4

Fueling the Flames: the Expansion of Credit

Axiom number one: You can’t have inflation without the rapid growth of money.

Axiom number two: You can’t have a real estate bubble without the rapid growth of credit.

Speculative manias gather speed through expansion of credit. Most increases in the supply of credit do not lead to a mania – but nearly every mania has been associated with rapid growth in the supply of credit to a particular group of borrowers. In the last hundred years, the increases in supplies of credit have been provided in part by the banks and in part by the development of new financial instruments. The mania in tulip bulbs in the seventeenth century occurred because the sellers of the bulbs provided credit to the buyers.¹ John Law had his Banque Générale, later the Banque Royale, as his source of credit while the South Sea Company relied on the Sword Blade Bank. In 1763 credit expansion in Holland was financed by the Wisselruiti, or chains of accommodation bills (think post-dated checks) from one merchant to another. The British canal mania of 1793 was fed by spending facilitated by loans from many newly established country banks to the entrepreneurs who were developing the canals.

In many cases the expansion of credit resulted from the development of substitutes for more traditional monies. In the United States in the first part of the nineteenth century, the innovation was that bills of exchange (again think post-dated checks) replaced silver in the triangular trade among the United States, China, and Britain. The United States had a bilateral trade deficit with China and China had a bilateral trade deficit with Britain. Previously the United States bought silver from Mexico which was then shipped to China to finance the US trade deficit; then the silver was shipped to Britain to finance China’s trade deficit. The institutional innovation was that American merchants sent bills of exchange that specified payment in the British pound for China’s imports; the Chinese then shipped these bills to Britain to finance their trade deficit. The transactions costs involved in making cross-border payments using bills of exchange were much smaller than those that involved the shipment of silver. The result of this innovation was that the silver stayed in the United States so that the US money supply did not decline.²

The global boom of the 1850s followed from the combination of new gold discoveries, the formation of new banks in Britain, France, Germany, and the United States, the establishment of clearing-houses by the banks in New York and in Philadelphia and the expansion of the London clearing-house. The growth in bank clearing-houses led to the increased use of credit in the transactions among the banks which were its members; payments imbalances among them were settled by transfer of clearing-house certificates – a new form of ‘money’. The expansion of credit in Britain in 1866 resulted from the increase in loans by the newly formed joint-stock discount houses. The boom in Central Europe in the 1870s was based on the gold reparations payments from France to Prussia and on the creation of Maklerbanken (brokers’ banks) in Germany, which spread into Austria, and Baubanken (construction banks) in Austria, which expanded into Germany.

One of the many different institutional avenues for the expansion of credit that occurred in France in 1882 was based on a system of fortnightly clearing of stock exchange transactions that provided credit to speculators through a system of delayed payments called reportage. The buyers of stocks had up to fourteen days before they had to pay for their purchases; in effect they got interest-free loans until the date of payment (although the value of the loans may have been reflected in the prices paid for the stocks).³ Similarly the expansion of credit in the call-money market in New York helped finance the stock market boom in the late 1920s. The catalyst for the expansion of credit in the United States in 1893 was the addition of silver coins to the US money supply; in 1907, the increase in the supply of credit resulted from the increase in the loans by the trust companies. In the years before and after World War I, the development of the gold exchange standard led to an expansion of the international credit base so that a larger volume of international trade could be financed with the existing stock of monetary gold. The rapid increase in installment credit in the United States in the 1920s contributed to the surge in automobile ownership (the dramatic increases in the number of automobiles and other consumer durables led to rapid growth in demand for staggered payments arrangements).

After World War II, the development of negotiable certificates of deposit (CDs) contributed to the expansion of credit. In the 1870s the Austrian banks had developed a new financial instrument (the Cassenscheine) that paid interest; the expansion in the demand for these instruments led to an increase in the supply of credit and hence to an increase in spending. In the 1950s and the 1960s the large US banks adopted the practice of liability management which meant that the growth of their loans no longer depended on the increases in their deposits; instead they funded their loans by borrowing in the inter-bank money market. Liability management enabled the banks to be much more aggressive in managing the growth of credit.

One unique form of the expansion of ‘bank credit’ occurred in Kuwait between 1977 and 1982 when shares and real estate were bought and sold on the Kuwaiti Souk al-Manakh (stock exchange) with post-dated checks. Eventually the value of these checks in circulation increased to billions of dinars – nearly US$100 billion. The values of post-dated checks written by the buyers of the shares and real estate were much, much larger than their bank deposits. The sellers of the shares and real estate increased their spending as their wealth increased; they hoped that there would be money in the bank accounts of the payors when the due dates of the checks that they had received for the sale of shares and real estate arrived. In July 1982 the checks bounced when some sellers of stocks tried to collect on checks on the due dates.⁴

These examples suggest that the expansion of credit is a systematic development that has continued for several hundred years as the participants in financial markets seek to reduce the costs both of transactions and of holding liquidity and money balances. The form each event takes may seem accidental – the substitution of bills of exchange for silver in payments to China, or the development of deposits in the Eurocurrency market because ceilings prevented US banks from increasing the interest rates that they could pay on deposits in New York and other US cities. The development of new credit instruments occurs episodically in response to changes in institutional arrangements. Monetary expansion is systematic and endogenous rather than random and exogenous.

