5

The Critical Stage – When the Bubble Is About To Pop

Changing expectations

The standard model of the sequence of events that leads to financial crises is that a shock triggers an economic expansion that morphs into an economic boom and then euphoria develops; asset prices increase rapidly, much more rapidly than GDP or some other income measure. Then there is a pause in the pace of these increases. A few savvy or lucky investors sell some of their assets to park their speculative gains. The slowing of the increase in asset prices may induce a more cautious approach by others. Distress is likely to follow as asset prices begin to decline. The pattern is biological in its regularity. A panic is likely and then a crash may follow. Lord Overstone, the leading British banker of the middle of the nineteenth century, saw a similar pattern and was quoted with approval by Walter Bagehot: ‘quiescence, improvement, confidence, prosperity, excitement, overtrading, CONVULSION [sic], pressure, stagnation, ending in aquiesence’.¹ Overstone identified five stages of euphoria before the financial crisis, or, in his words, the convulsion.

The theory of rational expectations assumes that investors’ expectations change more or less instantaneously in response to each shock and that investors immediately see through to the impacts of each shock on the long-run equilibrium prices for real estate and stocks and commodities. In contrast the insight from financial history is that expectations change slowly at some times and rapidly at others as various groups realize – sometimes at different moments and at other times more or less simultaneously – that the current forecasts of prices and values in the distant future differ from earlier views of these same prices and values.

The change in the mind-sets of investors from confidence to pessimism is the source of instability in the credit markets as some borrowers – individuals as well as firms – realize that their indebtedness is too large relative to their incomes. These borrowers adjust to their new perceptions about the future by reducing their spending so they will have the cash to reduce debt or to increase saving. Some firms may sell divisions and operating units to get the cash to repay loans. The lenders recognize that they have too many risky loans and they seek repayment of outstanding loans from the riskier borrowers and they become reluctant to renew these loans as they mature and they raise the credit standards for new loans.

The period of financial distress may last weeks, months, or even for several years, or it may be concentrated in a few days. The economic downturn that followed the collapse of the US stock market in 1929 continued for four years – until a new, more interventionist government came to power. Japan was in the economic doldrums for more than ten years after stock prices and real estate prices began to decline in January 1990. In contrast South Korea recovered from its 1998 economic shock by the beginning of 1999. The handover of sovereignty in Hong Kong in mid-1997 and its advent as a special economic region of China coincided with the onset of the Asian Financial Crisis; property prices declined by 40 to 50 percent during the next five years. The recession that began in the United States in 2001 was relatively mild and brief, even though stock prices declined by 40 percent from 2000 to 2003.

The United States experienced a much sharper decline in household wealth in 2008 and 2009; prices of residential real estate declined by more than 30 percent and stock prices fell by nearly 50 percent – before a partial recovery. Property prices began to decline at the end of 2006 and economic growth was positive throughout 2007 and in the first half of 2008; the first negative quarter was the third quarter of 2008 – and even then the decline was modest. Then the collapse of Lehman Brothers in September 2008 triggered a panic and a crash. The credit system froze; interest rate spreads surged, and LIBOR increased by 500 basis points relative to the Federal Funds rate whereas the traditional spread had been 10 to 20 basis points. The combination of the more restrictive supplies of credit and the greater cautiousness by households led to a very sharp decline in GDP.

Should a government intervene to moderate the cycle? Government policies play a vital role in the formation of expectations. Can the government head off a financial crisis by dampening the expectations that develop in the euphoria? Should the government seek to moderate the impacts of the decline in prices of stocks and real estate and commodities after the bubble has imploded?

Bear Stearns, Fannie Mae and Freddie Mac, and Lehman Brothers

One of the first signs of adjustment in the United States to the extraordinary increase in home prices and the large excess supply of homes that began in 2002 was a surge in the bankruptcies of mortgage brokers that started in the last few months of 2006; these firms were ‘middlemen’ and wholesalers who acquired mortgages from home buyers, after having indicated the terms of the mortgages that they would acquire. Then they sold the mortgages to the investment banks, which would place mortgages with similar terms in a trust that they would use as collateral for issuing mortgage backed securities (MBSs). The investment banks retained the right to ‘put’ these loans back to the mortgage brokers if the borrowers stopped making their monthly payments within 12 or 18 months. As the borrowers fell behind in their payments, the investment banks returned these mortgage loans to the brokers and asked for the return of their money, and the firms that were unable to repay went bankrupt. The investment banks were stuck with tens of billions of dollars of failed mortgages after the brokers went bankrupt.

The combination of the default by those who had recently borrowed to buy homes and the failure of the mortgage banking firms led to a sharp decline in the supply of credit available for mortgages. The pace of home purchases slackened. The property developers that had built homes in anticipation of future sales experienced a surge in their inventories of unsold properties as home prices declined.

House prices in some of the regional markets peaked toward the end of 2005, and at the national level at the end of 2006. The decline in 2007 of nearly 15 percent led to a dramatic decline in new housing starts.

The first major casualty was Bear Stearns; two of the leveraged hedge funds that it managed tanked, and the investors in these funds lost most of their money. Then there was a run on Countrywide Financial, the largest US mortgage lender, in mid-August 2007. At the same time, there was a run on Northern Rock, which was the largest mortgage lender in Britain. Both firms had been aggressive in their efforts to increase their market share, both had relied on the commercial paper market for more than 30 percent of the money that they used to buy mortgage loans. In effect they had been selling short-term IOUs to firms that had extra cash to get the money to buy more mortgages. Countrywide Financial was rescued by the Bank of America, which initially made a capital investment through the purchases of preferred shares and subsequently acquired Countrywide. Northern Rock was rescued by the Bank of England, which initially provided liquidity; subsequently the British government acquired shares in the firm and became its majority owner.

The next striking development were the runs on Bear Stearns in March 2009, one by the owners of its shares and the other by those who owned fixed-price claims on Bear and were reluctant or unwilling to acquire new claims on Bear as the ones they owned matured. The Federal Reserve facilitated the acquisition of Bear by JPMorganChase by agreeing to accept up to $29 billion of securities of dubious quality that Bear owned.

