numerous
variables and estimates, making this a very complex process.  In the
event of a deposit payoff and asset liquidation, FDIC's total
resolution cost is basically the sum of payments to insured
depositors and secured creditors minus the net realizable value on
the bank's assets. 

Creditor analyses are also to be done at this time.  Upon receiving
DOL's estimate of the amount of uninsured deposits, DOR is to
estimate the amount of insured deposits in the bank.  FDIC's legal
staff estimates the amount of contingent liabilities.  These
liabilities may include standby letters of credit, unfunded loan
commitments, or liabilities from pending litigation. 

After agreeing on a marketing strategy, DOR officials are to hold an
information meeting with potential acquirers.  The acquirers are
invited from the previously mentioned DOS bidders list.  The proposed
transaction is to be discussed and those with continuing interest in
the institution are to sign a confidentiality agreement, receive the
information package, and secure permission to perform due diligence. 
Due diligence is the bidder's on-site inspection of the books and
records of the institution and an assessment of the value of the
assets and liabilities done for the purpose of preparing a bid. 

During the information meeting, DOR is to advise potential acquirers
when bids should be submitted.  Bids received are evaluated and
compared with each other and FDIC's estimated cost of liquidation in
order for DOR to arrive at the least costly resolution alternative. 
Once the least cost test is complete, a decision package is to be
prepared for FDIC's Board of Directors requesting approval of the
transaction.  The decision package is also to include information
about the share of the estimated loss that should be absorbed by
uninsured depositors, and whether an advanced dividend should be paid
to uninsured depositors so that they have a portion of their
deposited funds, while FDIC proceeds with the resolution and
disposition of the remaining assets. 

The FDIC Board is responsible for making the final resolution
decision.  Resolution decisions involving institutions with assets of
$300 million or more are to be scheduled for discussion by the Board. 
The Board may accept the cost analysis provided by DOR, seek
additional information, or request DOR to provide a new cost analysis
based on factors identified during the Board meeting.  Resolutions
involving institutions with less than $300 million in assets are
placed on the Board's summary agenda.  FDIC officials indicated that
resolutions on the summary agenda may be approved by the Board
without discussion.  At times, for these smaller resolutions, the
Board may delegate authority to DOR to select the final resolution
alternative.  In such cases, DOR must report its decision to the
Board. 

This entire resolution process is generally carried out between the
time the chartering authority advises FDIC that a bank is in imminent
danger of failing and the time the charterer appoints FDIC as
receiver.  This time period can be as short as a few weeks or as long
as several months.  FDIC is required to make a least-cost
determination at the time it makes a determination to provide
financial assistance to the institution.  Thus, the time available
and access to bank information are critical to the precision of the
numerous estimates that must be made in determining the least costly
resolution.  However, some imprecision will always remain in
least-cost determinations because of factors such as complex deposit
insurance rules and changes resulting from future market and economic
events. 


--------------------
\10 A receivership is the functional equivalent of bankruptcy for a
failed bank.  FDIC, as receiver, is responsible for the disposition
of the assets of a failed bank and repayment of its creditors to the
best extent possible.  For national banks, OCC is required to appoint
FDIC as receiver. 

\11 DOS generally consults with other federal regulators on the
bidders list and with the primary regulator for input for the case
memorandum on national or state member banks. 


   OBJECTIVE, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:5

The overall objective of our review was to determine the extent to
which FDIC complied with FDICIA requirements to select the least
costly method of resolving institutions.  In this first report on
FDIC's compliance, we surveyed the FDIC resolution process, which
continues to evolve, to identify the key controls and potential
vulnerabilities.  From our survey, we concentrated our detailed
analysis on FDIC's process for (1) calculating the cost of resolution
alternatives and (2) documenting its evaluation of those alternatives
considered in selecting the least costly resolution alternative. 

To address our objective, we judgmentally selected for review 42 of
the 122 banks resolved by FDIC during 1992.  The 42 banks--all of
which were resolved after passage of FDICIA--had assets of $29.8
billion and had an estimated resolution cost to BIF of $2.2 billion,
while the 122 banks had assets of $44 billion and had estimated BIF
losses of $4.7 billion.  The 42 banks represented 22 resolutions,
since 20 banks were included in a resolution of First City
Bancorporation of Texas and 2 banks were included in a resolution of
American Savings Bank.  Our selection criteria included headquarters
and regional resolutions and varying attributes such as different
types of resolution methods.  Of the 22 resolutions, 7 major
resolutions were done at DOR headquarters in Washington, D.C., and 15
regional resolutions were done at DOR regional offices--5 each in
Boston, New York, and San Francisco.  See appendix I for profile
information relative to the 22 sampled resolutions. 

We developed a data collection instrument to document and track the
information gleaned from our perusal of the resolution case files. 
We collected data from the inception of resolution activity through
the final resolution decision.  In particular, we focused on DOR's
approaches to marketing the institution, performance of asset
valuations, documentation of the assumptions used, bids received and
costed via application of the cost test, and the treatment of
uninsured depositors. 

Since asset valuations are a crucial input to the least-cost
determination, we reviewed in considerable detail the computer
programs underlying the models developed by FDIC to value assets.  We
concentrated primarily on the assumptions and financial calculations
used in the models to determine whether they would result in
reasonable valuations.  We also discussed FDIC's program logic and
methodology with officials from large private organizations actively
involved in performing due diligence and purchasing bank-type or
similar assets.  Obtaining information on their asset valuation
approaches provided us a basis for assessing the adequacy of FDIC's
efforts.  We did not review the supporting documentation of FDIC's
valuations of thousands of individual bank assets to ensure their
accuracy.  Instead, we focused on the valuation process and how its
results were used in FDIC's cost tests for our sampled resolutions
with the limited time and resources available for this first annual
review. 

To further address our audit objective, we studied FDIC's resolution
process including reviewing its procedures and practices and
interviewing numerous FDIC officials, including those in the FDIC
Inspector General's Office responsible for reviewing resolutions
programs and activities.  We also reviewed the financial calculations
developed by DOR to determine whether the estimated cost of
resolution alternatives was accurately calculated. 

We assessed the adequacy of FDIC's resolution process to determine
the least costly resolution alternative based on the criteria
developed in our earlier report on CrossLand Savings Bank.\12 We did
not determine whether, in fact, the least costly resolution
alternative resulted.  The ultimate cost of a resolution cannot be
identified until all remaining assets are sold and liabilities are
paid by FDIC as receiver, which generally takes several years. 

We did our work between October 1992 and December 1993 at FDIC
headquarters in Washington, D.C., and DOR regional offices in Boston,
New York, and San Francisco.  FDIC provided written comments on a
draft of this report.  These comments are presented and evaluated in
chapter 3 and are reprinted in appendix VI.  Our work was done in
accordance with generally accepted government auditing standards. 


--------------------
\12 GAO/GGD-92-92, July 7, 1992. 


FDIC COMPLIED WITH FDICIA'S COST
CALCULATION AND DOCUMENTATION
REQUIREMENTS
============================================================ Chapter 2

FDIC improved its processes for calculating and documenting
resolution cost estimates throughout 1992.  By mid-year, FDIC had
gained experience with the new FDICIA requirements and had developed
processes to satisfy those requirements.  Also, FDIC's related
documentation procedures had been sufficiently improved to ensure
compliance with FDICIA's requirements for calculating estimated
resolution costs and documenting cost evaluations and their
underlying assumptions.  The least-cost provisions contained in
FDICIA were effective on December 19, 1991, immediately upon the
act's enactment. 

Resolution decisions involve many uncertainties, including
unanticipated gains and losses in the value of assets held in
receivership and the outcome of lawsuits pending at the time of a
resolution decision.  As discussed later in this chapter, other
uncertainties relate to estimates of uninsured deposit payouts and
future proceeds under loss-sharing and other asset disposition
agreements.  FDIC recognized the need and opportunity to reduce such
uncertainties and initiated positive efforts in 1992.  However, the
complexity of insurance coverage rules and unreliable bank-reported
information limit improvements that can be made in estimating total
uninsured deposits in failed banks.  Also, decisions to operate a
failed bank as a bridge bank also involve uncertainties and raise
policy issues about uninsured depositors. 


   FDIC COMPLIED WITH FDICIA'S
   REQUIREMENTS ON CALCULATING
   RESOLUTION COSTS
---------------------------------------------------------- Chapter 2:1

As outlined in chapter 1, one of FDIC's first activities to resolve a
failed bank is that DOR prepares an information package on the
institution's assets, deposits, and other liabilities.  Information
in the package is based on the institution's financial records as of
the date the package is prepared.  This package provides potential
acquirers and DOR with a common base of information to estimate (1)
the amount that potential acquirers may bid for the institution's
assets and deposits and (2) the net realizable value of the
institution's assets. 

Basically, DOR is to estimate liquidation costs by calculating the
difference between the result of DOR's asset valuation and the amount
to be paid to insured depositors.  Therefore, asset valuations are
critical to determining the cost of liquidation; the cost of
liquidation is, in turn, critical to FDIC's determination of the
least costly resolution alternative.  To make this determination, DOR
first is to compare the estimated liquidation cost with the bids from
potential acquirers to see if an acquisition is less costly than a
liquidation.  If so, then FDIC compares the bids received to
determine which would be the least costly to the Bank Insurance Fund
(BIF). 

While some types of assets can be valued with a fair degree of
precision, others must be valued judgmentally on the basis of a
number of often uncertain factors.  The amount of cash can be
precisely determined as can the value of marketable securities from
readily available financial sources.  However, the values of loans,
which typically make up the bulk of bank assets, are more difficult
to determine.  Loans such as mortgages and consumer loans must be
valued with consideration to the risk associated with repayment, the
value of any underlying collateral, and other factors.  Even more
imprecise are the valuations of loans that involve borrower defaults
and bank foreclosures. 

FDICIA requires FDIC to calculate the cost of each resolution
alternative on a present-value\1 basis and to use a "realistic
discount rate." This provision primarily affects the method that FDIC
uses to adjust the book value (or the value reflected in bank
records) of a failed bank's assets when valuing the assets.  A
failing bank's records generally reflect asset values based on
historical cost.  FDIC has found that these values generally
overstate the value that can be recovered through the agency's
liquidation of the assets.  FDICIA leaves to FDIC's discretion the
definition of realistic discount rates, as well as the specific
factors the agency considers in valuing assets such as the holding
costs that may be incurred by FDIC pending the sale of assets, which
can affect the amount FDIC may realize once assets are sold. 


--------------------
\1 Present-value analysis is used to calculate the current value of a
future payment or stream of payments.  That is, it recognizes and
adjusts for the fact that $100 received 5 years from now is worth
less than $100 today because of the time value of money, risks, and
inflation. 


      FDIC USED THREE MODELS TO
      ESTIMATE THE VALUE OF
      FAILING BANK ASSETS
-------------------------------------------------------- Chapter 2:1.1

During 1992, FDIC used three models to estimate the present value of
assets held by failing banks.  Two of the models, the Total Asset
Purchase and Assumption (TAPA) asset review and the Asset Valuation
Review (AVR), evaluated all types of assets held by a failing bank on
the basis of on-site reviews of sampled loan asset files and other
asset records at the bank.  AVR, which replaced TAPA, was an
improvement over the TAPA asset review in that AVR sampled across
more narrowly defined categories of asset pools and provided more
extensive analysis of the asset pools within the portfolio than the
TAPA asset review.  When an on-site review of a failing bank's assets
was not possible because of the bank's liquidity problems such as
limited ability to meet depositors' fund withdrawals or legal
problems, FDIC used a third model, a statistical research model based
on data from FDIC's recovery experience.  FDIC also used the research
model as a means of checking the reasonableness of TAPA and AVR
results, but relied on the TAPA and AVR results after any changes
were made resulting from the comparison to the research model
results.  Appendix IV provides further description of FDIC's three
asset valuation models. 

