New York free banks and the role of reputations.
During the American era of free banking (1838-1863) the banking industry operated in an environment in which several different firms each produced their own paper money which they attempted to circulate. This environment was similar to that cited in the reputation literature as being likely to produce incentives for firms to invest in name brand capital. Such investment will occur in those situations in which product quality is variable and difficult to determine ex-ante.
During this era a secondary market in bank notes emerged. "The size of the discounts quoted on notes presumably varied with the geographic distance to the issuing banks, the perceived riskiness of that institution, and the quantity of counterfeit notes of that institution believed to be in circulation relative to the total issue."(1) Brokers published information concerning counterfeits along with current quotes on various notes in bank note reporters. Brokers trading in the secondary market also had strong incentives to monitor the quality of the assets backing bank notes since they collected specie in bulk as the source of their profits. It is these conditions, I argue, which led to, and provided the source of, bank investment in name brand capital.
This paper explores the reasons for the stable behavior of banks during the free banking period. Such behavior is a function of the competitive issue of money, which spawns competition among banks, each of which is interested in maximizing profits. With the governmental rules that were in place, profit maximization required efficient long-run operation. When multiple monies circulate in an economy, long-run profit maximization dictates that banks maintain a viable currency. In such an environment, the banking industry is similar to the environment cited in the reputation literature.(2)
What I argue here is that free banks, seeking long-run profit maximization, had the incentive to invest in reputations in order to enhance their note circulation. The results of this research indicate that banks did use many of the methods suggested in the reputation literature in an attempt to invest in name brand capital, and that this investment was recognized in the market for bank notes, affecting the price at which the notes exchanged in the secondary market. The reputation value of a bank is a measure of the probability that its notes will be redeemed. Each bank had its name printed on the notes it issued, and the name "represented the probability that the particular firm would fulfill its convertibility contract."(3) This was reflected in the price that the market set for the bank notes that were exchanged. As the reputation value of the bank decreased, the discount at which the bank's notes were sold in the market increased, and vice-versa.
Free Banking in New York
The demise of the Second Bank of the United States ushered in the American era of free banking. The expiration of the charter of the Second Bank dropped the sole responsibility for the nation's banking system into the laps of the individual states. The response by some was a system of banking based on the Second Bank of the United States: a state bank with several branches; for others it was an experiment somewhat misleadingly referred to as "free banking." In all states that passed free banking laws, banking was never truly free, since the government established certain prerequisites to opening a bank.
New York passed the second free banking act on April 18, 1838. It removed politics from the charter granting mechanism by eliminating the need for a special act of legislation to charter a bank, replacing it with a general banking charter. Included among the provisions of the new charter was the requirement that notes be secured by five percent stocks of the various states, or mortgages on New York real estate in amounts not to exceed one-half the value of the land. No stock would be accepted above its par value. Mortgages were accepted only on "improved, productive, unencumbered (sic)" lands within the state of New York at no more than fifty percent of the value of the land independent of any buildings on it. The securities were to be deposited with the state comptroller in the full amount of note issue. In addition, a specie reserve requirement equal to twelve and one-half percent was imposed on note circulation, semi-annual reports to the comptroller were required, and shareholders were exempt from liabilities for the bank's debts.
The New York Free Banking Act allowed "any person or association of persons . . . legally transfer[ing] to the comptroller any portion of the public debt . . . approved by the comptroller . . . to receive from the comptroller an equal amount of such circulating notes . . . to loan and circulate the same as money, according to the ordinary course of banking business as regulated by the laws and usages of this state."(4) The new law made the chartering of banks an administrative, rather than legislative, function of state government. The law allowed anyone to set up a bank by complying with the conditions set forth in the act. Once these conditions were met, the appropriate administrative authority would then issue a charter. The result was that "it might be found somewhat harder to become a banker than a brick-layer, but not much."(5)
The passage of the free bank act created two types of banks in the state of New York. Safety Fund, or incorporated, banks were those banks that were incorporated prior to 1838. Each was chartered by an individual act of legislation, and belonged to the safety fund system. Free banks were chartered after 1838 (this included all safety fund banks whose charters expired and who wished to recharter).
The New York free banking act was subject to much fine-tuning over the years. Among the earliest changes in the law were provisions that limited the stocks eligible to back note-issue. In 1840 the use of any state bonds other than those of New York was prohibited. Two other significant changes were made to the original act in 1840. The twelve and one-half percent specie reserve on circulating notes was repealed in May, and a stringent deposit requirement for all country banks was instituted that same month. The new provision called for all banks outside of New York City, Albany, Brooklyn, and Troy to provide for the redemption of their notes in either New York City, Albany or Troy at no more than a one-half percent discount, changed to a maximum of a one-quarter percent discount in 1851.
