Lessons from the Panic of 1907

Lessons from the Panicof 1907

Ellis W. Tallman and Jon R. Moen

The Bank Panic of 1907was so serious that it became

a catalyst for the creation of America'scentral bank. This study, which examines

the circumstances leading to and the inter-vention measures taken during the panic,

particularly focuses on trust companies'function as a financial intermediary. Unequalregulation among financial organizations, the

authors find, led to a concentration of riskier assetsin less regulated intermediaries, primarily trusts.

Trusts' riskier asset portfolios made them the focalpoint from which the crisis spread to other segments of

the financial market. Allowing various types of insti-tutions comparable access to all assets and investmentopportunities, the authors conclude, might reduce the

risk that the collapse of one type of asset would threatenthe solvency of an entire class of financial intermediary.

F or the past two centuries recurrentcrises have shaken the banking sys-tem and financial markets in the

United States. One severe crisis, the BankPanic of 1907, disrupted financial markets tosuch an extent that it became an importantcatalyst for creating the Federal Reserve andthe U.S. banking system as it operates today.The panic involved several types of financialintermediaries, each distinct and playing aunique role in capital markets at the sametime that each operated under a different setof regulations. This regulatory framework cre-ated conditions that made a panic more likelythan if regulation had allowed uniform accessto all investment opportunities.

The authors are economists in the macropolicy and regional sections,

respectively, of the Atlanta Fed's research department.

What follows is a case study of an individu-al financial crisis, a record detailing eventsthat led up to, and the maneuvers that tookplace during, the Panic of 1907. The focus ison the condition of New York City trust com-panies, a financial intermediary that hadgrown rapidly in prominence at the turn of thecentury and experienced the most severe de-positor runs during the Bank Panic of 1907.Their growth can be attributed largely to freerinvestment opportunities that resulted frombeing less subject to regulation than state ornational banks. Although trust companieswere profitable, their specialization in collat-eralized loans, perceived as risky loans tofirms that could not obtain credit through na-tional or state banks, added to the severity ofthe panic.

This research has direct relevance for theregulation of intermediaries. Examining therole of the trust company as a financial inter-

ECONOMIC REVIEW, MAY/JUNE 1990

mediary in the Panic of 1907 helps to exposethe crucial role that uneven regulation playedin determining the composition of asset port-folios of banks and trusts. Because trusts tookadvantage of investment opportunities towhich banks had limited access, trusts hadrelatively undiversified portfolios.

Economic Conditionsbefore the Panic

How does a financial crisis begin? Whatprompts a panic? Most answers suggest thatfinancial calamities result from an unusualcombination of economic conditions andevents. In the case of the 1907 Panic the col-lapse of F. Augustus Heinze's attempt to cor-ner the market for copper stock apparentlytriggered the chain of events, but informedobservers agree that the same developmentsprobably would not have led to a panic in amore benign economic environment.1 OliverM.W. Sprague, writing for the National Mone-tary Commission in 1910, describes in detailthe economic conditions and special circum-stances that resulted in the Panic of 1907.Unusually severe liquidity problems in NewYork City emerge as a backdrop in the crisis.

Seasonal Liquidity Fluctuations. Duringthe National Banking Era the New York moneymarket faced seasonal variations in interestrates and liquidity resulting from the trans-portation of crops from the interior of theUnited States to New York and then toEurope. The outflow of capital necessary to fi-nance crop shipments from the Midwest tothe East Coast in September or October usu-ally left New York City money marketssqueezed for cash. As a result, interest ratesin New York City were prone to spike upwardin autumn. Seasonal increases in economicactivity were not matched by an increase inthe money supply because existing financialstructures tended to make the money supply"inelastic." The base money stock—gold,greenbacks, national bank notes, and goldand silver certificates—was also affected byunusual variations in gold flows through for-eign exchange markets. Recent research byFabio Canova offers evidence that external

disruptions to the movement of gold wereimportant determinants of bank panics.2

Atypical gold flows in 1907 seem to have con-tributed to the extreme seasonal tightness inNew York City's money markets in the fall.

The absence of finance bills during 1907substantially altered gold flows, contributingto the conditions that framed the crisis.Finance bills were contracts to extend cred-it—essentially bonds issued to borrow over-seas in hope of profit from anticipatedexchange-rate fluctuations. The dollar's ex-change rate varied over the year, strengthen-ing during the harvest season when foreigndemand for dollars to purchase crops washigh and weakening thereafter. Finance billswere most frequently drawn in the summer,two or three months before crop movement,when the dollar price of sterling was quitehigh (E.W. Kemmerer 1910). Banks and trustcompanies then sold sterling notes for dollarswhen sterling was stronger and repaid thenotes when the dollar value of sterling waslower, thus making a profit. Increased use offinance bills seems to have reduced thevolatility of exchange rates and the volume ofgold shipments overseas, enhancing the effi-ciency of the international exchange market,according to C.A.E. Goodhart and MargaretMyers. Finance bills also provided a crude fu-tures or forward market in foreign exchange.

International Gold Flows. Unlike the for-eign exchange market, domestic trade offeredno such contractual provision to smooth capi-tal flows. The New York money market trans-ferred funds to the interior of the UnitedStates to finance transport of agriculturalgoods to New York City ports. Without amechanism to arbitrage regional interest ratesor increase liquidity, interest rates in NewYork City generally climbed during the fall.This regular pattern signaled the increasedliquidity needs of New York City banks. Usu-ally, higher interest rates attracted sufficientfunds to offset the city's money shortage. In1907, however, aberrations in internationalgold flows created additional credit con-straints in the financial market that height-ened the probability of a panic.

