The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
SM-85-3 c at E E o u "O c $ *> o QC IM P LIC A TIO N S O F LARGE BAN K PRO BLEM S A N D INSO LVENC IES FO R T H E BANKING SYSTEM A N D EC O N O M IC PO LICY George G. Kaufman Loyola University and Federal Reserve Bank of Chicago m a 50 > 50 m c/i 50 < m 00 > Z 5* o ■n n o > n Q Implications of large bank problems and insolvencies for the banking system and economic policy George G . K au fm an * It has long been recognized, both before and after the introduction o f fed eral deposit insurance, that large, money center bank failures are different from small bank failures. Nevertheless, the available evidence suggests that bank regulators were not prepared to deal with a failure the size o f the Continental Illinois National Bank. There were no blueprints on the shelf! This m ost likely affected both the regulators’ handling o f the bank before the final agreement and the nature o f the final agreement itself. In addi tion, there also appears to be widespread agreement that, at least in theory, the introduction o f federal deposit insurance broke the close link between the failure o f an individual bank and that o f the banking system. Y e t, many o f the statements made by regulators during the Continental affair noted dangers to other banks. The fall o f the Continental might bring down other banks through a “ dom ino” or chain reaction effect and increase depositor losses further. After discussions with the Federal Reserve, F D I C and Treasury Departm ent, Com ptroller o f the Currency C onover warned that, i f the “ Continental had failed and been treated in a way in which depositors and creditors were not made whole, we could very well have seen a national, if not an international, financial crisis the dimensions o f which were difficult to imagine. N o n e o f us wanted to find o u t.” 1 This paper will reexamine why and when large bank troubles are m ore serious than those o f smaller banks; what the problems caused by such financial troubles are; the implications for other institutions, the financial sector, and the econom y as a whole; and possible efficient corrective actions by bank regulators. Business firms o f any kind fail economically when the market value o f their assets (including goodwill, franchise value, deposit insurance value, and other legitimate intangibles) falls below that o f their liabilities. I f liquidated or sold as a unit before this occurs, losses are experienced by shareholders, but not by creditors. In addition, there is likely to be disruption to em ployees, suppliers, and customers, who must scurry around to locate new affiliations. Unless the firm',, products are unique, the greatest discomfort ‘ Loyola University and Federal Reserve Bank o f Chicago. I am indebted for helpful comments in the preparation o f this paper to Herbert Baer, George Benston, Larry Mote, Harvey Rosenblum, and Steven Strongin. An earlier version o f this paper was presented at a Sympo sium on Issues and Options in Dealing with Large Banks’ Problems and Failures sponsored by the Bank Administration Institute; Dartmouth College; Hanover, New Hampshire; August 29-30, 1984 and is published in Issues in Bank Regulation (Winter 1985). The views expressed in this paper are the authors' and do not necessarily represent the views o f the Federal Reserve Bank o f Chicago or the Federal Reserve System. FRB CHICAGO Working Paper 1 may be expected to be felt by the employees, who may lose their job s, se niority arrangements. In perfect labor markets, however, any unemploy ment should be brief and other losses minimal. But insolvent firms need not be closed. I f the firm were permitted to operate in bankruptcy and experience further losses, the immediate losses to the employees may be reduced, but losses will now accrue to creditors. The longer the firm’ s losses are permitted to continue or the larger they are, the larger will be losses to creditors. I f the firm ’s fortunes do not reverse, losses will eventu ally accrue to all parties. Failure plays an important role in maintaining economic efficiency. It ef fectively is the econom y’s way o f saying that the firm’ s products no longer pass the market test and that its resources should be transferred to other firms. Thus, changes in consumer tastes and supplier technology are transmitted throughout the economy. Barriers to exit create barriers to entry and interfere with efficiency. Although painful, the losses associated with failure are a lesson that may be expected to influence the behavior o f market participants. The greater the penalty for failure, the more reluctant will those involved be to try again and the less likely will others be to follow the same strategy. Conversely, the smaller the penalty, the more likely is it that history will repeat itself. Com mercial banking is no exception. Failure is as important in banking as anywhere else. But history has demonstrated that, at times, bank failures may generate larger losses than failures o f nonfinancial firms o f comparable size. Thus, bank failures need to be analyzed separately.2 M an y o f the reasons that bank failures are both more serious and more difficult for regulators to handle were spelled out some 10 years ago by Paul H orvitz and Thom as M a y e r.3 M u ch o f the argument is familiar and will not be repeated here. Primarily, it deals with the greater likelihood o f spillover to other banks that ignites a chain reaction o f bank failures. But Horvitz also noted in passing that some more recent evidence suggested that the spillover may not be as great as frequently believed. The spillover was precisely what deposit insurance was designed to prevent. Full federal insurance coverage for all deposits certainly avoids any form o f bank run; the banks can effectively issue U . S. Treasury securities and there is no in centive for depositors to withdraw funds for reasons o f safety. But it is also likely to reduce market discipline on banks and, in the absence o f explicit risk-sensitive cash insurance premiums, implicit noncash risk-sensitive pre miums through regulation and examination or other risk-sensitive schemes, such as mandatory minimum short-term subordinated debt, encourage them to be riskier than otherwise.4 In case o f failure, only shareholders will experience losses and only up to the amount o f their investment. W e will restrict our analysis to the current system o f de jure m axim um $100,000 insurance