Policy Responses: Benign Neglect, Exhortation, and Bank Holidays
If many financial crises have a stylized form, should there be a standard policy response? Assume plethora, speculation, panic; what then? Should the government authorities intervene to cope with a crisis and if so when? Should they seek to forestall increases in real estate prices and stock prices as the bubble expands so the subsequent crash will be less severe? Should they prick the bubble once it is evident that asset prices are so high that it is extremely unlikely that increases in rents and in corporate earnings could be sufficiently rapid and large to ‘ratify’ these lofty prices? When asset prices begin to fall, should the authorities adopt measures to dampen the decline and ameliorate the consequences?
Virtually every country has established a central bank to prevent or minimize shortages of liquidity, especially during a financial crisis. Many countries have a deposit insurance arrangement to reduce the likelihood of runs on their banks and to forestall what otherwise could be a self-fulfilling prophecy that a shortage of liquidity would trigger a solvency crisis. Even when there is no formal insurance for bank deposits, individuals in many countries believe that their governments will ensure that they will not incur losses if the banks should fail – in effect, the banks will be nationalized and the government will become their owners.
Chairman Greenspan of the Federal Reserve initially took the view that the Fed should focus monetary policy on the achievement of a stable price level – or at least a low inflation rate – and the macro-employment objectives. Then as the US financial system imploded in the last few months of 2008, he seemed more open to the idea that the Fed should have paid more attention to increases in real estate prices.
This chapter centers on the management of financial crises – it first deals with regulatory measures that might be adopted to reduce the susceptibility of financial arrangements to manias and then with the approach to panics. The next chapter focuses on the domestic lender of last resort and the following one on the lender of last resort in an international context. This chapter considers the Austrian view that the best remedy for panic is to ‘leave it alone’ – to let it run its course, and to allow the economy to adjust to the decline in household wealth, the de-capitalization of the banks, and the slowdown in household and business spending that would follow from the declines in the prices of real estate, stocks, and commodities.
The moral hazard argument is the primary rationale for non-interference by the government; the view is that the more interventionist the authorities are in the current crisis, the more intense the next mania will be. The argument is that intervention skews the risk and reward trade-off in the minds of many investors – both the shareholders in banks and perhaps some of the creditors of the banks including those who own the bonds or other fixed price claims on the banks – by reducing both the likelihood and scope of future losses. A secondary rationale for this view is that if the adjustments to the imbalances created by the implosion of the bubble are delayed, the economic recovery will be slowed.
The moral hazard argument usually does not distinguish shareholders from bondholders. Shareholders in Lehman Brothers and Fannie Mae and Freddie Mac lost virtually all their money when these firms failed, as did the shareholders of the Icelandic banks. The shareholders of Northern Rock, the Royal Bank of Scotland, and Lloyds TSB lost most of their money – probably 90 to 95 percent. Similarly the shareholders of AIG and Bear Stearns lost more than 90 percent of their money. Moreover, a substantial part of the personal wealth of the employees of both Lehman and Bear was in the shares of these firms and in options to buy more shares, so some of these individuals lost 60 or 70 percent or more of their personal wealth.
At times the concern about moral hazard is expanded beyond the shareholders to include the creditors of the financial firms, including both the bondholders and those who have short-term claims on the banks; unlike the shareholders, these creditors have no significant upside gain if the banks’ strategies prove immensely profitable. The returns of these creditors are skewed – if the banks fail, these creditors will incur substantial losses and if the banks are successful, the creditors do not share in these gains. Creditors may be lulled into complacency about potential losses because of the view that the banks will not be allowed to fail – but the failure of the banks is not unambiguous. If the shareholders lose 90 or 95 percent of their money before the government provides financial assistance, has the bank failed? Most of the shareholders are unlikely to agree to the statement that they have been bailed out. These bondholders may receive somewhat higher interest rates because the managers of the banks believe that they are ‘too big to fail’ and undertake somewhat riskier investments – and they will be able to pay higher interest rates – at least for a while. The ‘too big to fail’ doctrine may have the same impact as deposit insurance – unless the authorities reduce the value of the claims of the creditors below their face value if the banks are closed and re-organized with government assistance. Thereafter the owners of the deposits that are much larger than the amounts covered by deposit insurance may be ready to withdraw their money whenever there is modest skepticism about the solvency of the bank.
The choice for each country is whether to design a system that would allow large institutions to fail, with losses to creditors, or instead to provide some form of effective insurance for all of those with claims on the banks to reduce the likelihood of runs.
Financial reform: bank regulation and supervision
Can financial crises be forestalled by strict regulation and supervision? Some observers advocate this approach. Others recommend deregulation. Most of the rules for sound banking are already incorporated in the regulations or are implicit in banking tradition. Many of the rules are ignored by both the banks and the regulators. Banks are supposed to ‘mark to market’, that is, to value their loans and investments each day (or each week or month) at the prices that would be realized if they were sold in the market rather than at their historic costs. Reserves should be established against ‘problem loans’ and write-offs against ‘doubtful’ ones. As a bank’s loan losses increase and its capital declines, the bank might be required to raise more capital or be closed under the traditional rules. As an illustration of the unusual character of banks following these rules, the press was full of the news in the spring of 1987 when Citicorp wrote down the value of its Third World loans and the FSLIC allowed 500 insolvent banks to remain open in the hope that they would become sufficiently profitable to rebuild their capital. As part of the regulation process, the Federal Reserve began to collaborate with other central banks in the Group of Ten countries to strengthen bank structures worldwide by applying the same risk-based capital requirements to banks headquartered in different countries.
