EH.net is owned and operated by the Economic History Association with the support of other sponsoring organizations.
The Panic of 1907
Jon Moen, University of Mississippi
The Panic of 1907 was the last and most severe of the bank panics that plagued the National Banking Era of the United States. Severe panics also happened in 1873, 1884, 1890, and 1893, although numerous other smaller financial crises cropped up from time to time. Bank panics were characterized by the widespread appearance of bank runs, attempts by depositors to simultaneously withdraw their deposits from the banking system. Because banks did not (and still do not) keep a 100% reserve against deposits, it paid to be near the front of the line of depositors demanding their money when a panic blew up. What sets 1907 apart from earlier panics was that the crisis focused on the trusts companies in New York City. The National Banking Era lasted from 1863 to 1914, when Congress, in part to eliminate these recurring panics, created the Federal Reserve System.
What Caused the Panic?
Why would a panic happen? One answer that is really not of much help is that all depositors suddenly became so concerned about the solvency or liquidity of their bank that they decided they would rather hold cash than deposits (Diamond and Dybvig 1983; Jacklin and Bhattacharya 1988). (Solvency refers to the relationship between assets and liabilities; an insolvent bank has liabilities greater than its assets. Liquidity refers to the ease with which assets can be converted to cash without loss of value; liquid assets are close to cash or have a market in which they can be easily and quickly sold.) Whatever the deeper psychological reasons might be, it is not hard to identify some immediate shocks to depositor confidence that sparked the Panic of 1907. Such a shock occurred on October 16, 1907, when F. Augustus Heinze's scheme to corner the stock of United Copper Company failed. Although United Copper was only a moderately important firm, the collapse of Heinze's scheme, exposed an intricate network of interlocking directorates across banks, brokerage houses, and trust companies in New York City. Contemporary observers like O.M.W. Sprague (1910) believed that the discovery of the close associations between bankers and stockbrokers seriously raised the anxiety of already nervous depositors.
During the National Banking Era the New York money market faced seasonal variations in interest rates and liquidity resulting from the transportation of crops from the interior of the United States to New York and then to Europe. The outflow of capital necessary to finance crop shipments from the Midwest to the East Coast in September or October usually left the New York City money market squeezed for cash. As a result, short-term interest rates in New York City were prone to spike upward in autumn. Seasonal increases in economic activity were not matched by an increase in the money supply because existing domestic monetary structures tended to make the money supply "inelastic." Usually gold would flow into the United States from Europe in response to the high seasonal interest rates, increasing the monetary base of the United States and easing the liquidity squeeze somewhat.
Under more normal financial conditions, the discovery of a scheme like Heinze's might not have sparked a panic, but conditions were not normal in the Fall of 1907. The economy had been slowing, the stock market had been in decline since early 1907, and the supply of credit had been contracting causing rising interest rates. Tight credit markets in Europe, particularly in England where the Bank of England had been raising its bank rate since December 1906, have been implicated in setting an especially precarious financial stage in 1907. Therefore, the normal seasonal inflows of foreign gold were not happening in 1907 as European interest rates rose. Because there was no central bank or reliable lender of last resort during the National Banking Era, there was no reliable way to expand the money supply in the United States.
Heinze's extensive involvement in New York banking was subsequently linked to one of his close and suspicious associates, C.F. Morse. Morse controlled three national banks directly and was a director of four other banks. After the failure of his attempt to corner United Copper stock, Heinze was forced to resign from the presidency of Mercantile National Bank, and worried depositors began a run on the bank. Depositors began runs on several of the banks controlled by Morse as well. The New York Clearinghouse, a private organization formed by banks to centralize check clearing (a check clears when it is finally presented to the bank on which it was originally written for payment in cash or reserves), had its examiner analyze the banks' assets. On the basis of the examination, the Clearinghouse authorities stated that they would support Mercantile and the other banks on the condition that Heinze and Morse retire from banking in New York. On Monday, October 21, Mercantile National resumed business with new management, and the runs on these national banks ceased.