In the spring of 1906 the U.S. TreasuryDepartment, under Secretary Leslie Shaw, de-vised policies to stimulate gold imports into

FEDERAL RESERVE BANK OF ATLANTA 3

the United States to combat what was per-ceived to be a shortage of gold. Subsidizinggold imports through the use of finance bills,the policy generated a significant inflow of $50million in a little more than a month, betweenApril and May 1906. In typical trade, financebills issued during the summer would haveprevented such substantial gold outflowsfrom England. As it was, large-scale exports ofgold from London nearly spurred a crisis inGreat Britain. To defend its domestic finan-cial markets, the Bank of England raised itsdiscount rate in late 1906 and threatened an-other increase if American finance bills werenot paid upon maturity without renewal(Myers).

Thus finance bills were suspended during1907, substantially constricting the system ofarbitrage that minimized actual shipments ofgold. In 1907, despite relatively high U.S. in-terest rates, the United States exported $30million in gold to London during the summer.As a result, the New York money market wasleft with an uncharacteristically low volume ofgold upon entering the fall season of cashtightness.3 New York financial markets werethus pressed by even less liquidity than usualat precisely the time when the need formoney intensified. Any shock to the financialmarkets could, and in 1907 did, spark a majorcrisis.

The Onset of the Panic

Such a shock occurred on October 16, 1907,when F. Augustus Heinze's attempt to cornerthe stock of United Copper Company failed.Although United Copper was only modestlysignificant, the collapse of Heinze's scheme,which came atop a slowing economy, a declin-ing stock market, and a tight money market,sparked one of the most severe bank panicsof the National Banking Era. Investigation ofHeinze's interests exposed an intricate net-work of interlocking directorates across banks,brokerage houses, and trust companies inNew York City. Contemporary observers likeSprague believed that the close associationsbetween bankers and brokers heightened de-positors' anxiety.

As Heinze's extensive involvement inbanking became apparent, along with that ofanother speculator associated with the cop-per scam, C.F. Morse, depositors' fears of in-solvency precipitated a series of runs on thebanks where the two men held prominent po-sitions. After the failure of his attempt to cor-ner United Copper stock, Heinze was forcedto resign from the presidency of MercantileNational, and worried depositors began a runon the bank. The New York Clearinghouse, aconsortium of banks in New York City, exam-ined the bank's assets, announced that it wassolvent, and stated that the clearinghousewould support Mercantile on the conditionthat Heinze and his board of directors resign.

During the reorganization of MercantileNational Bank, the New York Clearinghousebegan examining other banks that had in-terests related to Heinze and that had beenraising suspicion for some time. The restructur-ing of Mercantile revealed that Morse was oneof that bank's directors. Sprague (1910, 248)describes Morse as having "an extreme char-acter, even when judged by American specu-lative standards."

Morse was a director of seven New YorkCity banks, three of which he controlled com-pletely. He was also held in low esteem bymost other bankers. His connection withMercantile's difficulties worried depositors athis other banks, and two called for aid fromthe clearinghouse on October 19 in responseto large withdrawals of deposits. The clearing-house granted assistance on the conditionthat Morse retire completely from banking inNew York. During the weekend, both Morseand E.R. Thomas, another of Heinze's cohorts,were relieved of their remaining banking in-terests. The clearinghouse promised to sup-port those banks as well.

The assets of Heinze's banks totaled $71million, compared to over $2 billion in all NewYork City banks and trusts (Sprague 1910,249). Although this was a significant amount,depositors apparently considered the clear-inghouse's promise of a $10 million fund toaid former Heinze banks sufficient because nonotable run occurred on the banks. OnMonday, October 21, Mercantile National re-sumed business with new management, andthe run ceased. Similar action was taken at


Heinze's Copper Corner Attempt

F. Augustus Heinze, a key player in the ini-tial stage of the panic, rose rapidly to notorietyin the financial world after he won a highly pub-licized legal battle against AmalgamatedCopper in Butte, Montana. Amalgamated hadbeen organized a few years earlier by severalStandard Oil Company executives and fi-nanciers, including James Stillman of NationalCity Bank of New York. The purchasers ofAmalgamated reportedly earned a profit of $36million on an investment of $39 million, whichhad gone primarily toward the acquisition of theAnaconda mines in Montana (New York Times,October 17, 1907).

Heinze, who owned a copper mine near theAmalgamated Mines, claimed that veins of cop-per from his mine extended under land ownedby Amalgamated and that according to the"apex law" he had a right to mine it (Carosso1970, 112). The matter was pursued in an exten-sive legal confrontation that was eventually set-tled out of court in February 1906, when Heinzesold his copper interest to Amalgamated for areported $25 million, half in cash and half inAmalgamated securities.

Heinze took his gains to New York City,where he became involved in banking (Allen1935). In January 1907 newspaper articles had al-ready associated Heinze with E.R. Thomas andC.F. Morse, both New York bankers and ownersof the Mechanics and Traders State Bank andConsolidated Bank, Heinze was placed on theboard of directors of eight banks and two trustcompanies.1 Elected president of MercantileNational Bank in February 1907, he immediatelyreplaced the directorship with his associates.The Heinze group gained control of severalother banks quite quickly through "chain bank-ing," an organizational strategy similar to today'sbank holding company. The group would buystock in a bank and use it as collateral to borrowmoney, which was then used to buy stock in an-other bank or trust.