Emphasis on capital requirements as a percentage of assets or bank liabilities led some banks to develop ‘off-balance sheet’ operations, which generate fees and commissions and interest income, but the assets and liabilities are contingent and are shown only as a footnote on the balance sheet. The banks established special investment vehicles (SIVs) that circumvent formal capital requirements; hence they increased their effective leverage. The demand for the IOUs of the SIVs depended on the reputational capital of the banks. These off-balance sheet transactions include interest rate and currency swaps, futures contracts, options, underwriting risks, ‘repos’ (sales of securities with a guarantee to repurchase them at a later date), and note-issuing facilities. Each of these positions can be valued as an option and included among assets or liabilities when calculating the appropriate required capital.1
A strong case can be made for stricter regulations and supervision of banks to forestall lending in euphoric periods that may end in financial crisis. Responsibility for bank examination in the United States is divided among the Comptroller of the Currency, the Federal Reserve Banks, and the state banking commissions. In one view there is competition not in deregulation but in re-regulation.2 ‘Divided responsibility,’ said a famous German banker-politician, ‘is no responsibility.’3 The astute examination personnel needed when asset prices tumble are unwilling to submit to the boredom of long periods of calm between crises. The mismanagement of banks is hard to detect before a crisis. In booms, entropy in regulation and supervision permits the build-up of danger spots that only burst after asset prices decline. The question then is whether to liquidate, stall, guarantee, bailout, takeover, or rely on other means of last-resort lending.
The almost universal response to failure or near failure of banks is that there is need for more regulation or for more effective regulation. The problems of Bear Stearns, Lehman Brothers, and other US investment banks have been attributed to the decision of the US Securities and Exchange Commission to relax the restraints on the leverage of the investment banks. At times the collapse of some of the banks is attributed to the repeal of the Glass-Steagall Act – which conveniently neglects the failures of hundreds of firms in the 1980s and the failures of banks in Britain, Iceland, and Ireland.
Regulation imposes costs. As a result the history of regulation is that new types of institutions are developed that are outside the scope of regulations – and yet they are otherwise similar to regulated firms. Money-market funds were developed as a way to pay interest on demand deposits. The offshore deposit market developed to avoid the costs that domestic banks incur in the form of reserve requirement and deposit insurance premiums; the offshore branches of US banks could pay higher interest rates than their domestic branches.
Manias are macro phenomena and result from the excessively rapid growth of credit. Regulation is a micro phenomenon. Regulation is not likely to affect whether a mania develops, although it will impact some institutions and it might impact the pace of the development of the mania.
Many economists believe that the panic will work its own cure, and that ‘the fire can be left to burn itself out’.4 ‘Cool if not very imaginative heads in the Bank [of England] parlour thought it in the nature of panics to exhaust themselves.’5 Lord Overstone maintained that support of the financial system in crisis is not really necessary because the resources of the system are so great that even in times of the utmost stringency those that offer a sufficiently high rate of interest could borrow a large amount of money.6 In 1847 an increase in the discount rate to 10 and 12 percent in London stopped the flow of gold to the United States; a small sloop was sent to overtake a ship that had already sailed for New York and got it to turn around and unload £100,000 in gold.7 Testifying before the 1865 French Enquête (inquiry) into monetary circulation, Baron James de Rothschild stated that increases in interest rates could be relied upon to reduce speculation in commodities and securities. He added: ‘If speculators could find unlimited credit, one can’t tell what crises would ensue.’8
The moral hazard dilemma is that policy measures undertaken to provide stability to the system during this crisis may encourage speculation during the next expansion by those who seek exceptionally high returns and who have become somewhat convinced that there is a strong likelihood that the government will adopt measures to prevent the economy from imploding – and so their losses on the downside will be limited. A ‘free lunch’ for the speculators today means that they are likely to be less prudent in the future. Hence the next several financial crises could be more severe. The moral hazard problem is a strong argument for non-intervention as a financial crisis develops, to reduce the likelihood and severity of future crises. Will the policymakers be able to devise approaches that penalize individual speculators while minimizing the adverse impacts of their imprudent behavior on the rest of the country? Even then the cost-benefit question is whether the benefit to the economy from not allowing the panic to run its course is worthwhile in terms of the undeserved reward to the speculators.
The losses incurred by shareholders of Bear Stearns, Fannie Mae and Freddie Mac, AIG, Citicorp, and Bank of America were immense. Moreover the leadership and directors of these institutions were replaced. The firms have continued in business, but for all practical purposes they failed. Individuals who invested on the assumption that these institutions were ‘too big to fail’ may have incurred large losses.
The view that a panic should be allowed to pursue its course has two elements. One element takes pleasure in the troubles – or enjoys schadenfreude – that the investors or speculators encounter as retribution for their excesses; this puritanical view welcomes hellfire as the just deserts for the excessively greedy. The other sees panic as a thunderstorm ‘in a mephitic and unhealthy tropical atmosphere’ that clears the air. ‘It purified the commercial and financial elements, and tended to restore vitality and health, alike conducive to regular trade, sound progress and permanent prosperity.’9 One powerful statement of this position was made by Herbert Hoover as he characterized – without approval – the view of Andrew Mellon:
The ‘leave-it-alone liquidationists’ headed by Secretary of the Treasury Mellon felt that government must keep its hands off and let the slump liquidate itself. Mr. Mellon had only one formula: ‘Liquidate labor, liquidate stocks, liquidate the farmers, liquidate real estate.’ He insisted that when the people get an inflationary brainstorm, the only way to get it out of their blood is to let it collapse. He held that even panic was not altogether a bad thing. He said: ‘It will purge the rottenness out of the system. High costs of living and high living will come down. People will work harder, live a moral life. Values will be adjusted, and enterprising people will pickup the wrecks from less competent people.’10
The neo-Austrian economic historian Murray Rothbard added: ‘While phrased somewhat luridly, this was the sound and proper course for the administration to follow.’11 The conservative historian Paul Johnson commented: ‘It was the only sensible advice Hoover received during his presidency.’12
The opposing view conceded that while it is desirable to purge the system of bubbles and manic investments there is the risk that if the government does not intervene to staunch the crisis than a deflationary panic would spread and wipe out sound investments by the non-speculators who would not be able to obtain the credit they need.