The beginning of the end of the free banking era came with the passage of the National Bank Act in 1863. This act was passed in response to the government's need to finance the war. While the act effectively extended free banking to the entire country, it did so by establishing a national banking system, featuring a uniform currency, designed to replace the system of state banks issuing their own notes. The federal government established national banks with free entry given the attainment of criteria very similar to those set out in the New York free Banking Act.(6) Federal government control of the banking system was renewed for the first time since the expiration of the charter of the Second Bank of the United States.
Reputation Theory and Banks
The American free banking era illustrates a problem that exists when product quality is variable. If it is difficult for consumers to determine the quality of a good before and even after its purchase, theory predicts that firms will attempt to develop reputations to signal quality. In situations when quality is difficult to monitor ex-ante, the development of reputations can be especially valuable. There have been several articles detailing the economic role of reputations, among which the most influential are Akerloff (1970), Klein and Leffler (1981), Kreps and Wilson (1982), Rogerson (1983), Schmalensee (1978), Shapiro (1982), Smallwood and Conlisk (1979), and Stigler (1961).
Reputation theory outlines those instances in which the formation of name brand capital is necessary, and how such capital investment by firms will occur. Reputations are important in markets in which products are differentiable by quality and those differences are not easily or costlessly discernible to potential purchasers ex-ante. This is especially important for those goods for which quality is difficult to determine even after purchase. Bank notes are an example of such a good because their quality often cannot be determined even during use, i.e. the holding period. Rather, their quality may not be discernible until the holder attempts to dispose of them by purchasing goods, redeeming the notes at their place of issue, or trading them to a broker.
When a buyer receives a bad bank note, he loses money. Thus, to convince buyers to accept his notes of unknown quality, a banker must persuade potential buyers that the bank would suffer a substantial loss if its note is found to be of a lower than advertised quality. The most basic measure of quality from the point of view of the note holder is how easily and quickly the note can be exchanged for real goods or redeemed for specie. By backing the quality of the note with its brand name, the bank then faces the consequences that a bad note will tarnish the value of all other notes.
The name brand value of a firm will depend on its reputation, and the critical issue for any bank is the degree of confidence that it can establish. Money is not accepted unless people believe that they can exchange it for goods and services. Money does not have to be legal tender to be accepted, but rather, it must be "what one might call common tender, i.e., commonly accepted in payment of debt without coercion through legal means. Indeed, privately issued money to exist at all would have had to be common tender, and would have had to earn its acceptability in a market environment."(7) The belief that others will accept the money in exchange for real goods and services is the underlying principle which keeps all money systems operating. If a money lacks the confidence of the people that it will be accepted as a means of exchange by others, it will cease to circulate as a money (Claassen 1984, Menger 1892). Following this line of reasoning, the brand name value of a competitive money issue will be its lifeblood, because "for fiduciary money, consumers must rely solely on the brand name method of obtaining information about quality since the monetary service flow from a money is assumed to be independent of any technical characteristic of the money (e.g. the size or color of a currency)."(8) If a bank develops a reputation for issuing bad money or not redeeming its money as promised (if the money is a fiduciary currency), its money will be driven from the market. Lawrence White (1984) underlines this point by citing the experience of 19th century English country banks. Their weakness led to their frequent failure in good times as well as bad. This record in turn enhanced their cyclical instability, for it undermined public confidence in the banks.
In the case of a good whose quality cannot be determined prior to purchase, one indication of quality is the existence of some form of "bonding" on the part of the firm. Bonding by a firm is the process of investing in some firm-specific, non-salvageable asset that will be lost to the firm if it should shirk on its promise of high quality in order to take short-run profits and then go out of business. One such example is the physical assets of the firm. Large and imposing buildings, fancy interiors, engraved stationery, and ornately designed bank notes are all examples of non-salvageable investment that a bank would lose should it produce a good of less than expected quality. In addition, the use of such grandiose accouterments is intended to be awe-inspiring, instilling the customer with a sense of importance and sophistication deserving of a noble and trustworthy institution. This bond formed by the seller is an attempt to assure the customer that the firm has no intention of producing anything less than what it promises.
Banks recognized the importance of having a flashy image, and one way of promoting such an image was through the banknotes that circulated among the public. In an attempt to make an impression on potential customers, one Michigan banker gave the following instructions or the design of his notes: "Get a real furioso plate, one that will take with all creation -- flaming with cupids, locomotives, rural scenery, and Hercules kicking the world over."(9) Certainly, reasoned this banker, such an ostentatious note would not be issued by a fly-by-night bank.
Tibor Scitovsky (1945) points out some additional quality indices, such as the size of a firm, its length of service, and its degree of financial success. This is one reason for the importance producers often attach to goodwill and trademarks, hence the much advertised claims of being the biggest or oldest finn in their market. These are indicators of brand name quality because they are a form of bonding that the firm provides the customer. Such bonding serves as a guarantee to the customer that the firm has a commitment to excellent service in the long run and is not out to make a quick killing before going out of business.