During economic booms the money supply defined as means of payment has increased and the existing stock of money has been used more efficiently to finance both increases in economic activity and the purchases of real estate and securities. The efforts of central bankers to limit and control the growth of money have been offset in part by the development of new and very close substitutes for money. Such efforts have a long history, including the resumption of specie payments and the return to convertibility of national currencies into gold after the end of wars. The demonetization of the lesser metallic monies – initially copper was displaced by silver and subsequently silver’s monetary role was eclipsed by gold – was an effort to obtain greater control over the money supply. Central banks wanted a monopoly on the issue of currency notes, which restricted and then eliminated the rights of private, country, and joint-stock (corporate) banks to issue currency notes. Legislation and custom limited the amounts of deposit money that could be issued against primary bank reserves, starting shortly after the Bank Act of 1844 and continuing through the application by the Federal Reserve System of reserve requirements on both demand and time deposits (as embodied in the Federal Reserve Act of 1913) and then on certificates of deposit and subsequently on the borrowings by US banks from their branches in offshore financial centers including London. The process is Sisyphean, a perpetuum mobile; the history of money is a story of continuing innovations so that the existing supply of money can be used more efficiently and the development of close substitutes for money that circumvent the regulatory requirements applied to the creation of money.

The Eurocurrency deposit market surged in the 1960s as an end-run around the regulations imposed on US banks by the Federal Reserve and the Federal Deposit Insurance Corporation; the US dollar deposits produced by the branches of US banks in London and other offshore centers were not subject to interest rate ceilings, reserve requirements, and deposit insurance premiums. The US stock brokerage firms developed money market funds that were a close substitute for bank deposits in the 1970s and paid interest on these deposits (the deposits were not guaranteed by any agency of the US government – at least not until the financial crisis in the autumn of 2008).

Currency school vs banking school

One feature of the history of monetary theory is a continuing debate between two different views – the Currency School and the Banking School – about how to manage the growth of the money supply. The proponents of the Currency School advocated a firm limit on the expansion of money to avoid inflation. The adherents of the Banking School believed that increases in the supply of money would not lead to inflation as long as they corresponded with increases in business transactions. The same distinction in ideology and economic analysis separated the hard money school from the populists in the 1890s – the former were concerned with inflation while the latter believed prices would not change as long as economic activity increased. The debate between these two views of managing the growth in the money supply has continued for at least three hundred years.

The Currency School wanted a simple rule that would fix the growth rate of the money supply at 2 or 4 or 5 percent, much like today’s monetarists.⁵ Viner’s discussion of the nineteenth-century controversy is succinct:

The currency school tended also to minimize or to deny the importance of bank credit in other forms than notes as a factor affecting prices, or as in the case of Torrens, to claim that the fluctuations in the deposits were governed closely by the fluctuations in the note issues. They had a hankering also for a simple automatic rule, and could find none suitable for governing the general credit operations of the Bank. They also had laissez faire objections to extending legislative control of the banking system any further than seemed absolutely necessary.⁶

Neither the Currency School nor the Banking School paid much attention to the expansion of non-bank credit. The Bank of Amsterdam, founded in 1609, issued notes against deposits of precious metals; in effect these notes were warehouse receipts and the amounts of these notes were tied one-for-one to the deposits of the metal. Initially the Bank of Amsterdam did not expand credit; subsequently in the eighteenth century the bank increased its loans in the effort to rescue the Dutch East India Company during the fourth Anglo-Dutch War. The Bank of Amsterdam was also a Wisselbank where bills of exchange (Wissel in Dutch, Wechsel in German) were paid. Merchants kept deposits at the Bank of Amsterdam to meet bills presented for collection. Deposits of precious metals enabled the Bank of Amsterdam to earn seignorage – a type of profits – when it produced coins and it paid a low interest rate on deposits. In 1614 the Municipality of Amsterdam established a Bank of Lending (Huys van Leening), that enabled merchants to establish their own credit efficiently but it was not an active lender.⁷ This credit created by the merchants led to an excessive expansion of the Wisselruiti; when the chain of bills of exchange broke in 1763 because one of the merchants did not have the money to pay on a maturing bill, the DeNeufville bank failed.

The Swedish Riksbank, established in 1668, had two departments, a Bank of Exchange patterned after the Bank of Amsterdam (Wisselbank) and a Bank of Lending (Liinebank).⁸ These two departments foreshadowed the Bank Act of 1844 in Britain, which was a compromise between the two schools; an Issue Department would provide bank notes against deposits of coin or bullion above a specified fiduciary issue that represented the Bank of England’s holdings of British government debt and a Banking Department that would make loans and discounts up to a multiple of its reserves of bank notes that had been produced by the Issue Department. The establishment of the Issue Department was a victory for the Currency School, which had criticized the loans made by Bank of England after the suspension of the gold standard in 1797. (The defense of the Bank of England was that these loans did not lead to an increase in the inflation rate because they financed trade.) The establishment of the Banking Department was a victory for those who believed that an expansion of credit would help finance the initial upswing in the early stages of an economic recovery.

The view of the Currency School that the expansion of credit would eventually lead to inflation was correct. The Banking School’s view that an increase in the supply of credit was needed at the start of an economic expansion was also correct. The Currency School’s view that the discounts be limited to acceptances that were related to actual commercial transactions became known as the ‘real bills doctrine’. The larger the number of business opportunities, the greater the scope for discounting, and the larger the increase in the money supply and eventually the higher the inflation rate. The central policy questions, once an expansion of credit has started, are whether it is practicable to decree a stopping place and whether this limit could be determined by an automatic rule.

The core issue is that it is easier to define money than it is to measure the effective money supply. Walter Bagehot wrote ‘Men of business in England do not ... like the currency question. They are perplexed to define accurately what money is: how to count they know, but what to count they do not know.’⁹

The stylized fact is that every time the monetary authorities stabilize or control some quantity of money, M, either in absolute volume or at a predetermined rate of growth, more of the money and the near-money substitutes will be produced in periods of euphoria. If the definition of money is fixed in terms of designated liquid securities the euphoria may lead to the ‘monetization’ of credit in ways that are beyond the definition; the velocity of money (velocity is defined either as total spending or national income divided by the money supply) will increase even if the amount of money defined in the traditional way remains unchanged. The debate was over whether money should be defined as M₁, currency plus demand deposits; M₂, equal to M₁ plus time deposits; M₃, consisting of M₂ plus highly liquid government securities; or some other designation.