Heinze's attempt to corner the stock ofUnited Copper, a company of which he waspresident, eventually triggered the Panic of1907. The corner attempt, which probably ex-plains the steady and relatively high price ofUnited Copper stock despite a weak coppermarket during 1907, was not an unusual strategy.However, unlike other market corner schemes,this one seemed to be public knowledge, assuggested by several newspaper articles refer-ring to the intent of the Heinze group (seebelow). His reputation as a speculator was fur-ther reinforced when the respected investment

banking house J.S. Bache withdrew from its busi-ness relations with Heinze in February 1907(New York Times, February 15, 1907).

The alleged corner of United Copper stockcollapsed in October 1907. It was foiled in partby actions taken by an Amalgamated subsidiary,United Metals Selling Company, which appar-ently had been manipulating the market for rawcopper. Subsequent Pujo Committee investiga-tions revealed that United Metals SellingCorporation sold only 5 million pounds of cop-per from April to August 1907 (U.S. Congress, 734;see also 717-40). The normal amount rangedfrom 150 million to 250 million pounds. WhenCongressional Counsel Samuel Untermeyerpressed assistant manager Tobias Wolfson ofthe United Metals Selling Corporation for an ex-planation, he stated that no buyers could befound for copper. Untermeyer then quipped,"And all of a sudden, in October, they were in-terested in 93 million pounds in a singlemonth?" Wolfson responded, "Yes. They hadused up all that they had bought." As a result ofthese market manipulations, the price of rawcopper plummeted, and the price of coppermining stocks broke. Having reached a high ofnearly $121 a share in January 1907, Amal-gamated Copper fell from $56 1/4 to $41 3/4 inOctober. Although United Copper maintained asteady price throughout the first half of October,the following events led to the total collapse ofthe Heinze interests.

United Copper first reached the headlines ofthe New York Times on Tuesday, October 15. OnMonday its stock had risen from $39 to $60 dur-ing the first 15 minutes of trading on the CurbMarket.2 Buying was not done through Heinzebrokers. Curb brokers emphasized that Heinzebrokers had been taking great pains to keeptrack of all shares in United Copper that hadcome out since the high prices of January 1907,in an attempt to distinguish short selling.3 Shortpositions in United Copper were thus known tobe dangerous. Heinze was not interested in thetotal number of shares outstanding because hebelieved many shares were held in the westernUnited States, where, in those days, they couldtake a week or more to reach New York for sale;rather, Heinze was concerned about how manyshares were quickly accessible to the market.

Apparently thinking the time was right for acorner, Heinze purchased a large quantity ofshares on Monday through the brokerage houseof his brother, Otto Heinze. Many shares ofUnited Copper had appeared on the marketduring trading on Saturday, October 12, and


Heinze suspected that brokerage houses werelending out his shares of United Copper to sup-port short selling of the stock. He ordered Grossand Kleeberg, a brokerage house started in1904, to purchase 6,000 shares of the stock at as-cending prices, so that he was in effect buyinghis own shares again from short sellers at a high-er price to attract more short sales. Of course,the short sellers did not realize that Heinze al-ready owned the shares that they had borrowedand that he was now buying from them. This ac-tion, Heinze hoped, would force short sellers toa settlement in a technique known as a "bearsqueeze." The squeeze would result whenHeinze owned a large percentage of UnitedCopper stock, in which the majority of activelytraded shares were his own—shares that he hadpurchased from brokers who allowed largeshort-sale positions. Then, even though he hadbought shares at increasing prices, by demand-ing delivery of his shares Heinze intended toforce short sellers to come up with shares thatthey did not possess, and could not possess,unless they bought them from Heinze. Thus thesettlement between Heinze and the short sellercould be at almost any price and would clearlyprovide Heinze with a profit as long as therewas no other source for United Copper shares.

To punish the exchange houses that hadgone short in United Copper stock, Heinze putout an order calling in all his United Coppershares. However, Heinze's actions were ill-advised because his suspicions of short salesturned out to be unfounded. Gross andKleeberg found many shares not owned byHeinze for sale at high prices. More shares ap-peared on Tuesday after news of the high stockprice spread, so the anticipated time lag be-tween the increase in stock price and the addi-tional number of shares for sale was insufficientto support a corner

Heinze's corner was further thwarted by whatappeared to be the maneuvers of an unknowngroup of individuals determined to hinder hisscheme by controlling a large block of UnitedCopper stock. Newspaper sources reported thaton Tuesday the transfer agency of UnitedCopper, T. Buckingham, refused to transfer own-

ership of 17,830 shares of common stock(Commercial and Financial Chronicle, January 4,1908). The agency claimed that the block washeld in a joint account and could not be trans-ferred unless both parties agreed. A newspaperarticle reported that Heinze believed a "marketpool" of United Copper stock was being lent outto short sellers in violation of the agreement, al-though Heinze did not identify who was in thepool {New York Times, October 17, 1907). Had thepool been unwilling to cooperate, Buckingham'srefusal of the transfer order might have prevent-ed the market pool from upsetting the attempt-ed corner. Announcement of the refusalstrengthened United Copper stock on Tuesday,though it still closed down 16 from the previousday. When legal action was threatened againstthe refusal of transfer, the order was rescindedand the transfer went through.