The Panic at the Trust Companies
By October 21, nothing resembling a systemic panic, however, had yet stricken the New York banking system. Depositors at Mercantile Bank withdrew funds but redeposited them in other New York City banks. Many accounts of the Panic of 1907 cite Monday, October 21, as the beginning of the crisis among the trust companies and the true onset of the panic. Late that Monday afternoon the National Bank of Commerce announced that it would stop clearing checks for the Knickerbocker Trust Company, the third largest trust in New York City. Vincent Carosso (1987), however, suggests that the run on Knickerbocker began Friday, October 18, when Charles Barney, the Knickerbocker president, was reported to have been involved in Heinze's copper corner. Drawing from the private papers of J.P. Morgan, Carosso notes that the National Bank of Commerce had been extending loans to the Knickerbocker Trust to hold off depositor runs. National Bank of Commerce's refusal to continue acting as a clearing agent for Knickerbocker was interpreted as a vote of no confidence that seriously alarmed Knickerbocker depositors.
On Monday evening, October 21, J.P. Morgan organized a meeting of trust company executives to discuss ways to halt the panic. Morgan, along with James Stillman of National City Bank and George Baker of First National Bank, had earlier organized an informal team to oversee relief efforts during the panic at the national banks (Carosso 1987). Assisting them were several young financial experts responsible for evaluating the assets of troubled institutions and indicating which ones were worthy of aid. Chief among these investigators was Benjamin Strong of Banker's Trust Company, who would later become president of the Federal Reserve Bank of New York. Strong reported to Morgan that he was unable to evaluate Knickerbocker's financial condition in the short time before funds would have to be committed. Unwilling to act on limited information, Morgan decided not to aid the trust; this decision kept other institutions from offering substantial aid as well. It appears that at first Morgan was uninterested in aiding the trust companies in general, as he felt they should pay for their risky behavior. It is not clear that they were riskier; perhaps Morgan just did not want to aid intermediaries competing with the banks. On October 22 Knickerbocker underwent a run for three hours before suspending operations just after noon, having paid out $8 million in cash.
Ominously, next to the front-page article describing the run on the Knickerbocker Trust in the Wednesday, October 23, edition of the New York Times was a headline describing the Trust Company of America, the second largest trust company in New York City, as the current "sore point" in the panic. By attracting attention to the Trust Company of America, the newspaper article greatly exacerbated the serious run on it. Barney, who was president of Knickerbocker, was also a member of the board of directors of Trust Company of America.
On Tuesday, October 22, withdrawals from Trust Company of America were approximately $1.5 million; on the Wednesday when the ill-timed article was published depositors claimed another $13 million of nearly $60 million in total deposits. Withdrawals from Trust Company of America on Thursday, October 24, were a further $8 million to $9 million. During the span of the run, which lasted two weeks, Trust Company of America reportedly paid out $47.5 million in deposits.
Saving the Trusts
Realizing that the failure of Trust Company of America and Lincoln Trust, another trust company whose distress had been publicized, would endanger the New York money market, five leading trust company presidents formed a committee to assist trusts needing cash. Not all trusts were willing to cooperate, however, so the committee was not able to collect enough cash to provide reliable relief for a trust company facing a sudden run. They petitioned Morgan for more help.
Morgan, Baker, and Stillman knew that aid for Trust Company of America was not certain and saw that the collapse of several large trusts would be disastrous. Strong had arrived at Trust Company of America sometime after 2:00 A.M. Wednesday and had begun to appraise its assets. That afternoon he reported to Morgan that Trust Company was basically sound and deserved assistance. Morgan channeled about $3 million to Trust Company just before closing time, which allowed it to resume business the next day.
