Notes from the Vault

Warren E. Weber
May 2015

Today's institutions and technologies differ from those that existed in the past. Nonetheless, many past institutions and technologies are similar enough to those that exist today that much can be learned from studying their history. The notes issued by banks in the United States prior to 1933 are one such case. Bank notes were distinguishable by issuer and were payable to the bearer in gold or silver on demand. As a result, they share many of the characteristics of the financial instruments that are and would be classified as e-money. The only difference is that for e-moneys the monetary value is stored on an electronic device, and for bank notes it is "stored" on a piece of paper.1 This Notes from the Vault reviews the U.S. experience with bank-issued notes and draws some implications for government policy regarding e-money.

Overview of bank-issued notes
The U.S. experience with bank notes can be divided into two periods. The first, from 1786 to 1864, was a period in which bank notes were exclusively issued by banks chartered by the individual states. During this period, large numbers of distinct bank notes circulated simultaneously as there were a large number of banks in existence. For example, there were more than 250 note-issuing banks in 1820, more than 500 in 1835, more than 1,000 in 1854, and about 1,400 in 1860. Imagine living in a world with such a large number of different currencies. Notes issued by state banks were taxed out of existence in 1866.2

The second period, from 1866 to 1913, was a period in which bank notes were exclusively issued by banks chartered by the federal government. These banks were called "national banks," because they came into existence under the National Banking Act passed in 1864. This period provides an interesting contrast to the previous one because of the actions taken to correct the two major problems with state banks notes as a currency. These problems were that state bank notes were not safe and they were not a uniform currency.

State bank notes
State bank notes were not safe in the sense that these notes were typically not insured, so the holder could be subject to losses if the issuing bank went out of business. There was a large amount of turnover of state banks. Of the almost 2,400 state banks that went into business at one time or another during this period, approximately 40 percent went out of existence by 1860.3 More importantly, if a bank went out of business, there was approximately a 43 percent chance that the holders of its notes would suffer a loss. (see figure 1)

In general, note holders did not lose 100 percent of the value of the note if the issuing bank failed. For the most part, note holders lost between 13 and 40 percent of the value of the notes if the bank failed. But of course, the experience varied widely by bank. For example, for two banks, one in Indiana and one in New York, note holders only suffered a 3 percent loss. In contrast, for one bank in Vermont, the holders of its notes lost almost all (96 percent) when that bank failed.

The losses quoted above most likely understate the losses suffered by note holders. They are based on the rates at which holders of the note of a failed bank could ultimately get the note redeemed at a state banking authority. Such redemptions typically did not take place until well after the bank failed because of the time required to wind up a failed bank's affairs. If the holder of a note wanted to convert it to specie (gold or silver coin) sooner than that, the holder would most likely have had to take it to a note broker, who would make such an exchange. However, it is likely that the note would be discounted by brokers by more than the final redemption rate as a result of the time and trouble the broker would incur in getting the note eventually redeemed.

For state notes to have been uniform currency, the notes of various banks would have had to exchange at par (dollar-for-dollar) with each other, regardless of banks' locations and regardless of the location in which the exchanges took place. As an example, for state notes to have been a uniform currency, the notes of banks located in New York City would have exchanged one-for-one with notes of banks located in Atlanta regardless of whether the exchange took place in New York City, Atlanta, or some other location—say, Philadelphia.4

Such par exchange did not occur because of concerns that the issuer might fail and because of the cost of sending the note back to South Carolina if the holder wanted to redeem it for specie.

Notes of one state bank generally did not exchange with those of other state banks at par and the rates at which the notes of various bank traded depended on the locations of the trade. Further, the rates at which notes traded fluctuated over time. Figure 2 plots the discount on the notes of the Bank of South Carolina (located in Charleston, South Carolina), as quoted in New York (blue line) and in Philadelphia (red line) from 1840 to 1860.5 Figure 3 plots the discounts on the notes of the Suffolk Bank, located in Boston, and the Bank of South Carolina as quoted in Philadelphia from 1830 to 1858. The two figures show that the discounts on notes varied by the issuing bank, by the location in which the discount was quoted, and over time. For example, if the discount on Bank of South Carolina notes was 1 percent in Philadelphia and 5 percent in New York, it meant that a par value of $100 those notes could be exchanged for $99 of local bank notes in Philadelphia, but only $95 of local bank notes in New York.

National bank notes
There are two generally accepted reasons why the national banking system was established. The first is that it provided the Union with an additional means of financing the Civil War. The other was dissatisfaction with state bank notes for the reasons mentioned above. The National Banking Act addressed these problems as follows.