The process seemed endless; fix any M_i and in economic booms the market will create new forms of money and near-money substitutes to get around the limit. The Radcliffe Commission in Britain in 1959 claimed that in a developed economy there is ‘an indefinitely wide range of financial institutions’ and ‘many highly liquid assets which are close substitutes for money, as good to hold, and only inferior when the actual moment for a payment arrives’. The Radcliffe Commission did not use the concept of velocity of money because it ‘could not find any reason for supposing, or any experience in monetary history indicating, that there is any limit to the velocity of circulation’.¹⁰ The commission recommended that a complex of controls of a wide range of financial institutions be developed as a substitute for the traditional control of the money supply: ‘Such a prospect would be unwelcome except as a last resort, not mainly because of its administrative burdens, but because the further growth of new financial institutions would allow the situation continually to slip out from under the grip of the authorities.’¹¹

Economists have debated which assets should be included in ‘money’ for more than two centuries. One view is that the most appropriate definition is the one that provides the strongest correlation with changes in economic activity. Measuring economic activity is relatively unambiguous. The identification of the monetary variables that have the highest correlation with the economic activity variable might change over time and differ across countries. ‘In common parlance, bank currency means circulating bank notes – “paper money.” Yet some writers include checks and promissory notes, if not also loans and deposits’ (italics in original) in their measure of money.¹²

The debate was neatly summarized by John Stuart Mill:

The purchasing power of an individual at any moment is not measured by the money actually in his pocket, whether we mean by money the metals, or include bank notes. It consists, first, of the money in his possession; secondly, of the money at his banker’s, and all the other money due him and payable on demand; thirdly of whatever credit he happens to possess.¹³

One approach is to include all assets in the measurement of the money supply. But in theory the question is the one raised by Mill – how much credit can households, firms, and governments command at a given time? The amounts are certain to vary because access to credit depends on satisfying certain conditions, which is likely to be easier in euphoric periods than at other times.¹⁴ But in theory the analyst wants to know, with Mill, what credit a household, firm, or government would be able to command at a given time, and the amount is almost certain to vary over a wide range because the access to credit depends on satisfying certain conditions and both households and firms are better able to satisfy these conditions in euphoric periods. Banks and other lenders have often extended credit lines to firms and household borrowers, but the amount of credit available under the lines at each moment may require that the borrowers satisfy specific requirements.

Consider the rapid growth of US dollar deposits in London and other offshore banking centers in the 1960s, the 1970s, and the 1980s, which was a response to the increases in interest rates on these deposits relative to interest rates on bank deposits in New York and Chicago and Los Angeles that were subject to regulatory ceilings. The banks that sold these offshore deposits used the funds to make US dollar loans to American and non-American firms that they might otherwise have made from one of their US offices. The firms that borrowed US dollar funds from the offshore banks in London were as likely to spend these funds in the United States as if they had borrowed the US dollar funds in New York or some other US city. Should the US dollar deposits produced in London and other offshore banking centers be included in the US money supply?

The home-equity credit line is a recent financial innovation; banks and other lenders offer to lend homeowners an amount that may be equal to the value of the equity in their homes or in some cases a modestly larger amount. (At an earlier period the loans that used the equity in the home as collateral were known as second mortgages; a home-equity credit line represents potential borrowing until the homeowner draws on the line, while the second mortgage was an actual loan.) The availability of home equity credit lines means that homeowners economize on their holdings of money and near-monies; households engage in liability management of the type that banks developed thirty and forty years earlier, and the increase in these lines leads to higher levels of spending with the same money supply.

The purchasing power of the household cannot readily be extrapolated to that for a country since an increase in the amount of credit extended to one individual may or may not subtract from the amounts of credit available to others, depending on both banking institutions and on the scope of euphoria. As one novelist wrote about credit:

Beautiful credit! The foundation of modern society. Who shall say this is not the age of mutual trust, of unlimited reliance on human promises? That is a peculiar condition of modern society which enables a whole country to instantly recognize point and meaning to the familiar newspaper anecdote, which puts into the speculator in lands and mines this remark: ‘I wasn’t worth a cent two years ago, and now I owe two million dollars.’¹⁵

The basis for this generalization is the historical development of close substitutes for money that led to increases in the amount of credit and total spending. Consider only bills of exchange, call money, and the gold-exchange standard from a list that also includes bank notes, bank deposits, clearing-house certificates, the liabilities of specialized banks (for example, banques d’affaires, Maklerbanken, or Baubanken), the liabilities of trust companies, negotiable CDs, the Eurocurrency deposits, installment credit, credit cards, and NOW accounts.

Quality of debt¹⁶

The credit-rating agencies were established to rank the quality of the debt of individual borrowers – firms, governments, and even households. Minsky’s taxonomy of corporate debt used a three-part distinction based on the relationship between cash flows to the borrowers from their operating activities and their projected debt-servicing payments. ‘Hedge finance’ occurs when the cash from the firm’s operating activities is larger than the cash needed for its scheduled debt-servicing payments. ‘Speculative finance’ occurs when the cash from the firm’s operating activities is large enough to enable the firm to pay the interest on its debt on a timely basis; however, the firm would need to borrow to get the cash to pay some or all of the principal due on maturing loans. ‘Ponzi finance’ occurred when the cash from the firm’s operating activities is not large enough to pay all of the interest due on debt on a timely basis. The firms involved in Ponzi finance either will need to borrow or they will need to sell assets to get the cash to pay the interest.¹⁷ (This distinction between Ponzi finance and speculative finance is comparable to that used in the public finance literature between a ‘primary fiscal balance’ which involves the relationship between the government’s tax and other receipts and its total payments exclusive of those for interest. A government with a primary fiscal deficit does not have enough money from its fiscal revenues for all of its current scheduled interest payments.)