on private deposits. Thus, potential losses m ay, in theory, extend FR.B CHICAGO Working Paper 2 to large creditors (including depositors) and the penalty for failure be more severe. W h y , in contrast with the p re-F D IC environment, has there appeared to be little, i f any, spillover for large, uninsured deposits in the p o st-F D IC establishment environment that enlarged losses beyond those already suf fered at the initial distressed bank, even in the aftermath o f the failures o f large institutions in a market area? Perhaps because large depositors viewed their funds as de facto, although not de jure, fully insured. A n d , at least equally important, what can large depositors do with the funds? A careful analysis o f U . S. banking crises, up to and including the 1929-33 crisis, shows clearly that the primary reason for the severity o f the spillover effects was the attempt by depositors to convert first, notes and later, de posits, first into gold and later into currency.5 Under a fractional reserve banking system, an aggregate loss o f reserves in the form o f gold or o f currency causes a multiple contraction in deposits and thus money and bank credit. In addition, such a contraction is likely to cause forced, hur ried “ fire” sales o f assets by banks at below equilibrium market value and ignite a chain reaction o f bank failures. Spillover is said to exist when one bank failure sets o ff failures in other banks. This is difficult to measure directly. Because chain reaction failures m ay reasonably be expected to decrease aggregate bank deposits and m oney supply more than other bank failures, we can use reductions in the money supply associated with a cur rency outflow— that is, a ju m p in the currency-money ratio—as a proxy for chain-reaction bank failures. Unfortunately, the literature is not clear on whether large depositors with drew funds in gold and currency or shifted their funds to other, presumably more secure, banks upon receiving unfavorable inform ation on the financial condition o f their bank.6 In contrast, the evidence is quite clear that, in re cent bank crises, large depositors shifted their deposits to other banks. They have not withdrawn and held their funds in the form o f currency. N o r have the difficulties experienced by banks ignited currency drains by smaller, insured depositors. There were no lines to speak o f at the C onti nental or other depository institutions in Chicago, Possibly, this is because large bank difficulties are becoming old hat in Chicago. Only two years earlier, the two largest savings and loan associations (S L A s) in the city—the First Federal and Talman H om e—were effectively taken over by the federal government. A gain, there were no lines at that time. In between, a number o f smaller banks and S L A s had closed their doors and had been merged away.7 (The lack o f currency runs in these instances is, o f course, related to the m axim um de jure deposit insurance coverage. The lower this am ount, the more likely is a currency run. I f deposit insur ance were reduced to, say, $1,000 or even $5,000, net currency runs would very likely have developed. The question o f what is the minimum coverage FRB CHICAGO Working Paper 3 that is consistent with both preventing currency runs and preserving market discipline warrants further research). I f there is no currency outflow, then the spillover to other institutions may be expected to be significantly smaller and any destruction o f total deposits and credit alm ost insignificant in the aggregate. Some reductions may oc cur from losses on uninsured deposits and write-offs against net worth. Thus, it is highly unlikely that the Continental crisis would have ignited a significant contraction in aggregate money and credit as in earlier American banking crises.8 Instead, the run on the Continental caused other problems. Assum e, for the mom ent, that the Continental was the only bank rumored, correctly or incorrectly, to be in financial difficulties. Corporate treasurers would reasonably be expected to shift deposit balances that are not required at the Continental for loan and service compensation to other large banks as soon as possible. W h a t occurs at the Continental depends on a number o f factors including its liquidity and, most importantly, whether it is actu ally or perceived to be solvent in terms o f the market value o f its assets exceeding that o f its liabilities other than net worth. The greater the liquidity, the faster can a bank sell o ff assets to meet deposit losses without loss below their true market values and, thereby, preserve its capital. For an individual solvent bank that marks to market, deposit runoffs present little problems other than relatively small losses from hur ried sales. I f the bank is actually or is perceived to be solvent, the problem for the banking system is basically a recycling problem—that is, a problem o f cycling the funds back to the stricken bank. Throughout U . S. history, this has been undertaken by other large banks operating in a com m on in terest. A lthough the central bank’ s lender o f last resort function tradi tionally applies primarily to the banking system as a whole, it includes providing assistance to solvent individual large banks experiencing runs for one reason or another. W hile funds are lost to the Continental, they are not lost to the banking system. However, what if whatever started the run on the Continental also affected some other large banks, for example, fear o f default by a single large borrower whose loans are concentrated in a few large banks. Large depositors may be less willing to transfer their deposits to the other banks. But where will they park their funds? I f they shift them to European or other out-of-country banks, the funds still do not leave the U . S. banking system. The receiving bank would acquire the dollar bal ances at the paying bank or some other domestic bank. I f they use them to purchase Treasury securities or other high quality debt, the balances are transferred to the seller o f the securities, who will buy other securities or redeposit the funds in the perceived safest bank§. A gain, no funds are lost to the system, they are just redistributed. The “ flight-to-quality” , however, will affect interest rate spreads, lowering interest yields on Treasury securities and raising yields on bank C D s and FRB CHICAGO Working Paper 4 related securities. Such changes will change the profitability