Fannie Mae and Freddie Mac, the two very large government-sponsored enterprises that carried more than 50 percent of the credit risk attached to US home mortgages, were ‘nationalized’ by the US Treasury in the second week of September 2008. The owners of their common shares lost virtually all of their money. The owners of their preferred shares also were wiped out; many banks had purchased these preferred shares because that could be considered as part of their required Tier I capital. Thus the bankruptcy of Freddie and Fannie had a major whammy on the capital of other financial institutions.

The next week Lehman Brothers went bankrupt, because the US government had decided that it didn’t want to ‘save Lehman’. The government officials said that they didn’t have the authority to save Lehman. Perhaps. Or perhaps they could have found the authority if they are tried harder. Or they could have provided some temporary financial assistance to enable Lehman to avoid bankruptcy while they acquired the legal authority to save Lehman.

The decision not to ‘save Lehman’ triggered the most massive credit crisis in the last one hundred years; credit spreads soared. A large number of different firms relied on short-term funding for the money to buy long-term securities, and they found it extremely difficult – and costly – to refinance their maturing IOUs.

The decision not to ‘save Lehman’ reflected that government officials were concerned by the public outrage about ‘bailing out’ the rich bankers. They – like officials and analysts in the previous two hundred years – believed that saving a large firm would encourage risky financial behavior ten and twenty and thirty years in the future.

Then on Tuesday 14 September – the day after Lehman failed – the Federal Reserve lent $85 billion to AIG; the loan gave the Fed the option to buy 79.9 percent of AIG’s stock. The ‘too-big-to-fail’ doctrine had been revived. The concern was that if AIG tanked, runs on many other financial institutions would occur, and this dramatic rush for cash would lead to the collapse of the financial system.

The interplay among the decline in real estate prices and bank loan losses was inevitable and predictable, although the identity of the firms that would be pummeled could not have been predicted. However, those firms that had rapidly increased their market share – Countrywide, Northern Rock, Washington Mutual – had done so on the basis of short-term funds borrowed in wholesale markets, and they were more vulnerable to runs than the lenders that relied almost exclusively on deposits for the money to buy mortgages.

The money market funds – established primarily by brokerage firms as an alternative to government-guaranteed deposits – were subject to runs, and the US Treasury hastily extended the umbrella of deposit insurance to these firms.

The major question is how many of these developments were predictable in response to the decline in real estate prices. It was inevitable and hence predictable that the net worth of millions of households would decline as property prices fell. Several millions of these households would have ‘upside down mortgages’ – they would owe more on their mortgage loans than the market value of their homes. Millions would walk away from these properties, and leave the lenders with losses.

Warnings

One proposal is that a government should make its knowledge or publish its own forecast if it thinks it knows more than the speculators.² Thus the government might calm the concerns or fears of investors by making that knowledge publicly available.³ Many individuals within government have views about the economic and financial outlook but these are often at variance with each other so that developing a ‘government view’ could occur only if someone – the prime minister, the head of the central bank, the minister of finance – forged a consensus.

The historical record provides little support for the view that statements from government officials have had much of an impact in dampening euphoria. In some cases ‘a word to the wise’ may be sufficient but in others the warnings are not likely to be adequate to forestall further increases in asset prices. The likelihood that investors and speculators would heed the warnings of a government official when asset prices are increasing at annual rates of 20 to 30 percent a year is not especially high.

The first such warnings on the record were made about 1825. Although many writers have viewed the Bubble Act of June 1720 as a warning by Robert Walpole and King George II against speculation, the primary objective of that legislation was to repress competitors of the South Sea Company because the other bubbles were draining cash that the South Sea Company wanted and needed.⁴ The Bubble Act, which had been strengthened in 1749, made both swindles and starting a legitimate business more difficult and was not repealed until the nineteenth century.

The banking authorities began to warn about speculative booms in the nineteenth century. In the spring of 1825, Prime Minister Canning, the Chancellor of the Exchequer, Lord Liverpool, Sir Francis Baring, and W.R. McCulloch in The Scotsman warned against the excesses of speculation. The warning contributed to the crisis – but the crisis was probably inevitable. In the panic and crash of December 1825, Lord Liverpool did not rescue the speculators because nine months earlier he had said he would not.⁵ In 1837 the President of the Board of Trade, J. Poulett Thomson, excoriated the prevalent spirit of speculation, which differed from that of 1825 in that individuals were investing at home rather than abroad.⁶

In the fall of 1837 the gambling spirit crossed the Channel, and the Belgian and French authorities attempted to repress speculation by forbidding the quotation of the prices of notes and shares of corporations. Their efforts were futile; speculation had gone beyond the narrow framework of the Bourse and non-professionals such as rentiers and even ‘women and foreigners’ were becoming involved. Chambers of Commerce in Liège, Vervier, and Antwerp condemned stock market speculation. The Belgian king refused to charter a proposed bank, the Mutualité Industrielle. Investment slowed as a result of a decline in economic activity rather than in response to the statements from the administration and business establishment.⁷ In July 1839, Lamartine in the French Chamber of Deputies spoke against speculation and especially warned about the guarantees of railroad securities.⁸

The only suggestion that official condemnation of speculation may have been effective was from a French observer who commented on the 1857 crisis. In March 1856 the minister of the interior brought legal action against certain swindlers. Emperor Napoleon III congratulated O. de Vallée, the author of Les Manieurs d’Argent, which dealt sternly with dubious financial practices. The Senate passed laws. The Bank of France raised the discount rate to 10 percent. Napoleon III published a letter in Le Moniteur on 11 December that indicated that the government would provide support only for those catastrophes that were beyond human anticipation. According to d’Ormesson, speculative exuberance declined and the memory of the 1857 crisis reflected a certain glory on French commerce.⁹ But Rosenberg concluded that the warnings had been too late.¹⁰

Restrictions by the Austrian National Bank in 1869 produced a ‘great crash’ that proved to be a mini-crash compared with the one that followed in 1873.¹¹ The warnings and revelations of Eduard Lasker, a member of the Diet who in February 1873 had exposed the scandalous connections between the Prussian government and its commerce ministry and the railroad concessionaires, had no significant impact in quelling speculative sentiment.¹²

More timely were warnings issued by The Economist in 1888 against commitments to buy cedulas, the Argentinean land bonds. In April The Economist said that ‘the bonds ... might ... become a very inadequate security. Just at present all real estate at the River Plate commands inflated prices, but the occurrence of financial difficulties might easily render them unsaleable.’¹³ Then in May ‘A collapse of the “boom” in real estate, which is easily conceivable, would be sure to severely depress the value of the cedulas.’¹⁴ The warnings proved ineffective.