Minsky emphasized the ‘quality’ of debt to gauge the fragility of the credit structure; the terms ‘speculative’ and ‘Ponzi’ highlight fragility. The implication of the term ‘Ponzi finance’ is that the firm may not be able to make a debt service payment on a timely basis in the absence of a ‘miracle’.¹⁸ An edifice of debt contracted to finance risky ventures is inherently unstable. Buyers of homes in California and Nevada and Arizona who used subprime mortgages were betting the increase in the property prices would be sufficient so they could borrow against the increase in the equity of their homes to get the money to pay the interest. So were their lenders.

The model in the previous chapter emphasizes that in periods of economic euphoria the quantity of debt increases because the lenders and investors become less risk-averse and more willing – or less unwilling – to make loans that had previously seemed too risky. During economic slowdowns, many firms experience less rapid increases in their revenues than they had anticipated with the result that some that had been in the hedge finance group are shunted into the speculative finance group while some firms that had been in the speculative finance group move into the Ponzi finance group.

Bills of exchange

Bills of exchange were claims for future payment written by a seller of goods and were a form of vendor financing.²¹ The seller provided a 90 or 120-day credit to the buyers to facilitate the sales. These bills of exchange were frequently discounted with the banks that provided the holder of the bills with cash in the form of bank notes or coin and then in the nineteenth century, bank deposits. The bills of exchange often were used directly in payment. Once the seller of the goods had received a bill of exchange from the buyer, the seller transferred the bill to someone else in payment. Each recipient of a bill would add its name to the bill, much like endorsing a check; there were five or ten endorsers on some bills. ‘The bill was now money.’ Ashton said that even if some of the parties in the chain of endorsers were of doubtful credit, the bill would still circulate as if it were a bank note.²² In the first half of the nineteenth century, some bills for as little as £10 circulated with fifty or sixty names.

Payment practices differed. Bank notes were disliked in Lancashire, and at the beginning of the nineteenth century coins and bills of exchange were the primary items used for payments.²³ Because of the increase in the use of bills of exchange in payments, Bank of England note circulation declined by £9 million during the economic expansion from 1852 to 1857, even though the demand for money increased. The deposits of five banks in London increased from £18 million to £40 million. The average volume of bills of exchange in circulation, however, expanded from £66 million to £200 million, according to the contemporary estimates of Newmarch.²⁴

Initially the bills of exchange were issued in connection with specific transactions and the amount of the bill matched the exact value of the sale. Subsequently the link between the sale of goods and the issue of a bill of exchange was relaxed. In 1763 in Sweden, Carlos and Claes Grill bills on Lindegren in London could not be identified with particular shipments, which often were made in rapid succession, but were drawn when the firm needed to pay creditors.²⁵ Thus the credit of a firm or individual was gradually separated from particular transactions and the bill had become ‘accommodation paper’ or a promissory note – an IOU.

Some economists were firmly opposed to ‘accommodation paper’ because it was believed to be of lower quality than self-liquidating, commercial bills since there was less assurance that the firms that issued the bills would have the cash to pay the holders when the bills matured.²⁶ In a period of falling prices, however, the merits of the higher quality commercial bills were exaggerated, since the buyers of the goods might not have the cash to settle their obligations because they might not be able to sell the goods at a profit.²⁷ The ratio of the debt to the debtor’s income or wealth is a more meaningful measure of the quality of credit.

The bill of credit, as Franklin said,

is found very convenient in Business; because a great Sum is more easily counted in them, lighter in Carriage, concealed in less Room, and therefore safer in Travelling or Laying Up, and on many other Accounts they are very much valued. The Banks are the General Cashiers of all Gentlemen, Merchants and Great Traders ... This gives Bills a Credit; so that in England they are never less valuable than Money, and in Venice and Amsterdam they are generally more so.²⁸

The statement that in Britain bills were ‘never less valuable than Money’ is optimistic, but the efficiency of bills when they were as good as money is clear. During the first half of the nineteenth century there was a continuous debate as to whether bills of exchange were ‘money’, ‘means of payment’, or ‘purchasing power’. The members of the Currency School agreed that only the supply of bank notes needed to be controlled, and that there was no need to control or limit the amounts of bills of exchange and of bank deposits.²⁹

Problems were likely to arise when the ratio of the debt represented by the outstanding value of bills of exchange issued by a borrower became large relative to the borrower’s wealth, which often happened in periods of euphoria. Drawing of bills of exchange in chains was infectious. Described by Adam Smith as a normal business practice, it could easily be overdone.³⁰ A draws on B, B on C, C on D, and so on, which increases the supply of credit. The vice of the accommodation bill, according to Hawtrey, was its ‘use for construction of fixed capital when the necessary supply of bona-fide long-run savings cannot be obtained from the investment market’. Hawtrey claimed the system was particularly abused in the London crisis of 1866 and in the New York crisis of 1907.³¹ Each endorser on the bill was liable for the full payment. The spectacular failure of the DeNeufvilles in Amsterdam in 1763 has been noted, which produced panics in Hamburg and in Berlin, and to a lesser extent in London and Amsterdam because a particularly impressive chain of bills was unraveled. If one bank in the chain of banks that had endorsed the bill failed, the chain collapsed and might bring down good names, those with a relatively low ratio of outstanding IOUs to capital as well as those with much higher ratios. Accommodation bills enabled traders with limited capital to borrow large amounts of money, and these short-term loans stretched into longer-term loans because they often were rolled over when they matured. During the period when the gold standard was suspended at the beginning of the nineteenth century, there was no need to be concerned about the impacts of the expansion of credit on the value of the pound. Sir Francis Baring knew of clerks worth less than £100 who were allowed discounts of £5000 to £10,000. The ‘phrenzy of speculation’ during this period strongly influenced the Currency School.³² In 1857 John Ball, a London accountant, reported knowing firms with capital under £10,000 and obligations of £900,000 and claimed it was a fair illustration.³³ In Hamburg during the same boom, Schäffle reported a man with capital of £100 and £400,000 worth of acceptances outstanding.³⁴