On Wednesday Heinze's corner attempt suf-fered the final blow. Gross and Kleeberg wereforced to sell United Copper stock to pay for theshares purchased earlier on margin. UnitedCopper fell from $36 to $10 a share, and the firmhad to suspend operations. The same day, thebrokerage house Otto Heinze and Companyclosed. It was said at the stock market that Heinzeand his brokers were "taken to the wall." The bro-kers had bought large amounts of United Copperstock on margin at increasing prices resulting al-most entirely from their own purchases. Whenthey stopped buying, the price fell, threateningtheir financial position. As Heinze interests wereforced to sell their shares purchased on margin,the stock price broke dramatically.

The newspaper attributed the failure of thecorner to the market pool of stock held by un-known individuals whose transactions Heinzehad attempted to block through the transferagent. One commentator suggests that Amal-gamated Copper interests, namely H.H. Rogers,Stillman, and other powerful financiers, were"waiting in the wings" to deny Heinze an oppor-tunity to corner the market in his stock (RobertSobel). This analysis is feasible. If the stock wastraded infrequently, Heinze would not havebeen aware of how much stock existed to be un-loaded during his corner maneuver.

Notes1Banks were Mercantile National, ConsolidatedNational, Mechanics and Traders, Union, Bank ofDiscount, Riverside, Northern National, andMerchants Exchange National; trusts were Hudsonand Empire (New York Times, January 21, 1907).

2The Curb Market in those days actually took placeoutdoors on the curb of the street. It later moved in-doors and is now the American Stock Exchange.

3 A short sale is a maneuver in which the seller, expect-ing prices to fall, offers stock he or she does not yetown to be delivered at a future date, taking profitsfrom the difference between current (high) pricesthey would be paid and the future (low) prices theywould face to acquire the stock.


several small banks operated by associates ofthe Heinzes, and by October 21 reorganiza-tion of the national banks was complete.

The Run on Trusts

By October 21, nothing resembling a sys-temic panic had yet stricken the banks, asSprague points out (1910, 250). Depositors atMercantile Bank withdrew funds but rede-posited them in other New York City banks.The conditions of the economy, however, wereuncertain. The apparant lack of liquidity in thefinancial markets, as discussed above, set thestage for a major financial crisis to erupt fromcircumstances that in other times might nothave sparked concern.

Many historical accounts of the Panic of1907 cite Monday, October 21, as the begin-ning of the crisis among the trust companies.On that day the National Bank of Commerceannounced that it would stop clearing checksfor the Knickerbocker Trust Company, thethird largest trust in New York City. However,Vincent Carosso (1987, 535) suggests that therun on Knickerbocker began October 18, whenCharles Barney, the Knickerbocker president,was reported to have been involved inHeinze's corner maneuver. Drawing from theprivate papers of J.P. Morgan, Carosso notesthat the National Bank of Commerce hadbeen extending loans to the KnickerbockerTrust to hold off depositor runs. National Bankof Commerce's refusal to continue acting as aclearing agent for Knickerbocker was inter-preted as a vote of no confidence that seri-ously alarmed Knickerbocker depositors.

Morgan, along with James Stillman ofNational City Bank and George Baker of FirstNational Bank, had organized an informalteam to oversee relief efforts during the panicat the national banks (Carosso 1970, 129; 1987,538-39). Assisting them were several young fi-nancial experts responsible for evaluating theassets of troubled institutions and indicatingwhich ones were worthy of aid. Chief amongthese investigators was Benjamin Strong ofBanker's Trust Company, who would later be-come president of the Federal Reserve Bankof New York.4

On Monday evening, October 21, Morganhad organized a meeting of trust company ex-ecutives to discuss ways to halt the panic.Strong reported to Morgan that he was unableto evaluate Knickerbocker's financial condi-tion in the short time before funds wouldhave to be committed. Unwilling to act on lim-ited information, Morgan decided not to aidthe trust; this decision kept other institutionsfrom offering substantial aid as well. OnOctober 22 Knickerbocker underwent a run forthree hours before suspending operationsjust after noon, having paid out $8 million incash.

Ironically, next to the front-page article de-scribing the suspension of the KnickerbockerTrust in the Wednesday, October 23, editionof the New York Times was a headline describ-ing Trust Company of America, the secondlargest trust company in New York City, as thecurrent "sore point" in the panic. By attractingattention to Trust Company, the newspaperarticle greatly exacerbated the serious run onit. Barney, who was president of Knicker-bocker, was also a member of the board of di-rectors of Trust Company of America.

It has been argued that the statement inthe New York Times by George W. Perkins, oneof Morgan's partners, citing Trust Company'sproblems as the current "sore point" was anattempt to isolate the panic at an important,fundamentally sound institution that wouldpresumably be aided through the run by themajor financiers (Frederick Lewis Allen 1949,248-49). Trust Company of America was nearthe Morgan and Company offices, making it alikely candidate for such a maneuver. Duringthe panic, the newspapers described frequentexchanges of big leather boxes betweenMorgan offices and Trust Company offices, sig-naling the exchange of money and securities.However, H.L. Satterlee, Morgan's son-in-law,later emphasized that no banker would havepurposely started a run on any bank for fearthat the panic might eventually engulf his owninstitution as well (470).