One feature of many liquidity crises is that interest rates seem extremely high for private borrowers, especially because they are expressed as annual percentage rates they are really premiums for liquidity for one, two, or at most a few days. The more fundamental question is whether money is available at these high interest rates or whether the credit market is cleared by non-price rationing. The evidence from many crises is that it is difficult and sometimes impossible to borrow in a panic and that the quoted interest rates are irrelevant because money is not available.
• After first Arend Joseph and then DeNeufville failed in 1763, and panic broke out on 22 July, a succinct, not very informative or convincing report commented: ‘Panic even on securities and on goods, no money was to be had.’13
• The statement in 1825 was: ‘A panic seized upon the public, such as had never been witnessed before: everybody begging for money – money – but money was hardly on any condition to be had. “It was not the character of the security,” observes the Times, “that was considered, but the impossibility of producing money at all”.’14
• The interrogation of Thomas Tooke before the Select Committee on the Commercial Crisis in 1847: Question 5421: ‘For several days, if not some weeks, the Bank of England was the only establishment that was discounting?’ Answer: ‘Yes.’ Question 5472: ‘The Governor of the Bank of England said he could not sell £1 million of stock [English government bonds] in the week after October 14, if there had been no letter. Do you think possible?’ Answer: ‘No, perfectly impossible, taking the word impossible to signify with the exception of such a reduction in price as could not be contemplated.’15
• The evidence of Mr Glyn in the same inquiry: Question: ‘Are you aware that it was the opinion of the Bank broker that a very large sum might have been sold without materially affecting the prices of Consols [English perpetual bonds]?’ Answer: ‘I was not aware that the Bank broker had stated that. I should say, from what I saw at the time, that a sale of a million or two million, which were the figures talked of, would have been almost impossible without knocking down the funds to such a price as would have created a further panic.’
• Mr Browne, MP, did not think such sales could have been effected, unless at a great sacrifice, adding that ‘if the panic had been equal to what we might suppose it might have been, under such circumstances, I doubt whether they could have been sold at all.’16
• Then in 1857 ‘At one stage during the crisis it was impossible to negotiate paper at all, the charge under the most favorable circumstances being 12 and 15 percent.’17
• A letter from Liverpool: ‘Bills of exchange of the first Quality in themselves, and to which this and other Banks were willing to add their Endorsement, were absolutely inconvertible into Cash, and it is my Belief that many Houses, who were not merely solvent but able to pay 40s and 60s to the Pound, must have stopped had not the Government letter been issued.’18
• ‘Commercial confidence in Hamburg is entirely at an end. Bills of only three or four of the first houses are negotiable at the highest rate of interest ... A government bond advance of 15 million marks banco payable by banks failed to help. The panic was so great that government bonds could not be discounted, and on no security whatsoever would capitalists part with their money ... When it was known on December 12 that assistance would help all, the panic ceased. Government bonds which had not been discountable at 15 percent on the first of the month were readily taken at 2 and 3 percent.’19
• Edward Clark in New York wrote to Jay Cooke in Philadelphia before Cooke left Clark’s firm to start his own firm: ‘Money is not tight – it is not to be had at all. There is no money, no confidence & value to anything. A week more of such times and the Bks [sic] will fail.’20
• Then in 1866 ‘The Bank court raised the discount rate to 9 percent and intimated that loans on Government securities were available at 10 percent. Before that announcement it was impossible to sell either Consols or Exchequer bills. Jobbers in other securities refused to deal.’21
• During the 1873 crisis in New York the ‘National Trust Company of New York had eight hundred thousand dollars worth of government securities in its vaults, but not a dollar could be borrowed on them; and it suspended.’22
• And finally, in 1883, ‘The growing demand for money finally led to a money famine. Time loans were unobtainable, call loans were 72 percent in June, 72 percent on July 28th, 51 percent on August 4. First-class commercial paper was quoted 8 to 12 percent nominal, with a very small amount of money available.’23
The evidence is not unambiguous, since the sale of government bonds is somewhat qualified by such remarks as ‘with the exception of such a reduction in price [i.e., increase in the rate of interest] as could not be contemplated’. Moreover, there is occasional information on the other side of the argument, especially in the United States under the national banking system, in which a lender of last resort was unavailable. In 1884:
To add further to the discomfiture of dealers, money became exceedingly stringent, and at one time commanded as much as 4 percent for 24 hours’ use. This caused a further sacrifice of stocks since few could afford to pay the high rate asked. The exorbitant charge was, of course, the direct result of the distrust prevailing, since there was no actual scarcity ... It was to ... the desire to realize and obtain cash that the large decline on Thursday and Friday of nearly 7 percent on United States Government bonds is to be attributed. There was no loss in confidence in these, nor was there in good railroad bonds and stocks.
One result of the phenomenal and temporary rise in rates for money was to bring a vast amount of foreign capital to the market. Some of it was sent here to buy stocks at their depressed prices, and more to loan on stocks or on any other good securities at the high rates of interest. The effect of this was to completely turn the foreign exchanges which had been running so heavily against the US.24
This statement in its turn is not unambiguous, since the panic started before an acute liquidity shortage had developed.
The International Monetary Fund acted as a lender of last resort in the Asian crisis of 1997 – although long after the Thai baht, the Indonesian rupiah, and other currencies had depreciated sharply – and insisted that the government in each of the Asian countries balance its budget and that the central bank in these countries increase interest rates. A number of economists objected to these measures because they would lead to an increase in unemployment, especially among the poor, while the financial problems had been set in motion by comfortable officials and well-to-do bankers.
Some support for orthodoxy, however, came from the Japanese experience in the 1990s when the combination of an expansive monetary policy and the depreciation of the yen produced a ‘liquidity trap’. Both Japanese interest rates and bank loans decreased after the declines in stock prices and real estate prices, and the inference was that there was a ‘credit crunch’. Banks were reluctant to lend because their loan losses had eroded their capital and firms were reluctant to borrow because of the sluggish growth in the demand for their products.