The length of service and financial success are signs of trustworthiness, and hence have some brand name capital value. A firm that has been in existence for a long time can point to its past record of performance as proof that it will still perform tomorrow. Past performance is one of the most valuable forms of name brand capital by virtue of the fact that it is evidence of the ability and willingness of a firm to deliver on its promises. The performance level of the good can be estimated by observing its past performance levels, and the longer the record of quality service, the more convincing the claim that the same type of performance will continue in the future. This is especially true if that service record shows evidence of having performed through some tough times without having failed to live up to expected quality levels.
Successful past performance is particularly important for a money. As an example, White (1984) cites the case of the Ayr bank in Scotland. On the day before the struggling Ayr bank failed, the Bank of Scotland and the Royal Bank advertised that they would accept the notes of the defunct bank. The potential benefits of this action to the two banks are clear: it would bolster public confidence, attract depositors, and help put their own notes into wider circulation. According to Claassen (1984), the money producer has costs of selling (creating confidence in) his money as well as costs of physically producing it. These selling costs are equivalent to the investment costs of building up a brand name capital which assures a degree of money confidence.
In Wisconsin, even after the legislature repealed its charter, George Smith's Wisconsin Marine and Fire Insurance Company continued to receive deposits and lend money. It had already compiled a record of eight years of stellar service, and after revocation, it announced publicly that the repeal was illegal and it would continue to do business and meet its obligations. Farmers, merchants and dealers went on using the company's obligations and there was nothing the legislature could do about it.
Past performance records of banks can be analyzed in a number of ways, the most significant of which are the length of time the bank has been in business, and the frequency with which it has failed to redeem its notes. Of additional importance is the length of time since the bank last failed to redeem its notes and the conditions under which the bank has been in service. Difficult financial times may either excuse the bank somewhat for not redeeming (i.e. if most banks temporarily suspended then the relative damage is minimized) or reward the bank that redeemed despite difficult times. In the future the bank can lay claim to having maintained redemption, or point to the fact that its notes never failed to sell at or near par in the markets, even though other banks did not fare as well during tough times.
The Chemical Bank in New York City is an example of the importance of past performance in establishing a reputation. The Chemical Bank was known for a long time as "Old Bullion" because it was the only New York City bank that did not suspend specie payments during the panic of 1857.(10) The fact that the Chemical Bank was able to maintain specie redemption while all other banks in the city suspended was not overlooked by the public, and came to be a bragging point for the bank.
Note-issuing banks benefit by having more customers because this translates into a larger number of their notes circulating. Profits are to be made on these circulating notes because the assets that are purchased with them (stocks, bonds, and loans) are earning the bank a return, and as long as the notes continue to circulate, the bank can earn this return. The more often the notes are redeemed, the larger the non-interest earning amount of specie the bank must keep on hand to cover these redemptions. It is somewhat paradoxical in that banks prefer not to have their notes redeemed, yet one of the most effective ways of preventing frequent redemption is to make it convenient to do so. A note that a consumer believes to be readily redeemable is one that he has confidence in holding and does not feel the need to redeem quickly. This is directly related to the degree of acceptability that others show in the notes. If they believe the notes are easy to redeem (a measure of the soundness of the notes) then they will be more willing to accept them.
The easier banks make it for their notes to be redeemed, the lower are the associated transaction costs for holding those notes. Transaction costs of holding a note include the cost of calculating exchange rates between it and other notes or goods, the cost of determining whether the note is genuine or a counterfeit, the search costs associated with finding someone who is willing to take the note in an exchange, and the cost of redeeming the note, including the time and distance one must travel to find a place of redemption. The easier it is to exchange a bank's notes for other goods and notes, the more valuable that note becomes, because it is more widely accepted and has lower transactions costs associated with it. This in turn leads to an increase in the value of the bank's brand name. A bank note that is difficult to exchange will have a poor reputation, which will reflect negatively on its name brand value.
Banks recognized the advantages of having balances on deposit at other banks, especially those in New York City. By maintaining balances in New York City, the center of most trade at the time, they were able to allow for their notes to be redeemed in the city, and thus keep them circulating at or near par. State bank notes tended to accumulate in large cities, especially those on the eastern seaboard, New York in particular. The net flow of goods from New York City to the rest of the country necessitated a flow of funds in the opposite direction, thus state bank notes tended to accumulate in the cities. Deposits made it easier to redeem, thus keeping the notes at or near par. These balances constituted interest earning reserves for the country banks. The city banks could afford to pay this interest because they would loan out these funds. A report filed by the Duchess County Bank of New York notes:
Since the last annual return, as before, the bank has kept an account in the North River Bank, in the City of New York, in order to have its bills receivable and current in New York, so that such bills might pass in the state and elsewhere, without discount; and the directors have accordingly ordered such surplus funds, as were not needed at the bank for its ordinary business, to be sent to the said North River Bank, to redeem the paper there, and the sum above stated as being in that bank are the funds of this bank, placed there for the aforesaid purpose, with the ordinary means of the bank; and these deponents also say, that they have used such means whenever requested by their customers and persons holding their bills, by giving drafts on New York at sight, and without premium.(11)
The fact that a bank maintains deposits in other banks serves as a means of fostering confidence and hence investing in reputation. A country bank which deposits money in a bank in a major commercial center such as New York, does so "for the fact of its bills being current in [New York] inspires confidence in the soundness of the bank, and its bills are consequentially less liable to be presented for specie by their holders."(12) This was recognized at the time as a method of inspiring confidence, which as a result kept the notes circulating at or near par and kept them from being presented for redemption at a high rate, thus being more profitable to the bank.