On Tuesday, October 22, withdrawals fromTrust Company of America were approximate-ly $1.5 million; on the Wednesday when theill-timed article was published depositorsclaimed another $13 million of nearly $60 mil-lion in total deposits. Withdrawals from Trust


Company of America on Thursday, October 24,were a further $8 million to $9 million. Duringthe span of the run, which lasted two weeks,Trust Company of America reportedly paidout $47.5 million in deposits.5

Rescue Efforts

Realizing that the failure of Trust Companyof America and Lincoln Trust, another institu-tion whose distress had been publicized,would endanger the New York money market,a committee of five trust company presidentsformed to assist trusts in trouble. Not alltrusts were willing to cooperate, though, sothe committee was not able to collect enoughmoney to provide reliable relief for a trustcompany facing a sudden run. They peti-tioned Morgan for help.

Morgan, Baker, and Stillman knew that aidfor Trust Company of America was not certainand saw that the collapse of several largetrusts would be disastrous. Strong had arrivedat Trust Company of America sometime after2:00 A.M. Wednesday and had begun to ap-praise its assets. That afternoon he reportedto Morgan that Trust Company was basicallysound and deserved assistance. Morgan chan-neled about $3 million to Trust Company justbefore closing time, which allowed it to re-sume business the next day.

Aid began to come from several othersources. J.D. Rockefeller deposited $10 mil-lion with the Union Trust to help the trustsand announced his support for Morgan.Secretary of the Treasury George Cortelyouand the major New York financiers met on theevening of Wednesday, October 23, and dis-cussed plans to combat the crisis. Cortelyoudeposited $25 million of the Treasury's fundsin national banks the following morning.Between October 21 and October 31, theTreasury deposited a total of $37.6 million inNew York national banks and provided $36million in small bills to meet runs. By the mid-dle of November, however, the U.S. Treasury'sworking capital had dwindled to $5 million.Thus Treasury could not and did not con-tribute much more aid during the rest of thepanic (Timberlake, 173-78).

Crisis on the Stock Exchange

Meanwhile, by Thursday, October 24, callmoney on the New York Stock Exchange wasnearly unobtainable. Call money was moneylent for the purchase of stock equity, with thestock serving as collateral for the loans. Callloans could be called in at any time. Theopening rate for call money was 6 percent, butexchange president Ranson H. Thomas no-ticed a serious scarcity of money. At one pointthat morning a bid of 60 percent went out forcall money. Yet, even at that exorbitant rate,no money was offered. The last recordedtransaction of the day was at the opening rateof 6 percent (U.S. Congress, 355). Fearing atotal collapse of the stock market, Thomascalled Stillman for aid. Stillman referredThomas to Morgan, who was in control of mostof the available funds. While Thomas traveledto Morgan's office, the call money rate on theexchange reached 100 percent.

In his testimony to the Pujo Committee, es-tablished in 1912 by Congress to investigatethe possible existence of New York Citymoney cartels and their potential conspiracyto precipitate the panic, Morgan's partnerCharles Steele described efforts to providefunds to the stock market during the crisis.Morgan, who reportedly discussed the situa-tion at the stock exchange with other bankersbefore his meeting with Thomas, told Thomasto announce that $25 million would be avail-able on the exchange floor. After a short time,Steele arrived at the exchange with a list ofnational banks which, as a group, promised toloan $25 million to the exchange, including $4million from First National and $8 million fromNational City. The market borrowed a total of$18.95 million that day (U.S. Congress, 457).

Indirect use of Treasury funds to forestallcollapse of the market during the panic alsocame under scrutiny during the Pujo investi-gation. Legally restricted to national banks,Treasury deposits were channeled toward thebanks that most quickly presented acceptablecollateral, which for the most part meantTreasury bonds. Direct use of Treasury de-posits in the stock market was prohibited. Intestifying to the Pujo Committee, however,Treasury Secretary Cortelyou explained that


the use of Treasury funds was not specifiedbefore they were credited; rather, the major fi-nanciers determined the most appropriateapplication for the money (U.S. Congress,439). Thus, in effect, nearly all the funds con-tributed to aid the panic were controlled byMorgan, who decided how much moneywould be used and where.

Trying to determine whether governmentfunds were used directly to ease the creditstrain on the stock market, the counsel forthe Pujo Committee, Samuel Untermeyer,pressed Cortelyou for information about thespecific amount of government deposits re-ceived by each national bank from the total$25 million allocated. Cortelyou had no recol-lection of the transactions and did not knowwhether the Treasury had records of them.

Estimates of available cash reserves inNew York national banks indicate they werehigh enough to provide funds to the stockmarket had government funds been denied tothe exchange. On August 22, 1907, New Yorknational banks held $218.8 million. Cash re-serves in the "big six" national banks were$140.7 million.6 On December 3, 1907, re-serves had fallen to $177 million for all NewYork national banks and $112.5 million for the"big six." During the worst period in the pan-ic, reserves were probably lower. However, tooffer their own funds to the stock market,banks would have had to drop below the legal25 percent reserve requirement. Thus Unter-meyer's concerns were not without a basis de-spite the apparent availability of funds frombanks. The congressional testimony suggeststhat Morgan simply allocated the governmentdeposits in national banks to the stock ex-change in the same amounts that the govern-ment deposited them.

On October 25 another money pool was re-quired. About $10 million came from the Mor-gan group, $2 million from First National, and$500,000 from Kuhn, Loeb, and Company. Thistime, however, Morgan allowed the market todetermine the call money rate, which re-mained at nearly 50 percent most of the day.The Morgan funds had restrictions designedto stifle speculation. First, no margin saleswere allowed—only cash sales for investment.Also, the full amount of Morgan money wasnot released until afternoon. Morgan's partner,

Perkins, noted that the money collected forthe Friday stock exchange pool was about themost that could be collected that day and yetwas barely enough to keep the market open(Allen 1949, 255). Throughout the stock ex-change crisis, both Trust Company of Americaand Lincoln Trust were supported by Morgan'sefforts.

