United States: New York Clearing House Association, the Panic of 1873

According to the Comptroller of the Currency (OCC 1873), the Panic of 1873 started to build on September 8, with the failure of some New York City companies. However, there was little to no media attention dedicated to the slowly building panic (Gorton and Tallman 2018, 27). The closure of Jay Cooke and Co., one of the most prestigious merchant banks, on September 18 shocked the City and unleashed the Panic of 1873 (Wicker 2000, 20). On this day, and intensifying through September 20, depositors rushed to withdraw their funds in a series of bank runs in New York City. By September 20, several prominent banks had failed and the New York Stock Exchange closed for the first time ever for 10 days (OCC 1873).

In this era, banks typically had to liquidate assets to handle withdrawals during a bank run. New York banks invested a significant amount of deposits in the call loan market, which was the main source of funds for the New York Stock Exchange. These loans could be called back on demand by the lender in times of crisis, although this would cause drastic consequences for stock prices and the stock market in general (Anderson, Hachem, and Zhang 2021). On September 12, New York banks had committed $199 million in loans, with 30% in call loans (Sprague 1910, 84).

Instead of forcing member banks to liquidate their assets to process withdrawals and cause deleterious effects to the stock market, the NYCH provided liquidity to member banks through clearinghouse loan certificates (Anderson, Hachem, and Zhang 2021).

There has been some controversy over the years regarding the legality of the NYCH’s issuance of CLCs during panics (Timberlake 1984). The law during the National Banking Era effectively banned state-chartered banks from issuing their own notes by applying a steep 10% tax, which the CLCs did not pay and which could establish them as illegal.

However, contemporary sources and scholars tend to agree that CLCs used entirely for interbank transactions—such as those issued in 1873, 1884, and 1890—did not break this law because they did not circulate as currency. The National Banking Act of 1864 appears to sanction CLCs that circulated only between banks: “Clearing-house certificates, representing specie or lawful money specifically deposited for the purpose of any clearing-house association, shall be deemed to be lawful money” (National Bank Act of 1864 1864, 13:109). A former clearinghouse chairman, James Cannon, wrote in 1910 that the term “clearing-house certificates” used in the act would refer both to the certificates that clearinghouses issued in normal times, which were backed by specie or legal tender, and to clearinghouse loan certificates that they issued in crises, which were backed by securities (Cannon 1910).

Starting in 1893, some clearinghouses issued CLCs in smaller denominations and encouraged their use as currency, raising more directly the question of their legality under the National Banking Act.

Separate from CLCs were certified checks. In 1873, the NYCH guaranteed certified checks that banks issued to depositors in lieu of cash when they had partially suspended the convertibility of deposits, as noted in Key Design Decision No. 3, Part of a Package. Such checks were “quasi-currency” and circulated at a discount to currency (Timberlake 1984). This “currency premium” became a measure of market confidence in banks during crises. Gorton and Tallman (2016) say that certified checks were not considered “money” and were not subject to the 10% tax, but that this was “essentially a fiction.”

Suspension of deposit convertibility was also effectively illegal because demand deposit contracts gave depositors the right to withdraw cash any time a bank was open. But those contracts were never legally enforced, as suspension was seen as a necessary tool to calm banking panics at the time (Gorton and Tallman 2018, 181).

Within the institutional framework of the NYCH, members needed to first vote on the provision of CLCs as to “whether the crisis was significant enough to warrant their use” (Gorton and Tallman 2016). For example, the NYCH members unanimously approved the provision for CLCs immediately as the crisis of 1884 unfolded in an emergency meeting on May 14, 1884 (OCC 1884).

Of the five main banking crises of the National Banking Era, the Panic of 1873 is unusual for the measures the NYCH took for the last time (reserve pooling) and for the first time (certifying checks issued by banks). As in most other such panics, the NYCH suppressed information about individual banks and conducted examinations of banks that were troubled.

Reserve Pooling

In 1873, the NYCH implemented a tool that it did not use again in the National Banking Era: reserve pooling. The National Currency Act of 1864 required New York City banks to maintain a minimum reserve ratio of 25%, using the ratio of cash to total public liabilities (Cannon 1910). According to the plan adopted by the NYCH:

In effect, this condition created a single pool containing every member bank’s cash reserves. In a situation in which a troubled bank faced an extremely low level of reserves, the NYCH could shift reserves from stronger banks to support the weaker bank. Reserve pooling is distinct from the NYCH’s temporary lending, as that initiative was optional and contained an interest rate as an incentive to, in effect, lend their surplus reserves (Moen and Tallman 2013). On September 25, the NYCH utilized reserve pooling for the first time during the crisis, and the New York Times (NYT 1873d) described this as, “[T]he banks [were] virtually made one solid and compact body.” This condition made the structure and actions of the NYCH akin to a proto-central bank. This tool was used prior to the National Banking Era but never utilized again by the NYCH (Moen and Tallman 2013).