Holding excess specie reserves also serves as a means of investing in name brand capital. If the excess holdings are advertised to the public, it serves as a bonding mechanism. Banks seem to assure the public that sufficient liquidity is being maintained. "Under the system of free competition, every bank would be obliged to rely upon its own unassisted credit; and in order to obtain a circulation, it would be necessary either to have capital, or to have reputation of it."(13) Even if the public is not aware of the bank's holdings of excess reserves, they serve as an indirect method of investment in name brand capital via their potential effect on the performance record of the bank. The holding of the reserves is a form of insurance against suspending note redemption during a panic. The excess reserves can be used during periods of unrest, when notes are usually brought in for redemption in greater quantities. This pays off in the long run for the bank by allowing it to stave off suspension in the face of heavier than normal redemption demands, keeping intact its record of maintaining note redemption. Reference can once again be made to the case of the Chemical Bank of New York.
A history of successful past performance helps to build confidence in the bank and its notes. In competition, banks without successful performance records would, ceteris paribus, have lower brand name values, or would not be able to build up as much brand name capital, which would reduce their chances of surviving in a market depending so much on confidence.
Reputations in the Marketplace
Banks will invest in name brand capital only if the value of such investment is recognized in the marketplace. An attempt to make such a determination is carried out in the following simple regression which focuses on the discount rate at which New York bank notes sold in the Philadelphia market, and the factors affecting that discount rate. The discount rate of an individual bank's note is not equal to its name brand value, but is affected by it. The discount rate reflects the bank's reputation, that is, the probability that the note will be redeemed, as well as other things, most notably the cost of transporting the notes to the bank of origin for redemption.
The dependent variable is the discount rate, calculated as the annual average of the monthly discount rate quoted in bank note reporters for notes sold in the Philadelphia market. The rate is quoted as a percentage of par (100) in terms of local Philadelphia notes. A note with a discount of .25 will have a discount rate of 99.75, and a note with a discount of 5 will have a discount rate of 95.
TRANSPO is the transportation cost which is proxied by the rail distance of the bank from Philadelphia. These distances were obtained from contemporary railroad timetables and maps. AGE is the length of time the bank has been in service, measured in months. LAST SUSPEND is the amount of time since the bank last failed to redeem its notes for specie. The frequency of suspensions is represented by the variable SUSPEND FREQ. This is measured as the ratio of the total number of months during which a bank suspended specie payment to the total number of months the bank has been in service.
The term "suspension" refers to the failure of a bank's notes to sell in the secondary market. The failure of a bank's notes to sell during a given month could have been for a number of reasons, all of which, however, were related to the bank's failure to redeem its notes: a temporary shutdown, a closure of a longer duration, or a failure to have enough specie on hand. If the bank note reporter did not list a discount rate for a bank, it usually gave a specific reason for the omission, such as: the notes were not selling because there was no demand for them because it was believed they could not be redeemed, or the buying agent refused to buy them because the bank was believed to be unstable and it was likely the notes would not be redeemed, or the bank was in a weak condition and had suspended specie payments.
Although this definition of suspension rests on the actions of market agents, the act of suspending is still a function of individual bank behavior. To the extent that the bank is responsible for its own reputation, it can influence the level of confidence agents have in its ability to redeem its notes. If the bank has established a reputation for successfully redeeming its notes in the past, it is less likely that the notes will be suspect. The more confidence the bank can imbue in the consumer regarding the redeemability of its notes, the less likely it is that anyone will believe otherwise.
The quality weighted measure of service is QWMS, where the quality of a month is measured by the percentage of bank failures and suspensions during that month. This variable is based on the number of months of uninterrupted service: a period of time during which the bank continued to honor its notes by redeeming them for specie without suspending. The weighting measure is determined by the total number of banks whose notes did not trade during the period of time in question. The bank can make the claim that while other banks suspended, or their notes failed to trade, during this period of difficulty, it did not. It did whatever was necessary in order to meet its redemption commitments. The tougher the climate the more impressive the claim that one's bank maintained redemption throughout. To maintain service during a boom period is relatively easy, but to do so during a contraction is not. A bank able to remain in service during a given period of time when a large percentage of other banks could not, makes a credible reputation statement concerning its performance record.