Aid began to come from several other sources. J.D. Rockefeller deposited $10 million with the Union Trust to help the trusts and announced his support for Morgan. Secretary of the Treasury George Cortelyou and the major New York financiers met on the evening of Wednesday, October 23, and discussed plans to combat the crisis. Cortelyou deposited $25 million of the Treasury's funds in national banks the following morning. Between October 21 and October 31, the Treasury deposited a total of $37.6 million in New York national banks and provided $36 million in small bills to meet runs. By the middle of November, however, the U.S. Treasury's working capital had dwindled to $5 million. Thus Treasury could not and did not contribute much more aid during the rest of the panic (Timberlake 1978, 1993).
The Connection to the Stock Market
Meanwhile, by Thursday, October 24, call money on the New York Stock Exchange was nearly unobtainable. Call money was money lent for the purchase of stock equity, with the stock itself serving as collateral for the loans. Call loans could be called in at any time. The opening rate for call money was 6 percent, but exchange president Ransom H. Thomas noticed a serious scarcity of money. At one point that morning a bid of 60 percent went out for call money. Yet, even at that exorbitant rate, no money was offered. The last recorded transaction of the day was at the opening rate of 6 percent. Fearing a total collapse of the stock market, Thomas called Stillman for aid. Stillman referred Thomas to Morgan, who was in control of most of the available funds. While Thomas traveled to Morgan's office, the call money rate on the exchange reached 100 percent.
On October 25 another money pool was required. About $10 million came from the Morgan group, $2 million from First National, and $500,000 from Kuhn, Loeb, and Company. This time, however, Morgan allowed the market to determine the call money rate, which remained at nearly 50 percent most of the day. The Morgan funds had restrictions designed to stifle speculation. First, no margin sales were allowed-only cash sales for investment. Also, the full amount of Morgan money was not released until afternoon. Throughout the stock exchange crisis, both Trust Company of America and Lincoln Trust were supported by Morgan's efforts. The Trust Company of America and Lincoln Trust required further aid, and Morgan convinced other trust presidents to support a $25 million loan for the troubled institutions. The funds were provided on November 4 after several nights of negotiation. The panic began to ease when the trust company presidents organized by Morgan agreed to form a consortium to support trust companies facing runs.
The most severe runs on deposits in New York City were limited to the trust companies, not the state or national banks. Deposits contracted at all the trusts in New York, not just the prominent ones like Knickerbocker (Moen and Tallman 1992). This raises a question. If only the trust companies were being run by depositors, why would the banks want to help their competitors? The stock market provides a key link. Runs on deposits forced trusts to liquidate their most liquid assets, call loans on the stock market. Large-scale liquidation of call loans depressed the value of stocks because the stock serving as collateral for the call loan had to be sold quickly to pay off the loan. The sudden increase in the supply of stock would depress stock prices. Given the predominance of national banks in the call loan market, extensive liquidation of call loans by trusts threatened the assets of national banks. National banks and the clearinghouse were aware that they were economically linked to the trust companies through the call loan market. They realized that runs on the trusts could spread to the national banks through the call loan market, giving the banks a strong financial incentive to help the trusts stop the panic, even if they had no legal interest.
The New York Clearinghouse Association Steps In
While financiers were working out the crises with the trusts and the call loan market, money and reserves had become increasingly tight at banks. On October 26 the Clearinghouse issued clearinghouse loan certificates as an artificial mechanism to increase the supply of currency available to the public, a tactic it had used in earlier financial crises in 1873 and 1893 (Timberlake 1984; Gorton 1985; Tallman 1988).
Although the national banking system offered no legal mechanism to increase the supply of currency quickly, loan certificates provided an informal (if unlawful) way to free up a sizable amount of cash. In normal business banks used currency as reserve assets and as the medium to clear accounts with each other. Clearinghouse loan certificates enabled banks to convert their noncash assets into cash during a crisis: banks would substitute loan certificates for currency in their clearings, thus releasing the currency to pay depositors who demanded cash. In effect, loan certificates were IOUs between banks that were backed by eligible assets of the bank. Loan certificates were not recognized as currency by the public or by depositors, and they could legally circulate only among banks, not the public. A. Piatt Andrew (1908) noted, however, that during the 1907 Panic, a number of substitutes for cash were employed in transactions.