According to Sprague (1910, 20–21, 120), reserve pooling was necessary because seven of the largest banks in the NYCH had insufficient reserves to satisfy deposit withdrawal requests from rural banks. These banks paid interest on those deposits. Non-interest-paying banks, which didn’t source deposits from out-of-state banks, went along with reserve pooling on the promise that there would be reform after the crisis passed.

Guarantee of Certified Checks Issued by Banks and Suspension of Convertibility

On September 24, the NYCH announced that all checks certified by any bank would be certified as “payable through the Clearing House.” This action implicitly recognized that banks had partially suspended cash payments to depositors and were satisfying large withdrawals with certified checks instead. Member banks continued to redeem small deposits in New York and deposits for banks outside New York. This was the first time the NYCH used certified checks in this way (Gorton and Tallman 2016; Timberlake 1984).

It appears that New York banks that suspended convertibility resumed payment on November 1 (OCC 1873).

Suppression of Information

As in other crises during the National Bank Era, the NYCH also suppressed reporting of individual banks’ balance sheet information to promote market confidence, as described further in Key Design Decision No. 18, Disclosure, and No. 19, Stigma Strategy.

Examination of Banks’ Books

The NYCH did examine the books of one member bank, after its director pleaded for the NYCH to get involved. It does not appear that a report on the bank in question was ever published, and it subsequently failed during the Panic of 1873 (Gorton and Tallman 2016). It was liquidated with no losses to its creditors, and Sprague (1910, 43) reports that “[i]ts failure, therefore, was not one of the serious factors in the situation” as a result of long-standing solvency issues. Additionally, the NYCH examined three nonmember banks during the Panic of 1873 (Gorton and Tallman 2016).

Government Bond Purchases

The US Treasury supported market recovery by purchasing bonds during the crisis. Cannon (1910) notes that, as the first issue of CLCs was made on September 22:

Sprague (1910, 42) notes that the government paid a premium for the bonds, so it really injected $13 million. However, he notes that the purchases weren’t effective in supporting banks and the stock market because most cash went to savings banks, which hoarded it.

The NYCH assigned the Clearing House Loan Committee, comprising five representatives, to oversee the lending program (Gorton and Tallman 2018, 43; OCC 1873). This committee would “manage and monitor the loans,” while keeping track of the value of the committed collateral (Gorton and Tallman 2018, 43). The authorization of the committee required the vote of 75% of the member banks, and it passed unanimously on September 20 (OCC 1873).

Prior to the opening of business during the panic, members banks had to provide a daily statement to the manager of the NYCH with: (1) the amount of loans and discounts; (2) amount of CLCs; (3) amount of US certificates of deposit and legal-tender notes; and (4) amount of deposits, minus special gold deposits (OCC 1873).

During crises, the NYCH issued CLCs, which were temporary collateralized loans that could be used only to settle clearinghouse balances (Cannon 1910). In some sense, CLCs resemble Federal Reserve discount window loans, although those come from a standing facility (Moen and Tallman 2013). Member banks would request CLCs by depositing collateral with the Loan Committee, which would decide whether to issue the requested CLCs (OCC 1873). The Loan Committee could issue CLCs in denominations of $5,000 and $10,000 (OCC 1873). See Figure 1 for an official CLC.

It appears that all 60 member banks could access CLCs, as the initial plan stated that “any bank in the clearing-house association” could receive them (Moen and Tallman 2013; OCC 1873). However, Gorton and Tallman (2018, 44) cannot rule out the possibility that the Clearing House Loan Committee considered member bank solvency in its lending decisions. Both state and national banks could join.

As for who constituted the membership, the NYCH’s report in the aftermath of the panic stated:

There is anecdotal evidence from Cannon (1910), who was a leader of the NYCH, that all member banks felt a “patriotic” duty to request CLCs regardless of their expressed need, as a way to project solidarity and unity. However, quantitative research by Moen and Tallman (2013) contradicts this universal characterization. Of the 60 member banks at the time, only 33 took out CLCs during the Panic of 1873. Recent research shows that all these CLCs taken out by member banks in 1873 were used for settling clearing balances. This shows that no NYCH banks took out CLCs merely to project confidence and unity. Rather, all of the borrowing banks utilized their CLCs in clearing transactions (Anderson, Hachem, and Zhang 2021).