More memorable is the Cassandra-like utterance of Paul Warburg, a partner of Kuhn, Loeb & Co. and one of the founders of the Federal Reserve System, who in February 1929 warned the American public that US stock prices were too high and showed symptoms reminiscent of the 1907 panic; his statement followed a similar statement from the Chairman of the Federal Reserve Board. Investors paused briefly and then stock prices again increased. The ineffectiveness of Federal Reserve Chairman Greenspan’s statement about ‘irrational exuberance’ in December 1996 about the high level of US stock prices was noted earlier. In August 1999 Greenspan stated that the Fed would consider the level of stock prices when it set interest rates after the discount rate had been increased by 0.25 percent in August. Again the stock market barely noticed.

If prices of real estate and of stocks continue to increase despite these warnings, then it appears that the warnings lack credibility. Economic forecasters may know – or at least think they know – the long-run equilibrium values for real estate and for stocks, but the ability to foretell when market prices will move back toward these long-run average values rather than away from them is modest. Roger Babson sold his clients’ stocks in 1928 and he appeared foolish for more than a year as stock prices continued to increase.

The timing issue is complex. If the government authorities want their warnings to be effective, they need to provide their cautionary statements early enough to forestall some of the excesses of the euphoric period and late enough so the statements are credible. One metaphor from a former chairman of the Federal Reserve is that authorities are reluctant to take the ‘punch bowl away from the party just as the party is getting going’ because of the unfavorable public reactions.

The modern tradition is that the central banks develop their monetary policies to moderate the increases in the consumer price level or some other price level index; ‘inflation targeting’ became their new mantra. The policy question is whether the central bankers should ignore the increases in the prices of real estate and stocks if they are far above their long-run equilibrium values. Attempting to convince the speculators through statements alone generally has been futile.

Financial distress

Distress is widely discussed in connection with financial crises. The term is imprecise: one meaning is a state of suffering and another is of a hazardous situation. Commercial distress reflects the first definition, financial distress the second. Commercial distress implies that prices and economic activity and profitability have declined and that many mercantile and industrial firms have gone bankrupt – or are on their way to bankruptcy. Financial distress for an individual firm means that it is incurring significant losses and that there is a non-trivial probability that it will not have enough cash to pay the interest due on its debt according to the agreed-upon schedule.¹⁵ The credit panic in the last several months of 2008 led to a sharp decline in automobile sales, and both General Motors and Chrysler became bankrupt. Financial distress for an economy also has a prospective significance and implies the need for economic adjustments; firms may be on the verge of bankruptcy and banks may need additional capital to compensate for the decline in capital that has resulted from exceptionally large loan losses. The US government developed a $700 billion-plus Troubled Assets Relief Program (TARP) that would make money available to banks and other financial firms. Many investment projects become stalled because the developers cannot obtain the money needed to complete construction.

Other words used to describe the interval between the end of euphoria and the onset of what classic writers called revulsion and discredit (or crash and panic) are uneasiness, apprehension, tension, stringency, pressure, uncertainty, ominous conditions, fragility. More colorful expressions include an ugly drop in the market¹⁶ or a thundery atmosphere.¹⁷ Meteorological metaphors have been used frequently: ‘One feels again the oppressive atmosphere that precedes a storm.’¹⁸ And geological metaphors have been used: two years before the panic of 1847 Lord Overstone wrote to his friend G.W. Norman (the grandfather of Montagu Norman, the Governor of the Bank of England in the 1920s): ‘We have no crash at present, only a slight premonitory movement of the ground under our feet.’¹⁹ The seismographic metaphor was also used by Michel Chevalier who wrote from America about President Jackson’s war against the Second Bank of the United States: ‘A general collapse of credit, however short the time it lasts, is more fearful than the most terrible earthquake.’²⁰ Another French writer noted the ‘presentiment of disaster’.²¹ A German metaphor spoke of the ‘bow being so bent in the fall of 1782 that it threatened to snap’.²²

Distress is not an easily measured condition for an economy. Investors may have become apprehensive when the values of certain variables diverged significantly from average values; some of these variables include the gold reserve ratios of central banks, the ratio of debt to capital of a large number of firms or individuals, the losses of banks in relation to their capital, the ratios of a country’s external debt service payments to its export earnings, and the price-earnings ratios for stocks and the rental rate of return on real estate. Investors may be increasingly aware of the approach to a limit, such as the ceiling on note issues by the Bank of England stipulated in the Bank Act of 1844, the $100 million gold minimum requirement of the US Treasury in 1893, the ceiling on advances from the Bank of France to the French Treasury in 1924, the gold reserve ratio of the Reichsbank under the Dawes Plan in June 1931, or the free gold available to the Federal Reserve System prior to the passage of the Glass-Steagall Act in February 1932. A ratio of external debt to GDP of 60 percent for a country has been viewed as a premonitory indicator by investors; the ‘ice is thin’ for a country when this ratio is much higher. Similarly a ratio of government debt to GDP of significantly more than 60 percent is viewed as too high. Limits excite, as one Chancellor of the Exchequer noted in 1857:

Now when you impose a limit, there is no question that the existence of that limit, provided it makes itself felt at a time of crisis, must increase the alarm. People feel at the moment that a peril presses on them, they begin to calculate how much remains of that fund to which they look for assistance in times of commercial difficulty, and in whatever way you fix the limit, whether by Act of Parliament, or, as Mr. Thomas Tooke [a leader of the Banking School] proposed, by a sort of usage, or, as in France, by the discretion of the Government acting on the Bank of France, there is no doubt that in moments of crisis the limit must aggravate the alarm.²³

A French official expressed the same idea two decades later when he defended the tradition that the Bank of France should maintain a specie reserve of about one-third of its demand liabilities but without a hard and fast legal requirement: ‘A fixed ratio is not required. That would be unwise ... the terror of a limit fixed and absolute.’²⁴ Overshooting of the limit may have psychological significance. In March 1924 although sophisticated bankers knew that a small increase in the French money supply would not be dangerous, the public had come to regard the ceiling on Bank of France advances to the Treasury as an index of economic health. As one minister put it, the French were close to the upper limit of elasticity of confidence in their own currency.²⁵

Causes of distress and the symptoms of distress are observed at the same time and include sharply rising interest rates in some or all segments of the capital market, an increase in the interest rates paid by subprime borrowers relative to the interest rates paid by prime borrowers, a sharp depreciation of the currency, an increase in bankruptcies, and an end of price increases in commodities, securities, and real estate. These developments are often related and show that the lenders have become over-extended and are trying to reduce their exposure to risks and especially to large risks.