o f institutions depending upon both the duration and quality com position o f the two sides o f their balance sheets. In addition, in reaction to the crisis that ignited the depositor’ s flight to quality, banks are likely to restructure their portfolios to emphasize Treasury and other safe securities. This will reinforce the widening o f the risk premia on “ risky” securities and make it costlier for m ost but the Treasury to borrow and may dampen private investment spending. Even though the recycling and spread problems are real, they are unlikely to be nearly as serious as the net reserve outflow and multiple deposit contraction problem, and require different solutions. Some banks and borrowers may be losers, but others will be winners. The assets sold by the losers should be bought by the winners at near their market values, although there will be some decline in the value o f securities other than Treasury securities from before the start o f the initial run. Riskier bor rowers are more likely to be rationed out o f the market. The flights to quality that have occurred after the failure o f the Franklin N ational, C o n tinental Illinois National Bank, and other large depository institutions in recent years had no noticeable effects on aggregate economic activity. The potential magnitude o f depositor losses may be estimated by examining depositor losses from bank failures before the introduction o f the F D IC . From 1865 through 1933, commercial banks were estimated by the F D IC to have experienced losses o f $12.3 billion.9 O f this am ount, $7.7 billion (62 percent) was charged against net worth, $2.5 billion (20 percent) was borne by shareholders, including $0.5 in assessments on national bank share holders under provisions for “ double-liability” , and $2.2 billion (18 percent) was borne by depositors. A s a percent o f total deposits, total depositor losses were 0.21 percent. But $1.3 billion, or 60 percent, o f the $2.2 billion losses occurred between 1930 and 1933. This reduced depositor losses to only 0.08 percent in the 65 years from 1865 to 1929. Even in the 1920s, when, on average, more than 500 banks failed per year, depositor losses accounted for only 0 .14 percent o f total bank deposits. Depositor losses between 1930 and 1933 averaged 0.81 percent o f total deposits, close to the average 0.78 percent loss in 12 years classified by the F D IC as “ crises” years between 1865 and 1940. In the 64 “ noncrisis” years, depositor losses averaged only 0.07 percent. These losses appear small in comparison to losses by bond holders from price movements or defaults, even on short term bonds. Thus, bank failures, particularly when not accompanied by a currency outflow, may not have been as serious as popularly believed. So far, we have abstracted from the thorny question o f how many large banks can be or be perceived to be in trouble at the same time without ig niting a currency run or, more likely, a very severe flight to quality. The public may perceive concurrent problems at a number o f banks if the banks have similar loans, for example, loans to foreign countries or, less likely, to major firms experiencing severe financial difficulties.10 A currency drain can FRB CHICAGO Working Paper 5 be envisioned if the public doubts the ability o f the federal insurance funds to pay o ff fully insured deposits at failed banks. W hen this is likely to happen is uncertain because the public is unlikely to understand that the cost to the insurer is only the difference between the insured deposit value and the market value o f the bank’ s assets, which for large, nearly solvent banks should be quite small; e.g., 10 cents or less on the dollar. The cost is the total uninsured size o f the bank only if the bank is totaled, a highly unlikely event. M oreover, some part o f any loss is charged against the bank’ s remaining net worth. Thus, the public may envision exhaustion o f the fund well before there is any serious threat. I f the public acts on this, the collapse will become a self-fulfilling prophecy. This threat could be defused m ost easily by changing the deposit insurance system to a deposit guaranty system backed by the full faith and credit o f the federal govern ment as Congress started to do with its Joint Resolution in 1982. C onfusion about the relationship o f the size o f the insurance fund and p o tential losses from bank failures is also prevalent among bank renovators and public policy officials. C . T. C onover stated in his testimony on the Continental crisis before the H ouse Banking Com mittee that: Sixty-six banks ... had deposits in Continental in amounts in excess of the total net worth of the bank. Another 113 banks had deposits in Conti nental amounting to between 50 and 100 percent of their net worth. If Continental had failed and been treated as a payoff, certainly those 66 banks would have failed and probably a goodiy number of the other 113 would have failed.11 Deposits include Fed funds sold to the Continental. The W orking G roup o f the President’s Cabinet Council on Econom ic Affairs concluded in its recent report that the present size o f the federal deposit insurance funds were too small and: it seems rational that the funds should be able to handle at least ongoing operations and the potential failure o f one or tw'o of the largest depository institutions. Yet, the largest depository institutions’ deposits are many times the size of the funds, giving rise to concern about the funds’ ade quacy. For example, as of year-end 1983, 6 FSLIC-insured and 13 FDIC-insured institutions’ deposits each exceeded the size of their respec tive fund ... Increasing the size of the funds should increase the perception o f stability compared with the status quo.12 These interpretations are not only incorrect but are dangerous as they re inforce the very public concerns and anxieties o f self-feeding cumulative bank failures that these officials are trying to prevent. Although an exact accounting cannot be made without access to the books o f the bank, it appears doubtful that, excluding Latin American loans, the market value o f Continental’ s overall asset portfolio was much less than 90 cents on the FRB CHICAGO Working Paper 6 dollar. Som e o f this loss would be charged against Continental’s remaining net worth. A study by the staff o f the H ouse Com m ittee on Banking, F i nance and Urban Affairs concluded that if a 90 percent recovery rate is assumed, no banks would have