If no banks suspended during a month the weight is equal to zero. If all banks suspended during a month the weight is one. The QWMS is summoned over all months during which the bank's notes sold. If a bank's notes did not sell during a month, then its QWMS for that month was zero. The greater the percentage of banks suspending, the higher the "quality" of the month, thus the more impressive it is for the bank to have "survived" through it. This quality weight provides a measure of the rigors of the environment in which the bank performed.
CIRC is the bank's total note circulation. This variable is a proxy for the size of the bank's operations. CIRC RATIO is the ratio of the notes the bank actually had in circulation to the amount of notes which it was authorized to circulate. The authorized level of circulation was equal to the face value of the securities that the bank had on deposit with the comptroller for free banks and twice the value of paid-in capital for incorporated banks.
Finally, two dummy variables are included. INC has a value of one for those banks incorporated prior to the passage of the free banking act, and belonging to the safety fund system, and a value of zero for free banks. The dummy variable YEAR takes on a value of one for those banks which were only in service during the final year of the sample, and zero for those banks in operation prior to 1855.
The sample, which includes all banks in operation from 1851 through 1855, comprises it subset of the New York free banking era. This is a brief time period involving the state which is considered to have had the best free banking system in the country, and thus provides for the weak form of the test. The results of the empirical investigation in this initial foray into the role of reputations in free banking environments is meant only to test for the existence of reputation effects in the discount rate. A much more thorough test must be made in the future using longer time periods and a variety of free banking systems in order to test the limits of the role of reputations. The evidence here merely indicates that they do exist, at least in the short run and in the best free banking environment. As limited as the results are, they are still important for the fact that they are the first such results and establish a base from which to work in the future.
The sample set is created by pooling the data for each year for both incorporated (safety fund) and free banks. All banks are combined to form one data set for the five year period, which was then corrected for heteroscedasticity before the regression was run.(14) The results of the regression are as follows, with t-statistics in parentheses:
DR = 145.49 + .07 TRANSPO - .0001
(267.42) (11.43) (-1.33)
CIRC - .03 AGE + .14 LAST SUSPEND (-2.83) (23.23)
+ .24 SUSPEND FREQ + .20 QWMS (2.40) (10.30)
+ .46 CIRC RATIO - 10.92 INC (20.94) (-3.34)
- 5.90 YEAR (-1.84)
[R.sup.2] = . 842 N = 1197 F = 704.6
The expected sign for the transportation variable is negative. The hypothesized negative relation between distance and discount rates is widely cited in the literature (Gorton and Mullineaux 1987, Hammond 1957, Madeleine 1943, Sylla 1975). It is cited by many as the-major or even only reason that bank notes sold at a discount, and the reason why discount rates among different bank notes varied.
The farther the town is located from Philadelphia, the greater a note should be discounted. The greater distance means a greater expense of transporting notes back to the bank of origin from Philadelphia for specie. In addition, the more remote the bank, the more difficult it is for the bank to disseminate information concerning its operations, and the more difficult it is for individuals to gather that information.
While the transportation variable is significant at the .05 level, it has a positive sign. This suggests that transportation does not have the hypothesized effect on the discount rate. In fact, it seems to suggest just the opposite, that the discount on the note will be greater the closer the bank is to Philadelphia.
One possible explanation for this inconsistency is that the transportation costs actually have an uncertain effect on the discount rate because of the 1840 law requiring all banks to provide for the redemption of their notes in either New York City, Albany, or Troy. I tested for this possibility by regressing dummies for these banks, as well as recalculating transportation distances for all banks. In recalculating the transportation distances, I used the distance from Philadelphia to the redemption agency for each country bank, rather than the distance to the actual location of the bank. In neither case however, did the new variables prove to be significant.
Another possible explanation could be the spatial relation of production centers and markets in the 19th century. It could be that the positive sign on the transportation variable reflects the effect of notes trading from important trading centers at points distant from Philadelphia. For example, if bank notes from Buffalo dominate the "long distance" notes, while notes from less distant cities are from smaller, insignificant trading centers, then the fact that the transportation variable is positive reflects the fact that these bank notes circulate at a smaller discount in part because of the volume of trade that occurs with this distant city. An examination of the available data does not, however, support this theory. Table 1 shows the mean discount rate and the standard deviation for all New York banks, as well as four subsets. The average discount rate (measured as percentage of par) is lower for Buffalo banks than for all other country banks. That is, Buffalo bank notes traded at a deeper discount than other country banks. In fact, the Buffalo banks display a lower discount rate and a higher standard deviation than any other subset. This does not support the hypothesis that "long distance" notes from Buffalo dominated the trading and circulated at a lower discount rate.