The decline in short-term interest rates in Tokyo to 1 percent and below led to a surge in the ‘carry-trade’; US hedge funds borrowed yen, which they sold to buy US dollars which were used to buy US dollar securities that yielded 3 or 4 percent. The ‘carry-trade’ transactions led to an increase in the flow of funds from Tokyo to New York and to a decline in the value for the yen that in turn led to an increase in the Japanese trade surplus and increases in Japanese output and employment. The increase in the Japanese trade surplus was a useful response to the liquidity trap and an effective supplement to expansive Japanese fiscal and monetary policies. The carry-trade transactions would remain profitable for the US hedge funds as long as any appreciation of the yen was smaller than the excess of the interest rates on US dollar securities over the interest rates they were paying on their yen loans.
Moral suasion and other exhortatory devices
The dominant argument against the a priori view that panics can be cured by being left alone is that they almost never are left alone. The authorities feel compelled to intervene to forestall a further decline in asset prices. In panic after panic, crash after crash, crisis after crisis, the authorities or some ‘responsible citizens’ try to halt the panic by one measure or another. The authorities may be unduly alarmed and the position might correct itself without serious harm. The authorities may be stupid and unable to learn. (The Chicago School assumes that the market participants are always more intelligent than the authorities, in large part because the authorities are motivated by short-term political objectives.) The uneven distribution of intelligence cannot be tested against crisis management because the authorities and leading figures in the marketplace both exert themselves to halt the spread of falling prices, bankruptcy, and bank failures. If there is a learning process at work – and the assumption of rationality requires one – the lesson has been that a lender of last resort is more desirable and less costly than relying exclusively on the competitive forces of the market.
One insight from the historical record is that there are many examples when the authorities initially were resolved not to intervene but eventually did so reluctantly. Lord Liverpool threatened to resign as Chancellor of the Exchequer in December 1825 if an issue of Exchequer bills was provided to rescue the market after he had warned against excessive speculation six months earlier.25 William Lidderdale, Governor of the Bank of England at the time of the Baring crisis, refused categorically to accept a ‘letter of indemnity’ to permit the Bank to exceed its lending limit.26 On both occasions face was saved by finding some other approach to avert the panic. The initial strong stand not to intervene was reversed on many other occasions as the panic escalated. These included the intervention of Frederick II in the Berlin crisis of 1763,27 the Bank of England’s refusal to discount for the ‘W banks’,28 and the US Treasury’s decision in 1869.29
The view of the US authorities in September 2008 was that Lehman Brothers should be allowed to fail, even though the government had been centrally involved in JPMorganChase’s purchase of Bear Stearns and the conservatorship of Fannie Mae and Freddie Mac. One explanation is that the then Secretary of the Treasury did not want to become known as ‘Mr. Bailout’. A day after Lehman closed, the ‘policy’ was reversed in response to the massive run on AIG. In effect a new policy of 100 percent deposit insurance was implicitly provided to all large financial firms.
In a run, each depositor rushes to get his or her money from the bank before the doors are closed because its money holdings have been exhausted. Banks are often reluctant to pay the depositors that want their money because the banks’ money holdings are much smaller than their short-term deposit liabilities. During the Great Depression, banks took their time to pay off depositors, hoping, like Micawber, for something to turn up. The technique goes back to the eighteenth century.
Macleod’s Theory and Practice of Banking describes how the Bank of England defended itself in September 1720 against a run brought on by its reversal of a promise to absorb the bonds of the South Sea Company at £400. The Bank organized its friends in the front of the line and paid them off slowly in sixpence coins. These friends brought the coins back to the Bank through another door. The money was deposited, again slowly counted, and then again paid out. The run was staved off until the feast of Michaelmas (29 September). When the holiday was over, so was the run, and the Bank remained open.30
A second story, which may well have the same origin and is likely to be more accurate, is that the Sword Blade Bank, a supporter of the South Sea Company, resisted attempts to redeem its paper with silver coins. When the run started on 19 September, the bank brought up wagonloads of silver that it paid out ‘slowly in small change’. One depositor is reported to have received £8000 in shillings and sixpences before the bank closed its doors on Saturday 24 September.31 The circumstances suggest one story; the dates, two. Since the Sword Blade Bank and the Bank of England were mortal enemies, it is unlikely they cooperated.
The lessons of 1720 were not lost on the Bank of England a quarter century later. The Young Pretender (Charles Edward, grandson of James II) landed in Scotland in July 1745, unfurled his banner in September, invaded England in November, arrived in Carlisle on 15 November, and reached Derby on 4 December. Panic broke out on Black Friday, 5 December 1745. British Consols fell to 45, the lowest price on record, and a run began on the Bank of England. The Bank resisted, partly by paying off its notes in sixpence coins. The time gained was used to induce London merchants to proclaim their loyalty and readiness to accept Bank of England notes. The second half of the prescription, collecting pledges of faith in notes, was used again in similar circumstances when the French landed at Fishguard in 1797. On that occasion, 1140 signatures of merchants and investors in government stock were collected in a single day.32 The time gained in 1745 by both the slow payout and the petition of support enabled the government to organize the army that defeated the Young Pretender at Culloden in April 1746.
Complete shutdowns and bank holidays
One way to stop a panic is to close the market so trading stops. Trading on the New York Stock Exchange was halted in 1873, and in London and many other centers at the outbreak of war in 1914. In both cases the motivation was to stop a run by providing the market participants with more time to think through whether it was necessary or desirable to sell at depressed prices.
However, shutdowns may drive the trading underground and intensify the panic. Moreover, short-run and long-run goals are in conflict. Closing the stock market during one panic may exacerbate the next, as investors sell their stocks or withdraw their money from the call money market because they are fearful that trading will be halted. The New York Stock Exchange was closed in a panic in September 1873, but a financial editor suggested that fear that trading on the exchange might be halted in October 1929 hastened the withdrawal of call money by out-of-town banks and other market participants.33 The closing of stock exchanges in Pittsburgh and New Orleans for two months in 1873 had fewer serious consequences, since they traded only the securities of the local firms that had brought on their own difficulties.34
The New York Stock Exchange and the other US stock exchanges were closed for a week after the bombing of the World Trade Center towers on 11 September 2001 because the communications and the technical support systems were inoperative. Many of the same securities could have been traded on the regional stock exchanges in the United States although these exchanges would have been overwhelmed.