Long-Term Capital Management borrowed more than $125 billion; its capital was $5 billion. Its leverage ratio of 25 to 1 was much higher than the ratios of most other hedge funds, which generally were less than 10 to 1. In the eighteenth century, however, many firms, according to Wirth, speculated for ten to twenty times their capital during the boom of 1763.³⁵ Lehman Brothers had capital of three percent in the several years prior to its collapse – and it tended to transfer some of its assets to affiliates at the end of each month to reduce its leverage ratios.

Finance or accommodation bills could lead to excessive credit expansion. From time to time fictitious names were introduced into the chain so they would appear more creditworthy. Moreover, such bills were written for odd amounts to suggest an underlying commercial transaction. Claims were sometimes made (for example, by German banks drawing on Dutch and American banks after the halt in American lending) that the banks knew it was finance paper disguised as commercial bills.³⁶

Securitization

Securitization contributed to the bubble in US real estate between 2002 and 2006; the investment banks created new Asset Backed Securities (ABSs) which were claims to the interest and principal of securities with similar attributes that had been placed in trusts.

Mortgages, credit card debt, and student loans were securitized. Mortgage Backed Securities (MBSs) were one type of ABS; they were much more liquid than the individual mortgages that were in the trust. Moreover they provided a way for firms to diversify the credit risks attached to individual mortgages. Because of these advantages, the supply of credit for mortgages was much larger.

Call money

The expansion in the use of call money was important prior to the crash of 1882 in France, which was classic mania financed by call money, or money lent to stockbrokers by banks ‘on call’, that is, for one day (in French, reports).³⁷ The stockbrokers used the money to finance their holdings of an inventory of stocks and they anticipated that they would be able to renew the one-day loans day after day.

The Union Générale was a bank started by Eugène Bontoux, an engineer who had worked with Rothschild and then left to initiate rival operations in Austria, Serbia, and southeastern Europe. An earlier Union Générale, founded in 1875, was not a success. Bontoux started his Union Générale in Paris in 1878 as France was entering a boom based on the expansion of the railroads and the construction of the Suez Canal. The boom peaked in December 1881 and the crash followed in the next month. Bouvier’s interest was in whether Bontoux, a Catholic, failed because of his mistakes as a lender or whether he was ‘done in’ by a conspiracy of establishment Jewish and Protestant bankers that resented an intruder. Bouvier concluded with a Scottish verdict of ‘not proven’.

Bontoux’s Union Générale initially was capitalized at 25 million francs, which was increased in the spring of 1879 to 50 million francs and increased again in January 1881 to 100 million francs; a third increase, planned for January 1882, would have raised the capital to 150 million francs. Only one-fourth of the capital was paid in.³⁸ With each increase in capital, the investors had to pay a premium above the par value of 500 francs into the reserves of the bank in response to the increase in the market price of the stock. These premiums were 20, 175, and 250 francs, respectively. Shares were registered in the names of buyers even though they still owed three-quarters of the par value for the purchases of shares; nevertheless, roughly half of the 200,000 original shares floated in the trading market.

Trading in securities was conducted through fortnightly settlements in both Paris and Lyon. A purchaser would pay 10 percent down, borrow 90 percent from an agent de change or broker who in turn borrowed the money in the call-money market. Money was invested in reports by banks, by special caisses (funds created especially by banks and other investors for this outlet) and by individuals. A bank and caisse moreover could favor brokers who specialized in trading in a particular stock. Thus banks such as the Union Générale and the Banque de Lyon et de la Loire – not to mention three or four less successful if less spectacular banks created during the boom – could support their own stock indirectly. When changes in stock prices were modest, speculators realized small gains and small losses, and the brokers would typically pay out or receive very little money. If stock prices increased, some speculators would withdraw money from the market. Then more capital would need to be invested. Assume a speculator bought a share for 100 francs, paid 10 francs down and borrowed 90 francs. If the speculator sold the share at 110 francs and withdrew his 20 francs, 11 francs of the 20 francs withdrawn from the market would come from the new speculator, and 9 francs had to be new report and borrowed from the call market. As stock prices increased the interest rates on call money (taux des reports) increased to attract new money; the interest rate on call money increased from 4 to 5 percent at the end of 1880 to 8 to 10 percent in the spring of 1881, and reached a peak of 12 percent in the autumn of 1881.³⁹

When the share prices declined the speculators needed more money to comply with margin requirements. If the speculator bought a share at 100 francs, again with 10 francs down and 90 francs in reports, and the share price fell to 90 francs, the speculator had to produce 9 more francs to comply with the 10 percent margin requirement. If the speculator had been fully leveraged earlier and did not meet the margin call, the broker sold the speculator’s stocks. If the price dropped below 90 francs, the broker, bank, or individual that had made the loan lost money. The stock of the Union Générale went from 1250 francs in March 1881 to a peak of 3040 on 14 December as the mania gathered speed. Thereafter a period of distress followed with quotations at 2950 francs on 10 January 1882, and 2800 francs on 16 January. On 19 January the price declined to 1300 francs. Brokers were 18 million francs short on that day, as speculators could not produce the cash needed to meet margin requirements, and 33 million francs short when it came to the month-end liquidation on 31 January.⁴⁰

