Home > Industry Analysis > Research & Analysis > The First Fifty Years




The First Fifty Years

Chapter 5
Handling Bank Failures

An important consideration in setting up the FDIC was the establishment of an agency that, in addition to providing deposit insurance, would handle bank failures and liquidate failed bank assets in an orderly, inexpensive and nondisruptive manner. These latter functions have played an important role in the FDIC's 50-year history.

Procedures Used in Handling Failures - Early Years

The Banking Act of 1933 authorized the FDIC to pay up to $2,500 to depositors in insured banks that failed. The only procedure to be used to pay depositors was a Deposit Insurance National Bank (DINB), a new national bank chartered without any capitalization and with limited life and powers. Twenty-four insured banks were placed into receivership and their deposits were paid off through a DINB by the FDIC during the period of the temporary insurance plan, January 1, 1934 to August 23, 1935.

The 1935 Act gave the FDIC authority to pay off depositors directly or through an existing bank, and once that additional authority was granted, the FDIC ceased using the DINB for the next 29 years. During the past 20 years, the FDIC has used a DINB five times, the last occasion being the failure, in 1982, of Penn Square Bank, N. A., in Oklahoma City. The DINB essentially provides a vehicle for a slow and orderly payout, and its use in recent years has been confined to situations where only limited banking services were available in the community or where, as in the case of Penn Square, a regular payoff would have been substantially delayed.

In addition to broadening the ways in which a payoff could be effected, the 1935 Act gave the FDIC the authority to make loans, purchase assets and provide guarantees to facilitate a merger or acquisition. This authority had been sought by the FDIC because of its concern that many of the banks that had been granted insurance might not survive, and paying off insured depositors in these banks would be too expensive. In addition, most banking observers felt that there were too many banks in operation and that it would be desirable if the FDIC could facilitate an orderly reduction in their number through increased mergers.

Between 1935 and 1966, the procedure used by the FDIC to merge out failing banks did not actually involve a pre-merger closing or the establishment of a receivership. Acquiring banks assumed all the deposits of a failing bank and an equivalent amount of assets. In early assumption transactions, the FDIC determined the volume of sound assets of the failing bank and made a demand loan for an amount equal to the difference between deposits and sound assets, the loan being collateralized by the remaining assets. The FDIC would demand payment and foreclose on the remaining assets. Thus, the acquiring bank obtained cash and sound assets equal to assumed deposit liabilities. The FDIC would liquidate the acquired assets and repay itself for its cash advance from these proceeds. If collections exceeded the FDIC's advance plus interest, excess collections went to stockholders of the merged-out bank.

After several years in which loans were used to effect assumption transactions, it became apparent that certain legal problems that complicated the transaction (these related to bank borrowing limits and collateral foreclosure procedures) could be avoided if, instead of lending to the failing bank, the FDIC purchased assets from it. Consequently, direct purchase of assets became the standard procedure for facilitating a merger and the same general result was accomplished.

Beginning in 1935, the FDIC had two options in handling bank failures: payoffs or assumptions. When banks were paid off, depositors received direct payments from the FDIC up to the insurance limit. Uninsured depositors had a claim on the receivership for the uninsured portion of their deposits along with the claims of other general creditors, including the FDIC, which stood in the place of the insured depositors that it had paid.1 In these transactions uninsured depositors frequently did not receive the full amount of their deposits, and even when they did, there typically were long delays resulting in some loss through foregone interest. In assumption transactions, uninsured as well as fully insured depositors received all of their funds in the form of deposits in the acquiring bank. Once the FDIC began using the assumption transaction, it appears that the decision on which procedure to be used depended primarily on whether a potential, interested acquirer existed. Most payoffs occurred in states that did not permit branching so that an acquisition could not be easily effected.

It should be kept in mind that throughout its history the FDIC has not had the authority to close banks. That has rested with the Comptroller of the Currency in the case of national banks and with the state banking authorities in the case of state-chartered banks. Generally, the FDIC has worked closely with the primary supervisor in disposing of failing banks.

FDIC as Receiver

Prior to 1934, national bank liquidations were supervised by the Comptroller of the Currency, who had authority to appoint the receiver and had a permanent staff of bank liquidation specialists. Liquidations of state banks varied considerably from state to state and before 1900 were most often handled under the provisions for general business insolvencies. By 1933, most state banking authorities had at least some control over state bank liquidations.2 The increased incidence of national bank failures from 1921 through 1932 created a shortage of experienced receivers. Complaints were heard that receiverships, both national and state, had been "doled out as political 'plums', the recipients of which attempt to make as much commission as possible, and to keep the job going as long as possible."3 There were also conflicting concerns that depositors had to wait too long to recover their funds and that liquidators were causing undue hardship in the community by dumping acquired assets. When the FDIC was established, insured depositors could receive their funds more quickly without requiring rapid asset liquidation.

When a national bank is closed, the FDIC is automatically appointed receiver by. the Comptroller of the Currency. When an insured state bank is closed, a receiver is appointed according to state law. In 1934, 30 states had provisions by which the FDIC could be appointed receiver but, in practice, most often it was not. In the first 63 state bank liquidations, the FDIC was named receiver only seven times. Today, however, it is the exception when the FDIC is not appointed.

Before the FDIC can pay off insured depositors certain tasks must be performed. These include: posting and balancing individual deposit accounts up to the day of closing; computing and crediting interest on deposits up to the closing; merging of deposit accounts where multiple accounts exist to determine insurance liability; separating claims of depositors who have past due obligations to the bank; and preparing checks for payment. In some instances, the determination of precise insurance coverage may be a matter for subsequent litigation.

On July 5, 1934, Mrs. Lydia Lobsiger received the first federal deposit insurance disbursement, following the failure of the Fond Du Lac State Bank in East Peoria, Illinois. Photo: UPI
On July 5, 1934, Mrs. Lydia Lobsiger received the first federal deposit insurance disbursement, following the failure of the Fond Du Lac State Bank in East Peoria, Illinois. Photo: UPI

Every effort is made to begin the payoff as soon as possible, and in many instances the delay is only a few days.4 Depositors have 18 months in which to establish a claim with the FDIC. Customers whose deposits exceed the limit of coverage become general creditors for the balance due them, except in a few states where depositors are preferred over other creditors.