The declaration of a legal holiday by the government is another technique for closing the market, which was used during the panic of 1907 in Oklahoma, Nevada, Washington, Oregon, and California.35 The device was the forerunner of the bank holidays that started at the local level in the fall of 1932 and were generalized throughout the country on 3 March 1933, the day that Franklin D. Roosevelt was inaugurated as president. (A bank holiday closes only the banks, while a legal holiday shuts down all businesses.)
Another device is to suspend the publication of bank statements, as in 1873, in the hope that ‘what you don’t know won’t hurt you’. The technique was designed to hide the large losses of a few banks since the fear was that accurate news would further reduce depositor confidence.36
Some commodity and financial markets set daily limits on the maximum change in prices; when the limit is reached, trading is suspended for the rest of the day. Specialists in individual stocks have often taken a ‘time out’ whenever the imbalance between the buy orders and the sell orders has been exceptionally large. This ‘circuit breaker’ was recommended for US stock markets after the meltdown of Black Monday, 19 October 1987. The proposal for the New York Stock Exchange was to postpone trading for a stated interval – such as twenty minutes – in those stocks whose prices increased above or declined below the limit.
Time has been gained by moratoriums on payment of all debts or on particular types of obligations, such as bills of exchange that have less than two weeks to run. The most ubiquitous measure of this sort is that the bank examiners ignore the bad loans that banks own as long as they can – an implicit moratorium on marking the loans to their market value. Regulatory forbearance was used in the US savings and loan debacle in the 1980s. The International Monetary Fund and the World Bank continued to allow the indebtedness of many African countries to increase by the amount of the interest that was due; if these institutions had declared the loans in default, they would have had to recognize the losses on the loans. The banks that were the lenders to the Real Estate Investment Trusts (REITs), landlords of mortgaged shopping centers and the owners of mothballed Boeing 747s allowed the interest on their bank loans to compound because they wanted to delay the recognition of the loan losses to a more propitious moment when their capital would be larger. But the lenders need forbearance from the bank examiners.
Official moratoriums may be less effective than informal ones, however. A moratorium on the settlement of differences in payments due on the 1873 Vienna Stock Exchange lasted a week, from the stock market collapse to 15 May. A guarantee fund of 20 million gulden was put together by the Austrian National Bank and the solid commercial banks; these imitations of earlier measures were of little assistance.37 Another moratorium was noted in Paris after the July Monarchy when the municipal council decreed that all bills payable in Paris between 25 July and 15 August should be extended by ten days. This moratorium sterilized the commercial paper in banking portfolios and did nothing to discourage a run by holders of notes.38
The major device used in the United States to cope with bank runs prior to the establishment of the Federal Reserve System was the clearing-house certificate, which is a near-money substitute that was the liability of a group of large local banks. A bank subject to a run could pay the departing depositors with these certificates rather than with coin. The New York clearing-house was established in 1853 and the one in Philadelphia in 1858 after the panic of 1857. During the panic of 1857, New York banks failed to cooperate to halt the run. The Mercantile Agency of New York took the position that if four or five of the strongest banks had come to the assistance of the Ohio Life and Trust Company, so it could meet its obligations, the business and credit of the country would have been preserved.39 By 1873 the New York banks were ready to accept payment on cleared checks in clearing-house certificates. The advantage of these certificates was that they reduced the incentive for any bank to bid deposits away from its competitors. Sprague insisted that this system had to be accompanied by an agreement to pool bank reserves; otherwise, a bank that was not subject to a net drain might be forced to suspend payments after it paid cash to its own depositors if it had not received cash in settlements from other banks.40 In 1873 reserves were pooled.
One serious drawback of clearing-house certificates was that they were acceptable only in the local area – New York, Philadelphia, Baltimore. Thus these certificates helped maintain domestic payments such as payrolls and retail sales within a city but they dampened the effective flow of payments between cities. In the 1907 panic, 60 of the 160 clearing-houses in the United States used these certificates to facilitate local payments. Nevertheless Sprague claimed that the dislocations of the domestic exchanges were no less complete and disturbing than on previous occasions. The prices of New York funds in Boston, Philadelphia, Chicago, St Louis, Cincinnati, Kansas City, and New Orleans between 26 October and 15 December 1907 varied from a discount of 1.25 percent in Chicago on 2 November to a 7 percent premium in St Louis on 26 November, an increase from 1.5 percent the previous week.41 In December 1907 Jacob H. Schiff wrote: ‘The one lesson we should learn from recent experience is that the issuing of clearinghouse certificates in the different bank centers has also worked considerable harm. It has broken down domestic exchange and paralyzed to a large extent the business of the country.’42
Other devices of the same general character were clearing-house checks and certified checks that were both close substitutes for money and increased the means of payment in circulation.
Non-bank groups can also organize to mitigate a panic. Consider, for example, the stock market consortium. On 24 October 1907, a bankers’ pool, headed by J.P. Morgan, loaned $25 million at 10 percent in call money in an attempt to stem the collapse of the stock market.43 Twenty-two years to the day later, on Black Thursday in 1929, Richard Whitney went from post to post on the floor of the New York Stock Exchange and placed bids to buy stocks on behalf of a syndicate headed by J.P. Morgan and Company – once again.44
Banks have also collaborated through rescue committees (as in Vienna in May 1873 and earlier), loan funds, funds for guarantees of liabilities, arranged mergers of weak banks and firms, and other devices whereby the strong banks support the weak and failing banks.45 Three examples include the role of the Paris banks in the 1828 crisis in Alsace, various devices employed by Hamburg in meeting the difficulties of 1857, and the Baring Brothers loan guarantee of 1890.