The collapse of Banque de Lyon et de la Loire was more spectacular as the share price went from a peak of 1765 francs on 17 December 1881 to 1550 francs on 28 December, when it was supporting its own stock, to 1040 francs on 4 January 1882, 650 francs on 10 January, and finally 400 francs on 19 January, the day after it closed its doors.⁴¹ The signal for the collapse was the approval given to Bontoux to establish a Banque de Crédit Maritime at Trieste, which was announced on 4 January 1882. The coup boomeranged.⁴² Investors with losses on their holdings of shares in Banque de Lyon et de la Loire sold their Union Générale shares. The combination of highly leveraged speculation and a credit mechanism that rested on bank and personal credit cycled through the call loan market and spread collapse among banks, caisses, brokers, individuals, and businesses in a few days. Economic activity was adversely affected in advance of the collapse of the stock market, not because of a change in the money supply but because at the peak of the fever the business world of Lyons turned to speculation in Union Générale: ‘silk merchants, cloth manufacturers, industrialists, tradesmen, dry-goods merchants, grocers, butchers, people with fixed incomes, janitors, shoemakers’; ‘A lot of capital was diverted from regular business to stock market both in securities and in call money.’⁴³

Some of the features of the US stock market crash of 1929 were similar to those in the collapse of the French banks nearly fifty years earlier: a preoccupation with speculation and a decline in economic activity as stock prices peaked; more money was needed to support the higher level of stock prices and less money was available for economic activity. Another similarity is that as stock prices peaked, call loans of ‘all others’ as opposed to New York banks and banks outside New York increased from just under $2 billion at the end of 1926 to nearly $4 billion two years later and to more than $6.6 billion on 4 October 1929. Margin credit was extensive. Brokerage firms might require that the buyers of stocks pay 10 percent down; the remaining 90 percent was borrowed. Meanwhile, brokers’ loans of New York banks declined from $1.6 billion at their height at the end of 1928 to $1.1 billion on 4 October 1929.⁴⁴

With the crash, ‘all others’ and banks outside New York took call funds from the market. They were fearful that the Stock Exchange might be closed as it had been in 1873, which would have frozen their liquid day-to-day loans.⁴⁵ At this stage, New York banks slightly increased their brokers’ loans. Similarly, in 1882, one consortium of Paris banks headed by the Banque de Paris et des Pays-Bas (Paribas) advanced 18 million francs in five loans to the Union Générale directly, while another group headed by the Rothschild bank loaned 80 million francs to the company of brokers to get them through the end of January settlement and allow them time to work out their payments arrangements. In both crises many brokers, clients, and (in 1882) banks and their caisses went bankrupt. The money market banks eased the adjustment, but in Paris in 1882 they did not save the Union Générale.

In the United States buying stocks on margin was curbed in the 1930s by a Federal Reserve regulation that fixed margin requirements at 50 percent. Financial institutions were able to make end runs about this regulation. The regulation applied to organized exchanges for shares, including the New York Stock Exchange, but not to the Chicago Mercantile Exchange, which dealt in Standard and Poor’s (S&P) 500 stock index futures and where the margin requirement was 10 percent on the value of positions in the futures contracts. Arbitrageurs linked prices in two markets so in effect they were one market. An investor who bought an S&P 500 futures contract in Chicago with a 10 percent margin was in effect buying stocks in New York with a 10 percent margin; as the price of futures contracts increased in Chicago, arbitrageurs would sell the futures contracts and at the same time buy a representative basket of stocks of the companies that dominated the index.

In the aftermath of the 19 October 1987 collapse, there was some support for the idea of regulating the Chicago and New York markets by a single agency – the Federal Reserve Board or the Securities and Exchange Commission – and some for tightening margin regulations in the futures market, and some for banning futures trading in stocks altogether.

The gold-exchange standard

A third example of the expansion of credit on a fixed money base is represented by the gold exchange standard, which involved central banks’ holdings of liquid assets denominated in the British pound and the US dollar. This practice began before World War I, although it is sometimes thought to have been based on the recommendations of the Genoa Conference of 1922 and the Gold Delegation of the League of Nations. This innovation had been strongly supported by Governor Montagu Norman of the Bank of England who sought to increase foreign holdings of British pound securities to provide relief for the Britain’s balance of payments and especially its modest holdings of gold.⁴⁶ The boom in world lending of 1913–14 was financed by increases in central bank holdings of securities denominated in the British pound, the French franc and the German mark.

Just as it is less costly to use currency notes and bills of exchange rather than coins in payments, ownership of securities denominated in the currency of a country with a large capital market dominate holdings of monetary gold as long as financial conditions are stable. These assets are easier to use in transactions, free of the need for transport, safeguarding, and assay, and directly useful without conversion into national money. A country can increase its holdings of international reserve assets by selling bonds in London or in New York and then use the money receipts to buy short-term securities denominated in the British pound or the US dollar. Since the British or US monetary authorities are not likely to regard the increase in the foreign holdings of liquid securities denominated in their currencies as a reason to contract their own credit superstructure, these transactions lead to an expansion of credit.

International lending on the gold standard may have the same unstable character as on the gold-exchange standard. Before countries borrowed foreign money that they would use as reserves for domestic monetary expansion, they borrowed gold. During the nineteenth century, American firms and state governments borrowed in London during economic upswings both to finance imports and to increase the gold base of the US banking system. Transactions in gold helped with the transfer of goods and services; gold loans enabled the borrowers to expand credit without inducing credit contractions in the lending countries.⁴⁷

Instability of credit and the Great Depression

The notion that manias and crashes result from the instability of the supply of credit is an old one. Alvin Hansen, writing on business cycles, discussed this view in a general survey of ‘early concepts’ and in a chapter on mid-nineteenth-century economists – John Stuart Mill and Alfred Marshall – entitled ‘Confidence and Credit’.⁴⁸ Hansen believed that these views had become obsolete because they neglected the investment and savings decisions of large firms. Perhaps. But theories that attach importance to the instability of credit persisted into the twentieth century. Hawtrey was a classic economist in this vein, and so was A.C. Pigou, whose book Industrial Fluctuations (1927) has a chapter dealing with panics.⁴⁹ The paradox is that the role of the instability of credit began to be neglected about the time of the Great Depression of the 1930s.