Following the first issue of clearinghouse loan certificates on October 26 during the 1907 Panic, loans initially increased by about $11 million. During the next three weeks more than $110 million in certificates were issued in New York City. Over the entire course of the Panic, nearly $500 million in currency substitutes circulated throughout the country as a "principal means of payment," according to Andrew (1910, 515). Sprague has criticized the clearinghouse for delaying the use of loan certificates until after the panic was well under way. He believed that issuing certificates as soon as the crisis struck the trusts would have calmed the market by allowing banks to accommodate their depositors more quickly. Aid would have gone directly to troubled banks and trusts, and the cumbersome device of money pools could have been avoided. Fewer loans would have been called in, thus reducing the tension at the stock exchange (Sprague 1910, 257-58).
The clearinghouse also restricted the convertibility of demand deposits into cash -- an action, which, like issuing loan certificates to the public, was illegal. The restriction, referred to as "suspension of payments," increased the costs of doing business by making payments more difficult. Nevertheless, banks continued other business activities such as accepting deposits and clearing checks. The suspension of payments spread across the country through the system of correspondent banks. Although convertibility was widely restored by the beginning of January, in a few instances loan certificates and other substitutes for cash circulated as late as March 1908.
Why Were There Runs on Trust Companies?
There were three main types of financial intermediaries during the National Banking Era: national banks, state banks, and later in the period trust companies. It is not surprising that trust companies were the focal point of the panic. In New York, assets at the trust companies had grown phenomenally between 1890 and 1910, increasing 244 percent during the 10 years ending in 1907, from $396.7 million to $1,394.0 million. In contrast, national bank assets had grown 97 percent, from $915.2 million to $1,800.0 million, while state-chartered bank assets had grown 82 percent, from $297 million to $541.0 million (Barnett 1911, 234-35). Thus the manner in which trust companies used their assets greatly affected the New York money market (Moen and Tallman 1992).
Trust companies were much less regulated than national or state banks in New York. In 1906 New York State instituted a requirement that trusts maintain reserves at 15 percent of deposits, but only 5 percent of deposits needed to be kept as currency in the vault. Before that time trusts simply kept whatever reserves they felt necessary to conduct business. National bank notes were adequate as cash reserves for trusts while national banks in central reserve cities like New York were required to keep a 25 percent reserve in the form of specie or legal tender (greenbacks or treasury notes but not national bank notes).
Trusts were originally rather conservative institutions, managing estates, holding securities, and taking deposits, but by 1907 trusts were performing most of the functions of banks except issuing bank notes. Many of the larger trusts specialized in underwriting security issues. Others wrote mortgages or invested directly in real estate activities barred or limited for national banks. New York City trusts had a higher proportion of collateralized loans than did New York City national banks. Conventional banking wisdom associated collateralized loans with riskier investments and riskier borrowers. The trusts, therefore, had an asset portfolio that may have been riskier than those of other intermediaries.
National and private banks found the investment banking functions of trusts so useful that many of them gained direct or indirect control of a trust through holding companies or by placing their associates on a trust's board of directors. In many instances a bank and its affiliated trust operated in the same building.
Trusts appear to have provided intermediary functions different from those of banks. Although the volume of deposits subject to check at trusts was similar to that at banks, trusts had many fewer checks (in number and value) written against their demand deposits than did banks. The check clearings of trusts were only about 7 percent of the volume of those at banks. Trusts were not then like commercial banks, whose assets are used as transactions balances by individual depositors or firms. National banks were part of a network of regional banks that had correspondent relationships to expedite interregional transactions (James 1978, 40). Trusts were not part of the correspondent banking system, so their deposits were more local and less directly subject to the recurring seasonal strains on funds.