suffered losses in excess o f their net worth and only two banks would have suffered a loss o f between 50 and 100 percent o f their net worth amounting to a total loss o f $1 m illion.13 The F D I C insurance funds were very adequate to absorb F D I C losses o f this magnitude to insured depositors. O n the other hand, the F S L IC fund may not be large enough to absorb insured depositor losses at S L A s if assets and liabilities at these institutions were marked to market and the negative net worth charged to the fund. Y et, the public has not panicked because o f their faith in the Treasury’s ultimate guarantee. But without formal legis lation this faith m ay on occasion be shaken and concern over the size o f the funds could ignite net currency runs. W ith a large number o f large banks under suspicion, corporate treasurers would experience significant difficulties in selecting banks above suspicion that could also satisfy their needs in terms o f service quality. They may search out strong regional banks to supplement the remaining strong m oney center banks, but the number suitable is likely to be relatively small. They are also likely to invest more heavily in Treasury securities. H ow well could the private banking sector deal with the recycling o f deposits? T o the extent that the deposit-losing institutions have sufficient marketable in vestments, there should be little trouble. I f the deposit-losing banks have to sell loans, larger losses from hurried sales m ay be expected and the problem becomes more severe. The m ore tailored the loan, the m ore severe the problem. These are not easily nor cheaply sold to other banks even in the same market area. It is possible that without remedial action a liquidity problem could develop into a solvency problem for some banks. In addi tion, banks build up expertise in servicing loan accounts that their custom ers expect and pay for and, in the process, acquire inform ation about the financial condition o f the borrower. The breaking o f bank-custom er credit relationships is likely to have substantial undesirable implications for eco nom ic activity even in the absence o f a contraction in the m oney supply.14 O f course, as long as the deposit-losing banks are perceived to be solvent—that is, are experiencing liquidity and not solvency problems—the deposit-receiving banks could redeposit the funds back into those banks through the Fed funds market, possibly at somewhat higher interest rates, rather than purchase their loans and investments. Unlike in the 1930s, such lending could be supplemented by the central bank making funds available through the discount window. The Fed would offset the effect o f the in crease on total reserves through open market sales. There should be no net effect on m onetary policy. FRB CHICAGO Working Paper 7 Thus, it would appear that even with doubts about a number o f large banks, deposit losses need not result either in spillover effects to other banks nor in further financial problems for the banks in question as long as they are perceived to be solvent.15 The banking system, with the support o f the central bank, should be able to recycle the funds or assets. I f loans have to be sold, more disruption may be expected, but it should not be o f crisis proportions. N o r is it likely that with appropriate action by the other banks or the central bank a run would transform a solvent bank (in terms o f market value) into an insolvent bank. Truly, “ unfounded" rumors that ignite runs are likely to be self-defeating as the recycling occurs. For the solvent bank, asset and liability management and core and purchased funding appear to carry equivalent risks. The ability to fund with pur chased funds depends on the perceived quality o f the bank’ s assets. It has little to do with any characteristic o f “ h o t" money other than that such money takes risks into account in evaluating returns. Liability funding is an asset quality problem! But, what if the original bank, say, the Continental, was operating but was neither solvent nor widely perceived to be solvent? It could continue to operate if and as long as the federal government guaranteed all o f its de posit liabilities. But recycling would be more difficult. Losses after the depletion o f the market value o f a bank’ s capital are borne by the federal deposit insurance agency. Maintaining a facade o f solvency is at least as likely to produce further losses as gains.16 Thus, there appears to be no reason for not declaring the institution insolvent, regardless o f its size, and turning it over to its chief creditor—the F D IC —for further disposition. This may be viewed as not vastly different from bankruptcy court control o f insolvent nonflnancial firms. Indeed, because the public may view the fi nancial position o f the bank as strengthened, this strategy may have fewer unfavorable ramifications for the banking system as a whole than contin uation o f the insolvent institution. (Unlike bankruptcy for nonfinancial firms, legal insolvency for banks is not a clear-cut event, and its timing is subject to considerable discretion on the part o f the chartering agency.) N o r is it likely to be viewed by the public as “ nationalization" in the narrow sense. The largest creditor just happens to be an insurance com pany owned by the federal government. (M y reading is that the public did not appear to view the “ phoenix” SL A s established by the F S L IC as nationalized in stitutions.) A s long as there are doubts about a bank’s solvency, the recy cling process engaged in by other private banks, discussed earlier, breaks down. Only the Fed, other government agencies, or other creditors who could secure their loans (deposits) would risk such recycling. The regula tory agencies’ delays in dealing decisively with the Continental, Seattle First N ational, Penn Square and others appear to have created larger rather than smaller losses borne by old creditors. M an a gement generally did not change their strategies materially and, on occasion, particularly at thrift institutions, actively increased their risk exposure in the hope o f recouping FRB CHICAGO Working Paper 8 their losses in one shot before it was too late. rector o f Research at the F D I C , has noted: A s Stanley Silverberg, D i If a bank becomes insolvent but remains liquid and open, it is generally in the interest o f managers and owners to gamble in an effort to recoup, es pecially if that can be done