The monetarist view of the Great Depression is set out in a monumental work by Milton Friedman and Anna Schwartz. They maintained that the sharp decline in economic activity in the first half of the 1930s was the result of policy mistakes by the Federal Reserve; they focused on the decline in the money supply from August 1929 to March 1933. Whenever the question turns on the start of the depression, they note that the money supply did not increase in 1928 and 1929, and that it declined 2.6 percent from August 1929 to October 1930 when it should have been increased to offset the weakness in the economy. Friedman and Schwartz contend that the stock market crash of October 1929 had little or nothing to do with the intensity of the decline in output and that the depression was the consequence of US domestic policies and had only a tangential connection with international capital movements, changes in currency values, and deflation abroad.⁵⁰ Their view of the depression has held sway in the United States for an extended period.⁵¹

Peter Temin challenged this monetarist view from a Keynesian point of view. He asked whether the decline in spending resulted from a decline in the money supply or whether instead the decline in the money supply followed from the decline in spending; he used sophisticated econometrics to choose between these two views. Much of Temin’s argument centered on how much actual consumption deviated from forecasts of the levels based on the relationships among consumption, income, wealth, and similar variables that should be specified in forecasting a ‘normal’ trend in consumption. In addition he examined the relationship between the timing of changes in the money supply and changes in interest rates; thus if the increases in spending preceded the increase in the money supply then interest rates should have increased whereas if the increase in the money supply preceded the increase in spending then interest rates should have declined. Since interest rates declined sharply after the 1929 crash (except for interest rates on high-risk bonds which increased because of the greater concern with default risk) he concluded that the decline in spending preceded the decline in the money supply. Temin also examined the changes in real money balances (the nominal money supply adjusted for changes in the consumer price level) and concluded that real balances increased between 1929 and 1931 by 1 to 18 percent, depending on the choice between M₁ and M₂ and between a wholesale price deflator and a consumer price deflator. Annual averages further dampened the movement; on the basis of monthly figures, and averaging M₁ and M₂ expressed as percentages or relatives of a base year and of the two price indexes, the money supply increased 5 percent between August 1929 and August 1931, if seasonal influences are minimized. Temin concluded that there is no evidence that changes in the money supply between the stock market crash and the British departure from the gold standard in September 1931 caused the depression.⁵²

Temin’s analysis did not provide an explanation of the depression even though it was a strong challenge to the monetarist view. One analyst claimed that the stock market crash led banks to ration credit to borrowers and thus started the depression without reducing the money supply.⁵³ Another suggested that the sharp decline in share prices reduced nominal wealth and household spending; the implicit assumption was that the decline in consumption followed from the decline in the real value of wealth.⁵⁴

These arguments ignore the speed of the decline in industrial production in 1929 that began four or five months before the stock market crash. Industrial production fell from 127 in June to 122 in September to 117 in October, 106 in November, and 99 in December; automobile production declined from 660,000 units in March 1929 to 440,000 in August, 319,000 in October, and 92,500 in December. These declines are much too large to be explained by changes in the money supply.

Instead these declines are best explained by the instability in the supply of credit. As the stock prices increased in the first ten months of 1929, funds were channeled to the call-money market from consumption and production; the volume of call money rose from $6.4 billion at the end of December 1928 to $8.5 billion in early October 1929. Moreover, first the New York banks and then banks headquartered in other US cities became more cautious lenders to the stock market participants and to other borrowers. When the stock market crashed, the credit system suddenly froze. Loans to finance imports declined sharply, in part because of the sharp decline in the prices of imports.

The debate between the monetarists and the Keynesians ignores the instability of credit and the fragility of the banking system and the negative impacts on production and prices when the credit system became paralyzed because declines in the prices of many commodities and goods caused many borrowers to default on their loans – which explains the events in the early stages of the 1929 depression. This view was largely ignored except by Minsky and Henry Simons, the Chicago economist who thought the Great Depression was caused by declines in business confidence that led through an unstable credit system to changes in liquidity and consequent effects on the money supply.⁵⁵

Simons’s views were set forth in Economic Policy for a Free Society,⁵⁶ written after World War II under the strong influence of the 1930s depression. He recommended a 100 percent currency reserve against bank deposits to prevent changes in deposits arising from changes in the willingness of the public to hold currency and a vigorous effort to stamp out the variability of credit elsewhere in the system. He proposed restrictions on open-book credit and installment loans as well as limitation of government debt to non-interest-bearing money at one end of the spectrum and very long-term debt (ideally, perpetual obligations) at the other. Simons advocated a system in which all financial wealth would be held in equity form with no fixed money contracts so that only banks could create effective money substitutes. He was concerned about the speculative temper of the community and the ease with which short-term non-bank borrowing and lending made society vulnerable to changes in business confidence.

Simon’s recommendations to limit the character of money and financial assets were very different from Friedman’s liberal propensity to let market forces determine the demand and supply for different types of financial securites.⁵⁷ Friedman was confident that control of the growth of the money supply would prevent major business cycles and that there was no need to fear that the credit mechanism would not be stable. Even if Simons’s positive proposals were desirable (and there is grave doubt on this score) they are utopian and hence not practical. Nevertheless Simons’s diagnosis of the tendency of the system toward unstable short-term borrowing and repayment is on target.