Actions of the New YorkClearinghouse Association

While financiers were working out thecrises with the trusts and the call loan market,money and reserves had become increasinglytight at banks. On October 26 the clearing-house issued clearinghouse loan certificatesas an artificial mechanism to increase the sup-ply of currency available to the public, a tacticit had used in earlier financial crises in 1873and 1893 (see Richard Henry Timberlake,Gary Gorton, or Ellis Tallman).

Although the national banking system of-fered no legal mechanism to increase the sup-ply of currency quickly, loan certificatesprovided an informal (if unlawful) way to freeup a sizable amount of cash. In normal busi-ness banks used currency as reserve assetsand as the medium to clear accounts witheach other. Clearinghouse loan certificates en-abled banks to monetize their noncurrencyassets during a crisis: banks would substituteloan certificates for currency in their clearings,thus releasing the currency to pay depositorswho demanded cash. Loan certificates werenot recognized as currency by the public or bydepositors, and they were supposed to be cir-culated only among banks. However, A. PiattAndrew (1908) noted that during the 1907Panic, a number of substitutes for cash wereemployed in transactions.

Following the first issue of clearinghouseloan certificates on October 26 during the1907 Panic, loans initially increased by about$11 million. During the next three weeks morethan $110 million in certificates were issued inNew York City. Nearly $500 million in currencysubstitutes circulated throughout the countryas a "principal means of payment," accordingto Andrew (1910, 515). Sprague has criticized


the clearinghouse for delaying the use of loancertificates until after the panic was wellunder way. He believed that issuing certifi-cates as soon as the crisis struck the trustswould have calmed the market by allowingbanks to accommodate their depositors morequickly. Aid would have gone directly to trou-bled banks and trusts, and the cumbersomedevice of money pools could have beenavoided. Fewer loans would have been calledin, thus reducing the tension at the stock ex-change (Sprague 1910, 257-58).

The clearinghouse also restricted the con-vertibility of deposits into cash—an actionwhich, like issuing loan certificates, was ille-gal. The restriction, referred to as "suspensionof payments," increased transaction costs.Nevertheless, banks continued other busi-ness activities such as accepting deposits andclearing checks. The suspension of paymentsspread across the country through the systemof correspondent banks. Though convertibilitywas widely restored by the beginning ofJanuary, in a few instances loan certificatesand other substitutes for cash circulated aslate as March 1908.

Distress Spreads

New York City government was also near-ing a financial crisis of its own. It needed $30million in new funds but had delayed a bondissue because of the situation in the financialmarkets. The city had attempted to float abond issue in the summer of 1907, but eventhen the bonds had not found a market.Though no source specifies how the New YorkCity Comptroller financed city expendituresfor the interim, it seems the city used short-term loans to pay its expenses until anotherbond issue could be attempted. The Mayor ofNew York, George McClellan, approachedMorgan on Monday, October 28, with the city'sfinancial problems. Short-term obligationswere coming due, and the city had no fundswith which to pay them. Morgan recognizedthat if the city defaulted on its loans, the crisiscould become completely unmanageable.

Morgan, Stillman, and Baker thus agreedon October 29 to underwrite a $30 million, 6

percent bond issue of New York City. Morgandevised a plan in which the major bankswould take pro rata shares of the issue anddeposit them with the clearinghouse. Theclearinghouse would then issue clearinghouseloan certificates in an equal amount and cred-it them to the city's accounts at First Nationaland National City.

Meanwhile, the lack of money to the callloan market was threatening the brokeragehouse of Moore and Schley. The firm had bor-rowed $25 million from New York banks, plac-ing a large block of Tennessee Coal, Iron, andRailroad Company stock as collateral. Theloans were about to come due. To complicatematters, the brokerage was already using thesame stock as collateral on other loans it hadgranted to its senior partner, Grant B. Schley,Baker's brother-in-law.

If Moore and Schley liquidated the stock topay off its loan, the price of the stock wouldhave tumbled, causing the call loan market tobecome even tighter. In the face of an alreadyweak stock market, such a disruption couldhave been disastrous, undermining confi-dence even further.

Morgan eventually solved the problem bygiving his support to a plan designed by hisattorney and friend, Lewis Cass Ledyard.Ledyard proposed that U.S. Steel buy Mooreand Schley's shares of Tennessee Coal, Iron,and Railroad, paying for them with its ownhighly rated 5 percent gold bonds. Carosso(1970) has noted that this maneuver was im-portant for several reasons. Moore and Schleywould be saved without depressing the stockmarket, and U.S. Steel would be able to ab-sorb a competitor. The innovative aspect ofthis arrangement was that it involved no cur-rency in a market that was already cash-shortfrom the runs on the trust companies. Thedeal went through on Monday, November 4,after President Roosevelt agreed not to op-pose it on antitrust grounds.

The crisis at the trust companies continuedduring the Moore and Schley episode. TrustCompany of America and Lincoln Trust re-quired further aid, and Morgan convincedother trust presidents to support a $25 millionloan for the troubled institutions. The fundswere provided on November 4 after severalnights of negotiation. The panic began to ease

10 ECONOMIC REVIEW, MAY/JUNE 1990

when the trust company presidents organizedby Morgan agreed to form a consortium tosupport trust companies facing runs.