The transportation variable may also be picking up some demand factor by Pennsylvania brokers for New York notes vis-a-vis other bank notes. The New York system was the most stable state banking system of the era, and the New York notes were in great demand both for this reason and because of the large amount of commerce that took place in New York City. Evidence for this can be found in the premium rates paid for some New York bank notes by Philadelphia brokers. These premiums occurred because the quotes were given in terms of Philadelphia notes, not specie, in the bank note reporters published in Philadelphia. There were, therefore, some instances in which brokers were willing to pay a premium in terms of local notes for the privilege of holding New York notes.
TABLE 1 Discount Rates of New York Free Banks, 1851-55, By Location All Banks New York City Buffalo Country I Country II Average 99.43 99.85 98.16 99.32 98.98 Std Dev 2.22 0.33 7.81 2.44 2.27 N 916 176 53 740 687 Source: Philadelphia Reporter, Counterfeit Detector, Philadelphia Price Current, and General Advertiser, 1981-55. Notes: Country I includes all banks outside of New York City. Country II includes all banks outside of New York City and Buffalo. Average is calculated as annual average of monthly discount rates for the five year period 1851-55.
The wrong sign on the transportation variable is puzzling, but not damaging to this analysis. My claim is that the discount rate is affected by the reputation of the bank issuing the note. Whether the note is affected negatively by transportation costs does not directly affect my claim that discount rates will be effective in transmitting reputation information. If the discount on the notes is not a result of transportation costs, as so much of the literature claims, then it must be a result of something else. "Something else" includes the bank's reputation.
The variables AGE, LAST SUSPEND, SUSPEND FREQ, QWMS, CIRC RATIO, and INC are significant at the .05 level. However, AGE and SUSPEND FREQ do not display signs consistent with the theory. That is, the age of the bank should operate to increase the discount rate closer to par as the age of the bank increases, and the frequency of suspensions should put downward pressure on the discount rate as that frequency increases, but the regression results seem to indicate the opposite.
The variables QWMS and LAST SUSPEND have positive signs, consistent with the above hypothesis that as the length of time since the last suspension increases, so does the price (discount rate) individuals are willing to pay to hold the note, and as the quality of a bank's past performance increases, the discount rate of the note approaches par. The performance of these two variables lends credence to the theory that bank investment in reputation was recognized by the participants in the banknote market, and appropriately affected the price at which bank notes were exchanged. Those banks which evidenced high quality behavior in the past (as measured by QWMS and length of time since last suspension) saw their notes exchange at a higher discount rate than the notes of those banks with lower quality past performances.
CIRC RATIO, the measure of the ratio of a bank's actual circulation to its legal maximum tested significant with a positive sign. The positive sign on this variable may be picking up the demand for the bank's notes. The greater demand for a bank's notes would result in a larger percentage of those notes remaining in circulation, and would put upward pressure on the discount rate, resulting in a note trading at closer to its par value. This explanation supports the contention that a bank with a strong reputation enjoys a larger circulation of its notes at a price closer to the par value. This is a result of the strong demand for the notes which stems at least in part from the confidence in their being widely acceptable because of the belief that they are easy to redeem for specie whenever necessary.
The variable CIRC, though insignificant, has a negative sign, contrary to the hypothesis. This would seem to indicate that it is not the size of the bank which matters as much as the reputation that the bank has exhibited over time. A large bank by itself will not necessarily evoke confidence from the populace. Rather, it is what the bank has done, especially in the recent past, that drives the current demand for its notes.
Finally, the negative sign on the dummy variable INC indicates that free bank notes tended to trade at a higher discount rate than did the notes issued by incorporated banks. An explanation for this is not immediately apparent, though it could have been because free banks had deposited securities with the comptroller as an ultimate mode of note redemption should they fail. The value of these securities could be tracked by agents by monitoring the bond market, thus providing an accurate measure of the ability of bank notes to be redeemed in the face of crisis. Incorporated banks, however, though they belonged to the safety fund system, provided no such means by which agents could independently verify the value of the portfolio held to back circulating notes.
Another possible explanation for the negative sign on INC could be a residual fear that brokers had of the safety fund system's solvency. The fund had experienced losses in the early 1840s which it was unable to completely cover even as late as the period under study here. The fear that a rash of bank failures would finally bankrupt the system may have made brokers wary of these banks. It is possible that brokers felt the bond-backed system of free banks was relatively more stable than the safety fund system.
The results of the regression, while providing some support for the argument that bank reputations were recognized in the market and affected the discount rate, are tempered somewhat by the counter-intuitive signs on the transportation, age, and suspension frequency variables, and the use of the New York free banking system as a test case.