When the FDIC pays off insured deposits, it becomes a creditor of the receivership for the amount of its advances. Its claims against a receivership arise from its role as an insurer, and it essentially stands in the place of insured depositors. When appointed receiver, the FDIC assumes a fiduciary obligation to all creditors of the receivership and stockholders of the bank, with the responsibility to maximize the amounts recovered for them in as timely a manner as possible. The Federal Deposit Insurance Act, in Section II(d), requires that liquidations be conducted "having due regard to the condition of credit in the locality." This means that liquidations should be conducted in an orderly manner, avoiding a forced-sale dumping of assets. This requirement not only lessens the impact on the community, it is also conducive to realizing the greatest possible value on recoveries.

As assets of the receivership are liquidated, proceeds are periodically distributed as dividends to creditors, on a pro rata basis. If sufficient recoveries are made so that all creditors are fully paid, the remaining assets are turned over to the bank's stockholders. While this has occurred on occasion, the more typical receivership finds that the assets are not sufficient to satisfy all claims. In these instances, the receivership remains in existence until all recoverable assets have been liquidated or until the expected cost of recovery exceeds the value of the remaining assets.

Cost Test

Improved economic conditions in the late 1930s and during World War II significantly reduced the number of bank failures. Beginning in the mid-1940s, the FDIC ceased paying off banks. In its 1944 Annual Report, the FDIC reviewed disbursements and collections in payoffs and assumption transactions and suggested that the latter were a more efficient means of handling failing banks. Moreover, it suggested that the assumption method "provides a more flexible method of liquidating the affairs of an insolvent bank than does placing it in receivership. Depositors were fully protected; there was no break in banking service... and the community does not suffer the economic dislocations which inevitably follow a bank suspension."5

There was one payoff in 1944 and none between 1945 and 1953. During this latter period there were 24 assumptions, including cases in Illinois, Missouri, Texas, and Wisconsin - all essentially unit banking states. The FDIC was able to arrange assumption transactions with newly chartered banking groups in several of these cases. In its 1950 Annual Report, the FDIC boasted that "for nearly seven years receiverships of insured banks in difficulty have been avoided, and no depositor of any insured bank has lost a single penny because of bank failures. This constitutes an all-time record in the nation's history for bank solvency and safety of deposits."6

In Senate hearings on the confirmation of FDIC Directors in the fall of 195 1, Senator Fulbright, then presiding subcommittee chairman, questioned the FDIC policy of providing 100 percent de facto insurance to banks. While FDIC representatives defended their policies, Senator Fulbright argued that the FDIC was going beyond the scope of the insurance protection that Congress had contemplated and that the FDIC record suggested that its decisions to avoid receiverships did not reflect any substantial analyses or cost calculation.7 In October 1951, FDIC Chairman Maple Harl wrote to Senator Fulbright and indicated that in the future the FDIC would undertake a cost calculation to determine whether an assumption would be cheaper than a pay-off. Thereafter, the FDIC began to use a cost test in determining how to handle failing banks, and the prevailing thinking within the FDIC shifted to the opinion that the wording "such action will reduce the risk or avert a threatened loss to the Corporation" in Section 13(e) of the FDI Act required the FDIC to make an explicit cost calculation in deciding to facilitate a merger rather than paying off a bank. This is not a universally held interpretation.8

While the legal basis for requiring the cost test may have been in doubt, the FDIC continued to use it during the next 31 years. The Garn-St Germain Depository Institutions Act of 1982, which significantly revised Section 13 of the Federal Deposit Insurance Act, explicitly inserted a cost test.9

Closed-Bank Purchase and Assumption Transactions

The FDIC began to shift to payoffs in the 1950s, and between 1955 and 1958 there were nine payoffs and only three assumption transactions. From 1959 through 1964 there were 18 pay-offs and no assumptions. By the mid-1960s, the FDIC had rediscovered assumption transactions and it was recognized that there were advantages to having a bank closed by the Comptroller or the state, creating a receivership, and effecting a purchase and assumption transaction out of the receivership. This procedure eliminated the need for stockholder approval and, in certain instances, reduced the potential exposure of the acquiring bank and, indirectly, the FDIC.

In open-and closed-bank transactions the FDIC sometimes had several options with respect to assuming banks, and limited negotiations occurred with respect to such matters as loans to be assumed by the acquiring bank and the valuation of banking premises. However, it was not until January 1966 that the FDIC received an explicit premium in a purchase and assumption transaction, in connection with the failure of Five Points National Bank in Miami, Florida. By 1968 the FDIC had developed an explicit bidding process for handling closed-bank purchase and assumption transactions (P&As), and this was the way most bank failures, including practically all of the larger ones, were handled during the next 15 years.

A bank is closed and a uniform package is offered to bidders. This package consists of deposits and other nonsubordinated liabilities and a like amount of assets, less the amount of the premium bid. In its simplest form the assets consist of bank premises (subject to subsequent appraisal), cash assets, securities valued at market, performing consumer loans and cash furnished by the FDIC to equate acquired assets (less the premium paid) to assumed liabilities.

With the use of an explicit premium, the FDIC established a more formal procedure for its "cost test" and made it more likely that a P&A would be cheaper than a payout. When a bank was closed the FDIC estimated the cost of a payout by determining the shortfall in likely asset collections, the share of nonsubordinated liabilities accounted for by insured deposits and the expense associated with the actual payoff. Since the FDIC made all general creditors whole in a P&A, its share of the likely loss would be increased by the use of a P&A. However, that might be more than offset by the premium bid so that a minimum premium necessary to justify a P&A could be calculated beforehand and compared with the best bid received. In practice, the estimates of likely loss and even the level of insured deposits were not very precise so that there was a considerable margin of error in this calculation.

Using this procedure, the FDIC handled most commercial bank failures and practically all large failures through purchase and assumptions during the next 15 years, except where certain circumstances prevailed. These generally fell into two categories: (1) situations typically in nonbranching states where there was virtually no interest in acquiring the failed bank, and (2) situations where substantial fraud or other factors indicated the likely presence of significant unbooked liabilities or contingencies, which made it difficult to estimate the ultimate loss in the transaction and hence made it difficult to apply the cost test.