The Alsatian crisis of 1828
Three firms in textiles failed in December 1827 at Mulhouse. The Paris banks then refused to accept any Alsatian paper, and the Bank of France set a limit of 6 million francs on the amount it would support, a figure that was ‘scarcely the fortune of two Alsatian houses’. The Bank of France then decided against accepting any paper with Mulhouse or Basel endorsements and that decision precipitated a panic. On 19 January two more Mulhouse merchants failed. On 22 January in Paris there were rumors of the failure of two Schlumberger firms. The Paris banks sent Jacques Laffitte as an emissary to Mulhouse; he arrived on 26 January and offered to lend 1 million francs on the consignment of merchandise. Before he came, however, two textile merchants, Nicholas Koechlin and Jean Dollfuss, had left Mulhouse for Paris. To raise cash, these merchants had been selling inventories on the market at discounts of 30 to 40 percent from the traditional market prices for these goods. Nine houses failed from 26 January to 15 February. Lévy-Leboyer wrote that it could have been worse. At the last minute a syndicate of twenty-six Paris banks, presided over by J.-C. Davillier, extended a credit of 5 million francs to Koechlin and Dollfuss, who returned to Alsace on 3 February and distributed 1 million francs to those of their colleagues who offered guarantees and kept 4 million for themselves. These measures restored confidence.46 Those who qualified for neither the Koechlin-Dollfuss fund nor the Basel money failed.47
The Hamburg crisis of 1857
The background of the crisis of 1857 in Hamburg was that trade had expanded, particularly because the Crimean War had led to an expansion of credit. Hamburg was the ‘all-English’ city of Germany, but had close relations with the United States in sugar, tobacco, coffee, and cotton, and with Scandinavia. When the deflationary tidal wave swept across the Atlantic, Hamburg was inundated. The panic touched off by the failure of Ohio Life on 24 August arrived in Hamburg three months later (following price declines of 30 percent) with the suspension of Winterhoff and Piper, a firm that was engaged in the American trade.48 Daily dispatches from the British consulate in Hamburg by date tell the story:
21 November: Some of the leading merchant houses and two banks plan for relief.
23 November: Two major houses engaged in the London trade fail, and the Discount Guarantee Association grows more cautious in endorsing Hamburg bills.49 On one authority the Discount Guarantee Association is exhausted in three days.50
24 November: A Discount Guarantee Association (Garantie-Diskontverein) is formed, initially with a capital of 10 million marks banco, later raised to 13 million (about £1 million), of which the sum of 1 million marks is to be paid in immediately.
28 November: The chamber of commerce and leading merchants induce the Senate to call parliament (Burgerschaft) to arrange to issue government bonds in order to lend 50 to 66 2/3rds percent of the value of hypothecated goods, bonds, and shares to merchants in distress.
1 December: With the suspension of Ullberg and Cremer, ten to twelve houses in the Sweden trade have gone down. The Discount Guarantee Association will not issue any more guarantees. Business is at a standstill.
2 December: A suggestion is made to change the laws of bankruptcy to enable creditors to share in attachment of goods.
7 December: A proposal is made to establish a state bank for discounting good bills to the amount of 30 million marks banco (about £2.4 million). Parliament rejects this, wanting instead to issue 30 million marks banco of paper currency as legal tender. The senate rejects this, insisting on clinging to the silver standard.
In the end, a compromise was reached for a State Loan Institute fund of 15 million marks that included 5 million marks banco of Hamburg government bonds and 10 million in silver to be borrowed abroad.51 The story of the silver train (Silberzug) is recounted in Chapter 12 as an example of an international lender of last resort.
One observer totaled the sums available for rescue operations to 35 million marks banco, which included 15 million in the Discount Guarantee Association, 15 million in the State Loan Institute, and 5 million from the chamber of commerce. He compared this amount with 100 million marks banco of protested bills and noted that if merchants speculated with capital equal only to one-sixth the value of their goods, a 17 percent decline in prices of these goods would be sufficient to wipe out their capital position. To the suggestion that the senate was 300 years behind the times, he reported with approval the senate’s answer: The merchants have been 300 years ahead of the times in issuing debt. State help in these cases, he insisted, merely means assistance for speculation and perpetuation of higher prices at the cost to the consumer.52
Guarantees of liabilities: the Baring crisis
The most famous guarantee of liabilities was that worked out by William Lidderdale when he was Governor of the Bank of England during the Baring crisis of 1890. Similar guarantees had occurred earlier in Britain. In December 1836 the private bank Esdailes, Grenfell, Thomas and Company, which served as London agents for seventy-two country banks, was in financial difficulties. The view was that this firm could not be allowed to fail because of its relationships with the country banks; moreover, its paper included all the best names in the City. The assets of the bank far exceeded its liabilities, and the London bankers offered guarantees. The Bank of England led the list with £150,000. Esdailes survived, but only for two years.53
The guarantee was worked out as an alternative to a letter of indemnity that permitted suspension of the Bank Act of 1844. The letter was offered by the Chancellor of the Exchequer, Lord Goschen, to Lidderdale, who refused on the ground that ‘reliance on such letters was the cause of a great deal of bad banking in England’.
If Lidderdale refused to quiet the market by the usual means that had been employed in 1847, 1857, and 1866, he was not one to let the market take its medicine. In August 1890 he warned Baring Brothers that the firm would have to moderate its acceptances for its Argentine agent, S.B. Hales. Baring Brothers revealed its acute distress to Lidderdale on Saturday 8 November. Fearful of a panic when Barings’ condition was made public, the Bank of England met with the Exchequer on Monday 10 November, turned down the letter of indemnity, prepared for the problem by seeking assistance from foreign countries (a subject for Chapter 12), and formed a committee headed by Lord Rothschild to address the question of the large overhang of Argentinean securities.