legally. If the bank rolls the dice and loses, the FDIC typically bears the loss. Keeping an insolvent bank open, further more, allows uninsured depositors to flee. By the time a bank is closed, the FDIC is the only creditor. It thus becomes academic whether deposits are paid off or a P&A (purchase and assumption) is effected.17 Once a bank is perceived, correctly or incorrectly, to be insolvent in an environment o f less than 100 percent deposit insurance, the choice o f asset or liability management and core or purchase funding becomes critical. A ll deposits in excess o f $100,000 not otherwise comm itted will leave as quickly and quietly as possible (demand deposits immediately and time deposits at the earliest permissible data o f withdrawal). I f the perception is incorrect according to market values, then those in the know may be able to correct the situation. The bank regulators, or even the bank itself, could invite other banks to review the books and draw their own conclusions. Alternatively, the regulatory agency could inform the Fed o f the bank’s true solvent condition and request the Fed to accom modate fully the bank’s re quest for funds at the discount window. Such action should change the market’ s perception. However, for these strategies to be effective, it is necessary that both the banfs’ books and the regulators’ evaluations be in terms o f market values and that disclosure be as complete as possible to reduce the likelihood o f later, larger, unpleasant surprises.18 W h at are the implications o f one or more large banks being declared in solvent and temporarily operated by the F D IC in trusteeship while awaiting liquidation, merger, or reestablishment as an independent entity? Liqui dation is obviously not an efficient solution, nor is it generally practiced for larger nonfinancial bankrupt firms. Large banks are no longer, if they ever were, physically closed and boarded up. It is not that large banks are too large to fail, but that they are too large to liquidate or to merge or sell im mediately. Thus, continued operation by the F D IC is the m ost likely tem porary solution. It would appear reasonable that the F D IC would not elect to involve itself in day-to-day operations, but would limit its role to over sight and guidance. Unless the reasons for a bank’ s insolvency are totally outside the bank’ s control, it stands to reason that the F D IC would prefer to make changes in senior management and possibly on the board o f di rectors. The pain o f failure would be spread beyond shareholders and large creditors. The transfer o f the insolvent bank to F D I C trusteeship should be completed as quickly as possible for two reasons. One, the quicker the transfer, the FRB CHICAGO Working Paper 9 smaller the losses o f the F D IC . Indeed, if the bank were declared insolvent as soon as the true market value o f its net worth came to zero, there would be no losses whatsoever to any depositor or creditor and, therefore, also no losses to the F D I C .19 T w o, any interruption in banking services either on the loan or deposit side is disruptive and costly to the community. I f the chartering regulatory agency declared the institution insolvent and the F D I C was able to intervene as soon as the market value o f net worth be came zero but before it turned negative, no losses would accrue to uninsured depositors or creditors and the institution would effectively re open the next day without any changes on the balance sheet. If, however, the bank were declared insolvent after its true net worth be came negative, losses accrue to uninsured depositors and creditors equal to the differences between the market value o f the bank’ s assets and its liabil ities. A t the transfer o f ownership, say, at the close o f business o f the day insolvency is declared by the chartering regulatory agency, the value o f these deposits would be written down by the prorata amount, say, five cents on the dollar. The bank would reopen the next morning with uninsured liabilities at their new, lower values. The bank would be solvent and in business. M o st bank customers would be almost totally unaffected by the transfer. Bank relations would be uninterrupted and continue as they were.20 Uninsured depositors would have full and complete access to the current value o f their deposits. Nothing is frozen. This transaction may be referred to as a “ modified trusteeship” analogous to the “ modified payout” arrangement used by the F D IC in some financially assisted pur chase and assumptions in which uninsured depositors are guaranteed only a percentage o f their deposit balances based on estimated equation values. Three problems may be raised with such an arrangement. One, it is difficult to value all asset and liability accounts at market to determine the accurate value o f net worth. This problem, however, confronts all firms, nonfinancial as well as financial, although it may be more difficult for financial firms. Nevertheless, it generally is solved. M any bank asset and liability accounts have reasonable marketable counterparts and most that do not can have their market values estimated within a reasonable margin o f error with some creativity. M ore creativity may be required for some loans, such as dollar loans to foreign governments. But all mutual funds, even junk bond funds, have to mark to market daily. Errors are bound to occur, but it is unlikely that the estimated number value will be further from the true market value than is book value. The faster banks move to market value accounting, the more accurate will recorded balance sheet values be at any time in the future. It should be noted that the problem entails not only market value accounting but also appropriate monitoring to obtain timely and current observations. M o st bank failures, particularly for smaller banks, result from fraud or theft which are both difficult to detect and change the values o f accounts quickly. FRB CHICAGO Working Paper 10 T w o, there may be legal challenges to the estimates o f the market value made and thus to the amounts by which the noninsured liabilities are marked down. This is a major reason for declaring the bank insolvent as soon as possible so that losses to uninsured creditors are nonexistent or as small as possible. But such legal challenges are normal occurrences in all failing bank arrangements in which uninsured creditors suffer losses, e.g., F D I C assisted “ haircut” mergers. Three, does the F D I