For all 22 resolutions we reviewed, FDIC used these models in its
valuations of assets to estimate the net present value that the
agency could recover through liquidation and to estimate resolution
costs.  Of the 22 resolutions we reviewed, 5 used the TAPA asset
review; 12 used AVR; and 5 used the research model.  All resolutions
that used TAPA asset reviews were done in early 1992, while the
research model was used occasionally throughout the year, as
circumstances dictated. 


      FDIC'S ASSET VALUATION
      METHODS DESIGNED TO USE
      REALISTIC DISCOUNT RATES IN
      ESTIMATING PRESENT VALUES OF
      ASSETS
-------------------------------------------------------- Chapter 2:1.2

We determined that the asset valuation methods FDIC used in our 22
sampled resolutions complied with the FDICIA calculation rules.  We
also determined that the bases of FDIC's TAPA and AVR valuation
methods--its calculation of present value and the discount rates used
in the calculations--were consistent with those generally used by
private sector firms experienced in valuing and acquiring bank assets
as well as by academicians who have studied such valuations. 

FDIC's selection of discount rates varied according to a variety of
factors, for example, market rates at the time of the resolution and
risks associated with the assets, such as credit risks.  FDIC's
research model also uses a present-value calculation method using a
discount rate reflecting FDIC's cost of funds. 

To determine if FDIC's valuations of assets were calculated based on
present value with realistic discount rates, we reviewed FDIC's asset
valuation methods and FDIC's documentation for determining realistic
discount rates in estimating the present value of a failed bank's
assets.  Also, to better understand the selection of discount rates
and the use of present-value analysis, we reviewed the academic
literature on asset valuation and interviewed officials of five
private sector firms, including banks and securities and investment
firms involved with valuing asset portfolios similar in type to those
of many banks. 

Private sector officials said that present-value analysis requires
assets to be judged in terms of a variety of aspects and market
conditions.  The analysis often requires estimating the timing and
amount of cash flows from an asset to an investor.  The cash flows
generally represent an investor's return (yield) on the asset.  If
the investor's yield differs from current market rates for similar
assets, the asset value--or selling price--is adjusted or discounted
by an amount to essentially provide an investor a yield comparable to
the current market rate. 

The officials said that they used discount rates based on a variety
of factors.  Some of the factors they mentioned included (1) the
firms' investment objectives, (2) the relative quality of the asset
to be acquired, (3) the length of time the asset would be held, and
(4) how the quality of the asset might change during the period held. 


   FDIC IMPROVED OVER TIME THE
   DOCUMENTATION OF ASSUMPTIONS
   UNDERLYING ASSET VALUATIONS
---------------------------------------------------------- Chapter 2:2

FDICIA requires FDIC to document the agency's evaluation of the costs
of resolution alternatives considered, including the assumptions on
which the evaluations are based.  FDICIA specifically requires
documentation of any assumptions that relate to interest rates, asset
recovery rates, asset holding costs, and payments of contingent
liabilities.  It also requires such documentation to be retained for
at least 5 years.  In reviewing the 22 sampled resolutions, we
attempted to identify the underlying assumptions of asset valuations,
including the assumptions specified in the least-cost provisions.  We
focused primarily on DOR's process for documenting its AVR valuations
because in 1992 FDIC developed and implemented this methodology in
response to FDICIA's requirements. 

We generally found that FDIC's capability to document these
assumptions improved after the agency implemented its AVR
methodology.  We had greater difficulty and limited success in
identifying the underlying assumptions made in the research model and
TAPA, which were designed before FDICIA was enacted.  Asset
valuations, using the research model and TAPA, were based on DOL's
asset disposition experience with assets held by failing banks. 
Although holding costs, holding periods, and recoveries on assets are
inherent in that experience, the resolution case files we reviewed
generally lacked documentation that explained specifically how these
factors figured in asset value calculations. 

We found that as DOR gained more experience with AVR, the division
improved required procedures for preparing AVR reports as well as the
procedures for organizing and maintaining resolution case records. 
For example, AVR procedures required documentation of methods used to
value assets.  The AVR reports we reviewed complied with these
procedures by documenting the factors and assumptions underlying the
calculations used to determine the values and losses in the value of
assets held by a failing bank.  We also found that the AVR reports we
reviewed generally met our criteria, as discussed in chapter 1, for
documenting the valuations FDIC expects to use in evaluating the
costs of resolution alternatives. 

DOR made further improvements to the AVR documentation. 
Specifically, it required clear identification, in a separate report
section, of key assumptions about holding periods, holding costs, and
selling expenses.  This section of the report, along with its
discussion of asset valuation methodologies, allowed us to identify
the underlying assumptions of the asset valuation process. 

We found that the underlying assumptions used in AVRs we reviewed
were often based on DOL's asset disposition experience, and in some
cases, on surveys of local firms dealing with the same types of
assets being valued.\2 In valuing the assets for one major bank, for
example, the valuation team contacted local liquidation firms to
determine the current market values for office equipment assets such
as desks and furnishings.  The valuation team also contacted local
art dealers to determine the potential value of an art collection
held by the failing bank. 

To provide an assessment of the effect of a failed bank's letters of
credit, lawsuits filed against and on behalf of a failing bank, and
other contingent liabilities, AVRs we reviewed usually incorporated
estimates from a separate valuation report prepared by a team
composed of FDIC liquidation and legal personnel.  Like recoveries on
assets, the outcome of lawsuits and other contingent liabilities may
not be known for several years after the resolution of a failed bank. 


--------------------
\2 See Financial Audit:  Federal Deposit Insurance Corporation's
Internal Controls as of December 31, 1992 (GAO/AIMD-94-35, Feb.  4,
1994).  We have found internal control weaknesses relative to FDIC's
asset management and disposition information.  FDIC officials have
acknowledged and agreed to initiate actions to overcome those
reported weaknesses. 


   FDIC RESOLVED BANKS IN THE
   LEAST COSTLY ALTERNATIVE OF ALL
   THOSE CONSIDERED
---------------------------------------------------------- Chapter 2:3

In all 22 cases we reviewed, the FDIC Board chose the resolution
alternative that FDIC's DOR had determined to be the least costly of
the alternatives considered.  We found one instance where the Board
revised the cost estimates presented by DOR; however, this revision,
which lowered the estimate of FDIC's loss in this resolution, did not
change DOR's recommendation of the least costly resolution
alternative.  The case was a complex resolution involving a holding
company that controlled 20 banks.  The recommended least-cost
resolution alternative was a bridge bank for each of the 20 banks. 
The Board estimated that FDIC's recovery on assets would be greater
than DOR estimated, mainly because of the market interest in the
banking franchise.  In this resolution case, we were unable to trace
FDIC's analysis of resolution alternatives to documents providing
empirical evidence to support the Board's revised estimate of
recovery on assets.  However, we encountered no similar difficulties
in any other of the 21 resolutions we reviewed.  The closure and
resolution processes for this banking franchise are the subject of a
separate GAO study, which will be issued later this year. 

Table 2.1 shows the resolution alternatives that the FDIC Board
selected as the least costly alternative for the 22 bank failures we
reviewed. 



                          Table 2.1
           
           Resolution Alternatives Determined Least
           Costly by the FDIC Board in 22 Cases, in
                             1992

                            Banks with    Banks with
Least costly resolution   assets under   assets over     All
alternative                 $1 billion    $1 billion   banks
------------------------  ------------  ------------  ------
Whole bank P&A                       0             2       2
P&A                                  8             3      11
IDT\a                                3             1       4
Bridge bank                          0             1       1
Liquidation                          4             0       4
============================================================
Total                               15             7      22
------------------------------------------------------------
\a IDTs included the purchase of assets in three cases.  The acquirer
typically purchased the failed bank's most marketable assets, such as
cash, securities, and some performing loans. 

Source:  GAO analysis of sampled resolution cases. 


      DOR GENERALLY EVALUATED ALL
      BIDS RECEIVED
-------------------------------------------------------- Chapter 2:3.1

In the 22 cases we reviewed, we found that FDIC generally evaluated
all bids received for a failing bank.  FDIC received a total of 206
bids\3 from 89 potential acquirers in the 22 cases we reviewed.  Of
the 206 bids, 34 were nonconforming bids. 

Bids submitted to FDIC fall in two categories--conforming bids, which
match FDIC's marketing strategy and nonconforming bids, which differ
from FDIC's marketing strategy.  A nonconforming bid may or may not
be significantly different from FDIC's recommended conforming bid. 
For example, in one of our cases, several bidders for the institution
did not want to purchase a certain subsidiary and excluded it from
their bids.  In another case, an unsuccessful bidder rejected FDIC's
proposed loss-sharing structure and instead submitted a bid to
service an asset pool of classified loans and purchase only
performing assets. 

FDIC received 34 nonconforming bids for 12 of our sampled cases.  We
found that 14 nonconforming bids contained minor differences from
DOR's marketing strategy, while 20 nonconforming bids contained major
differences from DOR's marketing strategy.  In four cases, the
nonconforming bid was determined to be the least costly resolution
alternative with one nonconforming bid containing major differences
from DOR's marketing strategy.  For the 34 nonconforming bids, DOR
was able to evaluate all but 5 bids in the cost test.  For the
nonconforming bids, which DOR could not evaluate, we found that DOR
provided a reasonable explanation for not analyzing these bids.  For
example, in one case, DOR staff was unable to accurately determine
the cost of providing the tax indemnification requested by the
bidder.  Appendix II contains summary information on the bids
received by FDIC. 


--------------------
\3 Of the 206 bids received, 111 bids were submitted by 32 bidders
for the 20 First City Bancorporation banks. 


   UNCERTAINTIES IN RESOLUTION
   DECISIONS
---------------------------------------------------------- Chapter 2:4

In making resolution decisions, the FDIC Board is faced with a great
many uncertainties.  The uncertainties include, among others,
unanticipated gains or losses from subsequent receivership asset
sales and lawsuits filed against or on behalf of the failed bank that
can affect actual recoveries on assets estimated at the time of the
resolution decision.  While many failed bank's assets placed in
receivership are sold during the first 5 years, 10 years or more may
pass before asset sales are essentially completed and a receivership
is terminated.  Asset recoveries are subject to uncertainties because
of changing economic and market conditions affecting asset values. 
Such changes could affect actual recoveries from the level estimated
at the time of the resolution decision. 

Other sources of uncertainty in resolution decisionmaking are the
estimates of uninsured deposit payouts and future proceeds under
loss-sharing and other asset disposition agreements.  In addition,
bridge bank decisions by FDIC involve unique uncertainties. 


      ESTIMATES OF UNINSURED
      DEPOSITS
-------------------------------------------------------- Chapter 2:4.1

As discussed earlier in this report, the amount to be paid to
depositors is a key factor in determining the cost of resolution of a
failed bank.  Even so, FDIC did not have precise estimates of insured
and uninsured deposits at the time of the resolution decision because
of the complexity of the insurance coverage rules and the related
inadequacy of bank financial information about deposit accounts that
is reported to FDIC.  Various aspects of uninsured deposits are
further discussed in appendix V. 