Financial distress in the nineteenth century was compounded by the arrangements for purchases of newly issued stock that provided for a series of payments by the buyers in response to ‘calls’ from the issuers of stocks as construction work proceeded. In 1825 and 1847 in Britain and in 1882 in France some of the buyers of the stocks did not have the cash to meet the calls; they may have counted on selling the security at a profit before the next call. Thomas Tooke describes this embarrassment in 1825 as acute because the call for cash payment was immediate and pressing while prospects for earnings on the stock were remote and uncertain.²⁶ Distress developed in January 1847 when railroad calls amounted to £6.5 million in one month.²⁷

The chain-letter aspect of security issues became evident in the finances of the South Sea Company in June, July, and August 1720 with repeated attempts to raise cash through new issues of stock. In 1881 more than 125 new issues with a market value of 5 billion francs were sold in Paris when France’s annual savings were estimated at 2 billion francs.²⁸ Nor was this an era when private companies were going public in large numbers, as in the late 1880s in Britain and the late 1920s in the United States; in both cases there was no need for an increase in savings because the newly issued publicly owned shares were exchanged for privately owned shares.

The end of a period of rising asset prices leads to distress whenever a significant number of investors have based their purchases of these assets on the anticipation that these prices will continue to increase. Some of these investors then may have a ‘negative carry’ because the interest rates on the funds borrowed to buy the assets exceed the cash income on the assets; these investors anticipated that they would be able to use the increase in the value of the assets as collateral for new loans that would provide them with some of the cash needed for interest payments. When asset prices stop increasing, these investors are shunted into distress mode since they have no ready way to get the cash they need to pay the interest on their indebtedness.

Distress enveloped the market for subprime mortgages toward the end of 2006 and 2007 when real estate prices began to fall. The hallmark of these mortgages is that the homebuyers’ equity often was less than ten percent; when prices began to fall, they then had ‘upside-down mortgages’ – their mortgage indebtedness was greater than the market value of the properties. One response to the distress was ‘jingle mail’ – to ship the keys to the properties to the lenders.

Distress may arise from an increase in the flow of funds from the country – a bad harvest may require an increase in imports and an increase in interest rates in a major international financial center may attract funds from domestic financial markets. Credit in the domestic credit market might become less readily available – tighter – because of a reduction in the reserves of the banking system.

The money outflow may be potential; there was distress in the London money market in 1872 when the French payment of reparations to Prussia meant that the Reichsbank acquired substantial money market securities denominated in the British pound that could be readily converted into gold. Similarly there was a flow of money to London prior to April 1925 in anticipation that the British pound would appreciate to its pre-war gold parity; once the pound was again pegged to gold, the owners of these deposits had modest incentives to keep them in London. The Bank of England’s interest rate policy was constrained by the concern that some of these funds might leave London. The Bank of France acquired large British pound balances in the effort to limit the appreciation of the French franc after the successful monetary stabilization at the end of 1926; the likelihood that the French might use part or all of these balances to buy gold was a bargaining chip that increased nervousness in London.

The essence of financial distress is loss of confidence. What comes next – belief in the future as imbalances in the economy are corrected, or runs on banks, panics, and the collapse of prices?

The issue is concisely posed by James S. Gibbons:

Bank officers are not always insensible to alarm when respectable merchants, failing in their best endeavors, are driven into a corner and assume an air of desperation. They know the danger that hangs over the market. Credit is prodigiously extended; the public excitement is wrought up to a high pitch of apprehension, and there need be but a single failure of a ‘great house’ to explode the ‘mighty bubble.’ Who knows that it is a bubble? Who knows that the highest point of pressure is not reached today, and that tomorrow the waters will not begin to subside? And then gradually things fall into their old channels, confidence revives and it is proved that there was no bubble to burst after all.²⁹

How long does distress last?

Financial distress may subside as in France in 1866 and in Britain in 1873 and 1907 or a panic may follow the distress. In the United States there was great concern over the failed attempt by Bunker Hunt to corner the silver market in 1979, the failure of the Continental Illinois Bank in 1984, and the debacle of Long-Term Capital Management in 1998. There were extended periods of distress after August 1982 over bank loans to Mexico, Brazil, and other developing countries and over loans by the thrift institutions in Louisiana, Oklahoma, and Texas to borrowers who had based their exploration activities on the prospect that the oil price would increase to $80 per barrel. After the collapse of hundreds of US banks and thrift institutions at the end of the 1980s, the US Resolution Trust Corporation (RTC) acquired tens of billions of dollars of real estate that had been the collateral for mortgage loans that were in default. Eventually these properties were sold to the public but the uncertainty about the values of these properties depressed values.

Similarly there was distress in Tokyo for an extended period in the 1990s; in any ‘mark to market’ test most of the large Japanese banks were bankrupt and yet there were no runs on the banks because depositors understood that they would be made whole by the government if any of the banks closed. The Japanese government’s policies toward these failed institutions were a cause of distress: would the government close these failed banks or would the government provide new capital to them on favorable terms?

The sharp decline in stock prices on Monday 19 October 1987 proved to be a correction rather than a panic because it did not spread to other US markets, although there were nearly simultaneous sharp declines in most other national stock markets. Distress lasted for several weeks while investors waited to see whether the decline in stock prices would have significant impacts on other markets.