As the days passed, the rumors circulated and Barings’ bills were increasingly discounted at the Bank of England. By Wednesday Lidderdale had learned that although Baring was solvent it would still need £8 to £9 million. On Friday John Daniell, the leading man at Mullens and Co., used by the Bank of England in open-market operations, came to Lidderdale, crying ‘Can’t you do something, or say something, to relieve people’s minds: they have made up their minds that something awful is up, and they are talking of the very highest names – the very highest!’54
On 14 November Lidderdale met with two cabinet ministers who represented the Exchequer, Lords Smith and Salisbury. They agreed that the government would increase its balance at the Bank immediately and that the government would share with the Bank in any losses suffered on Baring Brothers’ paper discounted by the Bank between 2pm Friday and 2pm Saturday. On the basis of this agreement, Lidderdale met with eleven private banks to get them to contribute to a fund that would guarantee Barings’ liabilities, and he got the State Bank of Russia to agree not to withdraw its £2.4 million deposit at Barings. The private banks as a group contributed £3,250,000, but this included £1 million from the Bank of England as well as £500,000 from each of three lenders, Glyn, Mills & Co., Currie and Co., and Rothschilds. Lidderdale then used these commitments to obtain the agreement of the five London joint-stock banks to join the guarantee fund for another £3,250,000. On the basis of these assurances, The Times of 15 November announced that Baring Brothers would fail but that there would be no loss. The work of the guarantee fund continued on Saturday because the directors of the joint-stock banks had to meet to approve their subscriptions, which was done by 11am. Then other banks and financial institutions raised the fund from £7.5 million in the morning to £10 million by 4pm; the guarantee fund eventually reached £17 million and was taken as a measure of the strength of the London financial system. Martin’s Bank was in distress over its loans to Barings and to Murriettas, another bank involved in Argentina. Martin’s Bank joined the guarantee fund for £100,000 on 18 November (Tuesday), too late to afford much help to Barings, but early enough to demonstrate to the world its own strength.55 In summarizing the episode, Powell stated: ‘The Bank is not a single combatant who must fight or retire, but the leader of the most colossal agglomeration of financial power which the world has so far witnessed.’56
On 25 November a new firm, Baring Brothers and Co., Ltd, was formed as a joint-stock company with a capital of £1 million. The form of the guarantee may be of interest.
Guarantee Fund
Bank of England, November, 1890
In consideration of advances which the Bank of England have agreed to make to Messrs. Baring Brothers and Co., to enable them to discharge at maturity their liabilities existing on the night of the 15th of November, 1890, or arising out of business initiated on or prior to the 15th of November, 1890.
We, the undersigned, hereby agree, each individual, firm, or company, for himself or themselves alone, and to the amount only set opposite to his or their names respectively, to make good to the Bank of England any loss which may appear whenever the Bank of England shall determine that the final liquidation of the liabilities of Messrs. Baring Brothers and Co. has been completed so far in the opinion of the Governors as practicable.
All the guarantors shall contribute rateably, and no one individual, firm, or company, shall be called on for his or their contribution without the like call being made on the others.
The maximum period over which the liquidation may extend is three years, commencing the 15th of November, 1890.57
The rescue of Long-Term Capital Management in September 1998 provides a contrast to the rescue of Barings. William McDonough, the president of the Federal Reserve Bank of New York, induced fourteen large banks and investment banks, including Merrill Lynch, Morgan Stanley Dean Witter, J.P. Morgan, Chase Manhattan Bank, and the Union Bank of Switzerland to provide $3.6 billion of capital to prevent the collapse of LTCM; in exchange they acquired 90 percent of LTCM’s equity.58 These firms were large creditors of LTCM so the ‘bailout’ involved a change in the legal nature of their claim on LTCM. The Federal Reserve was concerned that if LTCM failed the markets would be paralyzed by the need to unwind its massive position in futures and options contracts and other types of derivatives.
Since 1934, federal deposit insurance in the United States has prevented bank runs by providing an ex ante guarantee of deposits. Initially the deposits were guaranteed up to $10,000 but gradually the ceiling on deposits was raised in stages and eventually reached $100,000 – and then $250,000 during the 2008 crisis. When large banks got into trouble, the FDIC deliberately removed all limits on the amounts of deposits covered by the guarantee to halt imminent runs and in practice it established that banks with significant deposits over $100,000 were ‘too big to fail’ (although the shareholders of these banks would probably lose all of their money). Similarly the owners of the subordinated debt of these banks might lose much or all of their money. The deposits of the foreign branches of these banks had implicitly become insured even though the banks had not paid deposit insurance premiums against their offshore deposits. Although deposit insurance was designed to prevent bank runs by taking care of the ‘little man’, this ceiling had in practice been raised to the sky.59
When the limit on deposit insurance was $100,000, ‘deposit brokers’ arranged for the placement of larger amounts of money in guaranteed deposits. John Doe could have an insured deposit of $100,000, his wife Jane could also have an insured deposit of $100,000, and together John and Jane could have a third insured deposit of the same amount. And the Does could follow the same strategy with the bank across the street.
The effect of this innovation was that it provided a guarantee to wealthy individuals and hence circumvented one purpose of the ceiling. Moreover it encouraged banks to make riskier loans since they were confident that they were protected against runs by the owners of the larger deposits – if these riskier loans proved profitable, the owners of the banks would benefit and if the loans went into default, the owners would not have to worry about bank runs (although the market value of their own shares might decline and even become worthless).
The implosion of the bubble in Japan in the 1990s caused the value of the loans of the banks headquartered in Tokyo and Osaka and various regional centers to decline sharply below the value of their liabilities. Nevertheless, there were no bank runs; the depositors were confident that they would be made whole by the government if any of the banks were closed.
Deposit insurance has limited both bank runs and contagion in the runs from one troubled bank to other banks in a neighborhood. What accounts for the reluctance to provide insurance at an earlier date?