Before the passage of FDICIA, FDIC or the acquirer of the entire
deposit base of a failed bank generally paid uninsured depositors. 
This mitigated, to a large extent, FDIC's need for an accurate
determination of the insurance status of deposits in making a
resolution decision.  Under FDICIA, however, the cost of paying off
uninsured depositors must be considered in least-cost decisions; and
payments of uninsured deposits are generally restricted to cases
where such payments can be made consistent with a least-cost
resolution decision.\4

Depending upon available time and information, FDIC generally relied,
and continues to rely, on an on-site review to determine the
insurance status of deposit accounts.  The information reported by
banks basically only identifies those accounts over $100,000, which
may or may not be covered under the complex deposit insurance rules. 
Our review of sampled major resolution cases showed that following
the enactment of FDICIA, FDIC has experienced difficulty in
accurately determining the amount of uninsured deposits held by a
failing bank.  In some cases, FDIC has had to wait until a bank had
been closed and resolved to make this detailed account determination
and has had to rely primarily on the bank's reported deposit
information. 

In October 1992, DOR changed its method of accounting for uninsured
deposits to enable FDIC to better estimate this amount at the time
the FDIC Board evaluates resolution alternatives.  The new method,
based on FDIC's historical resolution experience, is to estimate
uninsured deposits for which no insurance determination can be made
from available information as 15 percent of the total account
balances for which an insurance determination is not completed. 

Another factor that contributes to FDIC's difficulty in completing
timely estimates of uninsured deposits is that all depositors can
continue to make account deposits and withdrawals until the bank is
closed.  In one of the resolution cases we reviewed, the failing bank
experienced deposit outflows of about $180 million, including about
$14 million in uninsured deposits, in the 2 weeks before its closure. 
DOR officials told us that changes of this magnitude occurring so
near to the time of the FDIC Board's resolution decision could not
always be reflected in DOR's cost evaluations of resolution
alternatives. 

The FDIC Board is likely to continue to use the new method because of
difficulties in determining the amount of uninsured deposits.  These
difficulties are caused not only by the withdrawals that can occur
before a failing bank is closed, but also by the complexity of the
insurance coverage rules and the associated limited reliable account
information at failing banks, as discussed in appendix V. 


--------------------
\4 In early 1992, the FDIC Board was concerned about the possibility
of the systemic risk that could result from runs on failed banks if
uninsured depositors sought to withdraw their deposits.  FDIC sought
to minimize the impact of losses to uninsured depositors by providing
to such depositors advance dividends approximating the amounts
estimated to be recovered from asset sales.  This provided uninsured
depositors with immediate access to a portion of their uninsured
funds while DOL commenced asset sales.  The concerns of the Board did
not materialize in 1992, even though cumulatively, uninsured
depositors absorbed about $80 million of estimated losses in the 1992
resolutions, which is about 2 percent of the losses expected to be
incurred by the deposit insurance fund.  The Board's concerns about
uninsured depositors may also have been somewhat mitigated by the
passage in August 1993 of depositor preference legislation, which
places uninsured depositors ahead of unsecured creditors.  Now, in
the event of a bank failure, FDIC would make payments to uninsured
depositors before it would make payments to unsecured creditors,
which thereby reduces uninsured depositors' risk of loss. 


      ESTIMATES OF PROCEEDS UNDER
      LOSS-SHARING AGREEMENTS
-------------------------------------------------------- Chapter 2:4.2

For five of the seven resolutions of major banks we reviewed, FDIC
offered a loss-sharing agreement to potential acquirers.  In each of
the five cases, FDIC documented the assumptions about estimated costs
under loss-sharing agreements in terms of an expected upper limit of
losses.  The limits were on the basis of potential credit losses
generally associated with the assets covered by the agreement rather
than on FDIC's historical experience of losses on such assets when
covered by loss-sharing agreements.  The bases of FDIC's assumptions
were reasonable in that FDIC's experience in loss-sharing agreements
is limited.  However, we believe that a more certain basis for
estimating costs associated with loss-sharing agreements would come
from historical data.  As discussed in a later section of this
chapter, FDIC is taking steps to improve its assumptions about the
costs of its loss-sharing agreements. 


      BRIDGE BANK DECISIONS
-------------------------------------------------------- Chapter 2:4.3

While only two resolutions in 1992 involved a bridge bank, they raise
policy issues for FDIC.\5 In deciding to operate a failing bank as a
bridge bank until an acceptable acquirer can be found, the FDIC Board
is faced with a number of uncertainties.  Some risk is inherent in
bridge banks because of changing market conditions during the time
FDIC operates the bank while seeking an acceptable acquirer.  Bridge
banks are also based on assumptions about expected market interest
and likely bid price rather than on actual bids. 

In addition, an issue arises as to whether uninsured depositors in
the bridge bank should be expected to share in any losses from bridge
bank operations at the time it is sold to an acquirer, considering
uninsured depositors were already required to assume a share of
estimated losses when the bridge bank decision was made.  In the two
bridge bank decisions, FDIC did not place uninsured depositors at
further risk.  The two bridge banks were operated by FDIC for quite
brief periods, generally less than 6 months, with the banks'
activities remaining essentially the same during the period. 
Consequently, little if any additional BIF losses were estimated
because of bridge bank operations and uninsured depositors were not
assessed any further amounts.  FDIC has decided on a case-by-case
basis how to handle the resolution of bridge banks, including whether
uninsured depositors should be expected to share in any additional
losses.  FDIC intends to use its experience with bridge bank
decisions to develop relevant policy positions on these issues. 


--------------------
\5 One of these two resolutions involved First City Bancorporation of
Texas, which represents 20 banks. 


   FDIC HAS TAKEN STEPS TO IMPROVE
   LEAST-COST DECISIONMAKING
---------------------------------------------------------- Chapter 2:5

During 1993, DOR initiated efforts that may help to reduce some of
the uncertainties in resolution decisions.  Using as a baseline the
AVR results used in its 1992 resolutions, DOR has started to track
and compare actual asset recovery results to estimates of asset
recoveries used to calculate resolution costs.  Specific tracking and
monitoring provisions were included in resolutions with loss-sharing
agreements so FDIC could ascertain its proportionate share of losses
as well as recoveries.  DOR plans to use the results of these efforts
in creating "feedback loops" to compare actual results with earlier
estimates and improve its asset valuation techniques. 

The development of such feedback loops, if effectively implemented,
should eventually enhance FDIC's process for resolving banks.  DOR
officials advised us that they have also taken initiatives to work
with liquidation staff to improve data relating to asset disposition
activities to enhance DOR's decisionmaking. 

DOR has also started to analyze its 1992 resolution transactions as a
basis for developing more sophisticated, actuarial-based asset
valuation models in the future.  DOR anticipates such models will
facilitate the development of additional strategies and transaction
alternatives for resolving a failing bank. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 2:6

During 1992, FDIC continued to improve its evolving resolution
process.  These improvements, as reflected over the 22 resolutions we
reviewed, have enhanced FDIC's documentation of the cost assumptions
underlying the resolution decisions. 

FDIC's initiatives to improve estimates related to uninsured deposits
and loss-sharing agreements should allow the agency to further
improve its resolution process and provide a better analytical base
for future resolution decisions.  We are encouraged that FDIC
continues to analyze its resolution process to reduce uncertainties
in loss estimates, bid evaluations, and loss-sharing agreements.  We
believe that these efforts, if effectively implemented, should
increase the level of certainty in resolution decisionmaking. 


FDIC NEEDS TO IMPROVE ITS
DOCUMENTATION OF MARKETING
STRATEGIES
============================================================ Chapter 3

While FDICIA requires FDIC to resolve failed banks in the least
costly manner, the statute does not prescribe the way in which FDIC
must consider the realm of all possible resolution methods.  Thus,
FDIC has considerable discretion to construct a strategy by selecting
from among a large number of variations within the basic resolution
methods in any given resolution.  Because the marketing of failed
banks is central to FDIC's consideration of possible resolution
methods, we reviewed the marketing strategies FDIC used in 22
resolutions to determine the agency's rationale for those strategies
and the process used to implement these strategies. 

In most cases, we found the agency's rationale for judgments made in
selecting marketing strategies was not documented.  FDIC marketing
decisions, which determine how failed banks are presented to
potential acquirers, may affect the range of alternatives considered
by FDIC in its least-cost test.  FDIC needs to document its marketing
decisions, specifically addressing the agency's rationale for
selecting certain resolution methods and rejecting others.  We
believe that DOR should submit its marketing decisions to the FDIC
Board for consideration to provide greater assurance to Congress and
the public that resolution costs are being minimized. 


   FDIC DID NOT DOCUMENT ITS
   RATIONALE FOR MARKETING
   STRATEGIES
---------------------------------------------------------- Chapter 3:1

Generally, FDIC identifies resolution alternatives through its
process for marketing a failed bank to potential acquirers.  The
marketing effort requires FDIC to develop a strategy for marketing
the failed bank.  The marketing strategy includes selection of the
basic resolution methods as well as the packaging of a failed bank's
assets, deposits, and other liabilities for sale.  Further, FDIC's
marketing strategy typically includes determining whether (1) FDIC
will offer loss-sharing or other incentive agreements on certain
assets and (2) some poor quality assets will not be offered for sale. 


      DECIDING HOW TO MARKET A
      FAILED BANK
-------------------------------------------------------- Chapter 3:1.1

DOR's interim procedures, set out during 1992 and finalized in 1993,
require DOR to consider alternative resolution methods.  According to
DOR officials, DOR considers two basic resolution methods in deciding
the marketing strategy for a failed bank--purchase and assumptions
(P&A) involving essentially the whole bank or some portion of a
bank's assets and insured deposit transfers (IDT).\1 The P&A
resolution methods are discussed further in chapter 1. 

The starting point for DOR's development of a marketing strategy is
the division's development of an information package containing
financial and nonfinancial information about the failed bank.  Also,
DOR's guidance requires the division to review the structure of the
bank, competition and economic conditions in the geographical area,
past resolutions, and any other relevant information.  From these
reviews and interviews with bank and regulatory officials, DOR
officials attempt to identify the bank's problem assets as well as
the market interest in acquiring those assets.\2 DOR officials may
decide to exclude from the bid package certain risky assets that the
officials believe the market would not purchase.  In this case, FDIC
would transfer the risky assets to the receivership, and DOL would
sell them piecemeal to the private sector.  In addition, DOR may
decide to offer an incentive to bidders to encourage the purchase of
certain problem assets.  The incentive could be an agreement whereby
FDIC (1) would allow the acquirer to return (or "put back") assets to
FDIC if the acquirer could not sell the assets in a prescribed period
or (2) would share in losses incurred in the acquirer's disposition
of problem assets. 

For failures of regional banks (that is, banks with assets under $1
billion), DOR's regional office analyzes alternative resolution
methods and submits its recommended marketing strategy to Washington,
D.C., for review and approval.  DOR officials in Washington said that
they review the marketing strategy recommendation to ensure that it
is reasonable and that alternative resolution methods have been
considered.  The approved marketing strategy is presented to those
attending the bidders conference.  DOR also includes the approved
marketing strategy in the decision package provided to the FDIC Board
for its consideration in making the least costly resolution decision. 

DOR senior officials in Washington develop the marketing strategy for
failures of major banks (that is, those with assets over $1 billion)
with input from regional office personnel.  As in the regional banks,
DOR officials said that they consider and analyze alternative
resolution methods.  However, unlike regional banks, DOR does not
have a process that requires the formal review and approval of
recommended marketing strategies before presentation to the FDIC
Board to ensure that it is reasonable and that available resolution
methods have been considered. 

Although the least-cost provisions of FDICIA require FDIC to choose
between resolution alternatives on the basis of cost, the passage of
FDICIA did not change the basic types of resolution methods available
to FDIC or FDIC's general preference for whole bank P&As, which limit
the assets liquidated and held for liquidation by FDIC.  For the 5
years before FDICIA's enactment, FDIC officials preferred the whole
bank approach, and a senior FDIC official told us that the agency's
marketing strategy during that period sent a clear and explicit
message to potential acquirers that FDIC was primarily interested in
receiving bids for whole bank resolutions.  FDIC officials told us
that they continue to prefer this resolution approach when it is
feasible and complies with FDICIA requirements. 