Bank Failures Since 1970

The early 1970s were relatively prosperous and there were only 17 bank failures between 1971 and 1974. Nevertheless, they included the first comparatively large failures encountered by the FDIC. Banking was becoming more competitive and the economic environment was becoming less forgiving. The first oil price shock occurred in 1973 and contributed to a rising inflation rate and new highs in interest rates in 1974.

The severity of the 1973-1975 and the 1981-1982 recessions led to a sharp increase in commercial bank loan losses and an increase in the number of bank failures. The 1973-1975 recession led to substantial real estate loan problems. In many instances these persisted well beyond the onset of economic recovery and, as a result, the bank failure rate remained high, peaking in 1976 at 16, the highest number since 1940.

The 1981-1982 recession was severe and it followed a weak recovery. The economy experienced its worst performance of the post-World War II period from the standpoint of unemployment, capacity utilization and business failures, and in 1982 there were 42 bank failures, including eight mutual savings banks. Despite the turnaround in the economy during the first half of 1983, there were still 27 bank failures during this period.

The first $100 million-plus failure handled by the FDIC was the $109 million Birmingham Bloomfield Bank (1971), located in a Detroit suburb. That bank was affiliated with the same management group whose policies brought the billion dollar Bank of the Commonwealth in Detroit to the brink of failure. Both institutions had invested heavily in long-term municipal bonds, relying considerably on purchased deposits, in anticipation of expected interest rate declines. When interest rates rose, the institutions incurred losses and found themselves locked into low-yielding, depreciated securities. The experience of these institutions did not prevent other banks from subsequently getting into situations where they became vulnerable to high and rising interest rates. To some extent that problem existed for the Franklin National Bank, which failed in 1974, and the First Pennsylvania Bank, N.A., which received financial assistance from the FDIC in 1980.

When interest rates rose dramatically in 1979-1980 and again in 1981-1982, most FDIC-insured mutual savings banks found themselves locked into long-term, low-yield assets (primarily mortgages) while their deposit costs rose substantially. Most incurred operating losses, and in 1981 and 1982 a total of 11 mutal savings banks failed. Throughout the FDIC's history, here have been 25 failures of commercial banks with assets over $100 million, all of which occurred since 1971. All buy transactions. (This excludes the three surviving $100 million-plus banks that received financial assistance to avert failure. See Table 5-1.)

Large Bank P&As

While the handling of smaller bank failures has tended to become routine, those involving larger banks have frequently involved special circumstances and sometimes included bidding situations that were tailored for the specific case. In October 1973, the $1.3 billion United States National Bank (USNB) in San Diego became the first billion dollar failure, and it was followed, in 1974, by the failure of the Fanklin National Bank in New York, the country's 20th largest bank, with assets of about $3.6 billion. Both of these failures involved special problems. USNB had outstanding a substantial volume of standby letters of credit that the FDIC sought to isolate from the transaction by considering them contingent claims with lesser creditor standing than depositors, and hence the FDIC. Holders of standby letters of credit of USNB sued the FDIC and won,10 the court decision coming almost five years after the bank failure.11 The FDIC could not discriminate against equivalent classes of creditors, and in this case the court ruled that the claimants in question had general creditor status. This case meant the FDIC would have to take account of contingent claims in applying the cost test to determine whether to pay off a bank or use a P&A. Contingent claims might include - in addition to standby letters of credit - outstanding lawsuits and claims arising from loan participations and failure to meet loan commitments. Since it is frequently difficult to assess liability on such claims at the time of a bank failure, additional uncertainty was injected into the decision process and influenced subsequent behavior of the FDIC.

Table 5-1. Twenty-Five Largest Banks Requiring FDIC Disbursements ($ Millions)
Rank Name/Location Date Assets Action1
1 First Pennsylvania Bank, N.A.0
Philadelphia, PA
Apr-01 $7,953 13(c)
2 Franklin National Bank
New York City, NY
Oct-74 3656 P&A
3 New York Bank for Savings
New York City, NY
Mar-82 3403 FAM
4 Greenwich Savings Bank
New York City, NY
Nov-81 2491 FAM
5 Dry Dock Savings Bank2
New York City, NY
Feb-83 2452 FAM
6 Western Savings Fund Society
Philadelphia, PA
Jan-82 2113 FAM
7 First National Bank of Midland
Midland, TX
Oct-83 1547 P&A
8 Union Dime Savings Bank
New York City, NY
Dec-81 1365 FAM
9 United States National Bank
San Diego, CA
Oct-73 1266 P&A
10 Bank of Commonwealth
Detroit, MI
Jan-72 1257 13(c)
11 Western New York Savings Bank
Buffalo, NY
Jan-82 1022 FAM
12 Farmers & Mechanics Savings Bank
Minneapolis, MN
Feb-82 980 FAM
13 Central Savings Bank
New York City, NY
Dec-81 899 FAM
14 United Mutual Savings Bank
New York City, NY
Dec-81 833 FAM
15 United American Bank
Knoxville, TN
Feb-83 760 FAM
16 Banco Credito y Ahorro
Ponce, PR
Mar-78 713 P&A
17 Fidelity Mutual Savings Bank
Spokane, WA
Mar-82 689 FAM
18 United States Savings Bank
Newark, NJ
Mar-82 675 FAM
19 Penn Square Bank, N.A.
Oklahoma City, OK
Jul-82 517 Payoff
20 Abilene National Bank
Abilene, TX
Aug-82 446 FAM
21 Hamilton National Bank
Chattanooga, TN
Feb-76 412 P&A
22 Farmers Bank of the State of Del.
Wilmington, DE
Jun-76 370 13 {c}
23 American City Bank
Los Angeles, CA
Feb-83 319 P&A
24 Oregon Mutual Savings Bank
Portland, OR
Aug-83 266 FAM
25 City & County Bank of Knox County
Knoxville, TN
May-83 262 P&A

1 13 {c}: financial assistance from the FDIC to avert failure.
FAM: financially assisted merger of a failing but open bank.
P&A: financially assisted merger of a failed (closed) bank.
Payoff: payment of insured deposits in a failed bank.