The New York free banking system is noted for its stability among free banking systems, and for this reason it provides for the weak test of the argument that reputations were effective. Indeed, Table 2 indicates that for three of the years in the sample there was very little deviation in discount rates. In the years 1852 and 1855 the standard deviation of discount rates among all banks was less than one-quarter of one percent, as all notes traded at no more than a one percent discount. In 1853, the minimum discount rate was 96 percent of par, but the standard deviation for the sample was only . 17 percent of par. In 1854, there was considerably more deviation to explain, with notes trading at discounts ranging from 33 percent of par up to par. In 1851 the minimum discount rate was only 90%, but the standard deviation was 9.67%. Nevertheless, the results do lend support to the contention that investment in reputations occurred, and was recognized in the market. The focus of this paper was to identify the existence of reputation investment and to test for its recognition in the market. In two of the five years of the sample, the variance among discount rates was significant enough for reputations to have made a noticeable impact. The fact that in the other three years there was little to explain is a testament to the stability of the New York banking system, and an indication that future research needs to be carried out using other free banking systems to provide a more strenuous test of the theory.
TABLE 2 Discount Rates of New York Free Banks by Year All Free Banks 1851 1852 1853 1854 1855 Average 97.63 99.45 99.61 99.17 99.54 Maximum 100 100 100 100 100 Minimum 90 99 96 33 99 Std Dev 9.67 0.24 0.17 5.09 0.18 N 106 121 162 170 179 Country Banks 1851 1852 1853 1854 1855 Average 98.74 99.33 99.51 98.98 99.45 Maximum 100 100 100 100 100 Minimum 90 99 96 33 99 Std Dev 4.40 0.07 0.05 5.86 0.05 N 83 93 118 127 141 Source: Philadelphia Reporter, Counterfeit Detector, Philadelphia Price Current, and General Advertiser, 1851-55. Note: Country banks are those banks outside of New York City. Average is calculated as annual average of monthly discount rate for all New York banks in operation each year.
In summary, the regression results indicate that the past performance of banks is reflected in the price at which bank notes trade in the secondary market. In the case of these reputation proxies, all are under the bank's control, and all represent investment by the bank in their name brand capital. The successful past performance of the bank, especially during difficult times, is an important and effective method of instilling confidence in prospective bank note holders. A bank that has proven that it will honor its notes will be more likely to do so in the future, and its notes will therefore be in greater demand and sell closer to par.
The American free banking era illustrates a problem that exists when product quality is variable. If it is difficult for consumers to determine the quality of the goods before and even after their purchase, theory predicts that the firms will attempt to develop reputations to signal quality. In situations when quality is difficult to monitor ex-ante, the development of reputations can be valuable. This research examined whether those reputations were enough to give the correct behavioral incentives to banks by asking the questions: given the banking structure that existed at the time, did banks use those methods described in the reputation literature to establish reputations, and were those methods effective?
There is significant anecdotal evidence suggesting that attempts were made by banks to establish reputations during the American free banking period. Contemporary market participants recognized that note-issuing banks had the opportunity to profit at the expense of their customers, and that establishing a credible claim to honesty was important. The regression results suggest that bank reputations were recognized in the market. The value of a bank's reputation was reflected in the discount at which the bank note traded.
Despite various state regulations, both the opportunity and the incentive existed for banks to engage in fraudulent banking practices, and this fact was not lost on the banking public of the era. For this reason, those banks that intended to honor their commitments found it worthwhile to invest in their name brand capital in an effort to signal their intentions to the public and increase their circulation and hence their long-term profit potential.
The use of limited data from the state of New York represents the weak, but important, case that banks invested in their reputations, which were in turn reflected in the discount rate. It is the weak form of the argument because it is a short time period and a best case scenario, but it is nevertheless important because it serves to establish a case where reputation theory can be used to suggest that free banks were not necessarily prone to instability and wildcatting, but incentives existed for them to exercise self-discipline.
Future research in this area is to be encouraged by the findings here, specifically in the area of testing free banking systems of states not renowned for the quality of their free banking systems, over longer time periods, and to measure the value of reputations and the time it takes to create and destroy them. This test however, serves as a starting point for all of that. At least it exists in the weak case. The task before us now is to test its limits.
1. Gorton and Mullineaux (1987), p. 458.
2. See Akerloff (1970), Klein and Leffler (1981), Kreps and Wilson (1982), Nelson (1970), Shapiro (1982), and Smallwood and Conlisk (1979).
3. Klein (1974), p. 440.
4. An Act to Authorize the Business of Banking, as appears in Hildreth (1968).
5. Sylla (1975), p. 33.
6. National banks were required to deposit U.S. government bonds with the comptroller of the currency to back their note issue. After depositing the bonds, the bank would then be given national notes (with its name stamped on them) equal to ninety percent of the face value of the bonds. There were also minimum capital requirements according to the size of the town in which the bank was to be located. These ranged from $200,000 for a town with a population of 50,000 or more down to $50,000 for a town of 6000.
7. Timberlake (1987), p. 438.
8. Klein (1974), p. 430.
9. Hammond (1957), p. 600n.
10. Klebaner (1974), p. 32.
11. Margaret G. Meyers, The New York Money Market: Its Origins and Development, (New York, 1931), p. 105.
12. Merchants Magazine and Commercial Review, 5 (September, 1841), p. 262.
13. Hildreth (1837), p. 155.
14. Bartlett's test was used to determine the existence of heteroscedasticity occurring as a result of different mean squared errors across pooled groups. The correction is made by standardizing each group by its standard error of the estimate before pooling. The pooling tests for the data is an F test comparing the restricted and unrestricted sums of squared errors.