In 1992, a number of conditions worked against whole bank P&As.  FDIC
officials told us that, during 1992, acquirers were generally not
interested in whole bank transactions because of the poor condition
of the economy and the banking industry.  Also, banks were reluctant
to purchase additional assets because of regulatory capital
requirements.  Furthermore, the FDIC officials said that the rise in
the failure of financial institutions, which began in the late 1980s,
had saturated the market for failed bank assets. 


--------------------
\1 FDIC may also temporarily control and operate a failed bank as a
bridge bank until an acceptable acquirer is selected and approved. 

\2 Problem assets can be defined as those loans that are
nonperforming or for which repayment is otherwise in doubt.  Riskier
assets may include those loans in which the collateral values have
fallen below the amount owed or real estate owned by the failing bank
from foreclosures on defaulted loans. 


         MAKING DECISIONS ABOUT
         FAILED BANK ASSETS
------------------------------------------------------ Chapter 3:1.1.1

Some bank assets may be withheld from the bidding process or offered
on an optional basis if FDIC officials believe, on the basis of their
knowledge of the market, that those assets are of little or no
interest to potential acquirers.  FDIC officials stated that bank
assets may also be withheld because of the uncertainty of the value
of the assets, which would most likely result in a lower bid premium
from potential acquirers.  Such assets have included real estate
acquired by a failed bank as the result of foreclosures on defaulted
loans, known as real estate owned, and loans with real estate
collateral whose value has fallen below the delinquent loan balance,
known as in-substance foreclosure loans. 

Fixed assets, which include the bank's premises, may also be of
little interest to potential acquirers, according to FDIC officials. 
For example, in 15 of the 22 cases we reviewed, we found that real
estate owned and/or in-substance foreclosure loans were not offered
for sale to the market.  In the remaining seven cases, we found six
in which real estate owned was offered on an optional basis (in the
remaining case, the bank had no real estate owned or in-substance
foreclosure loans).  Also, for 16 of our 22 cases, the bank's fixed
assets were offered to bidders on an optional basis.  Concerning the
remaining six cases, FDIC retained the bank's fixed assets in three
cases and in the remaining three cases, the fixed assets were
included in the bid offering and ultimately passed directly to the
acquirer. 

FDIC officials continue to believe that better returns on failed bank
assets ultimately result if those assets are managed by the private
sector rather than the government until they are eventually sold. 
Accordingly, FDIC devises its marketing strategies so that, to the
extent possible, more assets are passed to the private sector.  In
some cases, FDIC may offer incentives for an acquirer to purchase
certain assets.  For example, FDIC may offer loss-sharing agreements,
which assign FDIC a portion of losses the acquirer incurs in
disposing of these assets within a prescribed period.  Loss-sharing
agreements typically cover assets that FDIC has found difficult to
liquidate during receivership, such as commercial loans delinquent
for 90 days or more.  Other techniques that FDIC used to keep assets
in the private sector in a market made resistant by the many bank
failures of the 1980s included the establishment of FDIC-assisted
"collecting banks," which provided the acquirer with capital to
dispose of problem assets, and "put-back" agreements, which enabled
acquirers of assets that did not sell within a prescribed period to
return those assets (put them back) to FDIC. 

For the 1992 resolutions we reviewed, we found that loss-sharing
agreements were the most frequently offered incentive, particularly
for resolutions of major banks.  In five of the seven major
resolutions we reviewed, FDIC determined that the resolution
alternative having the loss-sharing incentive was the least costly
resolution alternative.  (For all 22 cases we reviewed--including
resolutions of major and regional banks--8 offered loss-sharing
agreements as an incentive.) The loss-sharing agreements for the five
major banks covered total assets of approximately $4.4 billion, about
26 percent of the banks' combined $16.8 billion in assets.  The
loss-sharing agreements extended, on average, for 5 years. 


      FDIC CANNOT ASSURE THAT
      MARKETING PROCEDURES
      MINIMIZE RESOLUTION COSTS
-------------------------------------------------------- Chapter 3:1.2

Because the process by which FDIC selects its marketing strategy can
affect the range of alternatives that are later considered in
applying the least-cost test, we reviewed FDIC's selection of
marketing strategies in 22 resolution cases.  In our review, we
sought to determine the reasons, in each case, for FDIC's selection
of marketing strategies, including any exclusions of assets from bid
solicitations and any offerings of loss-sharing agreements.  We also
sought evidence that the marketing strategies selected were
consistent with achieving the purpose of resolving failed banks in
the least costly manner. 

FDIC has considerable discretion to construct a strategy for
resolving a failed bank in the least costly manner.  However, in most
cases we reviewed, we could not determine how FDIC arrived at its
marketing strategy.  Of the 22 resolution cases we reviewed, only
4--all regional banks--contained documentation of the rationale for
resolution approaches selected for marketing purposes.  None of the
seven resolution cases involving major banks had such documentation. 
We believe that without such documentation, FDIC cannot demonstrate,
with great assurance, that the fullest effect of FDIC's statutory
mandate to choose the least costly resolution is being met. 

The FDIC Office of Inspector General (OIG) also reported concerns
regarding DOR's documentation of its marketing decisions.\3 OIG
indicated that DOR's case files did not contain sufficient
documentation to support the selection of resolution types or the
rationale for offering an asset agreement.  OIG also indicated that
DOR would be unable to support the steps taken or alternatives
considered in designing individual resolution strategies and would
have difficulty assuring Congress and other parties that all
potential avenues of resolution were considered.  In July 1993, OIG
reported that DOR had corrected many of the problems identified in
its earlier audit report and was in the process of developing policy
memorandums and procedure manuals to address OIG's concerns.\4

DOR officials agreed with us that better documentation of the
rationale for marketing strategies is needed and assured us that they
will address this issue.  They also advised us that they are
continuing to improve the evolving resolution process, and the
diminishing resolution workload of the agency should give them an
opportunity to further improve the process. 


--------------------
\3 FDIC, Office of Inspector General, Audit of Division of
Resolutions Operational Controls, March 30, 1992. 

\4 FDIC, Office of Inspector General, Follow-Up Audit:  Division of
Resolutions Operational Controls, July 16, 1993. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 3:2

We were unable to determine from available documentation how FDIC
arrived at the marketing strategy for most of the 22 resolution cases
reviewed.  We found, in most of the cases we reviewed, documentation
was inadequate for us to determine FDIC's basis for making one or
more of the following decisions before soliciting bids:  (1)
selecting or rejecting various resolution methods, (2) offering or
not offering loss-sharing agreements on certain assets, or (3)
deciding not to offer poor quality assets for sale.  Because these
marketing decisions determine how failed banks are presented to
potential acquirers, they may affect the ranges of alternatives
considered by both FDIC and potential acquirers. 

As a practical matter, FDIC must make judgments regarding how best to
offer an institution for sale.  However, to give the fullest effect
to FDIC's statutory mandate to choose the least costly method, we
believe that marketing decisions should be thoroughly documented and
reviewed.  In particular, the record in each resolution should
address those methods that are potentially available and explain the
agency's rationale for selecting some and rejecting others.  Further,
the record should reflect a formal review of the marketing strategy
developed to ensure that all resolution possibilities have been
considered.  We believe that thoroughly documenting the record in
each case would both enhance the quality of FDIC's decisionmaking and
provide greater assurance to Congress and the public that resolution
costs are being minimized. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 3:3

We recommend that the Acting Chairman of FDIC require DOR to

  document the rationale for its marketing strategies for resolving
     all failing or failed banks and

  submit the documented record of the marketing strategies to the
     FDIC Board for its consideration in making the least-cost
     resolution decisions. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 3:4

FDIC provided written comments on a draft of this report, which
appear in appendix VI.  FDIC concurred with our findings and
recommendations and said that FDIC is updating DOR's internal
operating procedures in response to our recommendations.  If
implemented as described, we believe that the changes should provide
FDIC better assurance that its statutory mandate to choose the least
costly method is being met. 


SUMMARY DATA ON GAO SAMPLE OF
FDIC'S 1992 RESOLUTIONS
=========================================================== Appendix I

This appendix includes profile information on the DOR resolutions
included in our sample.  Table I.1 shows data from our analyses of
the 22 sampled resolutions, and table I.2 reflects the assets
retained by FDIC in the sampled resolutions. 



                                    Table I.1
                     
                      GAO Sample of FDIC's 1992 Resolutions

                              (Dollars in millions)

                                                Uninsure            Loss to
                               Total     Total         d  Estimate  uninsured
DOR office/     Date closed   assets  deposits  deposits    d fund  depositors
Failed bank     (1992)            \a        \a        \a    loss\b  (Yes/No)
--------------  ------------  ------  --------  --------  --------  ------------
Boston:
--------------------------------------------------------------------------------
Atlantic Trust  January 30       $21       $21      $\c)        $4  No
Company

Vanguard        March 27         407       408         9       102  No
Savings Bank

Winchendon      August 14         66        64         4         5  No
Savings Bank

Plymouth Five   September 18     216       182         5        10  No
Cents Savings
Bank

Guaranty-       November 13      326       313         4        55  Yes
First Trust
Company


New York:
--------------------------------------------------------------------------------
American        January 24        21        20       \c)         2  No
National Bank
of NY

Summit          April 3           90        89       \c)        23  Yes
National Bank

Brookfield      May 8             73        69       \c)        26  Yes
Bank

The Union       August 28        577       560        14        54  Yes
Savings Bank

Sailors and     December 11       33        32       \c)         6  No
Merchants Bank
and Trust
Company


San Francisco:
--------------------------------------------------------------------------------
United          March 20          29        28         1         8  Yes
Mercantile
Bank and Trust
Company, N.A.

The Bank of     April 3          119       115        12        30  Yes
Beverly Hills

The Financial   May 4            243       226        17        63  Yes
Center Bank,
N.A.

Statewide       November 13       10         9       \c)         2  No
Thrift and
Loan Company

Huntington      December 4        42        37       \c)         4  Yes
Pacific Thrift
and Loan
Association


Washington:
--------------------------------------------------------------------------------
Dollar Dry      February 21    4,028     3,733        57       574  No
Dock Bank

American        June 12        3,613     3,011       119       422  Yes
Savings Bank\d

First           October 2      1,638     1,361        15       122  No
Constitution
Bank

The Howard      October 2      3,612     3,392        49       117  No
Savings Bank

First City      October 30     8,789     7,879       410     507\f  Yes-4,
Bancorporation                                                      No-16\g
, Texas\e

Heritage Bank   December 4     1,316       985        33        15  No
for Savings

Meritor         December 11    4,501     3,197       225       \c)  No
Savings Bank

================================================================================
Total           n/a           $29,77   $25,731      $974    $2,151  n/a
                                   0
--------------------------------------------------------------------------------
\a Values as noted by DOR before the bank's closing. 

\b Loss reflects DOR's initial estimated cost of resolution, as
reflected in its cost analysis. 

\c Indicates values less than $1 million.  (Uninsured deposits at
these seven banks totaled $1,667,000:  Atlantic Trust Company,
$2,000; American National Bank of NY, $0; Summit National Bank,
$535,000; Brookfield Bank, $823,000; Sailors and Merchants Bank and
Trust Company, $135,000; Statewide Thrift and Loan Company, $24,000;
Huntington Pacific Thrift and Loan Association, $148,000.  The
estimated fund loss for Meritor Savings Bank was $0.)

\d American Savings Bank data include its subsidiary Riverhead
Savings Bank. 

\e First City Bancorporation data include all 20 banks. 

\f FDIC's initial resolution decision to bridge the First City
Bancorporation banks was estimated to cost $507 million.  However,
the final sale to acquirers of the banks resulted in no loss to the
Bank Insurance Fund. 