2 Merger was voluntary but with FDIC financial assistance.

The Franklin failure absorbed a substantial amount of FDIC personnel resources. There were negotiations over a five-month period among the FDIC, the Comptroller of the Currency, the Federal Reserve and the bidding banks. The transaction was complicated by the presence of foreign branches and foreign exchange speculation. As negotiations went on, Franklin experienced an enormous deposit outflow, which was funded by advances from the Federal Reserve Bank of New York. In the P&A transaction that was worked out, the winning bidder was required to take assets of Franklin equal to the remaining deposit liabilities less the premium bid. The trust activities of Franklin were sold separately to another institution. In contrast, the P&A bidding on USNB had been relatively simple. The FDIC agreed to remove the substantial volume of loans linked to that bank's management, and the transaction was effected quickly without significant deposit outflows.

By the time Franklin was closed, its borrowings from the Federal Reserve had reached $1.7 billion. The FDIC agreed to pay the amount due the Federal Reserve in three years, with periodic payments to be made from liquidation collections. The Federal Reserve released the collateral it held in connection with Franklin's borrowings. The FDIC had paid the Federal Reserve note down to about $600 million at the end of three years and, when it repaid the New York Fed in 1977, that represented the first significant cash outlay by the FDIC in that transaction. Subsequently, the FDIC recovered its cash outlay plus interest from additional liquidation collections.

The manner in which the Franklin P&A was handled significantly reduced the volume of assets to be liquidated by the FDIC. In several other large bank failures the FDIC sought to limit the volume of assets it took back by requiring winning bidders to take unclassified loans subject to certain limited buy-back arrangements. In smaller P&As, particularly where bidders were given little time to evaluate the condition of the failing bank, bidders generally received a "clean" bank. The winning bidder in the Franklin transaction was European-American Bank, a New York-chartered bank that was much smaller than Franklin, but a subsidiary of several very large European banks. In several subsequent P&A transactions, the FDIC invited foreign banks or subsidiaries of foreign banks to bid and in a few instances they were the winning bidder.

In two subsequent P&As, the FDIC accepted winning bids that involved two or more banks dividing up assets and liabilities of failing banks. These occurred in the case of Banco Credito in Puerto Rico in 1978 and American City Bank in California in 1983.

Bids received by the FDIC on failed banks have depended on the attractiveness of the franchise of the failing bank and its deposit mix, state branching laws and other considerations. An internal study done by the FDIC sought to explain the relationship between winning bids received by the FDIC and the volume of acquired deposits. Generally the explanatory variables were: (1) the volume of core deposits, essentially demand deposits and retail time and savings deposits (little value was given to large CDs and public deposits); (2) the number of bids submitted; (3) the attractiveness of bank franchises generally as measured by price-earnings ratios of bank stocks or the relationship between bank stock prices and book value; (4) the level of short-term interest rates (reflecting the fact that the FDIC typically provided a substantial volume of cash); and (5) the size relationship between the winning bidder and the bank acquired, a reflection of the likelihood that relative size of an acquisition is a good measure of the riskiness of the acquisition.

Until July 1982, every bank failure involving assets greater than $100 million had been handled through a P&A transaction. The largest payout was the Sharpstown State Bank in Houston, Texas, which failed in 1971 and had deposits of $67 million in 27,000 accounts. Litigation related to that bank's failure persuaded the FDIC that it could not reasonably assess the likely cost of a P&A transaction. Large bank failures were handled through P&As because that appeared to be the cheaper course. However, in most cases, precise cost calculations were difficult to make and close cases were probably resolved on the side of a P&A for several reasons. P&As were less disruptive to the local community and to financial markets generally. Moreover, the mechanical problems (balancing records, working out offsets and paying checks) of paying off a large bank with tens or hundreds of thousands of deposit accounts could conceivably take a month or longer.

Open-Bank Assistance

In 1950, the FDIC sought legislation to provide assistance to banks, through loans or the purchase of assets, to prevent their failure. Apparently there was concern that the Federal Reserve would not be a dependable lender to banks faced with temporary funding problems, particularly nonmember banks. The Federal Reserve opposed this recommendation, considering it an infringement on its lender-of-last-resort function. Congress did give the FDIC authority to provide assistance to an open bank, but it imposed restrictive language related to the circumstances under which such assistance could be given. Section 13(c) permitted such assistance "when in the opinion of the Board of Directors the continued operation of such bank is essential to provide adequate banking service in the community."

The FDIC did not use the authority of Section 13(c) until 1971, and it has only been used a total of five times. On one occasion ( 1974), open-bank assistance was given to provide temporary funding in order to buy time to arrange a P&A of American Bank & Trust (AB&T) in Orangeburg, South Carolina.12 This assistance was justified by the fact that AB&T was the only source of banking services in ten of the communities in which it operated, although other banks were located in nearby communities. It appears that this assistance could have been provided under Section 13(e), which allows the FDIC to provide financial assistance to facilitate the absorption of a failed or failing bank without a finding of "essentiality." AB&T was acquired by another bank 12 days after the assistance was given.

On the other four occasions that Section 13(c) was utilized by the FDIC, it was intended that the recipient bank would remain open and independent. Unity Bank and Trust Company in Boston (1971) and Bank of the Commonwealth in Detroit (1972) both served inner-city neighborhoods that were otherwise lacking adequate banking services. Farmers Bank of the State of Delaware (1976) was partially owned by the state and was its sole depository. The FDIC found the services provided by these three banks to be essential to at least a portion of the communities they served. In the most recent use of Section 13(c), assistance was given to First Pennsylvania Bank, N.A., in Philadelphia (1980). With assets of nearly $8 billion, First Pennsylvania was the city's largest bank, and its failure would have been the largest in U.S. history. In this case, the FDIC's determination of "essentiality" was based mainly on the bank's size. It would have been difficult to arrange a P&A, and the closing of such a large bank would have had serious repercussions not only in the local market but probably nationwide as well. This reasoning was also a factor in the "essentiality" finding for Bank of the Commonwealth, which had assets of $1.3 billion. In the Unity Bank and First Pennsylvania cases, other banks were partners to the assistance plan, agreeing to supply credit up to a certain amount. In the case of Farmers Bank, the State of Delaware joined the FDIC in aiding the bank.