Akerloff, G. (1970), "The Market for 'Lemons': Quality Uncertainty and the Market Mechanism." Quarterly Journal of Economics 84, 488-500.
Claassen, E. M. (1984), "Monetary Interpretation and Monetary Stability: The Economic Criteria of the Monetary Constitution." In P. Salin (Ed.), Currency Competition and Monetary Union. Boston: JIjihof. pp. 47-58.
Gorton, G. and Mullineaux, D. J. (1987), "The Joint Production of Confidence: Endogeneous Regulation and Nineteenth Century Commercial-Banking Clearinghouses." Journal of Money, Credit, and Banking 19, 457-68.
Hammond, B. (1957), Banks and Politics in America from the Revolution to the Civil War. Princeton: Princeton University Press.
Helderman, L. C. (1931), National and State Banks: A Study of Their Origins. Boston: Houghton Mifflin.
Hildreth, R. (1837), The History of Banks. Boston Hilliard, Gray & Co.
Hughes, J. R. T. (1987), American Economic History. Glenview: Scott Foresman.
Judge, G. et. al. (1980), The Theory and Practice of Econometrics. New York: John Wiley and Sons.
Klebaner, Benjamin (1974), Commercial Banking in the U.S.: A History. Hinsdale: Dryden Press.
Klein, B. (1974), "The Competitive Supply of Money." Journal of Money, Credit, and Banking 6, 423-53.
Klein, B. and Leffler, K. (1981), "The Role of Market Forces in Assuring Contractual Performance." Journal of Political Economy 90, 615-42.
Koutsoyiannis, A. (1973), Theory of Econometrics. London: MacMillan.
Kreps, D. and Wilson, R. (1982), "Reputation and Imperfect Information." Journal of Economic Theory 27, 253-79.
Menger, K. (1892), "On the Origin of Money." The Economic Journal 2, 239-55.
Merchants Magazine and Commercial Review. Various issues, 1841-57.
Meyers, M. G. (1931), The New York Money Market: Its Origins and Development.
Nelson, Phillip (1970), "Information and Consumer Behavior." Journal of Political Economy 78, 311-29. New York: Columbia University Press.
Philadelphia Reporter, Counterfeit Detector, Philadelphia Price current, and General Advertiser (1851-55). Philadelphia: J. Van Court.
Rockoff, H. (1975), The Free Banking Era: A Reconsideration. New York: Arno Press.
Rogerson, W. (1983), "Reputation and Product Quality." Bell Journal of Economics 14, 508-16.
Rolnick, A. and Weber, W. (1985), "Explaining the Demand for Free Bank Notes." Federal Reserve Bank of Minneapolis Staff Report 97.
Schmalensee, R. (1978), "A Model of Advertising and Product Quality." Journal of Political Economy 86, 485-503.
Scitovsky, T. (1945), "Some Consequences of the Habit of Judging Quality by Price." Review of Economic Studies 12, 100-05.
Shapiro, C. (1982), "Consumer Information, Product Quality, and Seller Reputation." Bell Journal of Economics 13, 20-35.
Smallwood, D. and Conlisk, J. (1979), "Product Quality in Markets Where Consumers are Imperfectly Informed." Quarterly Journal of Economics 93, 1-23.
Stigler, G. (1961), "The Economics of Information." Journal of Political Economy 69, 213-25.
Sylla, R. (1975), The American Capital Market. 1846-1914. New York: Arno Press.
Timberlake, R. (1987), The Origins of Central Banking in the United States. Cambridge: Harvard University Press.
White, L. (1984), Free Banking in Britain: Theory, Experience and Debate, 1800-1845. New York: Cambridge University Press.
Michael J. Haupert
The author is Associate Professor of Economics at the University of Wisconsin at La Crosse, La Crosse, WI 54601. I would like to thank Howard Bodenhorn, Barry Clark, Larry Daellenbach, Art Denzau, Jane Knodell, Gary Libecap, Don McCloskey, Douglass North, John Nye, the participants at the 1990 Economic History Association meetings and the 1991 Cliometrics Conference, as well as two anonymous referees for valuable comments and suggestions. All errors remaining in the analysis and interpretation are mine alone. Financial support was provided by the Bradley Foundation and a Cole Grant from the Economic History Association.
|Printer friendly Cite/link Email Feedback|
|Author:||Haupert, Michael J.|
|Date:||Sep 22, 1994|
|Previous Article:||Hyperlecturing and linkages.|
|Next Article:||A note on modelling OPEC behavior 1983-1989: a test of the cartel and competitive hypothesis.|