\g On the October 30, 1992, resolution of the First City
Bancorporation banks, FDIC imposed losses on uninsured depositors in
4 of the 20 banks.  Uninsured depositors at the remaining 16 banks
suffered no loss. 

Source:  GAO analyses of 22 sampled resolutions. 



                                    Table I.2
                     
                      Assets Retained by FDIC in GAO Sampled
                                   Resolutions

                              (Dollars in millions)

                                                                      Percentage
                                               Assets  Least costly    of assets
                                    Total    retained  resolution    retained by
DOR office/Failed bank           assets\a   by FDIC\b  alternative        FDIC\b
-------------------------------  --------  ----------  ------------  -----------
Boston:
--------------------------------------------------------------------------------
Atlantic Trust Bank                   $21         $21  P&A                    96
Vanguard Savings Bank                 407         402  P&A                    99
Winchendon Savings Bank                66          21  P&A                    32
Plymouth Five Cents Savings           216          77  P&A                    36
 Bank
Guaranty-First Trust Company          326         313  P&A                    96

New York:
--------------------------------------------------------------------------------
American National Bank of NY           21          21  Payoff                100
Summit National Bank                   90          88  IDT w/                 97
                                                        assets
Brookfield Bank                        73          68  IDT w/                 93
                                                        assets
The Union Savings Bank                577         351  P&A                    61
Sailors and Merchants Bank and         33          17  P&A                    53
 Trust Company

San Francisco:
--------------------------------------------------------------------------------
United Mercantile Bank and             29          24  IDT w/                 83
 Trust Company, N.A.                                    assets
The Bank of Beverly Hills             119         119  Payoff                100
The Financial Center Bank, N.A.       243         243  Payoff                100
Statewide Thrift and Loan              10           3  P&A                    27
 Company
Huntington Pacific Thrift and          42          42  Payoff                100
 Loan Association

Washington:
--------------------------------------------------------------------------------
Dollar Dry Dock Bank                4,028         420  P&A                    10
American Savings Bank\c             3,613       3,605  IDT--branch           100
First Constitution Bank             1,638         250  Whole P&A              15
The Howard Savings Bank             3,612         616  P&A                    17
First City Bancorporation,        8,789\e     8,789\e  Bridge bank         100\e
 Texas\d
Heritage Bank for Savings           1,316         126  Whole P&A              10
Meritor Savings Bank                4,501       1,342  P&A                    30
================================================================================
Total                             $29,770     $16,958  n/a                   n/a
--------------------------------------------------------------------------------
Legend

P&A = purchase and assumption
IDT = insured deposit transfer

\a Asset values as noted by DOR, before the bank's closing. 

\b Assets retained by FDIC upon closure of the bank. 

\c American Savings Bank data include its subsidiary Riverhead
Savings. 

\d First City Bancorporation data include all 20 banks. 

\e In FDIC's initial resolution decision to bridge the First City
Bancorporation banks, all assets were retained by FDIC.  In the
ultimate resolution of the banks, FDIC passed most of the banks'
assets to the acquiring banks. 

Source:  GAO analyses of 22 sampled resolutions. 


BID SUMMARY DATA ON GAO SAMPLE OF
FDIC'S 1992 RESOLUTIONS
========================================================== Appendix II

This appendix includes profile information on the bids received by
DOR for our sample.  Table II.1 shows bid summary data on the 22
sampled resolutions, and table II.2 is an analysis of nonconforming
bids on GAO sampled resolutions. 



                                    Table II.1
                     
                         Bid Summary Data on GAO Sampled
                                   Resolutions

DOR                                                                 Nonconformin
office/   Least costly  Winning bid    Total                 Total             g
Failed    resolution    nonconformin  bidder   Total  nonconformin      bids not
bank      alternative   g (Yes/No)         s    bids        g bids     evaluated
--------  ------------  ------------  ------  ------  ------------  ------------
Boston:
--------------------------------------------------------------------------------
Atlantic  P&A           Yes                1       1             1             0
Trust
Company

Vanguard  P&A           Yes                2       3             1             0
Savings
Bank

Winchend  P&A           No                 4       5             0           n/a
on
Savings
Bank

Plymouth  P&A           No                 4      10             4           3\a
Five
Cents
Savings
Bank

Guaranty  P&A           Yes                5       8             6             0
-First
Trust
Company


New York:
--------------------------------------------------------------------------------
American  Payoff        No                 0       0             0           n/a
National
Bank of
NY

Summit    IDT w/        No                 4       4             2             0
National  assets
Bank

Brookfie  IDT w/        No                 2       2             0           n/a
ld Bank   assets

The       P&A           No                 2       8             4             0
Union
Savings
Bank

Sailors   P&A           No                 3       6             0           n/a
and
Merchant
s Bank
and
Trust
Bank


San Francisco:
--------------------------------------------------------------------------------
United    IDT w/        No                 2       2             0           n/a
Mercanti  assets
le Bank
and
Trust
Company,
N.A.

The Bank  Payoff        No                 1       1           1\b             0
of
Beverly
Hills

The       Payoff        No                 1       1            \0           n/a
Financia
l Center
Bank,
N.A.

Statewid  P&A           No                 1       1             0           n/a
e Thrift
and Loan
Company

Huntingt  Payoff        No                 1       1             1             0
on
Pacific
Thrift
and Loan
Associat
ion


Washington:
--------------------------------------------------------------------------------
Dollar    P&A           Yes                2       3             3             0
Dry Dock
Bank\c

American  IDT--branch   No                10      12             2             0
Savings
Bank\d

First     Whole P&A     No                 3     6\e             4           2\f
Constitu
tion
Bank

The       P&A           No                 3      \8             0          \n/a
Howard
Savings
Bank

First     Bridge bank   No                32     111             0           n/a
City
Bancorpo
ration,
Texas\g

Heritage  Whole P&A     No                 2       2             0           n/a
Bank for
Savings

Meritor   P&A           No                 4      11             5             0
Savings
Bank

================================================================================
Total     n/a           n/a               89     206            34             5
--------------------------------------------------------------------------------
Legend

P&A = purchase and assumption
IDT = insured deposit transfer

\a In this case, a bidder submitted multiple bids, three of which
were nonconforming.  DOR was unable to evaluate the bidder's offer to
purchase residual interest in Federal Home Loan Mortgage Corporation
loans and the bank's premises with a $100,000 limitation on
encumbrances. 

\b In this case, the bidder's rating was reduced as a result of an
exam, and thus the bidder became ineligible to bid.  Since this was
the only bid received, FDIC did not perform a cost analysis. 

\c Dollar Dry Dock Bank represents the only case in which the winning
bid was a nonconforming bid with major differences from FDIC's
marketing strategy.  (See table II.2 for an analysis of nonconforming
bids on GAO sampled resolutions.)

\d American Savings Bank data include its subsidiary Riverhead
Savings Bank. 

\e The six bids received included one conforming bid, which was not
evaluated by FDIC because the bidder did not have regulatory
approval.  This bid represented an exception; all other conforming
bids in our sample were evaluated by FDIC. 

\f DOR was unable to evaluate these bids because it could not
accurately determine the cost of providing the tax indemnifications
requested by two bidders. 

\g First City Bancorporation data include all 20 banks. 

Source:  GAO analyses of 22 sampled resolution cases. 



                          Table II.2
           
            Analysis of Nonconforming Bids on GAO
                     Sampled Resolutions


                           Total                Nonconformin
DOR office/failed    nonconformi          Majo    g bids not
bank                     ng bids   Minor     r        costed
-------------------  -----------  ------  ----  ------------
Boston:
------------------------------------------------------------
Guaranty-First                 6       6     0             0
 Trust Company
Vanguard Savings               1       1     0             0
 Bank
Plymouth Five Cents            4       0     4             3
 Savings Bank
Atlantic Trust                 1       1     0             0
 Company

New York:
------------------------------------------------------------
Summit National                2       1     1             0
 Bank
The Union Savings              4       0     4             0
 Bank

San Francisco:
------------------------------------------------------------
The Bank of Beverly          1\b       0     1             0
 Hills
Huntington Pacific             1       1     0             0
 Thrift and Loan
 Association

Washington:
------------------------------------------------------------
Dollar Dry Dock                3       0     3             0
 Bank
American Savings               2       0     2             0
 Bank
First Constitution             4       1     3             2
 Bank
Meritor Savings                5       3     2             0
 Bank
============================================================
Total                         34      14    20             5
------------------------------------------------------------
\a We reviewed all nonconforming bids to determine the degree of
nonconformance.  Nonconforming bids very similar to the method
described to potential bidders in FDIC's marketing strategy were
considered to have a minor degree of nonconformance.  Nonconforming
bids significantly different from FDIC's marketing strategy were
considered to have a major degree of nonconformance. 

\b In this case, the bidder's rating was reduced as a result of an
exam, and thus the bidder became ineligible to bid.  Since this was
the only bid received, FDIC did not perform a cost analysis. 

Source:  GAO analyses of 22 sampled resolution cases. 


FDIC'S ANALYSIS OF RESOLUTION
ALTERNATIVES
========================================================= Appendix III

This appendix describes how FDIC, as set forth in its procedures and
guidance, uses its marketing activities and valuation of a failing
bank's assets to evaluate resolution alternatives and to select the
least costly resolution alternative. 


   PREPARING TO MARKET A FAILING
   BANK
------------------------------------------------------- Appendix III:1

FDIC generally begins its resolution process after receiving a
request for such assistance from a failing bank's chartering
authority.  One of its first steps is to develop a "snap shot" of the
failing bank's financial condition, which provides FDIC with the
initial information it needs to develop resolution alternatives. 

When FDIC first begins its financial analysis of a failing bank, it
develops a rather detailed breakdown of the failing bank's balance
sheet items.  That is, it looks at the failing bank's assets and
liabilities structure to develop detailed information on the amounts
and types of assets and liabilities held by the bank. 

The types of information FDIC develops on each failing bank can vary
based on each bank's business strategies as reflected in its asset
portfolio and liabilities structure.  For a failing bank primarily
involved in residential or mortgage lending, FDIC would develop
information on such assets based on the primary products or types of
loans the failing bank offered.  For example, FDIC would develop
mortgage loan data based on whether the mortgages were fixed or
variable rate and by the term of the mortgages, i.e., 15-year or
30-year mortgages. 

FDIC does a similar analysis of a failing bank's liabilities to
determine, for example, the amounts of insured and uninsured deposits
and other obligations of the failing bank. 

FDIC develops this analysis based on unaudited financial and other
data provided by the failing bank.  FDIC uses the analysis to prepare
an information package on the failing bank, which it provides to
potential acquirers who have expressed an interest in the assets or
deposits of the failing bank. 


   MARKETING THE FAILING BANK
------------------------------------------------------- Appendix III:2

As discussed in more detail in chapter 3, FDIC primarily relies on
its business judgment, based on factors such as the economic
condition of the geographical area, to determine the types of failed
bank transactions it will offer to potential acquirers. 

After FDIC has selected its preferred marketing strategy for
resolving a failing bank, it conducts an information meeting.  During
the information meeting, FDIC advises the potential acquirers and
discusses the FDIC terms for its selected resolution transactions. 
FDIC uses the information package data made available to potential
acquirers as its basis to discuss the details of the proposed
transaction.  This allows FDIC and the potential acquirers to discuss
the transaction terms based on the same financial data on the failing
bank. 

Typically, the transaction terms focus on the treatment of the
deposits and assets held by the failing bank.  That is, FDIC advises
the potential acquirers whether FDIC will accept bids on the basis of
all deposits or only insured deposits being assumed by an acquirer. 
FDIC also advises the potential acquirers about the (1) types and
amounts of assets that would be passed to an acquirer as part of the
transaction terms, (2) assets FDIC plans to retain, and (3) terms,
such as loss-sharing agreements or other significant conditions, that
are a part of the proposed transaction. 