C have the legal authority to undertake such an ar rangement? A lthough I am not a lawyer, the language o f Section 13 (c) appears to me sufficiently broad to support such an action. In addition, the F D I C has frequently been creative and imaginative in its interpretation o f this Section, e.g., in its handling o f the First Pennsylvania (1980) and the Bank o f the C om m onw ealth (1 972).21 Even the Continental Bank arrange ment could be considered innovative and imaginative. These arrangements have, on the whole, withstood numerous court challenges and legislative inquiries. M oreover, Sections 11(h) through (1) authorize the F D I C to “ organize a new national bank to assume the insured deposits o f such closed banks and otherwise to perform temporarily the functions herein provided for” for a m axim um o f two years.22 The permissible powers o f such a bank appear to be under the control o f the Com ptroller o f the Currency. Because the federal government’s credibility is on the line, it appears rea sonable that the value o f uninsured deposits and all new deposits received after the takeover will not decline. This guaranty m ay give such banks a competitive advantage over solvent banks. T o minimize the implications o f such an advantage, it m ay be useful to constrain the future deposit size o f the bank to a level no greater than on the day o f the takeover, to a growth rate no greater than, say, average for banks in the same geograph ical area, or by some similar rule. (Som e such rule was apparently imposed on the Continental Bank.) It is also understood that F D I C regency is only temporary until alternative arrangements can be completed under less hec tic conditions. Insolvency appears to have a number o f other potential advantages relative to the arrangement used at the Continental; it elimi nates the old com m on stock and does not establish a low -cost speculative vehicle; it prevents bank management from plunging and betting the bank in the form o f the F D I C ’ s and the taxpayers’ monies; it separates support o f the bank from support o f the holding com pany; it m ay permit abrogation o f “ excessive” severance or guaranteed future em ployment contracts; and so on. (Indeed, one may reasonably speculate whether con siderations such as these were not in the minds o f those involved in the Continental negotiations, particularly o f those representing Continental management. A lso , in the case o f the Continental, it would have avoided the need for the circus-like political lobbying that occurred to change Illinois state law to permit out-of-state acquisition o f the Continental. FRB CHICAGO Working Paper 11 G a m -S t. Germain permits such acquisitions for insolvent banks.23 U n doubtedly, there are disadvantages also, and these need to be carefully de fined and compared to the advantages. Like all difficulties, it is better to prevent bank difficulties before they start. After the start, no solution is perfect for all parties and all objectives. Given both the fragility and importance o f banking, it might be wiser in the long-run for regulators to err on the side o f caution.24 Excessive caution is likely to encourage entry and/or evasion that should, through time, produce a second-best solution. Insufficient regulatory caution, on the other hand, may produce serious harm that is not as easily offset by market forces. This is particularly true for forces that affect a number o f large banks, such as the sanction o f heavy concentrations o f loans to risky foreign countries. The private banking system is not an efficient tool for pursuing foreign macroeconomic or political policy. I f problems occur, they are likely to be relatively severe and perceived to be even more so. In such an environ ment, bank regulators and policymakers are likely to act quickly according to gut feeling and intuition, rather than to the logic o f the situation and long-run implications o f the action. Additional regulation is likely to be the result, regardless o f whether regulation, excessive or otherwise, was a con tributing factor to the crisis.25 Such a response to the crisis o f the early 1930s has been, in significant measure, responsible for many o f banking’ s recent ills. Conclusions This paper has argued that, while the existence o f less than 100 percent federal deposit insurance still leaves large bank difficulties more important than small bank failures, it has dramatically changed the consequences o f such difficulties. P o st-F D IC , bank runs no longer take the form o f cur rency drains out o f the system which, p re-F D IC , had led to multiple con tractions in aggregate deposits (m oney) and credit. Rather, they consist o f the redeployment o f deposits to other, presumably less risky, banks o f similar characteristics. A run on a bank no longer translates into a run on the banking system and instability at one or a group o f banks does not translate into instability for the financial system as a whole. The adverse externalities o f bank failure are based on pre-F D IC days and appear to be significantly overestimated.26 If the deposit-losing banks are in actuality or in perception viewed as solvent, the banking system itself with possible support from the central bank should be able to recycle the funds back to the original banks. This is likely to be accompanied, however, by a wid ening o f interest rate spreads between high and low quality credits, reflect ing a flight to quality. The magnitude o f the flight depends on the number o f deposit-losing banks. W hile serious, the problem o f widening spreads is FRB CHICAGO Working Paper 12 less serious than that o f cumulative deposit contraction and requires dif ferent solutions. There appears to be little need for a tightly held safety net that catches all large banks regardless o f financial condition or pros pects. Likewise, the destabilizing aspects o f partial de facto as well as de jure deposit insurance on aggregate econom ic activity m ay not outweigh the stabilizing market discipline aspects. Disclosure plays an important role in maintaining an efficient private banking system. I f the public’ s perception o f one or m ore banks’ solvency were incorrect, then the error can be corrected relatively easily by granting public access to the banks’ books. should start. Then, the previous solution process If, however, the banks are in fact insolvent, then there may be advantages to having them declared so quickly and having