In the last several decades still another viewpoint on the relations of money and banking to economic stability has come into prominence from the neo-Austrian school that wants to decontrol money and banking entirely. This group, led by Friedrich Hayek in Britain, Roland Vaubel in Germany and Richard Timberlake, Leland Yeager, Lawrence White, and George Selgin in the United States, seeks to abolish an active monetary policy. Any bank, company, or person would be allowed to issue ‘money’; they believe that market forces will determine which firms issue good money. Each firm will compete to ensure that its money is accepted, and so good money will drive out bad. White defends the position on the basis of the experience of the Scottish banks that were unregulated between the failure of the Ayr Bank in 1772 and the Bank Act of 1845 which applied the Bank Act of 1844 to Scotland.⁵⁸

During this period the leading commercial banks accumulated the notes of the lesser ones and were ready to convert them to specie if they thought the supply of currency notes issued by any bank was increasing too rapidly. The large banks served as informal controllers of the money supply. But several historical experiences – the country banks in England from 1745 to 1835, wildcat banking in Michigan in the 1830s, and the latest experience with bank deregulation in Latin America and especially in East Asia – do not support the view that ‘good money drives out bad’.

The global inflation of the 1970s resulted from a combination of expansive US monetary policy and expansive monetary policies in Europe and Japan as their large balance of payments surpluses led to rapid growth in their holdings of international Reserve assets. Central bank holdings of international reserve assets again increased rapidly in the mid-1990s and the late 1990s.⁵⁹ Three years later a French economist, Pascal Blanqué, wrote of a US credit bubble.⁶⁰ In a similar view, Graciela Kaminsky and Carmen Reinhart blame foreign countries for printing money and the United States for running a persistent balance-of-payments deficit.⁶¹

The central question is whether a central bank can restrain the instability of credit and slow speculation to avoid its dangerous extension. If the monetary authorities fix some proxy for the money supply or for liquidity, or if they focus directly on the rate of interest, can the upswing and decline of the crisis be moderated or eliminated entirely? There is no a priori way to determine whether a central bank policy of holding the money supply constant, limiting the liquidity of the money market, or raising the discount rate at the first sign of euphoric speculation would prevent the manias that lead to crises. Economists cannot conduct carefully controlled experiments. Nevertheless, the weight of historical evidence favors the view that a somewhat different monetary policy would have moderated booms even if it would not have eliminated them completely.

The Bank of England was severely attacked for its blindness to an approaching crisis in 1839 and its failure to raise interest rates sooner. The general view that it had been dilatory is said, in fact, to have been the proximate cause of the Bank Act of 1844.⁶² In the early 1850s the glut of gold led to declines in interest rates in 1852 and 1853. Thereafter interest rates increased although not by enough to forestall the severe crisis of 1857.⁶³ The number of bills of exchange in circulation kept on increasing as the discount rate rose, and declined as it fell, rather than responding to policy changes in the opposite direction; speculation based on bill creation seems to have been uninhibited by the increases in discount rates.⁶⁴ One mid-1850s proposal was that the Bank of England should vary its discount rate as its reserves change so the public will have a clear idea of what to expect – a suggestion Elmer Wood claims shows no grasp of Bank of England transactions.⁶⁵ In 1863 and 1864 the Bank of England twice raised the interest rate to 9 percent, which perhaps delayed but did not prevent the 1866 crash. Liquidation was completed in France in 1864; there were two shakeouts in Britain that year, but the main deflation was delayed.⁶⁶ In July 1869 the National Bank of Austria-Hungary raised interest rates but the increase came too late to forestall a crash in the fall of 1869 which was a pale forerunner of the Great Crash of 1873 in Vienna.⁶⁷ The Bank raised its discount rate again in 1872, although Wirth said the rate was too low.⁶⁸ In similar fashion, with the same timing and the same absence of result (unless it precipitated the crisis), the Federal Reserve Bank of New York raised its discount rate from 5 to 6 percent on 9 August 1929.

In 1873 the Bank of England changed its discount rate twenty-four times and avoided the financial crises that seized Austria and Germany in May and the United States in September. In November the rate was raised to 9 percent to prevent the Germans from withdrawing money that had been in London since the French-German reparations.⁶⁹ Whether this represented successful fine-tuning against the possibility of a crisis or merely increasing sensitivity to short-term capital movements cannot be determined from the secondary sources.

In the panic of 1907, the preparatory expansion had involved lending of unknown amounts by out-of-town banks in New York, together with large borrowing in London by New York banks – a combination of two of the methods of expansion discussed earlier, if one equates out-of-town banks lending in New York to the gold-exchange standard. Since the United States did not have a central bank, it was not possible to take discretionary actions to change interest rates. American borrowing led to gold exports from London and to increases in interest rates, and the Bank of England advised the market that further acceptance of American finance bills was a menace to stability and unwelcome.⁷⁰ This slowed the boom but failed to prevent the ‘rich man’s panic’ of March 1907 and the full-scale panic of October.

If central bankers were omniscient and omnipotent, they might be able to manage interest rates and reserve requirements to stabilize the credit system; they could then correct the instability in the supply implicit in the infinite expansibility of credit. But ‘there are no positive limitations to the expansion of individual credit’.⁷¹

Central banking developed to impose control on the growth and instability of credit. The evolution of central banking from profit-oriented private banking is a remarkable achievement. By 1825 there was implicit agreement on the division of labor between the private banks and the central bank; private bankers in London and the provinces would finance the boom while the Bank of England would ‘finance the crisis’ so it would not become self-justifying. In the United States, which was without a central bank after 1837, the major banks in New York were caught between the profit-making role that led them to contribute to the instability of credit and their role as holders of deposits of out-of-town banks that might be withdrawn and also contribute to instability. There was a conflict between the short-run concern with profitability and the long-run concern with financial stability, the private good with the public good. No one decreed that the New York banks should act responsibly in the public interest; it may or may not have been to their advantage to do so. The problem is a general one in politics and business and centers on who should look out for the public interest.