Rather than single loans between two member banks, CLCs were instead a transferable liability.^FNMoen and Tallman (2013) state that “as a transferable liability, [CLCs] implicitly involved a loan from the holding bank to the borrowing bank.” The NYCH issued CLCs and managed the collateral, in a structure similar to a modern tri-party repurchase agreement. To break down the balance sheet composition of CLCs, they counted as assets for the borrowing member banks and as a liability in the form of a loan from the member banks that accepted them for settlement (Moen and Tallman 2013).

There are two integral aspects of CLCs that made them an effective form of liquidity assistance (Gorton and Tallman 2018, 43). First, the adopted plan stated that member banks had to accept CLCs as specie or legal tender replacement for the settlement of balances, thus establishing CLCs as a valid form of payment within the membership (OCC 1873). Second, the NYCH collectively guaranteed the repayment of the CLCs. If a borrowing bank could not repay its fellow member banks, the NYCH paid for them and divided the cost amongst the membership relative to individual members’ share of membership capital and surplus (Gorton and Tallman 2018, 43).

The Panic of 1873 put extreme pressure on the reserves of the NYCH membership. On September 20 (the day the NYCH authorized CLCs), the legal tender reserves of member banks totaled $33.8 million. Only four days later, these reserves had declined by over 53% to $15.8 million on September 24. Reserve balances would reach their nadir on October 14 at $5.8 million (Wicker 2000, 32). The reserve pooling measure likely allowed some banks to draw down their reserves if needed, as they could rely on the reserve pooling measure to make up for the deficit during the crisis period.

The initial plan adopted by the NYCH on September 20 limited the Loan Committee’s issuance of CLCs to $10 million (OCC 1873). However, the NYCH raised this cap to $20 million on September 24 (NYT 1873c). On September 27, NYCH members said they would issue CLCs “to any amount that may be required” and further stated:

The effect of this action was to inspire the public with the belief that the banks intended to leave no measures untried which would restore confidence and protect the mercantile community in their transactions. (NYT 1873e)

In total, the NYCH issued nearly $26.57 million in CLCs, with a peak outstanding of $22.41 million (OCC 1907). It is unclear when or how the NYCH further increased the limit on issuance by the Loan Committee above $20 million.

The NYCH issued $22 million in CLCs from September 22–29, $3.2 million in October, and $1.3 million from November 1–20 (OCC 1907). See Figure 2 for a depiction of outstanding CLC usage and the reserve balance of NYCH member banks relative to the 25% required reserve ratio. As this chart shows, NYCH member banks ignored the 25% required reserve ratio during the Panic of 1873, while banks began canceling CLCs only once the reserve balance had recovered.

Source: Gorton and Tallman 2018, 47.

Research could not determine if there were limits on the amount of CLCs that individual NYCH member banks could borrow beyond the amount of securities they had available to serve as collateral. It appears that some banks did apply for more CLCs than they received.

Although member banks had to accept CLCs as a replacement for clearing balances, they received a 7% per annum interest rate for their participation, which was paid by the borrowing bank. The initial plan adopted by the NYCH called for the interest to be paid out on the earlier of the CLCs’ redemption or November 1, 1873, when the CLCs were supposed to expire (OCC 1873). However, the NYCH issued more CLCs beyond the initial $10 million without an expiry date, so it is likely that the interest was due whenever the CLCs were redeemed or canceled. Beyond the overall interest rate, there was an additional 0.25% fee charged for the administration of the CLCs (NYT 1873a).

According to Gorton and Tallman (2018, 46), the interest rates charged by the NYCH during the National Banking Era were consistent with market rates for commercial paper, although the significant haircuts added to this cost. According to Wicker (2000, 27), the rates on prime commercial paper in the three months prior to September hovered around 7%, before jumping to 12.5% in September and peaking in October at 17%.

The plan adopted by the NYCH stated that member banks could “deposit . . . an amount of its bills receivable, or other securities to be approved by [the Loan Committee]” (OCC 1873). A report in 1907 by the Office of the Comptroller of the Currency (OCC 1907) included information submitted by the manager of the NYCH stating that the NYCH accepted “[b]ills receivable, stocks, bonds, and other securities” during the Panic of 1873. According to retrospective studies such as Gorton and Tallman (2018, 43–44), the predominant collateral utilized for CLCs was commercial paper or commercial loans with short maturities.

The initial plan stated that the amount of CLCs granted to borrowing banks should not be “in excess of seventy-five per cent of the securities or bills receivable so deposited” (OCC 1873). This was, in effect, a haircut on collateral of at least 25%. However, CLCs could be issued upon the par value of the collateral if it was “United States stocks” (likely high-quality government bonds) or gold certificates (OCC 1873).

The Loan Committee could exchange the deposited collateral for other acceptable securities at the request of the borrowing bank. More important, it had the discretionary power to request additional collateral, whether by the previously mentioned exchange or an increased amount (OCC 1873). When member banks canceled CLCs, the Loan Committee returned their collateral to them.