The New York Clearinghouse had detailed knowledge of the quality of bank assets in New York. A similar, formal organization of trust companies would have had current knowledge of the assets and liabilities of its member trusts. Such an organization could have more readily assessed the situation at trust companies facing runs than the ad hoc consortiums and money pools organized by Morgan. The ability of a clearinghouse to shield its members from runs on deposits was clearly demonstrated by the Chicago Clearinghouse in 1907, where there were virtually no runs on deposits. In Chicago the trust companies, similar in structure to those in New York, were members of the clearinghouse and were not singled out by depositors. A lender of last resort covering all intermediaries in the payments system certainly adds stability to the system. JP Morgan and others, however, may have profited from earlier panics by lending money to otherwise desperate bankers. This is the popular view of their actions in 1907. The 1907 Panic, however, may have turned out to be far more severe than anticipated. Even if Morgan made money after the fact in 1907, the expectation of higher default risk made the possibility of lending in future panics unattractive. Perhaps this is what was realized by the New York bankers, causing them to abandon their role as de facto lenders of last resort and setting the groundwork for the establishment of the Federal Reserve System.
Much of this article is based on a review article from the Federal Reserve Bank of Atlanta, although I have updated some of our economic interpretations of the panic, particularly on matters related to liquidity and solvency. The complete reference to the review article is:
Moen, Jon and Ellis Tallman. "Lessons from the Panic of 1907." Federal Reserve Bank of Atlanta Economic Review 75 (May/June 1990): 2-13.
Other important references cited or used are:
Andrew, A. Piatt. "Substitutes for Cash in the Panic of 1907." Quarterly Journal of Economics 23, (August 1908): 497-516.
Barnett, George E. State Banks and Trust Companies since the Passage of the National Bank Act. Washington, D.C.: U.S. Government Printing Office, 1911.
Calomiris, Charles and Gary Gorton. "The Origins of Bank Panics: Models, Facts, and Bank Regulations." In Financial Markets and Financial Crises, edited by R. Glenn Hubbard. Chicago: University of Chicago Press, 1991.
Carosso, Vincent P. The Morgans: Private International Bankers, 1854-1913. Cambridge, MA: Harvard University Press, 1987.
The Commercial and Financial Chronicle. Various issues from November 7, 1907 through January 8, 1908.
Diamond, Douglas W., and Philip H. Dybvig. "Bank Runs, Deposit Insurance, and Liquidity." Journal of Political Economy 91 (June 1983): 401-19.
Friedman, Milton, and Anna 1. Schwartz. A Monetary History of the United States: 1867-1960. Princeton, N.J.: Princeton University Press, 1963.
Gorton, Gary. "Clearinghouses and the Origins of Central Banking in the United States." Journal of Economic History 45 (June 1985): 277-84.
Jacklin, Charles J., and Sudipto Bhattacharya. "Distinguishing Panics and Information-Based Bank Runs: Welfare and Policy Implications." Journal of Political Economy 96 (June 1988): 568-92.
James, John. Money and Capital Markets in Postbellum America. Princeton, NJ: Princeton University Press, 1978.
Moen, Jon, and Ellis W. Tallman. "The Bank Panic of 1907: The Role of the Trust Companies." Journal of Economic History 52 (September 1992): 611-630.
Moen, Jon, and Ellis W. Tallman "Clearinghouse Membership and Deposit Contraction during the Panic of 1907." Journal of Economic History 60 (March 2000): 145-163.
Sprague, Oliver M.W. "The American Crisis of 1907." The Economic Journal 18 (September 1908): 353-72.
Sprague, Oliver M.W. History of Crises under the National Banking System. National Monetary Commission. Washington, D.C.: U.S. Government Printing Office, 1910.
Tallman, Ellis W. "Some Unanswered Questions about Banking Panics." Federal Reserve Bank of Atlanta Economic Review 73 (November/December 1988): 2-21.
Timberlake, Richard Henry. The Origins of Central Banking in the United States. Cambridge, MA: Harvard University Press, 1978.
Timberlake, Richard Henry. Monetary Policy in the United States. Chicago: University of Chicago Press, 1993.
Timberlake, Richard Henry. "The Central Banking Role of Clearinghouse Associations." Journal of Money, Credit and Banking 16 (February 1984): 1-15.