In the long tradition of the United States, free banking, even wildcat banking, was the rule. Anyone could start a bank, and many did. Risks were large, banker turnover rapid. A guarantee of bankers’ deposits would have constituted a license to speculate, if not embezzle, and would have removed the threat of withdrawal of deposits, which was the major check on the irresponsibility of bankers. Deposit guarantees were rejected as conducive to bad banking as late as 2 March 1933, when the Board of Governors of the Federal Reserve was not prepared to recommend such a guarantee, or any other measures, on the eve of the national bank holiday.60
The Federal Deposit Insurance Corporation had an excellent financial record until the 1970s; the deposit insurance premiums that it collected were much larger than the amount it paid out to make good the losses when banks failed. From the beginning in 1934 through 1970, only one bank with deposits of more than $50 million failed, and most failures were of banks with deposits of less than $5 million. The FDIC had in most cases arranged for takeovers of the failed banks so that individuals and firms with deposits above the maximum insured amounts did not incur losses when banks failed.
Beginning in the late 1970s the financial problems of the FDIC and those of the Federal Savings and Loan Insurance Corporation (FSLIC) mounted sharply as more and more banks and thrifts were closed. The FDIC rescued a large number of banks including two giants, the Continental Illinois Bank of Chicago and the First Republic Bank of Dallas; honoring the deposit insurance guarantee cost these agencies billions of dollars. The usual operating procedure was to close the bank or the thrift institution when its capital had been depleted, and then to carve a ‘good bank’ from the rescued institution while the remaining assets of the failed institution would be retained for eventual sale by another newly created government agency, the Resolution Trust Corporation (RTC). Both insurance agencies incurred such large losses in honoring their guarantees that their capital and reserves were exhausted; they then borrowed from the US Treasury. In the early 1990s, the estimates were that the losses to the US taxpayers would amount to $150 billion but the increase in the US growth rate meant that the RTC received more money than anticipated from the sale of collateral and bad loans so the losses totaled a bit more than $100 billion.61 There was some question whether a portion of this cost to the taxpayers could be reduced by increasing the insurance premiums on bank deposits – a suggestion resoundingly opposed by sound banks.62
During the 2008 crisis the US government extended deposit insurance to the money market funds; the fear was that otherwise the owners of these funds would transfer their money to banks whose deposits were guaranteed by the FDIC. The money market funds then would be obliged to sell assets, and the rapid sale of assets would depress their prices. Some money market funds then would ‘break the buck’; the net asset value per share would decline below $1.00 – which would trigger a massive, self-justifying run.
One ancient device short of lending money to a firm in trouble was to issue marketable securities to the firm against appropriate collateral. (Of course, when markets break down, even the most liquid securities may not be sold readily.) The securities have been private and public; both types were part of the complex package put together by Hamburg in 1857. In 1763 and 1799, in an equally complex and jerry-built system of support, admiralty bills were an integral feature.63 The widest development, however, concerned the Exchequer bills issued in Britain in 1793, 1799, and (without enthusiasm) 1811, but sternly rejected in 1825.
The Exchequer bill was widely thought to have been the idea of Sir John Sinclair, although it may have originated with the Bank of England. On 22 April 1793, City leaders met with the Prime Minister, William Pitt, to devise means to combat the crisis that arose from the failure of 100 of the 300 country banks and the calamitous decline in commodity prices. The next day, eleven members of this group met at the Mansion House to formulate a scheme for state assistance. According to Clapham, there was no clear guide to what ought to be done. Then the idea emerged of having the government issue £3 million in Exchequer bills, a total that was later raised by parliament to £5 million, to be issued to merchants on the collateral of goods that they would deposit in the customs houses. An additional feature of the plan was to issue £5 notes – the previous minimum was £10 – to economize on the use of gold and silver coin. The Exchequer bills were issued by special commissioners rather than the Bank of England; £70,000 of these bills was immediately sent to Manchester and an equal amount to Glasgow. The device worked like a charm, according to MacPherson. Three hundred and thirty-eight firms applied for £3 million of the total amount. A total of £2.2 million was granted to 228 firms, only two of which subsequently went bankrupt. Applications for more than £1.2 million were withdrawn after the panic abated.64
In 1799 the panic in Hamburg had an echo in Liverpool, and Exchequer bills again were used. Parliament provided £500,000 in Exchequer bills, used solely in Liverpool, against £2 million of goods stored in warehouses.65
In 1811 the question arose again. A Select Committee on the State of Commercial Credit was appointed. Among its members were Henry Thornton, Sir John Sinclair, Sir Thomas Baring, and Alexander Baring. The committee’s report, completed in a week, recorded the distress of exporters to and importers from the West Indies and South America, as well as the piles of goods bound for the Baltic that had been cut off and stored in London warehouses, and recommended a new issue of £6 million of Exchequer bills. Support was moderate in the House of Commons because of the overtrading in South America; the opposition, while sympathetic to the distress, doubted the wisdom of bailing out the speculators. William Huskisson, who later made his mark as the president of the Board of Trade, claimed that the evil came from too easy credit:
Did gentlemen not see that the race of old English merchants, who never could persuade themselves to go beyond their capital, was superseded by a set of mad and extravagant speculators, who never stopped so long as they could get credit, and that persons of notoriously small capital had now eclipsed those of the greatest consequence; so that speculations now took place even in the lowest articles of commerce ... If the relief given was used for further speculation, it would only aggravate the evil – and he feared that this might be done – in which case the present measure would go only to add six million to the circulation and to raise the prices of all our commodities.66
Smart gave the fullest account of the debate and noted that many criticized the measure, though few were bold enough to deny it. In the end the proposal passed, but few applications were made and only £2 million was advanced. ‘Not many of those who were in embarrassed circumstances were able to furnish the desired security, and it is difficult to see what remedy there was in being enabled, by advances, to produce more goods when the radical evil was that there was no market for them.’67
Lady Luck once worked effectively to assist a bank in distress. Wirth wrote that the Brothers Kauffmann in Hamburg was failing during the crisis of 1799 when one of the brothers sent his bride a Hamburg city lottery ticket, which carried a first prize of 100,000 marks banco. She bought the same number in another lottery in the Duchy of Mecklenburg, the prize for which was an estate worth 50,000 Prussian thalers, then equal to 100,000 marks banco. She won both, and the Brothers Kauffmann was fully rehabilitated.68 The odds at Las Vegas may be higher for the solution to a global crisis.