The debacle of Long-Term Capital Management in the summer of 1998 occurred at the same time as the collapse of the Russian banking system and the ruble. The prospect of the impending disaster in Russia induced changes in interest rate and yield relationships that contributed significantly to the collapse of LTCM. LTCM was an unregulated bank even though it was usually considered a hedge fund. LTCM was considered a ‘very smart’ financial institution; two Nobel laureates in finance were among its top officers. The firm was much more highly leveraged than traditional banks and most other hedge funds and in addition it had tens of billions of positions in derivatives contracts like futures and options. Initially LTCM had been viewed as a money machine, and extremely clever in taking advantage of small deviations in prices of very similar securities. For example, the thirty-year US Treasury bond was extensively traded but the twenty-nine-year bond was less liquid and not extensively traded, so the interest rate on the twenty-nine-year bond was modestly higher than that on the thirty-year bond. LTCM bought hundreds of millions of dollars worth of the twenty-nine-year bond and sold short more or less the same amount of the thirty-year bond to profit from the interest rate differential. The excess of the interest rate earned on the bond with the twenty-nine-year maturity over the interest rate on the bond with the thirty-year maturity was modest, but the product of this small number and the position of hundreds of millions of dollars was large – and very profitable.

Some of the major banks that were large lenders to LTCM tended to mimic some of its portfolio positions. The positions of LTCM and these major banks in some securities dominated the markets for these securities.

In the spring of 1998, LTCM had a long position in emerging market bonds that it hedged by shorting US Treasury bonds. As investors became increasingly apprehensive about Russia’s financial future, the prices of emerging market bonds declined as the contagion effect cascaded. The Federal Reserve responded with greater monetary ease and the prices of US Treasury bonds increased. LTCM lost money on both legs of its hedge, which eroded its capital base. As the prices of these emerging market bonds declined, LTCM was between the proverbial ‘rock and a hard place’; if it sold any of its holdings of individual securities, their prices would fall further and its net worth would decline even more rapidly.

The Federal Reserve was concerned that if LTCM failed there would be an extended period of significant uncertainty – distress – in the capital markets while the firm’s positions in futures, options, and other derivatives were unwound and bond prices would fall further. The Fed used its muscle – more precisely the threat of its regulatory muscle – to induce the major banks that were lenders to LTCM to invest their own money in LTCM, and they then acquired 90 percent ownership.

The crash or panic that follows financial distress may do so immediately, in a matter of weeks, or with a delay of several years. John Law’s system peaked in December 1719 and collapsed in May 1720 – five or six months from glory to disaster. In the South Sea Bubble of 1720, the lunatic note sounded clearly at the end of April, the ugly drop in the market occurred in August, and the collapse came in the first days of September. In 1763 distress developed in March while the actual crisis precipitated by the failure of DeNeufville in Amsterdam occurred in July. In 1772 the Bank of England raised its discount rate early in the year; the Ayr Bank curtailed its operations in May, but it was too late. Fordyce absconded on 10 June and the news precipitated a panic in Britain on 22 June; the consequent distress in Amsterdam lasted until the failure of Clifford & Co. in December.

Timing of crises from 1789 to 1815 was dominated by individual apocalyptic events, including the guillotining of Louis XVI in January 1793 (losing one’s head always is apocalyptic), the landing of the French army at Fishguard on the southwestern tip of Wales in February 1797, and the penetration of the continental blockade in 1799. The periods of distress on these occasions were short because the panic was virtually immediate. In 1809–10 the setback arose from a tightening of the continental blockade and overtrading in exports to Brazil. Pressure mounted slowly from the middle of 1809, then more rapidly from mid-1810 to the climax of bankruptcies in January 1811.

Calls for further payments on subscriptions to the shares in railroads in January 1847 set the background of tension against which speculation in grain peaked in May, collapsed in August, and led to panic in November. The crisis of 1866 was the delayed result of the 1864 collapse of cotton prices that had brought panic to France in that year. Britain had two ‘critical moments’ in 1864, one in January – the real crisis that was related to the collapse in cotton prices – and another in the last quarter.³⁰ British treatment of the period tends to look more to speculative expansion that affected the discount houses and that started in the previous year, and to a series of firms that resembled the Credit Mobilier and were using the funds received from selling new share issues to buy back their own shares to pump up the investors. W.T.C. King wrote that one Albert Gottheimer used the name Albert Grant to float the imposing Credit Foncier and Mobilier of England, which ultimately achieved a paid-up or rather called-up capital of £1 million.³¹ The conversion of the discounting house of Overend, Gurney & Co. to a public company in July 1865 at the peak of the boom and ‘dividend race’ led to a 100 percent premium on the stock in October. The Bank of England responded by raising its discount rate from 3 percent to 7 percent; the crash did not occur until May 1866. The distress in Britain lasted seven months, from October 1865 to May 1866, while the distress in France lasted nearly thirty months.

Periods of financial stringency and crisis and panic (in the United States) occurred in the autumn when Western banks drew large sums of money from the East to pay for shipments of cereals.’³² Credit demand peaked in the autumn when banks in the grain-producing areas of the country borrowed large amounts of money from banks in the East that would be used to pay for the purchases and shipments of cereals. Sprague noted that the crisis of 1873 came in September because of the early harvest, that the outbreak of a crisis invariably came as a surprise to the business community and that the crisis of 1873 was not an exception.³³ The seasonal tightness of money was well known and the puzzle is that it was a surprise. The ‘excessive tightness’ of money from September 1872 to May 1873 caused the railroads to borrow short-term funds rather than issue bonds, which could have been seen as a sign of distress, and then the seasonal tightness precipitated the crash.³⁴

Distress may be continuous or it may oscillate in its own rhythm. The crash of the Union Générale in January 1882 was preceded by three separate tense periods, in July, October, and December 1881.³⁵ The panic of October 1907 was anticipated (although Sprague indicated its exact timing was not foreseeable) and preceded by a ‘rich man’s panic’ in March when Union Pacific stock, the security most widely used as collateral for finance bill operations, dropped 50 points.³⁶ Markets recovered from this blow and from the failure of an offering of New York City 4 percent bonds in June (only $2 million was tendered for an offering of $29 million of 4 percent bonds) and from the collapse of the copper market in July, and from the $29 million fine levied against the Standard Oil Company for antitrust law violations in August – only to succumb to the failure of the Knickerbocker Trust Company in October.³⁷ In 1929 distress lasted from June to the last week in October.

Financial distress in Japan began at the start of the 1990s and continued throughout that decade and into the next. Japanese industrial firms were extremely reluctant to downsize and make the other adjustments that were necessary to bring their costs below their current revenues; in the previous forty years these firms had been able to rely on the banks for the cash to finance their operating losses and investments. Japanese banks were extremely reluctant to stop making new loans, even to enterprises that would be considered bankrupt in any ‘mark to market’ test; and the regulatory authorities were reluctant to close banks that would be considered bankrupt. Traditionally the risk of financial losses in Japan has been ‘socialized’; Japanese society prefers to distribute these losses among taxpayers rather than to impose the costs of adjustment associated with closing failed enterprises on their employees.