Today, of the five 13(c) assistance cases, only First Pennsylvania has survived with the same ownership. Bank of the Commonwealth and Farmers Bank were sold but remain open, and AB&T and Unity Bank eventually failed.

The FDIC's authority under Section 13(c) was expanded by the Garn-St Germain Depository Institutions Act of 1982. At the discretion of its board of directors, the FDIC may provide necessary assistance to prevent the failure of any insured bank. Only if the cost of assistance would exceed the cost of closing and liquidating the bank does the FDIC have to make a finding of"essentia1ity." It is anticipated that the authorization of 13(c) assistance will continue to be the exception, though. The FDIC remains reluctant to use Section 13(c) because of its concern that the assistance would benefit stockholders, materially erode market discipline and keep afloat a weakened bank to the possible detriment of the local community.

As problem situations have become larger and more complex, the FDIC has been more inclined recently to make temporary loans under Section 13(e). This assistance provides the time necessary in the most difficult circumstances to arrange a P&A and minimizes disruption in the local market. Also, 13(e) advances can be secured, are short-term and do not require a finding of "essentiality." Temporary, subordinated loans of $25 million and $100 million were provided in 1983 under 13(e) to the United Southern Bank, Nashville, Tennessee, and the First National Bank of Midland, Texas, to provide time to work out an acceptable P&A for each bank. Also in 1983, a commitment was made to loan $250 million to Seattle First National Bank on a short-term, subordinated basis under Section 13(e). The bank was purchased by BankAmerica Corporation without FDIC assistance, so the 13(e) line was never utilized.

Various news titles regarding bank failures

Penn Square Bank

During the July 4th weekend in 1982, the Comptroller of the Currency closed the Penn Square Bank, N.A., in Oklahoma City, with deposits of $470 million, and the FDIC set up a DINB to pay off insured depositors. Penn Square had been an aggressive lender principally to small oil and gas producers. It had grown rapidly, relying heavily on purchased deposits and, to a much greater extent, on a program of participating the loans it originated to large regional and money center banks. As a result, when the bank failed it was servicing a loan volume almost five times the bank's liabilities. The loans were premised on extremely high oil and gas prices, and when the market weakened and production was curtailed, they went into default, and what collateral supported them had only limited value.

The FDIC paid off Penn Square primarily because it was not possible to assess the likely cost of alternatively arranging a P&A. Due to the heavy volume of loan participations and questions about the accuracy of information furnished to loan purchasers, a substantial volume of lawsuits was anticipated (and, in fact, have been filed). If those suits are successful, the cost to the FDIC of a P&A transaction would ultimately have been very substantial. By paying off insured depositors, the FDIC's maximum loss was the $250 million in insured deposits. This amount actually will be reduced by the FDIC's share of receivership collections. Had a P&A been effected, the FDIC would have had to agree to protect any acquiring bank from unbooked and contingent liabilities. To the extent that these were established in court, the FDIC would have had to pay full value on these claims. The way the failure was actually handled, claims established from lawsuits will have status in the receivership equal to other general creditors, including the FDIC.

The FDIC Board believed that the case for a payoff, as against a P&A, was overwhelming and that the FDIC would lose all credibility if it effected a P&A in the Penn Square case.13 That would have given financial markets a signal that all deposits, at least in banks above a certain size, were, for all practical purposes, fully insured. Discipline in the markets would have been seriously eroded, with deleterious long-term ramifications. Paying off Penn Square, though, had immediate repercussions. Uninsured depositors became more sensitive to the possibility of loss and could not assume that all but the smallest bank failures would be handled through purchase and assumption transactions. Some banks had difficulty rolling over large CDs. The business of brokers, who divide up large deposits and participate them to several banks, was significantly boosted. Depositors generally became more selective in their choice of banks, and the public's concern about the condition of banks was increased.

Recent Open-Bank Assumption Transactions

In the fall of 1982, the FDIC entered into two transactions where acquisitions of failing commercial banks were facilitated without the closing of these banks. These were essentially assisted mergers, but in each case (Abilene National Bank, Texas, and Oklahoma National Bank and Trust Company) the stock of the failed bank had been pledged as collateral to the acquiring institution. The stock was foreclosed, a merger was effected and the FDIC provided assistance. Stockholders of the failed bank obtained virtually no benefit from the transactions. In one instance the FDIC lent money on favorable terms to facilitate the transaction and in the other case the FDIC agreed to buy back loss loans when they surpassed a specified level. In both cases the FDIC Board believed that these transactions would be considerably cheaper than a payoff or a closed bank P&A. Other important considerations were that FDIC liquidation resources were considerably stretched at that time and the transactions (particularly Abilene) would not utilize any liquidation staff. At that time, banking in the southwest was still affected by the uncertainties from the Penn Square failure and additional failures could have had negative repercussions. While the initiative in both transactions came from the acquiring institutions, the FDIC went back to the pre-1966 procedure in working out negotiated pre-failure mergers of failing commercial banks. However, in both of these cases, special circumstances related to stock ownership helped make the transactions feasible for the FDIC in that shareholders received no subsidy and claims against officers, directors and others were preserved.

Assisted Mergers of Mutual Savings Banks

Mutual savings banks had been vulnerable to rising interest rates for several decades. Most of their asset portfolios consisted of long-term, fixed-rate assets, principally mortgages and mortgage-backed securities. An accelerating inflation rate in 1978 and a shift in the manner in which monetary policy was conducted in the following year led to an almost continuous rise in interest rates until the spring of 1980. Despite a sharp, though brief, break in interest rates in 1980 and a smaller decline in the fall of 1981, interest rates remained near record levels through mid-1982.

During this period interest ceilings on time deposits were raised several times and a variety of new deposit instruments were made available to banks and thrifts. Nevertheless, substantial amounts of deposits shifted from banks and thrifts to money market funds or to market securities, and depository institutions experienced both disintermediation and an increased cost of funds.

