The Real Bills doctrine holds that issuing money in exchange for real bills is not inflationary. It is best known as "the decried doctrine of the old Bank Directors of 1810: that so long as a bank issues its notes only in the discount of good bills, at not more than sixty days’ date, it cannot go wrong in issuing as many as the public will receive from it.'" (Fullarton, 1845)
Traces of the real bills doctrine can be found in the writings of John Law (1705), Simon Clement (1710), Adam Smith (1776), Charles Bosanquet (1810), Thomas Tooke (1845) and many others. It was at the heart of the Bullionist debates of 1810, the Banking School/Currency School debates of the 1840s, the Greenback debates of the 1870s, etc.
The Real Bills Doctrine was the cornerstone of the US Federal Reserve Act of 1913—which established the Federal Reserve System with the power to discount high-quality self-liquidating commercial paper; however it did not become a major policy tool of the Federal Reserve until after Benjamin Strong, governor of the New York Federal Reserve Bank died in October 1928. Since 1945, it has been regarded as "thoroughly discredited" (Mishkin, 2000) among mainstream economists.
Issuing money "in the discount of good bills" is a strange concept to the modern reader. The banker's T-account below will clarify the concept.
|100 oz. silver deposited||100 paper dollars|
|Farmer's IOU worth $200||200 paper dollars lent|
|Gambler's IOU worth $300||300 paper dollars lent|
In line 1, the banker receives 100 ounces of silver on deposit, and issues 100 paper receipts (“dollars”) in exchange. Each paper dollar is convertible at the bank into 1 ounce of silver. At this point each paper dollar will be worth 1 ounce of silver in the open market. Note that it is immaterial whether the dollars are issued as printed pieces of paper or as bookkeeping entries transferable by check or other means.
In line (2) we suppose that a farmer requests a loan of 200 paper dollars from the bank. Assuming the farmer offers adequate collateral and pays an adequate interest rate, any profit-seeking banker would agree to print 200 additional paper dollars and lend them to the farmer. The farmer, for his part, might write an IOU to the banker, promising to pay $220 after 1 year. At a 10% interest rate, this IOU or “bill” will be discounted to $200. That is, the banker will pay $200 in paper today for the farmer’s $220, 1-year IOU.
Can we say that the 200 paper dollars were issued “in the discount of good bills”? That depends. If the farmer offered only his future production of corn as collateral for the loan, then the farmer’s IOU satisfies the traditional idea of a real (i.e., good) bill: “Borrowers and banks agree that these forthcoming productions serve as collateral for the dollar value of the loans.” (Timberlake, (b) 2005, p. 3.) But if the farmer offered his farm itself as collateral, then there would be no direct promise of “forthcoming production” and the farmer’s IOU would not qualify as a real bill. Furthermore, the farmer’s IOU does not meet the condition of being due “at not more than sixty days’ date”.
The hair-splitting question of whether a bill is “real” or “short-term” would be irrelevant to the banker, and with good reason. The banker only cares that his loan will be repaid with interest, and the banker would view the farm as being at least as good collateral as the farmer’s future production. In fact, the banker might reasonably prefer to lend $300 newly-printed dollars to a gambler on his way to a casino (line (3)), as long as the gambler offers his house as collateral, and as long as the house is worth at least $300. In this case there is hardly any chance that the newly-printed $300 will result in any forthcoming productions at all, but that is irrelevant to the banker who has received adequate collateral for his loan.
The key question is this: After the bank has completed all the transactions shown in Table 1, having issued a total of $500 newly-printed dollars on loan, thus multiplying the original $100 six times, what is the value of a paper dollar? The answer is the same as it always was: one paper dollar is worth one ounce of silver. It is obvious that if the bank had issued only $100 against 100 ounces of silver, then each dollar would be worth 1 ounce. It is also obvious that if the bank issued the additional $500 without taking any additional assets in return, then the public would hold $600 against only 100 ounces of silver in the bank, and each dollar would be worth only 1/6 ounce of silver. But the banker did receive one dollar’s worth of assets for every dollar issued, and each dollar is adequately backed.
Two kinds of convertibility must be distinguished:
The importance of financial convertibility can be seen by imagining that people in a community one day find themselves with more paper currency than they wish to hold — for example, when the Christmas shopping season has ended. If the dollar is physically convertible (for one ounce of silver, let us suppose), people will return the unwanted dollars to the bank in exchange for silver, but the bank could head off this demand for silver by selling some of its own bonds to the public in exchange for its own paper dollars. For example, if the community has $100 of unwanted paper money, and if people intend to redeem the unwanted $100 for silver at the bank, the bank could simply sell $100 worth of bonds or other assets in exchange for $100 of its own paper dollars. This will soak up the unwanted paper and head off peoples’ desire to redeem the $100 for silver.
By conducting this type of open market operation — selling bonds when there is excess currency and buying bonds when there is too little — the bank can maintain the value of the dollar at one ounce of silver without ever redeeming any paper dollars for silver. In fact, this is essentially what all modern central banks do, and the fact that their currencies might be physically inconvertible is made irrelevant by the maintenance of financial convertibility. Note that financial convertibility cannot be maintained unless the bank has sufficient assets to back the currency it has issued. Thus, it is an illusion that any physically inconvertible currency is necessarily also unbacked.