   EVALUATING RESOLUTION
   ALTERNATIVES
------------------------------------------------------- Appendix III:3

As discussed in appendix IV in detail, FDIC does a valuation of a
failing bank's assets to estimate their liquidation value.  FDIC uses
the asset valuation to estimate its potential recoveries from
liquidating a failed bank's assets, which provides FDIC a basis to
estimate its cost to liquidate a failing bank.  FDIC also uses the
asset valuation results as its basis for evaluating resolution
alternatives.  Besides a liquidation, resolution alternatives
available to FDIC are the bids it receives for a failing bank.\1


--------------------
\1 FDIC can also choose to create a temporary bridge bank before
making a final resolution determination. 


      DETERMINING RESOLUTION COSTS
----------------------------------------------------- Appendix III:3.1

FDIC's payments to insured depositors represent its primary
resolution cost.  Because of their senior claims, secured creditors
of a failed bank represent another resolution cost because FDIC
typically honors their claims.  Potential payments to uninsured
depositors and general creditors represent additional elements of
resolution costs.  FDIC payments to these claimants depend on the
resolution results. 

FDIC (taking the place of insured depositors it has already paid),
uninsured depositors, and general creditors share in any resolution
proceeds estimated to be realized generally after payments to insured
depositors and secured creditors have been recognized.  FDIC
generally can make full payments to uninsured depositors at the time
of the resolution decision only as a part of a least costly
resolution.  For transactions involving the transfer of only the
insured deposits to an acquirer or a FDIC payment to only insured
depositors, uninsured depositors may receive an advance payment on
their uninsured amounts, depending on proceeds FDIC expects to
receive from the resolution transaction and any post-resolution asset
disposition activities. 

FDIC, uninsured depositors, and general creditors generally share in
resolution proceeds based on the relative or proportionate amount of
their claims.  That is, if each group held a $1 claim against a
failed bank, each would have a basis to claim one-third of the
proceeds FDIC realizes from the failed bank.\2

Proceeds from asset disposition--either at the time of the resolution
decision or subsequent to asset liquidation activities--represent
FDIC's major source of funds for offsetting costs to resolve a bank. 
In evaluating resolution alternatives, FDIC must compare its
estimated liquidation cost to any bids received.  On the basis of
this analysis, FDIC is to select the least costly alternative for
resolving a failed bank. 


--------------------
\2 In August 1993, depositor preference legislation was passed that
places uninsured depositors ahead of general or unsecured creditors. 
Now, in the event of a bank failure, FDIC would make payments to
uninsured depositors before it would make payments to unsecured
creditors. 


AN OVERVIEW OF FDIC'S ASSET
VALUATION PROCESS AND METHODS
========================================================== Appendix IV

During 1992, FDIC used three methods to estimate the present value of
assets held by failing banks.  Two of the methods, Total Asset
Purchase and Assumption (TAPA) asset review and the Asset Valuation
Review (AVR), were based on on-site reviews of sampled asset files
and records at the bank.  FDIC developed and used TAPA before
FDICIA's enactment and during the first half of 1992.  FDIC developed
and implemented AVR in the second half of 1992.  When liquidity
problems, such as limited ability to meet depositors' fund
withdrawals or legal problems at the failing bank, precluded an
on-site review, FDIC used a third method.  This method was a
statistical research model based on data from FDIC's recovery
experience for six broad asset categories of assets held by small
banks that failed between 1986 and 1990.  The research model was also
used by FDIC for comparison purposes in assessing TAPA or AVR
results. 


   ON-SITE REVIEWS
-------------------------------------------------------- Appendix IV:1

The TAPA and AVR models are similar in some ways.  TAPA calculates
asset values based on a sample of assets taken on-site, estimating by
various categories of assets the effect of current market conditions
and risk factors on the book value of the sampled assets for each
category.  AVR, like TAPA, involves an on-site review of assets;
however, the assets are drawn from more narrowly defined categories,
thus enabling greater specificity and reliability in book value
adjustments.  FDIC officials said that TAPA provided a reasonable
basis to pursue the agency's pre-FDICIA resolution strategy, which
focused primarily on arranging whole bank resolution transactions. 
The AVR model provides a more detailed analysis of a failing bank's
assets and also pays more attention than TAPA to how cash flows from
different asset categories may change over time.  FDIC officials said
the AVR model has improved FDIC's ability to evaluate whole bank and
other resolution methods. 


      ASSET VALUATION REVIEW
------------------------------------------------------ Appendix IV:1.1

The Division of Resolutions (DOR) is to use AVR in the least-cost
analysis of transactions proposed for resolving a failing bank.  FDIC
starts its AVR by using failing bank asset values from the
information package, which essentially extracts information from the
bank's unaudited financial statements. 

AVR includes several methods to compute the net present values of the
assets held by a failing bank.  Net present-value analysis provides a
basis to estimate the current value of assets that will be disposed
of in the future.  DOR is to use the model to estimate the
anticipated net cash recoveries that FDIC could obtain through the
liquidation of all the assets of a failing bank. 

DOR does not rely exclusively on financial models to estimate the
present value of all assets held by a failing bank, since this can be
done by other techniques.  Certain assets, such as cash and federal
funds sold, are to be valued through a reconciliation of the bank's
records.  Securities are priced to market by contacting brokers or
using publications such as the Wall Street Journal.  These activities
are to be done as part of AVR and determine the current values of
those assets, usually as of the date of the financial data used to
develop the information package.  The results are to be incorporated
into the final AVR report. 

DOR uses the financial models to estimate the present values of other
assets, such as loans, real estate owned, and subsidiaries.  AVR
relies on selecting samples of assets based, for example, on the
type, value, and performance status of loans and other
characteristics.  On the basis of analysis results for each sample of
assets, AVR projects the loss for all similar assets not included in
the sample. 

The financial models use one of two approaches to value assets,
depending on how FDIC anticipates the assets will be sold.  One sales
method involves selling groups of homogeneous assets, such as
one-to-four family residential mortgages, by securitizing the loans
and selling them in secondary markets.  The model computes the
present value of such assets using a discount rate built on the
secondary market's required yield, essentially market rates current
at the time of the valuation, adjusted for risk-related factors
(i.e., problems with loan documentation, underwriting standards, or
the remaining maturity of the loans).  As discussed below in more
detail, the model then estimates the present value of the proceeds of
the sales of the securitized loans, based on when the sales are
expected to occur, plus any proceeds FDIC realized while it held the
loans until they were sold, using a discount rate reflecting FDIC's
cost of funds. 

The second sales method assumes FDIC's recoveries come from any
payments a borrower continues to make until a loan is paid off or,
ultimately, the sale of the underlying collateral.  AVR estimates the
present value of the cash recoveries expected to be realized from
managing and eventually selling such assets as nonperforming loans,
real estate owned, fixed assets, and subsidiaries.  Figure IV.1 shows
FDIC's typical approach to valuing assets. 

   Figure IV.1:  FDIC's Typical
   Approach to Valuing Assets

   (See figure in printed
   edition.)

Source:  GAO analysis of FDIC procedures. 

DOR uses several sources, as needed, to develop its assumptions
related to current market rates, loan default and prepayment rates,
holding periods, and asset recovery rates.  DOR uses data obtained
from the following sources to complete its present-value analyses
using the AVR model: 

  Information on the quality and marketability of assets obtained
     through reviewing records maintained by the failing bank. 

  Local market condition data obtained by contacting other financial
     institutions located in the same general area as the failing
     bank. 

  Real estate property values obtained from local real estate
     appraisers. 

  Personal property values for assets such as cars, boats, and mobile
     homes from industry reference material. 

  FDIC asset liquidation results obtained from DOL Consolidated Field
     Offices responsible for managing and selling assets from failed
     banks. 

The AVR analysis can be done by a team composed of FDIC staff or by
contractors.  In addition to establishing and maintaining
documentation of asset valuations, AVR procedures require an overall
summary that identifies the methodologies and assumptions used during
the asset valuation process. 

DOR does not use the AVR model to estimate costs associated with
contingent liabilities, which can include financial instruments such
as letters of credit and lawsuits filed against or on behalf of the
failing bank.  A separate team composed of FDIC liquidation and legal
staff analyze the failing bank's contingent liabilities and estimate
FDIC's potential costs associated with those liabilities.  DOR
incorporates these results into its AVR report and in its cost
analyses of resolution alternatives. 


      AVR DISCOUNTING METHODOLOGY
------------------------------------------------------ Appendix IV:1.2

FDIC takes a two-step approach in applying discount rates to
determine the present value of an asset.  FDIC first estimates the
price at which it could sell the asset.  FDIC does this by estimating
the amount and timing of any net income, or yield, the property may
provide to the asset purchaser.  Using a private-sector discount rate
based on current market rates adjusted for credit risks or other
risks associated with the asset, FDIC adjusts the value of the asset
to determine a selling price that would provide an investor a rate of
return similar to rates on comparable assets. 

FDIC then determines its present value of the asset sale.  This
involves accounting for the asset's value as discounted in the first
step--the asset's estimated selling price--and any net income FDIC
may earn while holding the asset until it is sold.  FDIC uses its
cost of funds, which is pegged at the U.S.  Treasury Bill rates, as
the discount rate in its calculation to determine its present value
or net realizable value of the asset. 

This two-step calculation estimates the asset's present value based
on FDIC liquidation of the asset.  We believe that FDIC's process for
selecting and using discount rates overall appears reasonable. 


   RESEARCH MODEL
-------------------------------------------------------- Appendix IV:2

The research model uses (1) statistical methods to analyze a failing
bank's financial data and (2) asset disposition experience of the
Division of Liquidation (DOL) to estimate the values and losses
associated with assets held by a failing bank.  The historical DOL
asset disposition experience essentially serves as a proxy for market
and risk determinations done in TAPA and AVR.  The analysis is done
off-site, using financial data from the failing bank.  FDIC officials
said that the research model, which is based on FDIC's historical
experience, provides less reliable results than the models based on
the on-site reviews, which focus on current conditions found at the
specific bank to be resolved.  The research model served as the sole
source of valuation only when necessary due to legal problems, such
as fraud, or liquidity problems at the failing bank, which prohibited
on-site reviews of assets.  FDIC most often used the research model
as a rough means of checking results from TAPA and AVR asset
valuations. 

The research model also calculates the present value of the expected
recoveries or cash flows from the assets owned by a failing bank. 
However, the data it analyzes are drawn from regulatory filings such
as call reports, i.e., report of condition rather than from on-site
reviews.  The statistical model, which estimates the amount and
timing of cash flows and recoveries that could be achieved, is based
on FDIC's recovery experiences for six broad categories of assets
held by small banks that failed between 1986 and 1990.  The model
uses a discount factor based on FDIC's cost of funds to complete the
net present-value calculation. 


FDIC'S TREATMENT OF UNINSURED
DEPOSITORS UNDER FDICIA:  LOSSES
IMPOSED MORE FREQUENTLY, IN LARGER
AMOUNTS
=========================================================== Appendix V

FDICIA's least-cost provisions generally require FDIC to close banks
using methods least costly to the insurance fund.\1 Under FDICIA,
depositors over the $100,000 insurance limit and general creditors
have been more likely to incur part of the losses that formerly have
fallen almost exclusively to the Bank Insurance Fund (BIF). 