At the same time, yields on savings bank asset portfolios changed very little because of their lengthy maturities, and as the cost of funds rose, earnings disappeared and losses began to grow. By early 1982, aggregate savings bank losses were running at about a $2 billion annual rate, about 1.25 percent of assets. However, some of the weaker institutions in New York City were losing at a rate of 3.5 percent of assets. The problem faced by the FDIC from the standpoint of potential exposure of the deposit insurance fund was very different from any faced earlier in its history. Asset quality was not a problem. However, in the case of many large institutions that faced "book" insolvency, the market value of their assets was actually 25 to 30 percent below outstanding liabilities. Their failure could have resulted in enormous FDIC losses. The first failing savings bank transaction, involving the $2.5 billion Greenwich Savings Bank in New York, had an initial estimated cost of $465 million, more than the reported cost of handling all previous insured bank failures.

The FDIC's principal concern was how to keep the cost of handling failing savings banks at a reasonable level without undermining confidence in the industry or in the FDIC. Various devices were used to handle failures. One of the most successful was the income maintenance agreement. The FDIC agreed to pay an acquiring institution the difference between the yield on acquired earning assets (primarily mortgages and taxable bonds) and the average cost of funds to savings banks for some number of future years.14 This might be supplemented by an additional dollar payment in the future or by an up-front cash payment. The income maintenance was subsequently modified so that the FDIC defined the asset base according to existing asset maturities and yields on the failing bank assets and specified prepayment assumptions. Bidding banks would be paid the spread between defined asset yields and the cost of funds, whether they held the failed bank's assets or sold them.

The income maintenance covered any negative interest spread for acquiring banks regardless of what happened to interest rates and the cost of funds. Thus, the FDIC took the interest rate risk on the transactions. The FDIC was in a better position to assume this risk and potential acquirers were willing to bid more aggressively as a result of this. Income maintenance was used in nine of the 12 assisted mergers of failing savings banks between 1981 and early 1983.

The first savings bank transaction was handled through a mixture of bid and negotiation. In subsequent transactions, the FDIC defined certain bidding ground rules and indicated, generally, how bids would be priced, and then entertained bids in a variety of forms. This was in contrast to the way most commercial bank P&As had been handled, where everything was specified beforehand and bidding banks submitted a single number.

Failing savings banks were not actually closed. The transactions were assisted mergers. However, the FDIC insisted that senior management and most trustees could not serve with the surviving institution. Since there are no stockholders in mutual institutions, the FDIC did not have to concern itself with receivership interests of existing stockholders. In several of the failing savings banks there were subordinated notes that normally would have only a claim on the receivership in a purchase and assumption transaction on a closed bank. Generally, the FDIC negotiated with noteholders, forcing them to take a lower interest rate and/or an extended maturity. Thus, noteholders took a substantial "hit". In pursuing this policy the FDIC weighed the cost of not wiping out noteholders altogether, by closing the bank, against offsetting considerations. These included possible lawsuits to delay the transactions, greater flexibility for the acquiring institution in continuing leases and other contractual arrangements, cooperation from state supervisors and the possible impact on deposit outflows in other savings banks.

Two of the acquiring institutions were commercial banks and the remainder were other savings banks. Most of the latter were losing money at the time the transactions were effected, although they tended to be stronger than most of their peers. Traditionally, the FDIC has been reluctant to solicit bids from poorly performing institutions, but during this period stronger commercial banks were reluctant to bid aggressively on savings banks because of the asset depreciation and its impact on their balance sheets, and because of the potential impact on capital ratios. In order to keep its cost down the FDIC was willing to compromise on bidder standards and acknowledged the possibility, at least within the agency, that in an unfavorable interest rate environment, some of the acquiring banks could encounter difficulty in the future.

For the most part, classified assets were relatively unimportant in the failing savings banks, and after the first few transactions, when some problem assets were removed, virtually all assets were passed to the acquiring bank. As a result, the cost of the transactions was determined at the outset where FDIC assistance was confined to cash or notes, or else costs were dependent principally on future interest rate developments. Where the latter was the case, future costs were estimated by discounting projected future payments based on prevailing interest rates. The present value of estimated outlays was immediately determined. When interest rates subsequently declined, loss estimates were adjusted to reflect actual outlays and revised future outlays. Between the fall of 1981 and the end of 1982, there were 11 assisted savings bank mergers. The assets of the failing institutions totaled almost $15 billion, more than the total assets of all failed commercial banks since the FDIC was founded. Based on cost of funds projections made at the end of 1982, the cost of these transactions amounted to about 10 percent of assets. While this appears to be a higher cost than typical commercial bank failures, comparative figures may be deceiving. Until 1983 the FDIC did not take account of forgone interest in calculating its losses in commercial bank failures. If adjustment is made for this, then the cost of the savings bank transactions appears to be no higher than the relative cost of most commercial bank failures.

The Garn-St Germain Bill, which was passed in October 1982, included provisions, despite FDIC reservations, whereby savings banks and other qualifying institutions could apply for net worth certificates if they met certain conditions with respect to losses and low surplus ratios. In December 1982, the FDIC implemented a program enabling savings banks to apply for these certificates in amounts equal to a percentage of operating losses. The certificates count as surplus for regulatory purposes. The certificates involve essentially a paper exchange, enabling the institutions to continue to operate. By mid-1983, 24 savings banks with assets of about $37 billion were utilizing this program, and they had approximately $300 million in net worth certificates outstanding. The decline in interest rates has cut savings bank losses, increasing the possibility that many of these institutions will be able to survive or else be merged out with only limited assistance. The net worth certificate program has forestalled savings bank failures, at least temporarily. During the first half of 1983, there was only one assisted savings bank merger, and that was essentially a voluntary transaction that could have been forestalled through the use of net worth certificates.

FDIC Liquidation Activity

The two goals of a receiver - liquidating assets as quickly as possible and realizing the greatest possible value - can come into conflict because sometimes it is desirable to hold an asset until market conditions improve. An obvious problem, though, is that poor asset quality is a factor in virtually every bank failure, and liquidating assets is normally a very lengthy procedure.

In its first seven years of operation, the FDIC handled an average of 50 failures annually. As a result, the failure-related assets acquired by the FDIC increased, peaking at $136 million in 1940. Over the next three decades, failures averaged fewer than four annually, but these were generally larger banks than had failed in the early years. Still, the volume of assets in liquidation, which was only $2 million in 1952, did not again reach the 1940 level until 1971. FDIC liquidation activity has escalated dramatically in the past decade. The volume of assets in liquidation reached $2.6 billion in 1974, and stood at $2.2 billion at the end of 1982, and $4.3 billion by December of 1983. Through November of 1983, the FDIC had been involved in 665 receiverships, of which 170 were still active.