Argentina experienced an extended period of distress at the end of the 1990s and in 2000 before its currency collapsed in January 2001. At the end of the 1980s Argentina had suffered through two years of hyperinflation. The incoming government of President Carlos Menem pegged the new Argentinean peso to the US dollar at the rate of one Argentinean peso to one US dollar; at the same time Argentina adopted a currency board arrangement that meant that its central bank would increase the supply of peso liabilities only if its holdings of US dollars increased – more or less a very strict application of the Currency School doctrine. During the 1990s, the tax revenues of the Argentinean government were smaller than its expenditures and the excess of expenditures was financed partly with the receipts from privatization and partly by government borrowing. The hyperinflation of the 1980s had sharply reduced the real value of the Argentinean government debt and so investors purchased Argentinean government bonds denominated in the US dollar because the government looked like a good credit risk. As the ratio of Argentinean government debt to the country’s GDP increased, the interest rates that the Argentinean government had to pay to sell new bonds rose. There was a recession in Argentina toward the end of the decade, in part because the US dollar (to which the peso was pegged) was appreciating; and the Argentinean fiscal deficit increased as its tax receipts declined relative to its expenditures. The recession was intensified by the depreciation of the Brazilian real in January 1998. Brazil was Argentina’s major trading partner. The policy question was whether Argentina would be able to reduce its fiscal deficit while maintaining the established parity. (The Argentineans do not have an established tradition of paying taxes so tax rates tend to be high and tax collections low, while government salaries tend to be high and the performance of government officials low.) The efforts to raise taxes and cut government expenditures led to a series of political problems; Argentinean citizens were extremely reluctant to pay more taxes when the economy was doing poorly. The country was moving in slow motion toward a disaster. In the end Argentina both devalued and defaulted on the government debt.

Suppose the monetary authorities tighten credit to raise the cost of speculation. When commodity and asset markets move together, up or down, the direction that monetary policy should take is clear. But when share prices or real estate or both soar while commodity prices are stable or falling the authorities face a dilemma. The Federal Reserve encountered this dilemma in the 1920s; President Benjamin Strong of the New York Fed had agonized over the appropriate policy in 1925 and again in 1927. The dilemma is that the policymakers cannot kill two birds with one stone, or more precisely they cannot achieve two policy targets with one policy instrument, or in what is perhaps a better metaphor, it is difficult to pick off a target if it is standing next to another one that one wants to leave untouched and the weapon is a shotgun rather than a rifle. At the same time, any tightening of credit to dampen the boom in real estate would be likely to stall a remarkable economic expansion.³⁸

Fed Chairman Greenspan was concerned that US stock prices were too high or increasing too rapidly when he remarked about ‘irrational exuberance’ in December 1996. The Fed proved reluctant to raise interest rates to dampen the increase in stock prices because of the negative impacts on economic growth and employment. In 1999 the Fed became concerned (obsessed?) with the Y2K problem, the likelihood that US computer systems would collapse because so many software programs were not designed to recognize the transition to 2000. In the last several months of the year the Fed provided the monetary system with abundant liquidity to forestall problems associated with the end-of-the-millennium transition and in the meantime the money – which had to go somewhere – fed stock market speculation.

Onset of a crisis

Students of logic have discussed the damp squib thrown by A that lands at B’s feet which is then thrown from B to C and from C to D and so on only to explode after Y throws it in Z’s face. Who is to blame? A, causa remota? Or Y, causa proxima? Causa remota of any crisis is the expansion of credit and speculation while causa proxima is some incident that saps the confidence of the system and induces investors to sell commodities, stocks, real estate, bills of exchange, or promissory notes and increase their money holdings. The causa proxima may be trivial: a bankruptcy, a suicide, a flight, a revelation of fraud, a refusal of credit to some borrowers, or some change of view that leads a market participant with a large position to sell. Prices fall. Expectations are reversed. The downward price movement accelerates. To the extent that investors have used borrowed money to finance their purchases of stocks and real estate the decline in prices is likely to lead to calls for more margin or cash and to further liquidation of stocks or real estate. As prices fall further, bank loan losses increase and one or more trading firms, banks, discount houses, or brokerages fail. The credit system appears shaky, and there is a race for money and liquidity.

Identifying the original sellers is difficult. Conspiracy theories abound. One can single out bear speculators like Joseph P. Kennedy, Sr, or Bernard Baruch in 1929; the Protestant-Jewish cartel that allegedly did in Eugène Bontoux in 1882; or Thomas Guy, who liquidated £54,000 of South Sea stock over six weeks between April and June 1720, never selling more than £1000’s worth at a time. (He used his fortune to endow Guy’s Hospital in London, ‘the best memorial of the Bubble’.)³⁹

Someone sells. Occasionally it is a foreigner. In 1847, for example, the French (according to one S. Saunders, quoted by Evans) bought up surplus wheat and sent it to Britain in June and July where it was sold at prices much below the then-prevailing market prices, which fell from 96 shillings to 56 shillings per quarter (a quarter is equal to eight bushels) and brought about the bankruptcies of a large number of firms connected with the corn trade.⁴⁰ The story is not persuasive. The price of wheat had risen from 46 shillings in August 1846 to 93 shillings in May 1847 because of violent storms that ruined the crop and because of the potato blight in Ireland and on the Continent. The price dropped in July 1847 with favorable weather and the prospect of a good crop. Imports of wheat and flour rose from 2.3 million quarters in 1846 to 4.4 million in 1847, aided by the repeal of the Corn Laws;⁴¹ the 70,000 quarters are a trivial proportion of this sum. In 1846 France had its smallest crop of wheat in 100 years (a problem exacerbated by the potato crop failure); in 1847 the crop was the largest for 100 years. But the condition was general, and British wheat speculation had been excessive.