The effect of least-cost resolutions on uninsured depositors is
already becoming apparent during FDIC's first year of implementing
the FDICIA requirements.  In the 3 years before the passage of
FDICIA, uninsured depositors absorbed losses in about 14 percent of
the total bank failures.  Subsequent to FDICIA, uninsured depositors
absorbed losses in 49 percent of the total bank failures.  At the
time of resolution, FDIC estimated that losses absorbed by uninsured
depositors in 1992 bank failures would be approximately $80 million;
this amounted to about 2 percent of total expected losses in all 1992
failed banks. 


--------------------
\1 FDICIA provides for a systemic risk exception to the least-cost
requirement if a finding is made that compliance with the least-cost
requirement would have serious adverse effects on economic conditions
or financial stability and that a more costly alternative would
mitigate such adverse effects. 


   PRE-FDICIA RESOLUTIONS: 
   UNINSURED DEPOSITORS GENERALLY
   RECEIVED DE FACTO INSURANCE
   PROTECTION
--------------------------------------------------------- Appendix V:1

Before FDICIA, uninsured depositors generally received de facto
insurance protection when banks failed.  This occurred because (1)
resolution methods primarily focused on transferring to a healthy
bank all deposits held by a failing bank and (2) regulators believed
that such protection provided to uninsured depositors, and in some
cases general creditors, helped to maintain the stability of the
banking system. 

FDIC resolution officials explained that before FDICIA, the Federal
Deposit Insurance (FDI) Act, as amended, permitted FDIC to pick any
resolution option that was less costly than an insured depositor
payoff and liquidation of the failed bank's assets.  That is, until
FDICIA passed, FDIC did not have to consider all available
alternatives for resolving a failing bank.  This allowed FDIC greater
flexibility to pursue its preferred resolution strategies.  In the
years just before FDICIA, FDIC often focused on arranging whole bank
resolutions whereby a healthy financial institution assumed all of
the deposits along with all or almost all of the assets and other
bank liabilities.  As a result, uninsured deposits were assumed by
the buying bank, thereby providing de facto insurance protection. 

While uninsured deposits generally received protection, they
represented a small fraction of deposits at time of failure.  Table
V.1 shows that in the 3 years preceding FDICIA, 1989 through 1991,
uninsured deposits were about $2.5 billion, or about 2.7 percent, of
the $92.8 billion in total deposits held by failed banks. 



                          Table V.1
           
           Uninsured Deposits Held by Banks at Time
                    of Failure, 1989-1991

                    (Dollars in millions)

                                           Total     Percent
               Number of       Total   uninsured    deposits
Year         resolutions    deposits    deposits   uninsured
----------  ------------  ----------  ----------  ----------
1991                 127     $53,832      $1,423         2.6
1990                 169      14,837         715         4.8
1989                 207      24,097         348         1.4
============================================================
Total                503     $92,766      $2,486         2.7
------------------------------------------------------------
Source:  FDIC Division of Resolutions Report on Treatment of
Depositors, December 31, 1992. 

In the 3 years before the passage of FDICIA, FDIC imposed losses on
uninsured depositors in 69, about 14 percent, of the 503 failed banks
that it resolved.  It is not clear to what extent the coverage
provided to the uninsured depositors in the resolved institutions may
have increased losses to the deposit insurance fund. 

Before FDICIA, FDIC did not evaluate all of the bids that it may have
received for the deposits and/or assets of a failing bank.  FDIC
grouped the bids it received generally based on the amount of a
failing bank's deposits and assets that would pass to a potential
acquirer.  FDIC would then begin its bid evaluation process starting
with bid offers to acquire essentially the whole failing institution. 
If FDIC evaluated one or more whole bank bid offers that it estimated
would be less costly than an FDIC liquidation of the failing bank, it
would select the whole bank offer it estimated to be the best offer. 
It would not evaluate any other bids it may have received for a
portion of the failing bank's assets. 

By following this bid evaluation process, FDIC did not always
evaluate all bids it may have received for a failing bank. 
Therefore, it has no historical record that can be used to determine
whether its resolution methods achieved the least-cost resolution of
a failed bank.  Additionally, our review disclosed that FDIC's
records did not provide a basis for determining whether coverage
extended to uninsured depositors, as occurs in a whole bank
transaction, increased or helped to reduce FDIC's costs to resolve a
failing bank. 


   TREATMENT OF UNINSURED
   DEPOSITORS UNDER FDICIA
--------------------------------------------------------- Appendix V:2

Under FDICIA, FDIC has less discretion in providing coverage to
uninsured depositors.  FDIC can cover the losses that could be
imposed on uninsured depositors in a post-FDICIA resolution (1) only
as a part of the least costly resolution decision and (2) when the
amounts received from acquirers are sufficient to cover the losses
that could have been imposed on the uninsured depositors.\2

During calendar year 1992, FDIC resolved 122 failed banks with
deposits totaling about $41.2 billion.  Uninsured deposits, like
before FDICIA, represented a relatively small portion of deposits at
the time of resolution.  In 1992, uninsured deposits were $1.4
billion, or about 3.4 percent, of the $41.2 billion in deposits held
by the failed banks. 

Since FDICIA passed, FDIC has modified its bidding process.  FDIC now
accepts bids generally on the basis of an acquiring institution
assuming either insured deposits only or all deposits.  FDIC, to the
extent possible, also evaluates all bids received on a failing
institution. 

As a result of the application of the least-cost provisions, FDIC has
most frequently arranged transactions whereby an acquirer assumes
either the insured deposits only or all deposits of a failed bank. 
In selecting the least costly method to resolve a failing bank,
compared to prior years, FDIC, during 1992, more frequently chose
methods that resulted in uninsured depositors experiencing some
losses. 

Generally, FDIC depositor treatment at resolution has resulted in the
transfer of most deposits to a healthy institution.  In 111 of 122
resolutions, or 91 percent, FDIC transferred either all deposits or
insured deposits. 

In 62 of the 122 resolution transactions completed, uninsured
depositors were made whole as part of the resolution
transactions--FDIC did not impose any losses.  In the remaining 60
resolution cases, FDIC imposed initial losses, referred to by FDIC
officials as "haircuts," on uninsured depositors ranging between 13
percent and 69 percent of uninsured amounts.  FDIC determined the
amount of loss or haircut applied to uninsured amounts on a
bank-by-bank basis.  FDIC based the rate of haircuts it applied to
uninsured amounts held by each bank on the (1) estimated recoveries
it expected to realize from selling assets held by each failed bank,
(2) deposit premiums paid by acquiring institutions, and (3)
uninsured depositors' status as a creditor.  These haircuts were the
initial losses FDIC imposed on uninsured depositors and totaled about
$80 million.  Actual losses that uninsured depositors may experience
will be affected by the recoveries FDIC can achieve from asset sales. 

Uninsured deposit amounts not paid by FDIC at the time it resolves a
failed bank become claims against the receivership.  Receivership
claims entitle the uninsured depositors not made whole to share in
proceeds FDIC realizes from sales of assets in receivership.  While
FDIC generally sells a large amount of a failed bank's assets during
the first several years of receivership, FDIC may take 10 or more
years before it terminates the receivership.  Therefore, it may take
some time before FDIC can determine the actual amount of losses
imposed on uninsured depositors during 1992. 


--------------------
\2 FDIC has interpreted the least-cost requirements of FDICIA as
prohibiting the passage of uninsured deposits to the assuming
institution unless that particular resolution represented the least
costly resolution alternative.  Further, section 13(c)(4)(E) of the
FDI Act specifically prohibits the FDIC from taking any direct or
indirect action after December 31, 1994, concerning any insured
depository institution that would have the effect of increasing
losses to any insurance fund by protecting depositors for more than
the insured portion of their deposits or creditors other than
depositors.  This subparagraph makes clear that FDIC is not
prohibited from engaging in P&A transactions where uninsured deposits
are acquired, as long as the loss to the fund on those deposits is no
greater than the loss that would have been incurred concerning those
deposits had the institution been liquidated.  FDIC, as required by
this subparagraph, issued final regulations implementing this
provision on December 22, 1993.  As noted in the preamble to the
rule, FDIC believes that subparagraph (E) is "subsumed in the more
general least-cost provisions of section 13(c)(4)(A) and has no
independent operative effect." (58 Fed.  Reg.  67662, Dec.  22,
1993.)


   DIFFICULTIES EXIST IN RESOLVING
   UNINSURED DEPOSITS
--------------------------------------------------------- Appendix V:3

FDIC has encountered operational issues in resolving uninsured
deposits.  Specifically, these issues are (1) FDIC difficulties in
determining the level of uninsured deposits, (2) disparate treatment
in uninsured depositors across resolutions, and (3) public concerns
and/or confusion regarding insurance coverage. 


      LEVEL OF UNINSURED DEPOSITS
      NOT ROUTINELY MONITORED AND
      NOT EASILY DETERMINED
------------------------------------------------------- Appendix V:3.1

Tracking or determining the insurance status of deposits is not a
part of bank examinations.  As a part of their monitoring activities,
FDIC and the other federal regulators receive quarterly financial
reports on the condition and income of banks as well as other
financial information.  While these filings report the total amount
of deposits held, and the total amount of deposits exceeding
$100,000, the filings do not disclose the insurance status.  As
discussed later, an account exceeding $100,000 may be fully insured
depending, in part, on the rights and capacities of the account
holders. 

Further, banks do not routinely identify or monitor the insurance
status of their customers' funds.  FDIC's process for resolving a
failing bank usually begins after it receives a notification of
failure from the chartering authority.  As a part of the process to
resolve a failing bank, FDIC does an evaluation to determine the
amounts of both insured and uninsured deposits.  The time needed to
complete a determination depends on the (1) size of the bank in both
the volume of deposits and the number of bank subsidiaries and
branches, (2) number of deposit accounts held by the bank, (3)
quality of bank records, and (4) number of locations where deposit
account records are maintained. 

FDIC refers to the insurance determination as the aggregation of
accounts.  This aggregation of deposit accounts can be a laborious
task.  For example, in evaluating two individual accounts with the
same ownership name, FDIC would have to determine whether the two
accounts are held by the same individual or whether the accounts are
in fact separately held by two individuals who share a common name. 
FDIC may rely on social security numbers or tax identification
numbers to make the determination and, in some cases, review
signature cards completed when the accounts were opened, to make the
account ownership and insurance determination. 

For joint accounts, FDIC is to aggregate the balances of all accounts
with the same combination of owners irrespective of the sequence in
which the owners' names are listed or the social security numbers
attached to the accounts.  FDIC then allocates a portion of an
account balance to each account holder.  Once FDIC completes
allocating account balances to individuals or entities, it aggregates
or totals these balances.  Any such sums exceeding $100,000 are
considered uninsured.  FDIC cannot always make a final insurance
determination on all deposit accounts prior to a failing bank's
closure.  In such instances, FDIC generally places deposit accounts
for which an insurance determination could not be made into a
"pass/hold" category.  Essentially, FDIC delays making a final
insurance determination on some accounts at the time a failing bank
is closed and places such accounts "on hold" until a final insurance
determination can be made after the bank is closed. 

Further complicating the insurance determination is that depositors,
insured and uninsured, can withdraw funds from their accounts up to
the time of closure of a failing bank.  This situation exists because
the closure and resolution essentially occur concurrently.  That is,
a bank scheduled for closure continues its operations until it is
actually closed and FDIC is appointed conservator or receiver of the
failed bank.  Since depositors have continued access to the funds in
their accounts during this period, levels of insured and uninsured
deposits held by the failing bank can vary during this period and up
to the time it is closed and resolved.  Between the time the FDIC
Board approves a resolution alternative and the actual closure of the
failing bank, the amount of deposits may vary greatly from the
estimates used in pricing out resolution alternatives. 