Receivers of failed banks always acquire some loans which are in default. These result in litigation and, when secured, foreclosure on collateral. Many failed banks have been involved in what might euphemistically be referred to as "atypical" financial dealings, and the FDIC's liquidation portfolio has, from time to time during the past 50 years, included some rather unusual assets. In one instance, a bank failed because its president was illegally diverting bank funds to finance production of a motion picture. The failure occurred after filming had been completed but before editing. The FDIC then had to decide whether the movie, which had some name actors but was hardly an Academy Award threat, was likely to return the additional investment required to complete and distribute it.

The FDIC has also had interests in oil tankers, shrimp boats and tuna boats and has experienced many of the pitfalls facing the maritime industry. An oil tanker ran aground, a shrimp boat was blown by a hurricane onto the main street of Aransas Pass, Texas, and the tuna boats were idled when the price of tuna dropped sharply. Other liquidation assets have included several taxi cab fleets; a coal mine that was on fire the day the bank was closed; a horse training facility, two inept race horses and quarter horses valued at several million dollars; thousands of art objects, including an antique copy of the Koran; a collection of stuffed wild animals; and all forms of real estate, including churches and synagogues. Single bank failures have resulted in the FDIC's acquisition of 400 single-family homes and as much as $500 million in international loans. Assets have also included loans secured by distribution rights to a well known blue movie ("The Happy Hooker"), by the operation of a house of prostitution and by the warehouse inventory of a "King of Pornography. "

Assets require active FDIC management when, for one reason or another, their sale cannot be arranged quickly. This can necessitate additional investment by the FDIC, as well as development or acquisition of highly specialized expertise. Asset management has required purchasing wind machines to protect citrus orchards from freezing weather as well as beehives for pollination of almond trees. The FDIC's mortgage interest in a Chicago meat warehouse was abandoned when the refrigeration system failed, and one million pounds of meat spoiled. FDIC liquidators have been called upon to operate hotels, motels, condominiums, office buildings, restaurants, a bakery and a kennel. One management problem involved a residential real estate development, an attraction of which was a golf course that happened to be located in a flood plain (providing some insight into the developer's acumen). An investment of $1 million was required to improve the golf course and thereby enhance the overall marketability of the development. The FDIC also found itself in possession of an abandoned gold mine in Idaho. A buyer could not be found until the FDIC had transformed the property into a successful tourist attraction.

As predecessor to the FDIC's Division of Liquidation, the New and Closed Bank Division supervised seven receiverships in 1935 with a staff of 25 employees. It was also involved with 26 other liquidations for which the FDIC had not been appointed receiver but was a major creditor by virtue of having paid insured deposits. The personnel requirements of the Division have fluctuated widely from year to year, dictated by the number, size, complexity and duration of active receiverships. In the early 1940s, the Division employed more than half of all FDIC personnel, topping 1,600 in 1942, having had to handle nearly 400 failures from the time that deposit insurance became effective in 1934. In the early 1950s, by comparison, as few as 32 liquidation personnel were required as the number of failures had declined in the post-World War II period. Today, because of the recent increase in bank failures and a surge in the volume of assets in liquidation, the Division employs approximately 1,400 people, supplemented by scores of bank examiners on detail from the Division of Bank Supervision.

The occurrence of several bank failures within a short period of time - or even a single large bank failure - can create a sudden demand for experienced liquidators. Some personnel are retained from the failed bank, and many other clerical personnel are hired locally on a temporary basis. The FDIC also relies more heavily now on locally hired liquidation specialists to assist its professional staff.

Present Liquidation Procedures

When a bank is closed by its supervisor and the FDIC is appointed receiver, the first task is to take custody of the bank premises and all records, loans and other assets of the bank. In some instances, even this initial task has been formidable. Franklin National Bank in New York, for example, operated 108 branch offices, and its closing required a force of 778 FDIC personnel, most of whom were examiners on temporary assignment from the Division of Bank Supervision. When The First National Bank in Humboldt, Iowa was closed in 1982, weather conditions conspired to make it all but impossible for FDIC personnel to reach the bank. After first dodging tornadoes, they were confronted by a severe snowstorm that turned expected journeys of only a few hours into two-day ordeals. Happily, serious injuries were avoided, but these employees endured highway closings, vehicle abandonments and numerous accidents, completing portions of their trip by tractor trailer and state police car. That same weekend, in addition to monitoring these travails in Iowa, FDIC officials in Washington had to arrange the mergers of a failing $2 billion savings bank in Philadelphia and a small bank in Virginia, for which no buyer could be found until nearly midnight on Sunday (occasioning what may have been the latest FDIC board meeting).

Sometimes a banker is unwilling to accept his bank's insolvency. In an incident in Indiana, the president of a bank about to be closed had moved a cot into his office, threatening first not to leave and later to commit suicide. The situation was resolved peacefully.l5

After possession of the bank has been taken, notices are posted to explain the action to the public. Locks and combinations are changed as soon as possible, and correspondent banks and other appropriate parties are .notified of the closing by telephone and telegram. In a payoff all incoming debit items, such as checks, are returned marked "drawee bank closed." Deposits received after the closing are returned in full to the depositors.

A Liquidator-in-Charge is appointed by the FDIC to supervise the receivership. To provide some continuity, non-tainted" employees of the failed bank are hired by the receivership for as long as their services are required. As soon as possible, the liquidation activities are moved to nearby office space rented for that purpose, because in most instances the bank's premises are transferred to another banking organization. Thus, the FDIC has active liquidation offices scattered across the United States and its possessions. The five recently established Area Offices will enable earlier closing of on-site offices because the final stages of liquidations can be handled more efficiently on a consolidated basis. At the end of November 1983, all but 35 of the 170 active receiverships had been consolidated.

The time it takes to conclude a liquidation varies greatly according to the number and size of acquired assets as well as their salability. Markets can readily be found for most loans, which are often sold in blocks; but some assets, particularly those acquired in foreclosure, are more difficult to dispose of for reasonable value. Large bank failures occurring in the past decade have created receiverships so large and complex that some may take ten years or more to complete. The FDIC can serve as a lender-of-last-resort if additional investment is required to protect the interests of the receivership. Whenever possible, though, borrowers are required to establish new banking relationships.