the F D IC , in its role o f m ajor creditor, install new senior management, provide over sight, search for lasting solutions that may involve either merger or the banks’ maintenance as independent entities, and effectively guarantee all deposit values from that date until a permanent solution is achieved. This option needs to be explored more thoroughly, particularly for dealing with a larger number o f troubled banks. W h a t is clear is that, contrary to regulatory folklore, failure to disclose ac curately the financial situation o f a large bank is more likely to create un certainty and shifting o f funds than certainty. The old adage about accountants being willing to let two plus two sum to anything the client wants it to sum is probably true but such playing with numbers is unlikely to fool people for very long. Certainly, in the case o f the Continental, the alleged “ unfounded” rumors that were widely thought to have ignited the initial well-publicized deposit run turned out to be rather well-founded. (It is also likely that some o f the deposit withdrawals were motivated by the F D I C ’ s increasing use o f “ haircuts” on large deposits at failed institutions de facto as de jure uninsured.) This is not to argue that there are no technical and depositor fears that their deposits at large banks might be well as nor strategic problems with full disclosure whatsoever, but that they may be less serious than the problems associated with no or partial disclosure.27 But a society that safely sends persons into space should be able to estimate the market value o f almost all assets and liabilities reasonably accurately. A s soon as the public correctly believes one or more large banks to be in solvent, it appears preferable to declare the bank such and transfer its o p eration to the F D I C (or a new federal government agency organized for this purpose). This should prop up the bank’s credibility and buy valuable time to search for a more lasting solution. It also sends a valuable signal to the market that, at m inim um , bank shareholders will be held at risk and should increase their precautionary m onitoring to reduce their risk o f loss. FRB CHICAGO Working Paper 13 A lth ou gh not the subject o f this paper, the role o f less-than-full deposit in surance without risk sensitive insurance premiums needs urgent thought. It is this premium structure, as Professor K an e has convincingly argued, that is largely responsible for the high risk exposures assumed by depository institutions and possibly also for the alm ost sloppy monitoring o f risk.28 The risk monitoring systems o f both the institutions and regulators are in need o f careful reassessment and modernization. W h a t is certain is that considerably more thought should be given to developing an optimal sol ution to large, m oney center bank difficulties so that an efficient blueprint will be on the shelf when the next large bank(s) experiences m ajor financial problems. The longer the delay, the harder it will be to implement efficient solutions. Each solution, whether ad hoc or well thought out, transmits signals to the market that affects the behavior o f the market, institutions, and depositors. Large bank failures are traumatic and serious events that require careful public policy attention. But little is gained by blowing the consequences o f such failures out o f proportion and sending misleading and potentially dange rous signals through the econom y. 1 C. T. Conover, “ Testimony” in U.S. Congress, Subcommittee on Financial In stitutions Supervision, Renovation and Insurance o f Committee on Banking, Fi nance and Urban Affairs, Inquiry into Continental Illinois Corp, and Continental Illinois National Bank: Hearings, 98th Cong., 2nd Sess., p. 287-288. George Kaufman, Larry R. Mote, and Harvey Rosenblum, “ Consequences o f Deregulation for Commercial Banking,” Journal o f Finance (July 1984) and “ Consequences o f Deregulation for Commercial Banks,” Staff Memoranda, 84-5 ^Federal Reserve Bank o f Chicago), May 1984. Paul M. Horvitz, “ Failures o f Large Banks: Implications for Banking Super vision and Deposit Insurance,” Journal o f Financial and Quantitative Analysis (November 1975); and Thomas Mayer, “ Should Large Banks Be Allowed to Fail?” , Journal o f Financial and Quantitative Analysis (November 1975). 4 A review o f this literature appears in G. O. Bierwag and George G. Kaufman, “ A Proposal for Federal Deposit Insurance with Risk Sensitive Premiums,” Staff Memorandum, 83-3 (Federal Reserve Bank o f Chicago), 1983. Milton Friedman and Anna J. Schwartz, A Monetary History of the United States, 1867-1960, (Princeton; Princeton University Press), 1963; and Bray Hammond, Banks and Politics in America: From the Revolution to the Civil, War, (Princeton; Princeton University Press), 1957. 6 At least one large bank depositor, Henry Ford, threatened in 1907 to “ build a vault to take our money out of the banks and put it in the vault, so we can pay out men in cash.” Quoted in Susan E. Kennedy, The Banking Crisis o f 1933, (Lexington; University Press o f Kentucky), 1973, p. 92. No evidence o f contagious effect o f the failure o f the Franklin National Bank (1974), The United States National Bank o f San Diego (1973), and the Hamilton National Bank o f Chattanooga (1976) on the stock value o f other banks is re ported by Joseph Aharony and Iszhak Swary, “Condition Effects o f Bank Fail ures: Evidence from Capital Markets,” Journal o f Business, July 1983, pp. t KB CHICAGO Working Paper 14 305-322. Contagiousness has been observed, however, where failed thrift insti tutions were insured by state rather than federal agencies, e.g., Nebraska (1984) and Ohio (1985)," and depositors realized that, unlike the federal government, states do not own the printing presses. 8 Moreover, it is reasonable to assume that, unlike in the 1930s, the Federal Re serve would replace any aggregate flight to currency that might occur to maintain the money supply. 9 Federal Deposit Insurance Corporation, Annual Report, 1940 (Washington, D .C .), pp. 61-73. The data include recoveries but no accounting is made either for injury before recovery nor the length o f time that the deposit accounts may have been frozen before payoff. Thus, these figures understate the severity o f depositor and loan customer injury. 10 For savings and loan associations, the most likely common cause o f financial difficulties is unexpected increases in interest rates. On a market value accounting basis, almost all SLAs have been insolvent throughout much o f the 1980s. 11 Conover, “ Testimony” , pp. 287-288. 