Deposits flowing out of a failing institution can also be affected by
events beyond FDIC's control.  In one 1992 resolution in our sample,
DOR reported to the FDIC Board that a state-chartered failing bank
was under a formal enforcement action by its primary regulator. 
About 2 weeks before the FDIC planned date for resolving the bank,
the primary regulator apparently knew the failing bank could not meet
the target date and required the failing bank to assess its ability
to meet this capitalization requirement and to publicly release its
assessment results.  According to DOR staff, the bank publicly
acknowledged that it could not meet the capital requirements by the
target date set in the enforcement action and this acknowledgment
resulted in increased deposit outflows.  FDIC staff estimated that
between the time the failing bank disclosed its distressed financial
condition and when the resolution case was presented to the FDIC
Board for approval, the failing bank experienced deposit outflows
totaling about $180 million and that the level of uninsured deposits
dropped about $14 million, from about $63 million to about $49
million. 

During 1992, DOR staff adopted a new method for estimating the levels
of uninsured deposits held by a failing institution.  This new
estimating method does not address or account for the changes in
levels of uninsured deposits that can occur before a failing bank is
closed and resolved.  However, one component of uninsured deposits is
the amount of deposits placed in the pass/hold category that is
expected to be uninsured.  FDIC is to place a failing bank's deposits
in pass/hold when an insurance determination cannot be completed
prior to resolution.  FDIC reviewed its recent experience with
deposits placed in pass/hold and found that about 15 to 17 percent
eventually were treated as uninsured deposits.  Beginning with
resolution cases decided since October 1992, the resolution analyses
treat about 15 percent of deposits placed in pass/hold as uninsured
deposits and the remainder of these deposits as insured deposits for
the purposes of determining (1) the least-cost resolution and (2) the
proportion of the loss estimate that may have to be assumed by
uninsured depositors. 


      RESOLUTION OF FAILING BANKS
      CAN RESULT IN DISPARATE
      TREATMENT OF DEPOSITORS
------------------------------------------------------- Appendix V:3.2

As noted earlier, uninsured depositors can be fully covered from
losses during the resolution of a failing institution only as a part
of a least costly resolution decision in which the premium FDIC
receives from an acquirer is sufficient to offset any losses that
FDIC would have imposed on uninsured depositors had the institution
been liquidated.  FDIC is to treat each failing bank as a separate
resolution transaction.  Thus, FDIC is to decide the treatment of the
uninsured depositors of a failing bank on a bank-by-bank basis. 

These aspects of the least-cost provisions result in unequal
depositor treatment across resolutions.  For example, in 62 of the
122 resolution transactions completed during 1992, the first full
calendar year since FDICIA passed, uninsured depositors were made
whole as part of the resolution decision--FDIC did not impose any
losses.  In the remaining 60 resolution cases, FDIC imposed initial
losses on uninsured depositors ranging between 13 percent and 69
percent of uninsured amounts based on the estimated losses for
resolving the failing or failed institutions.  Uninsured deposit
amounts not paid by FDIC at the time a failed bank is resolved become
claims against the receivership.  These claims entitle the uninsured
depositors not made whole to share in proceeds FDIC realizes from
sales of assets in receivership. 

In one FDIC region, three failing institutions in the same general
market area were resolved at about the same time.  In one instance,
FDIC placed the failing institution into conservatorship, which to
some extent is similar to a bridge bank, and continued to operate the
institution.  Uninsured deposits were transferred in total to the
conservatorship bank which allowed the uninsured depositors to avoid
losses.  In resolving another failing institution, FDIC received bids
for the institution, which resulted in full coverage of all
depositors including those who were uninsured.  In resolving the
third institution, FDIC accepted a bid that resulted in the
assumption of insured deposits only.  Uninsured depositors were not
fully covered at the third institution, and FDIC imposed an initial
loss on the uninsured deposits of about 25 percent of the uninsured
amounts in their accounts. 

Similarly, uninsured depositors in failed banks within a multibank
holding company received different treatment at the time of
resolution.  This is because FDIC handled the resolution of each
failing bank as a separate transaction because each bank subsidiary
was a separate legal entity.  For example, resolution of the 20 First
City Bancorporation banks resulted in losses being imposed on
uninsured depositors in 4 of the 20 banks. 

Three variables determine how FDIC will treat the uninsured in any
particular resolution.  First, the bank's deposit accounts may have
sufficient market value to cover the uninsured pro rata share of
estimated loss.  In such cases, uninsured deposits are to be
transferred and no haircut will result.  Premiums are tied to the
franchise value, if any, of the failed bank.  For example, a bank may
be willing to pay the premium to get a "customer list" so that it has
the opportunity to sell financial services to them.  Whether a bank
will pay a sufficient premium is a function of matters such as its
business strategy, the amount of uninsured deposits, and the
stability of the uninsured depositor base. 

The second variable is FDIC's estimate of the expected loss on
assets.  Uninsured bear a pro rata share of estimated losses
resulting from asset disposal.  Thus, the amount of haircut varies
with the loss estimate.  To illustrate, when the 20 First City
Bancorporation banks failed, only 4 of the 20 failures involved
haircuts to the uninsured depositors.  The others had lower loss
estimates.  Therefore, if the existing equity and reserves exceeded
estimated losses then no haircut resulted. 

The third variable involves the uninsured depositors' status as
creditors.  Uninsured depositors are unsecured creditors.  However,
during the time of our study, some states had depositor preference
statutes--meaning depositors stand ahead of other general creditors. 
The passage of depositor preference legislation in August 1993 now
places all depositors ahead of other unsecured creditors.  If general
creditors take a larger portion of the loss, the uninsured will
absorb less. 


      COMPLEX INSURANCE RULES
      HINDER CONSUMER AWARENESS
------------------------------------------------------- Appendix V:3.3

The insurance status of an account can be affected by the basis on
which the account is established and the rights of the account
holder.  Before 1967, the various "rights and capacities" in which
funds were insured were determined by informal FDIC staff
interpretations of the FDI Act.  FDIC first promulgated deposit
regulations in 1967 and substantially revised the regulations in
1990. 

Under existing regulations, there are about nine different types of
accounts that reflect the different rights and capacities in which
funds are owned and may be separately insured.  For example, a
depositor may hold more than one single ownership account at an
insured institution.  For deposit insurance purposes, these accounts
would be added together (aggregated) and insured up to $100,000 since
they are maintained in the same rights and capacities.  If a
depositor has an individual account and a joint account with another
person at the same institution, those accounts would be insured
separately up to $100,000 each since the rights and capacities are
different (i.e., owned in different manners). 

The complexity of insurance rules makes it difficult for most
depositors to independently verify that their accounts are fully
protected or determine the extent to which their accounts may be
protected.  FDIC's aggregation of deposit accounts held by a failing
bank to determine the insurance status of funds in those accounts is
a complex process that FDIC, as the insurer, cannot always complete
before a failing bank is closed.  Thus, consumers cannot easily
determine the insurance status of their funds or know when their
funds could face increased risks from the potential failure of their
own bank or another insured depository controlled by the same holding
company (i.e., a multibank holding company). 

According to FDIC, the regulations involving the insurance of joint
accounts appear to be a frequent source of confusion for consumers. 
Regulations subject joint accounts to a two-part test to determine
insurability.  First, all accounts held by the same combination of
owners are aggregated.  Any amounts over $100,000 are uninsured. 
Next, the funds that pass the first test are deemed to be owned by
each co-owner on a pro rata basis--unless some other basis is
identified in the deposit account records.  Joint account funds
deemed to be owned by each individual are then aggregated.  No
individual is eligible for more than $100,000 in insurance for all
funds held jointly with other owners at a single insured institution
when owned in the same manner. 

FDIC has identified situations in which joint account owners appear
to have attempted to establish accounts in a manner that they thought
would maximize coverage but, in fact, left much of the funds
uninsured.  An example of this includes cases in which several
accounts were opened at one institution, but the account records
would list the owners's names in a different order for some of the
accounts or the social security numbers attached to the accounts
would belong to different owners.  Aggregating the balances of all
accounts in which the owners appear in the same sequence would
approach, but not exceed, $100,000.  However, in determining the
insurability of joint accounts, the FDIC would first aggregate the
balances of all accounts with the same combination of owners
irrespective of the order in which the owners' names are listed or
the social security numbers attached to the accounts.  Any such sums
exceeding $100,000 are not insured. 

Under the Financial Institutions Reform, Recovery, and Enforcement
Act (FIRREA) of 1989, FDIC may execute what is known as a
cross-guarantee provision.  When one bank among commonly controlled
institutions fails, such as a bank within a holding company, this
provision allows FDIC to offset potential losses to the insurance
fund by assessing the commonly controlled institutions for the
expected losses suffered by FDIC.  Applying the cross-guarantee
provision serves to protect the insurance fund and prevents commonly
controlled institutions from concentrating losses into one bank while
protecting others. 

In certain circumstances, however, the cross-guarantee provision
could increase the potential risks faced by uninsured depositors. 
For example, an uninsured depositor may deposit funds at a bank
perceived to be financially strong and solvent.  While the uninsured
depositor's bank may be adequately capitalized, it may be affiliated
with a bank, possibly in another city or out of state, that is in
serious financial trouble.  If the troubled affiliate fails, the
losses could be significant enough that FDIC's application of the
cross-guarantee provision could cause the solvent bank to fail. 

The major legislative changes affecting insured institutions enacted
by Congress over recent years, particularly FDICIA and FIRREA, have
served to protect the insurance fund but also have the potential for
exposing depositors--both insured and uninsured--to greater risks. 
Certain accounts, such as employee benefit plans, may experience
significant reductions in insurance coverage based on the financial
performance and condition of the insured institution holding such
funds rather than an investment decision made by the depositor.  In
other instances, funds placed in a financially strong institution can
face additional risks from problems that exist in financially weak
affiliated institutions.  With such uncertainty concerning these
potential risks and the insurance status of funds placed in a bank,
depositors have a greater need for adequate information to make
better informed judgments on where to maintain their deposit
accounts.  We have no basis at this time to determine or judge
whether depositors have access to or receive such types of
information. 

FDIC is aware of the potential that the current insurance rules can
lead to confusion among the public and within the banking industry. 
FDIC has started making more information available through public
announcement efforts, such as consumer pamphlets and a newsletter, to
increase the general awareness of the current insurance rules.  We
have no basis at this time to determine the effectiveness of these
FDIC efforts to disseminate information on deposit insurance
coverage. 




(See figure in printed edition.)Appendix VI
COMMENTS FROM FDIC
=========================================================== Appendix V


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


   GENERAL GOVERNMENT DIVISION,
   WASHINGTON, D.C. 
------------------------------------------------------- Appendix VII:1

Edward S.  Wroblewski, Project Manager
Vanessa Y.  Adams, Deputy Project Manager
James R.  Black, Senior Evaluator
Joe E.  Hunter, Evaluator
Edwin J.  Lane, Evaluator
Ned R.  Nazzaro, Evaluator
Mitchell B.  Rachlis, Senior Economist
Barry L.  Reed, Senior Social Science Analyst
Stephen J.  Saks, Senior Evaluator
Bonnie J.  Stellar, Senior Statistician
Susan S.  Westin, Senior Economist
Desiree W.  Whipple, Report Analyst


   OFFICE OF THE GENERAL COUNSEL
------------------------------------------------------- Appendix VII:2

Rosemary Healy, Senior Attorney


   BOSTON REGIONAL OFFICE
------------------------------------------------------- Appendix VII:3

Sally J.  Coburn, Evaluator
Kevin F.  Murphy, Senior Evaluator


   NEW YORK REGIONAL OFFICE
------------------------------------------------------- Appendix VII:4

John D.  Carrera, Senior Evaluator
Despina Hatzelis, Evaluator


   SAN FRANCISCO REGIONAL OFFICE
------------------------------------------------------- Appendix VII:5

Bruce K.  Engle, Evaluator