The New York Clearinghouse had detailedknowledge of the quality of bank assets inNew York. A similar, formal organization oftrust companies would have had currentknowledge of the assets and liabilities of itsmember trusts. Such an organization couldhave more readily assessed the situation attrust companies facing runs than the ad hocconsortiums and money pools organized byMorgan. As Sprague has argued and experi-ence supports, however, the legislative solu-tion to a major crisis is usually moregovernment regulation rather than improvedindustry self-supervision (1910, 273).

The Role of the Trusts

It is not surprising that trust companiesearly on became the focal point of the panic.In New York, trust assets had grown phenome-nally between 1890 and 1910, increasing 244percent during the 10 years ending in 1907,from $396.7 million to $1,394.0 million. In con-trast, national bank assets grew 97 percent,from $915.2 million to $1,800.0 million, whilestate-chartered bank assets grew 82 percent,from $297 million to $541.0 million (Barnett,234-35). Thus the manner in which trust com-panies used their assets greatly affected theNew York money market. (For a more detailedanalysis of the role of trusts in the panic, seeMoen and Tallman.)

Trust companies were much less regulatedthan national or state banks in New York. In1906 New York State instituted a requirementthat trusts maintain reserves at 15 percent ofdeposits, but only 5 percent of depositsneeded to be kept as currency in the vault.Before that time trusts simply kept whateverreserves they felt necessary to conduct busi-ness. National bank notes were adequate ascash reserves for trusts while national banksin central reserve cities like New York were re-quired to keep a 25 percent reserve in theform of legal tender or specie.

Trusts were originally rather conservativeinstitutions, managing estates, holding securi-

ties, and taking deposits, but by 1907 trustswere performing most of the functions ofbanks except issuing bank notes. Many of thelarger trusts specialized in underwriting secu-rity issues. Others wrote mortgages or invest-ed directly in real estate—activities barred orlimited for national banks. New York Citytrusts had a higher proportion of collateral-ized loans than did New York City nationalbanks. Conventional banking wisdom asso-ciated collateralized loans with riskier in-vestments and riskier borrowers. The trusts,therefore, had an asset portfolio that mayhave been riskier than those of other interme-diaries.

National and private banks found the in-vestment banking functions of trusts so usefulthat many of them gained direct or indirectcontrol of a trust through holding companiesor by placing their associates on a trust'sboard of directors. In many instances a bankand its affiliated trust operated in the samebuilding.

Trusts appear to have provided intermedi-ary functions different from those of banks.Although the volume of deposits subject tocheck at trusts was similar to that at banks,trusts had much less clearing activity than didbanks, registering clearings only about 7 per-cent of the volume of those at banks. Trustswere not then like commercial banks, whoseassets are used as transactions balances byindividual depositors or firms.

National banks were part of a network ofregional banks that had correspondent rela-tionships to expedite interregional transac-tions (James, 40). Trusts were not part of thecorrespondent banking system, so their de-posits were more local and less directly sub-ject to the recurring seasonal strains on funds.

The most severe runs in New York Citywere limited to the trust companies, not thestate or national banks (Moen and Tallman).Trusts' riskier asset portfolios in conjunctionwith their ambiguous relationship to the NewYork Clearinghouse signaled to depositorsthat the trusts were likely to become insol-vent during an economic and financial down-turn.7 Runs forced trusts to liquidate theirmost liquid assets, call loans on the stockmarket. Large-scale liquidation of call loansdepressed the value of stocks.


Given the predominance of national banksin the call loan market, extensive liquidationof call loans by trusts threatened the assets ofnational banks. Although trusts and nationalbanks were legally distinct, both intermedi-aries operating in the call market were eco-nomically integrated. It was because nationalbanks and the clearinghouse were aware thatthe runs on the trusts could spread to the en-tire financial system that they acted directlyto stop the runs.

Conclusion

Some important policy lessons emergefrom this case study of the 1907 Panic.Restriction of the types of investments nation-al banks could make in 1907 did not reducethe overall riskiness of the financial system'sassets; rather, the uneven regulation of trustsand banks concentrated riskier assets in a fewinstitutions, primarily the trusts. Negativeshocks to trust assets, notably collateralizedloans, raised the specter of their possible in-solvency. If regulations allowed intermedi-aries comparable access to all assets andinvestment opportunities, the potential for

adequate diversification of portfolios mightreduce the risk that the collapse of one typeof asset would threaten the solvency of an en-tire class of intermediary.

Nor is it certain that access to the New YorkClearinghouse could have averted insolvencyamong thrifts in 1907, given the high concen-tration of risk in their portfolios. Although theclearinghouse functioned to some extent as acentral bank, lack of explicit legal authority toissue clearinghouse loan certificates kept theclearinghouse from fully exploiting these func-tions. It did maintain records on the financialhealth of participating banks and made thisinformation available to members. Thus,when member banks requested aid, the clear-inghouse had the information necessary tomake a decision quickly. Trusts' limited affilia-tion with the clearinghouse made informationabout distressed trusts harder to obtain andprobably contributed to the destabilizing iso-lation of the Knickerbocker Trust.

Even with access to a lender of last resort,under conditions of uneven regulation trustcompanies would have had the incentives tomaintain portfolios with profitable but riskyassets. The potential for a financial crisis todrive a class of intermediaries into insolvencywould remain.

Notes1Kindleberger refers to "copper speculation" that in-volved more than just Heinze's corner attempt as aprime contributor to the panic. Analysis of the coppermarket during 1907 is interesting (see the testimony ofWolfson in U.S. Congress), but the direct links to thepanic are less clear. The connection is left for further re-search.