One view is that the Baring crisis of 1890 was triggered by the sale of Argentinean bonds by German investors, who had stopped buying these bonds two years earlier either because of general uneasiness,⁴² or because they were concerned about the instability of the Argentinean currency,⁴³ or because the domestic boom led them to sell other foreign bonds including Russian bonds.⁴⁴ German sales of the Argentinean bonds contributed to distress rather than to a crisis, since British investors then acquired more of the Argentinean bonds. In November 1888 a £3.5 million offering of the Buenos Aires Drainage and Waterworks Company failed, and Baring felt obliged to lend to Argentina through acceptance credits. Declining primary product prices in 1890 meant that the Argentine government did not have the money to repay these loans as they came due. The Baring crisis of November 1890, after two years of distress, resulted from a warning from the Bank of England to Baring Brothers to limit the level of its acceptances (which stood at £30 million in the summer of 1890), from the crisis in New York in October, and from the maturing of £4 million of acceptances in November.

New information may precipitate a crash, as when it was revealed that the Paris-Lyons-Marseilles railroad would cost 300 million francs instead of the projected 200 million.⁴⁵ Causa remota, and much more important, were the large balance of payments deficits from extensive imports of railroad materials and, especially, the crop failure of 1846 followed by the glut of 1847. The Granger movement helped precipitate the collapse in the United States in 1873. The Grangers, who in some ways resembled the environmentalists of today, started in the late 1860s and early 1870s as activists for legislation that would control intrastate transportation by prohibiting discriminatory freight rates, establishing regulatory commissions, and setting maximum freight rates.⁴⁶ A very large volume of railroad securities had been sold on credit – including a number of ‘superfluous and ridiculous’ enterprises like the Rockford, Rock Island and St Louis line which had been sold at par and then declined to 6 cents on the dollar as the prospect of local control of freight rates put an end to optimism and triggered sales and then liquidation of these bonds.

One ‘accidental’ detonator of a crisis has been the sinking of a ship. In 1799, when interest rates ranged between 12 and 14 percent and the price of sugar was 35 percent below the peak before the convoy had broken through the blockade, British merchants shipped £1 million on the frigate Lutine destined for Texel in an attempt to assist in the Amsterdam crisis. The ship sank in a storm off the Dutch coast and the hope of alleviating the crisis was dashed.⁴⁷ During the 1857 crisis in New York the news that the steamer Central America, bound from Panama to New York with a cargo including $2 million in gold, was overdue came at the time of extreme distress in Philadelphia, Cincinnati, and Chicago. Two days later it was learned that the ship had gone down, uninsured.⁴⁸

An accident may precipitate a crisis, but so may action designed to prevent it – or action by the authorities adopted to achieve other objectives. The matter was well put by H.S. Foxwell apropos the crisis of 1808–9:

To refuse accommodation altogether is always held to be dangerous. To make personal reference is invidious, especially for a National Bank. It is just possible that the Bank might have resorted to the expedient used in 1795–96, I mean the granting of pro rata discounts ... [In seeking to contract the circulation] it must have put severe pressure on the market and risked the creation of a panic ... The Bank was responsible for the solvency of this crowd of small, ill-managed institutions [country banks], but dared not call them to account, on peril of provoking a general collapse of credit.⁴⁹

Foxwell posed the dilemma neatly. Not to apply discipline will let the credit market expand to dangerous levels; to apply discipline may prick the bubble and induce collapse.

Policies adopted to deal with a crisis often encounter lags. Raising the discount rate in the face of an external drain of cash may induce a return flow. An increase in the discount rate of the Bank of England to 10 percent can ‘draw gold from the moon’ in the folklore of the City, but how long would it take to produce that result? The issue was debated between the Banking and the Currency schools in the context of the Bank Act of 1844 and the need either to suspend the Act or for the Bank to act as a lender of last resort. In 1825 and again in 1836, speculation in boom conditions led to an outflow of gold and financial stringency. On one interpretation, the boom broke before the Bank of England belatedly raised its discount rate in an effort to reduce its liabilities; thus, a combination of tight money and declining commodity prices produced the crisis and induced the Bank to reverse course and to reduce interest rates.⁵⁰ The Banking School believed that the increase in the discount rate produced an immediate inflow of money. The Currency School, on the other hand, had two wings; one believed there would be an immediate return flow and the other, represented by Lord Overstone, thought bank rate would have an impact only after a lag so that a lender of last resort would be required to fill the gap.⁵¹

Hawtrey pointed to a lag at the level of commercial banks and internal drains, based on backlogs:

Bankers may take proper steps, but panic because they work slowly: They may have really checked the fundamental danger of the position ... stopped the stress of new orders ... and yet the demand for fresh credits and the drain of cash may go undiminished. The consequence may be a state of panic among the bankers, who, unaware of the cause of the apparent ineffectiveness of the measures they have taken [the working off of the backlog of old orders], despair of saving themselves from failure, call in existing loans regardless of the embarrassment of debtors, and precipitate a series of bankruptcies among their customers and themselves.

The fact is that there is no golden rule for keeping the extension of credits within bounds.⁵²

Apart from lags and mistakes in discount policy, the authorities may precipitate a panic by brusque action in the early stages of distress. In the summer of 1836, with credit extended in acceptances drawn by American houses on British joint-stock banks, the Bank of England then refused to discount any bills that carried the name of a joint-stock bank and specifically instructed its Liverpool agent not to rediscount any paper of the three so-called ‘W banks’ (Wiggins, Wildes, and Wilson) among the seven American banks in Britain, an action that ‘seemed vindictive’⁵³ and led immediately to panic.⁵⁴ The Bank of England had to reverse its policies. It had long conferences with the ‘W banks’ in October, extended credit to them in the first quarter of 1837, but was unable to prevent their failure in June of that year. The Bank acted to dampen the extension of too large an amount of credit. But credit is delicate; expectations can quickly be altered.