CLCs did not exactly contain a maturity, as they were transferrable liabilities rather than direct loans (Moen and Tallman 2013). The NYCH did initially set an expiry date for CLCs, which is covered in Key Design Decision No. 20, Exit Strategy.

Member banks highly valued their membership in the NYCH and risked suspension or expulsion if they ignored any of the rules (Gorton and Tallman 2018, 21–22). A few days into the crisis (September 24), the NYCH membership met and unanimously adopted a resolution stating, “any member [not fully participating in the CLC arrangement] shall be reported to the manager of the Clearing-house, who shall forthwith expel him from this body” (NYT 1873c). The NYCH suspended Chemical National Bank in 1873 for three months when it refused to accept CLCs as payment (Gorton and Tallman 2016).

Since the NYCH was a private institution at the time, its actions during the Panic of 1873 did not have any impact on monetary policy transmission as they were an internal division and reallocation of liquidity.

According to Sprague, the reason the stock exchange had to close on September 20 was that, because of the tightness of money during the panic, banks were unable to continue to provide the extraordinary intraday credit that the market had come to depend upon for its transactions (Sprague 1910, 38–40). Certain banks gave stockbrokers large intraday “overdrafts” to allow them to purchase securities on behalf of their clients, with the understanding that the broker would deposit sufficient certified checks from clients by 3pm to repay their banks. The amount of credit could be 10 or 20 times a broker’s deposits. Congress passed a law forbidding national banks from conducting this activity in 1882 (Sprague 1910, 104).

On September 24, 1873, the New York Stock Exchange asked the NYCH to guarantee the payment of these overdrafts to allow the market to reopen. It also asked for a 10-day moratorium on collecting call loans. The NYCH committee declined, stating:

The NYCH was not the only clearinghouse in the country. Many other cities had clearinghouses, but the NYCH clearly had the most power and impact due to the size and stature of its member banks.

The Panic of 1873 was the first time other cities issued CLCs. Cannon (1910) lists Boston, Philadelphia, Baltimore, Cincinnati, St. Louis, and New Orleans. Wicker (2000, 23) also mentions Richmond. In 1873, the clearinghouse in Philadelphia issued similar CLCs in the amount of $6.8 million (OCC 1907).

The NYCH publicly announced its decision to issue CLCs and ensured that it became widely known and publicized (Gorton and Tallman 2018, 44). The New York Times (NYT 1873a) published the entire adopted plan of the NYCH on September 21. Later that night, as the New York Times reported the next day, the president of the NYCH appeared at a prominent hotel and answered questions about the crisis. He spoke forcefully about the stability of the NYCH membership and told the doubters that they were likely “ignorant of what they talked about” (NYT 1873b). He continued on to say, “People who haven’t a cent come here and put questions for the express purpose of damaging the credit of old-established firms” (NYT 1873b). In conclusion, he mentioned that the NYCH could increase the amount of CLCs from $10 million to $20 million, which it did a few days later.

The NYCH refused to publicize both the individual member banks that utilized CLCs and how many they received (Gorton and Tallman 2018, 44).

However, it seems that banks could themselves announce their own CLC usage. Fourth National Bank, one of the six largest banks in New York, announced in the New York Times (NYT 1873d) on September 26 that it “paid into the [NYCH] USD 800,000 worth of [CLCs] and received USD 800,000 in legal tenders.” The New York Times quoted the cashier of the bank praising the CLC program for allowing it to “afford much more accommodation to its customers than it might have otherwise done.”

As the public could not assess the condition of banks during the Panic of 1873, the OCC issued two extra calls for all national banks to report their condition on October 13 and November 1. The need to specifically request this information from banks implies that government regulators were also unsure of the actual conditions of many banks without taking additional disclosure measures. Although this data was not broken down to the individual bank level, the OCC utilized this data after the crisis in its 1873 Annual Report to analyze the panic (OCC 1873).

As in other crises, the NYCH suspended publication of any regular balance sheet information about individual banks. The suspension lasted more than two months. The NYCH also did not publicize individual CLC usage. This was due to a fear of stigma, as external observers would note potentially weak member banks and exacerbate a bank run (Gorton and Tallman 2018, 44). Cannon (1910) notes:

The first issuance of CLCs during the Panic of 1873 had an end date of November 1, 1873. However, it appears that later issues of CLCs did not have a set end date (Cannon 1910). The interest rate on CLCs likely served as a natural exit strategy, as they were reasonable during the crisis period but became costly as market financing became readily available (Gorton and Tallman 2018, 46).

All of the CLCs issued in 1873 by the NYCH were redeemed and canceled within four months of the first issuance (OCC 1907). The last CLCs were redeemed on January 14, 1874 (Cannon 1910).