The FDIC is usually quite successful in recovering the disbursements it has made. In the 495 insured bank liquidations that have been completed since 1933, the FDIC recovered about 93 percent of its outlays, faring somewhat better in deposit assumptions (95 percent) than deposit payoffs (89 percent), but in the 170 active cases, recoveries are expected to be lower. The historical recovery rates, however, do not fully take into account the foregone interest earnings on advances to receiverships. This interest was collected only on occasion, after disbursements had been fully recovered. Had this expense been acknowledged, and FDIC advances reduced by the present value of collections, it was estimated that for the period 1934-1980, insurance losses and expenses would have increased from four percent of failed bank assets to nine percent.16 Beginning in 1983, the FDIC's recovery and loss experience will more accurately reflect its money cost.

Until the 1970s, FDIC receiverships generally retained longterm performing assets. This tended to improve reported liquidation results since both interest and principal collections were included in recovery calculations. In recent years the practice has been to sell those assets (e.g., securities, mortgages) that are marketable without concern about boosting "apparent performance." In some cases, holding performing assets has benefited junior creditors and stockholders at the expense of the deposit insurance fund. Even where returns on assets exceed the FDIC's opportunity rate, FDIC policy has opted for early sale, recognizing that the FDIC is not an investment company and that its own investment portfolio is restricted to Treasury securities.

Summary

During its 50-year history the FDIC has handled bank failures by paying off insured depositors or merging the bank on an open- or closed-bank basis. In a small number of cases until the net worth certificate program was implemented, the FDIC has forestalled failures by assisting open banks. The specific manner in which failing banks have been handled has varied according to legislation, the experience gained by the FDIC and the specific nature of the problems faced. When confronted with major problems where traditional approaches may not have worked, the FDIC has been flexible and sometimes imaginative.

Throughout its history certain conflicts have emerged. Periodically the FDIC has had to question whether it is appropriate to raise de facto insurance coverage through P&As and assisted mergers when that approach is cheaper or less disruptive, and whether there is a cost associated with providing too much de facto insurance. When a bank is going to fail it is desirable to get the transaction done quickly. This argues for simple, clean P&As where P&As are appropriate. However, that means the FDIC must collect on more loans, a result that, in the long run, may be more disruptive to the community and more expensive.

A precisely defined bid situation where bidders submit a single number seems most fair, at least on the surface, and it exposes the FDIC to the least criticism. On the other hand, requiring everyone to bid on the same basis is not always likely to give rise to the best or cheapest solution, and it may favor a particular set of bidders. The FDIC may prefer an absolute ban on helping stockholders or subordinated creditors in assisted mergers or open-bank assistance. However, that may mean foregoing transactions that can save the FDIC a lot of money or forestall other failures. Concern on the part of the FDIC that acquiring banks not be exposed to excessive risk or that they meet certain capital standards or treat goodwill in a particular way can also increase the cost of transactions to the FDIC.

These and other conflicts have been faced by the FDIC during its history and have not always been resolved in the same manner by FDIC Boards. They will likely continue to confront future FDIC Boards.

FOOTNOTES

1 In receiverships prior to August 1935, the FDIC was a preferred creditor and was paid prior to uninsured depositors.

2 Cyril B. Upham and Edwin Lamke, Closed and Distressed Banks-A Study in Public Administration (Washington, D.C.: The. Brookings Institution, 1934), p. 30.

3 Ibid., p. 62.

4 It is generally conceded, however, that delays in the case of a large bank payoff could be considerably longer.

5 Federal Deposit Insurance Corporation, Annual Report, 1944 (1945), p. 18.

6 Federal Deposit Insurance Corporation. Annual Report, 1950 (195 1 ), p. 12.

7 U.S., Congress, Senate, Committee on Banking and Currency, Hearings before a subcommittee of the Senate Committee on Banking and Currency on the Nominations of H. Earl Cook and Maple T. Harl to be Members of the Board of Directors of the Federal Deposit Insurance Corporation, 82d Cong., 1st sess., Part 2, September 27 and October 1. 1951.

8 Golembe has argued, "Section 13(e) says nothing at all about a comparison of the use of the deposit assumption techniques with the deposit payoff procedures, nor does it require, in our view, that the former be less costly than the latter. But Senator Fulbright, who must long since have forgotten his little personal feud with the FDIC directors, still exerts his influence over the FDIC decisions!" Carter H. Golembe, Golembe Reports, vol. 1974-8: Memorandum re: Bank Failures and All That (Washington, D.C.: Carter H. Golembe Associates, Inc., 1974), p. 11.

9 In connection with revised provisions related to facilitating a merger, the Act states: "No assistance shall be provided . . . in an amount in excess of that . . . necessary to save the cost of liquidating . . . ."

10 First Empire Bank, New York, et nl. vs FDIC. 572 F.2d 1361 (9th Cir.), cert. den. 431 U.S. 919 (1978).

11 It appears that the FDIC anticipated an unfavorable decision on this case several years earlier and this seems to have entered into cost calculations.

12 The Federal Reserve had declined to lend to AB&T. a $150 million nonmember bank. In 1980 the availability of the Federal Reserve discount window to nonmember banks was made explicit by Congress.

13 The presence of a large volume of uninsured deposits in the bank and indications that liabilities substantially exceeded likely asset collections made it extremely unlikely that a P&A could have been cost-justified even if lawsuits were ignored.

14 Previously, the FSLlC had provided assistance along these general lines in connection with an assisted interstate merger. The FDIC's assistance to Bank of New Orleans in the closed-bank P&A of International City Bank in 1976 had also contained characteristics similar to the income maintenance agreement.

15 Interview with Neil Greensides, former Chief, Division of Examinations, "FDIC Pioneer Recalls 'Early Days'," FDIC News (June 1983), Vol. 37, p. 4.

16 Federal Deposit Insurance Corporation, Deposit Insurance in a Changing Environment (Washington, D.C.: Federal Deposit Insurance Corporation, April 15, 1983). p. V-6.

Last Updated 7/24/2006 insurance-research@fdic.gov