Panic in the form of a run on a bank or banks has usually been started by small depositors as happened in the 1980s in Ohio and Maryland and Rhode Island, where some of the state-chartered banks relied on state-chartered insurance arrangements because they charged lower premiums than the FDIC. (All federally chartered banks and thrifts were required to participate in the federal government’s deposit insurance programs.) In contrast stock market panic often resulted from selling by big-money insider speculators or institutional investors such as mutual funds, pension funds, and insurance companies, perhaps severally following similar program trading. The run on Franklin National Bank was initiated by other banks, and especially the large money-center banks in New York that refused to take the counterpart of Franklin’s forward foreign exchange contracts or to lend it federal funds or to buy Franklin repurchase agreements, except at rates of interest that reflected deep distrust.⁵⁵ Similarly the run on the Continental Illinois Bank in 1984 was triggered by the reluctance of other large banks to renew maturing deposits in the federal funds market and in the offshore deposit market. Smaller depositors were protected by the FDIC deposit insurance. In the stock market meltdown of October 1987, the Fidelity group of mutual funds of Boston was a large seller in the London market before the New York Stock Exchange opened on the nineteenth. These orders were communicated back to New York, where they had accumulated into a mountain of sell orders by the time the market opened for trading. Fidelity’s sales were a response to the redemptions by the holders of its mutual funds rather than for its own position, although it may have wanted to raise cash in anticipation of future redemptions – and before stock prices fell further.

The clumsiness of the International Monetary Fund triggered a run on many of the banks in Indonesia during the early days of the Asian crisis in 1997. The IMF induced the Indonesian government to take over and support fifteen of the large private banks, in effect guaranteeing their deposits. But the remaining fifty or so other private banks then were placed in a nether-world position and were subject to runs as the depositors scrambled to withdraw funds before the banks collapsed.

Crashes and panics

A crash is a collapse of the prices of assets, or perhaps the failure of an important firm or bank. A panic, ‘a sudden fright without cause’ (from the god Pan, known for causing terror), may occur in asset markets or involve a rush from less liquid securities to money or government securities – in the belief that governments do not go bankrupt because they can always print more money. A financial crisis may involve one or both and in either order. The collapse of the stock of the South Sea Company and the Sword Blade Bank almost brought down the Bank of England. The 1929 crash and panic in the New York Stock Market had adverse consequences in both the commodity and real estate markets, and the seizure in the credit markets led to sharp declines in income and employment and output. But there was no panic in the money market; interest rates did not increase because the Federal Reserve was providing liquidity to the market.⁵⁶ In 1893 lack of confidence in the ability of the United States to maintain the gold standard under pressure from the silver interests led to money market pressure, bank failures, and downward pressure on prices of securities.⁵⁷

The system is one of positive feedback. The debt-deflation cycles involve a decline in prices of assets and commodities which leads to a reduction in the value of collateral and induces banks to call loans or refuse new ones; firms sell commodities and inventories because their prices are in decline, and the decline in prices causes more and more firms to fail. Households sell securities and firms postpone borrowing and investing and prices fall still further. Further declines in the value of loan collateral lead to further liquidation. The failure of firms means that the banks incur loan losses and fail. As banks fail, depositors withdraw their money (this was particularly true before deposit insurance).

Deposit withdrawals require more loans to be called, more securities to be sold. Trading reforms, industrial firms, investors, banks in need of ready cash – their riskiest securities cannot be sold at any price, they are forced to sell their best securities and so their prices decline. The banks may carry the firms, corporations, and households that are known to be in trouble for a time in the expectation or hope that prices will increase again and refloat the frail bark of credit. Bank examiners may in extremis look the other way as banks continue to value loans and securities at cost rather than at market value, extend loans due, or add to loans of delinquent borrowers so they can pay some of the interest. But when bankruptcy occurs, the nettle of bad loans must be grasped. Prices, solvency, liquidity, and the demand for cash – in German Bargeld, in French numéraire – are related. Not only banking institutions: as Sprague stated, households, firms, and banks are ‘very similar to a row of bricks, the fall of one endangering the stability of the rest’.⁵⁸ The metaphor is a cliché, but nonetheless appropriate.

At the height of the panic, money is said to be unavailable. Descriptions are frequently exaggerated, not least about 1825:

Bankers in Lombard Street called on the Governor [of the Bank of England] on Sunday [after the panic of country banks had reached Pole, Thornton & Co. on 12 December] to warn that if such a house, drawn on by 47 country banks, were allowed to stop, a run would take place on every bank in London.

It was allowed to stop. 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.’⁵⁹

This was the occasion when the failure of seventy-three banks brought Britain, according to Huskisson, within twenty-four hours of barter.⁶⁰ ‘It was, as the Duke said of Waterloo, “a damned nice thing – the nearest run thing you ever saw in your life”.’⁶¹ Barter was avoided by exchanging silver for gold with the Bank of France, and by the luck of the Bank of England in finding, just as it ran out of £5 and £10 notes (which were then all it issued), a block of £1 notes left in the vault from 1797. With government approval these were issued on 17 December and ‘worked wonders’.⁶²

In 1857 New York Central stock went from 93 to 61, Reading from 96 to 36.⁶³ The price of pork fell from $24 a barrel to $13; flour, from $10 to $5 or $6.⁶⁴ In September interest rates rose from 15 to 24 percent as 150 banks in Pennsylvania, Maryland, Rhode Island, and Virginia failed in the last four days of the month. The panic reached a peak in October, when 1415 banks in the United States failed and interest rates rose from 60 to 100 percent per annum – for monies borrowed for a few days.⁶⁵ Very high rates of interest, such as 4 percent a day, have sometimes been quoted for a particular kind of loan, as for call money in 1884, when commercial discounts continued at 4.5 to 5 percent a day for first-class endorsed paper;⁶⁶ or 5 percent a day at the peak as a premium for cash at the onset of the panic of 1907.⁶⁷ Perhaps the apogee of the liquidity squeeze was recorded in 1907, when one bank paid $48 per $1000 for the cash gate receipts of the Harvard-Yale football game.⁶⁸ The mild and short recession in 2001 after the massive implosion in US stock prices resulted from the abrupt change in the policy of the Federal Reserve and its rapid and aggressive move to reduce interest rates. The result was a mortgage refinancing boom; millions of individuals refinanced their mortgages at lower interest rates and used some of the cash obtained in the refinancing to buy autos and other consumer durables and to go on vacations. The Fed reduced short-term interest rates to 1 percent and since the inflation rate was nearer 2 percent, real short-term interest rates were negative. The property market boomed; house prices increased sharply in New York, Boston, Washington, Miami, Phoenix, Las Vegas, and Los Angeles. Skeptics wondered whether the deflationary effects of the implosion of the stock price bubble had been largely offset by a bubble in the housing market.

Will the storm subside, the flood crest and fall? Or will boom and crash spread from one market and country to another and the steps taken locally and internationally fail to halt panic and reverse the damage?