12 The Working Group o f the Cabinet Council on Economic Affairs, Recom mendations for Change in the Federal Deposit Insurance System, (Washington, D .C .), January 1985, pp. iv and 53. 13 Continental Illinois National Bank Failure and Its Potential Impact on Corre spondent Banks, Staff Report To Subcommittee on Financial Institutions Super vision, Regulation and Insurance, Committee on Banking, Finance and Urban Affairs, October 4, 1984, pp. 16-18. 14 Ben S. Bernanke, “ Nonmonetary Effects o f the Financial Crisis in the Propa gation o f the Great Depression,” American Economic Review (June 1983). 5 Because almost all accounts are fully insured, most SLAs suffered neither cur rency or deposit runs when they experienced solvency problems in the late 1970s. 16 In his testimony before the House Banking Committee on October 4, 1984, William Isaac, Chairman o f the FDIC, stated that the “ Continental was not and is not insolvent in the sense o f its liabilities exceeding its assets... the bank had severe ... liquidity problems.” If this were so, one may wonder why he implied a less than 100 percent recovery rate for interbank deposits at the Continental or why an assistance program was, in fact, necessary. William M . Isaac, “ Statement on Federal Assistance to Continental Illinois Corporation and Continental Illinois National Bank” , Subcommittee on Financial Institutions Supervision, Regulation and Insurance; Committee on Banking, Finance, and Urban Affairs; U.S. House o f Representatives, October 4, 1984, pp. 2-3. 17 Stanley C. Silverberg, “ Restricting Brokered Deposits,” Housing Finance Review, January 1984, p. 100. Term in parenthesis mine. 18 It is apparent that Isaac’s reaffirmation o f the solvency o f the Continental (see footnote 15) was not shared by the market. This paper will not deal with the substantial problem o f marking all accounts, such as nonmarketable assets and, particularly, deposits, to market. Correct procedures involve marking both sides o f the balance sheet to market. Many ongoing bank accounting practices are, at minimum, misleading and, at maximum, economic fraud. For example, banks may accrue unpaid interest on loans past due until the loans are declared non performing. Only recently did the bank regulators define nonaccruing as no more than 90 days past due. However, these loans may continue to be valued at book. An excellent analysis o f such accounting practices for LD C loans on bank income appears in Suzanna Andrews, “ Accounting for LDC Debt,” Institutional Investor, August 1984, pp. 189-194. FRB CHICAGO Working Paper 15 19 G. O. Bierwag and George G . Kaufman, “ A Proposal for Federal Deposit In surance with Risk Sensitive Premiums” in Bank Structure and Competition: Con ference Proceedings, Federal Reserve Bank o f Chicago, May 1983, pp. 223-242. Accurate market values incorporate all expected further losses. 20 This appears to have been the case with FSLIC’ s takeover o f the SLAs turned into phoenixes. 21 Federal Deposit Insurance Corporation, The First Fifty Years, Washington, D .C .; 1984, pp. 81-108. 22 Federal Deposit Insurance Act, 1983. 23 For a good discussion o f some o f the F D IC ’s concerns in the Continental saga by a high F D IC official directly involved, see Stanley C. Silverberg, “ Resolving Large Bank Problems and Failures,” paper presented at Symposium on Issues and Options in Dealing with Large Banks’ Problems and Failures, August 20, 1984. 24 This may also be true for economists. A review o f the academic literature in banking over the past two decades reveals that economists both in and outside academe tended, on the whole, to believe that bankers were overly risk averse, bank capital was too high, and the F D IC ’s and FSLIC’s reserves were higher than necessary. 25 It is also inappropriate for a regulatory agency to justify its powers on the basis o f exaggerated public fear o f perceived spillover crisis. In a paper prepared for the Bush Commission, Fed Chairman Volcker argued that: A crisis in one limited part o f the banking system can quickly affect the strength and well-being o f other parts and the system as a whole .... In our view, it would not be workable or reasonable—indeed, it would be dangerous—to look to the Federal Reserve to “ pick up the pieces” in a fi nancial crisis without also providing the Federal Reserve with the tools to do the job and with adequate “leverage” in shaping the system so as to reduce the likelihood o f a crisis arising. Paul A . Volcker, “ The Federal Reserve Position on Restructuring o f Financial Regulation Responsibilities,” Federal Reserve Bulletin (July 1984), pp. 548-549. 26 I have argued elsewhere that most o f our fears and concerns about system-wide contagious bank failures and domino effects are based on the experiences o f the massive bank failures in 1931-33 rather than on the broad sweep o f U.S. banking history which is far less chilling. George Kaufman “ Consequences o f the Failure o f One or More Banks on Other banks, the Stability o f Financial Markets, and Regional and National Economic Activity,” Draft chapter for a study on Bank Safety and Soundness prepared for the American Bankers Association, March 1985. 27 It is more than unfortunate when a bank regulatory agency acquiesces in a practice that ties the compensation o f bank officers to changes in the book value o f the bank’ s net worth rather than in the market value, as the FD IC has done for the Continental. Such a system can only create incentives for the bank’s ac countants to use techniques to increase the bank’s book net worth, such as de ferring loan loss reserves (a practice that contributed to the severity o f the Continental’s crisis), regardless o f the true (market value) situation. A description o f this compensation arrangement appears in Jeff Bailey, “ Continental Illinois’ Auditor Revises Its Opinion, Qualifies Results for 1983,” The Wall Street Journal, August 28, 1984, p. 4. FRB CHICAGO Working Paper 76 IX See, for example, Edward J. Kane, “The Role o f Government in the Thrift Industry’s Net-Worth Crisis,” in George J. Benston, ed., Financial Services: The Changing Institutions and Government Policy (New York; Prentice-Hall), 1983, a n d :__________________ , The Gathering Crisis in D eposit Insurance (Cambridge, M a.; M IT Press), Forthcoming. FRB CHICAGO W orking Paper 17