Providing federal insurance provided a solution to the problem with note safety: The U.S. Treasury would redeem a failed national bank's notes for "lawful money" regardless of whether the assets of the bank were sufficient to cover its note liabilities.6

Although federal insurance was not explicitly contained in the National Banking Act, it was implicit given the process for redeeming failed national banks' notes, which was as follows:

Immediately upon declaring the bonds of an association [national bank] forfeited for non-payment of its notes, the Comptroller shall give notice the holders of the circulating notes of such association, to present them for payment at the Treasury of the United States; and the same shall be paid as presented in lawful money of the United States. (Section 5229 of the Revised Statutes)

Nowhere does this redemption provision mention that notes would be paid off by the Treasury only if the assets of the failed bank were sufficient to pay off all notes or that redemption of notes would only take place after the value of the failing bank's assets had been ascertained. The Treasury would pay off notes in lawful money and worry about how it would get reimbursed later. Further evidence that national bank notes were fully insured by the federal government is the statement in the Annual Report of the Comptroller of the Currency that "the United States guarantees the final payment of the notes" (1874, XI).

Two methods addressed the problem that the notes of state banks were not a uniform currency. First, the National Banking Act required every national bank "shall take and receive at par, [italics added] for any debt or liability to said association [national bank] any and all notes or bills issued by any association existing under and by virtue of this act." This provision had the effect of making the nonissuer public indifferent in choosing between the notes of individual national banks. Second, in June 1874 a clearing facility for national bank notes located at and run by the U.S. Treasury where a national bank could take the notes of other national banks and receive lawful money in return. Because national bank notes could not count as reserves against deposits, without such a facility banks would not necessarily have been indifferent between national bank notes and lawful money. What made this clearing system lead to national bank notes trading at par is that the costs of note redemption were borne by the issuers of the notes being redeemed, not by the bank presenting notes for redemption.

Lessons for e-moneys
One lesson from the U.S. experience with bank notes is that fractionally backed, privately issued e-moneys will not be perfectly safe. Failures will occur without some type of government intervention. One possible government intervention is to require issuers to back all issues 100 percent with government currency, short-term government bonds, or deposits at the central bank. The drawback to such regulations is that they would reduce the profitability of issuing e-money and would likely result in issuers imposing transaction fees, in much the same way that today fees are associated with many deposit accounts and transactions with M-PESA, an e-money issued by Safaricom in Kenya.

My expectation is that the safety of privately issued e-money will not be accomplished in this way, because of the inconvenience caused by transaction fees. This suggests that another type of government intervention will occur. This intervention is government insurance. The argument is that even if there is no explicit government insurance, given recent central bank interventions with respect to very liquid financial assets, the public expects that government will step in and bail out the holders of an e-money should the issuer fail and be unable to redeem its issues at par. In other words, I assume that government cannot credibly commit to not intervene should an issuer of e-money fail, especially if it were a "large" or "systemically important" (too-big-to-fail) issuer. If this assumption is correct, then the insurance should be made explicit and implemented in such a way as to minimize the attendant moral hazard problems.

The second lesson from the U.S. experience with bank notes is that it is unlikely that privately issued e-moneys would be a uniform currency without government intervention. Although a private par clearing arrangement for state bank notes existed in New England (the Suffolk Banking System), it only handled notes of banks in that region, and no similar systems arose in any other part of the country.

Further, there are strong incentives for e-money issuers to not make them a uniform currency. E-moneys, because they are digital instruments, have the potential to convey a large amount of information about both the holder and the issuer. This creates a method and strong incentives for e-money issuers to introduce features that induce potential users to use their e-money rather than another issuer's. An example of the form that such incentives might take are the loyalty points currently given by many credit card issuers when their cards are used in certain types of transactions.

Based on evidence with state notes and the information carrying potential of e-money, my assumption is that e-moneys are unlikely to be a uniform currency without government intervention. Such intervention would most likely have to take the form of requiring par acceptance of the e-moneys of various issuers in all transactions and a government clearing system along the lines of the mechanism that made national bank notes a uniform currency.

The bottom line: If a country wants a safe and uniform currency, then some form of government intervention will be required. Without it, currencies will either be subject to losses or be nonuniform or both.

Warren Weber is a visiting scholar at the Bank of Canada and the Federal Reserve Bank of Atlanta and also a visiting professor at the University of South Carolina. The author thanks Larry Wall for helpful comments on the paper. The views expressed here are those of the author and do not necessarily reflect the views of the Bank of Canada, the Federal Reserve Bank of Atlanta, or the Federal Reserve System. If you wish to comment on this post, please e-mail Weber directly at or


1 I do not classify Bitcoin and most other digital and crypto-currencies as e-moneys because they are not a liability of any issuer, which is part of the usual definition of an e-money.

2 Contrary to expectations, state banks survived by replacing note issuance with the issuance of deposits.

3 There were so many banks because branching was prohibited in all but a few Southern states, meaning that banks were limited to a single location.

4 The notes of district Federal Reserve Banks are a uniform currency today because they exchange dollar-for-dollar everywhere in the United States and no one pays any attention to which reserve bank is the issuer.

5 Discounts are in terms of notes of banks in the city in which the discount is being quoted.

6 The term lawful money has no precise meaning, but for national bank notes it is taken to mean specie (gold and silver coin) and United States notes (greenbacks).