2For a discussion of the money supply process in theNational Banking Era, see Goodhart or, for a more con-cise description, Tallman.

3The aberration of gold flows exacerbated the amount ofgold shipments to the United States when European im-porters paid for shipments of cotton and cereal from theUnited States during the panic.

4There he was recognized as a decisive leader during theearly years of the central bank. His untimely death in1928, which left the young Federal Reserve System with-

out focused leadership, has been argued by some asbeing the reason for the Fed's inept handling of thebank panics early in the Great Depression (seeFriedman and Schwartz).

5Carosso (1987), citing figures in J.P. Morgan's privaterecords. A run on Lincoln Trust, a smaller institution,began with withdrawals exceeding $1 million.

6Sprague (1910, 234). Sprague notes that the six nationalbanks (National City, National Bank of Commerce, FirstNational, Chase National, Park National, and HanoverNational) had grown from 30 percent to 60 percent of thetotal assets in New York national banks from 1873 to1907.

7Kindleberger suggests that the trusts were responsiblefor excessive credit expansion related to speculative ac-tivities prior to the Panic of 1907.

12 ECONOMIC REVIEW, MAY/JUNE 1990

References

Allen, Frederick Lewis. The Lords of Creation. New York:Harper and Brothers Publishers, 1935.

. The Great Pierpont Morgan. New York: Harper andBrothers Publishers, 1949.

Andrew, A. Piatt. "Substitutes for Cash in the Panic of1907." Quarterly Journal of Economics 23 (August 1908):497-516.

. Statistics for the United States: 1867-1910. NationalMonetary Commission. Washington, D.C.: U.S.Government Printing Office, 1910.

Barnett, George E. State Banks and Trust Companies Since thePassage of the National Bank Act. Washington, D.C.: U.S.Government Printing Office, 1911.

Canova, Fabio. "The Sources of Financial Panics: Pre- andPost-Fed Evidence." Brown University Department ofEconomics Working Paper 88-1, 1988.

Carosso, Vincent P. Investment Banking in America.Cambridge, Mass.: Harvard University Press, 1970.

. The Morgans: Private International Bankers, 1854-1913.Cambridge, Mass.: Harvard University Press, 1987.

Commercial and Financial Chronicle. November 7, 1907-January 8, 1908.

Diamond, Douglas W., and Philip H. Dybvig. "Bank Runs,Deposit Insurance, and Liquidity." journal of PoliticalEconomy 91 (June 1983): 401-19.

Friedman, Milton, and Anna J. Schwartz. A Monetary Historyof the United States: 1867-1960. Princeton, N.J.: PrincetonUniversity Press, 1963.

Goodhart, C.A.E. The New York Money Market and the Financeof Trade: 1900-1910. Cambridge, Mass.: HarvardUniversity Press, 1969.

Gorton, Gary. "Clearinghouses and the Origins of CentralBanking in the United States." journal of Economic History45, no. 2 (1985): 277-84.

Jacklin, Charles J., and Sudipto Bhattacharya. "Dis-tinguishing Panics and Information-Based Bank Runs:Welfare and Policy Implications." journal of PoliticalEconomy 96 (June 1988): 568-92.

James, John. Money and Capital Markets in Postbellum America.Princeton, N.J.: Princeton University Press, 1978.

Kemmerer, E.W. Seasonal Variations in the Relative Demand forMoney and Capital in the United States: A Statistical Study.

Washington, D.C.: U.S. Government Printing Office,1910.

. "Seasonal Variations in the New York Money Mar-k e t . " Quarterly journal of Economics 2 5 ( M a r c h 1 9 1 1 ) : 3 3 -4 9 .

Kindleberger, Charles P. Manias, Crashes, and Panics: AHistory of Financial Crises. New York: Basic Books, 1989.

Kinley, David. The Independent Treasury System of the UnitedStates and Its Relations to the Banks of the Country.Washington, D.C.: U.S. Government Printing Office,1911.

Moen, Jon, and Ellis W. Tallman. "The Bank Panic of 1907:The Role of the Trust Companies." Federal ReserveBank of Atlanta Working Paper 90-3, May 1990.

Myers, Margaret. The New York Money Market: Origins andDevelopment. New York: Columbia University Press,1931.

Neal, Larry. "Trust Companies and Financial Innovation,1897-1914." Business History Review 45 (Spring 1971): 35-51.

The New York Times. January 21 -October 17, 1907.Noyes, Alexander Dana. Forty Years of American Finance. New

York: G.P. Putnam and Sons, 1909.Satterlee, Herbert L. J. Pierpont Morgan: An Intimate Portrait.

New York: Macmillan and Company, 1939.Sobel, Robert. Panic on Wall Street. New York: Macmillan,

1968.Sprague, Oliver M.W. "The American Crisis of 1907." The

Economic journal (September 1908): 353-72.. 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 aboutBanking Panics." Federal Reserve Bank of AtlantaEconomic Review 73 (November/December 1988): 2-21.

Timberlake, Richard Henry. The Origins of Central Banking inthe United States. Cambridge, Mass.: Harvard UniversityPress, 1978.

U.S. Congress. House. Committee on U.S. Banking andCurrency. (Pujo Committee.) Money Trust InvestigationHearings on H.R. 429, 504. 62d Congress, 2d and 3dsess., 1913. 3Vols.


Documents

Lessons from the Panic of 1907