A related question concerns the timing of convertibility. A dollar that is instantly convertible into one ounce of silver will be worth one ounce on the market. If convertibility is delayed by 1 year, then for some interest rate R, the dollar will be worth 1/(1+R) ounces today, and will grow to 1 ounce next year. The annual cost of issuing a dollar must also be considered. These costs would include the cost of printing, periodic redemption, protection against counterfeiting, etc. (Note-issuing bankers in the nineteenth century generally claimed that these costs made it unprofitable for them to issue paper dollars, and the dollars were issued more as a form of advertising.) When the annual cost of issuing a dollar is C/year, a dollar that promises 1 ounce of silver in 1 year will be worth 1/(1+R-C) today and grow to 1 oz. after 1 year. If C=R, so that the cost of issue exactly equals the rate of interest, then the dollar will start the year worth 1 ounce and end the year worth 1 ounce. This makes it seem as if the dollar bears no interest, but in truth the interest on the dollar was offset by the cost of issue.
Define the exchange rate E as the value of the dollar, measured in silver (oz./$). Since assets (100 oz. + IOU’s worth 500E oz./$) must equal liabilities ($600 worth E oz./$), it must be true that
100+500E=600E, or E=1 oz./$.
(Here it is assumed that the IOUs are promises to deliver dollars, if necessary by the sale of specified assets, rather than promises to deliver the assets themselves.) If the bank loses some of its assets, then inflation will result. For example, the gambler might default on his loan, and his IOU might therefore fall in value from $300 to $200. The above equation would then become
100+400E=600E, or E=0.5 oz./$
The loss of assets has caused the value of the dollar to fall to half its original value. (If the bank had more non-dollar assets to call on, such as a claim on the farmer's house, the drop would be less.) Note that the real bills doctrine attributes inflation to inadequate backing, while the quantity theory of money, in contrast, claims that inflation results when the quantity of money outruns the economy's aggregate output of goods.
In normal times, the banker will be able to immediately redeem up to $100 for silver at the rate of 1 oz./$. If the banker expects a heavy demand for silver, he can sell the $500 worth of IOU’s for 500 oz. of silver and be ready to redeem all $600 at 1 oz/$. Even if the banker faces a run, where customers suddenly and unexpectedly demand silver for their dollars, the banker could survive the run by selling the $500 worth of IOU’s for $500 of his own paper dollars, and burning the paper dollars he receives. Then there would be only $100 of paper left in the hands of the public, which the banker could redeem with his 100 oz. of silver. At no point would the value of the dollar fall below 1 oz./$, but note that the banker could not survive the run if he did not have adequate assets backing the dollars he has issued.
Now let us reexamine the traditional view of the real bills doctrine: “that so long as a bank issues its notes only in the discount of good bills, at not more than sixty days’ date, it cannot go wrong in issuing as many as the public will receive from it.” According to what can be called the “backing view” presented above, it is only necessary that a bank issues its notes for assets of sufficient value. It is irrelevant whether the assets are due in sixty days or sixty years. It is also irrelevant whether the assets in question are “productive” (like a farmer’s IOU based on forthcoming productions) or “unproductive” (like a gambler’s IOU). It is even irrelevant whether the asset is a “bill” or not. One paper dollar could just as well be issued for a dollar’s worth of land as for a commercial bill worth $1.
Once the real bills doctrine is stripped of these irrelevancies, we can restate it as follows:
So long as money is only issued for assets of sufficient value, the money will maintain its value no matter how much is issued.
This statement is clearly true of the paper dollars described in table 1. It is also true of financial securities in general. For example, economists all recognize that if GM stock is currently selling for $60 per share, then GM can issue 1 new share, sell it for $60, and there will be no change in the price of GM shares, since assets will have risen exactly in step with the number of shares issued. One of the weaknesses of the quantity theory of money is that it claims that money is valued for entirely different reasons than any other financial security. One virtue of the backing version of the real bills doctrine is that there is no need for any “special” theory of money. The value of money is determined on exactly the same principles as any other financial security.
Suppose that the gambler’s IOU falls in value from $300 to $200, so that the market value of the dollar is now given by the equation: 100+400E=600E, or E=0.5 oz/$. If the bank tries to maintain convertibility at the original rate of E=1.0 oz/$, it will face a bank run. Customers see that a dollar will fetch only 0.5 ounces of silver on the market, and so they all rush to the bank for the chance of getting 1 ounce for that dollar. The bank, for its part, loses 0.5 ounces for every dollar redeemed at this rate, and this loss of assets causes the value of the dollar to fall still more. For example, if the bank redeems $40 for 40 ounces, then setting assets equal to liabilities yields 60+400E=560E, Or E=.375 oz./$ If the bank pays out its last 60 ounces of silver for $60, then the equation becomes 0+400E=500E, or E=0.
At this point the bank will be out of business and the dollar will have lost all value. But the worst problem is that the real value of the community’s money supply would have fallen. Originally, the total money supply consisted of $600, with an aggregate real value of 600 ounces. After the gambler’s IOU fell in value, the $600 in circulation had an aggregate real value of only 300 ounces. Money would be “tight”, and economic activity would be hampered. By the time the bank collapsed, the $500 still in the hands of the public would be worthless, and the only usable money left would be the 100 ounces of silver in the hands of the public. The resulting restriction of the money supply would be recessionary, since people would not have enough money to conduct their business conveniently.
Note that the bank’s loss of assets leads to inflation, while the restriction of the money supply leads to recession. There are three solutions to the bank run:
As a practical matter, bailouts have two main pitfalls:
Both of these pitfalls appear to have been at work during the Asian currency crises of the late 1990’s. Bailouts seem particularly ill-advised from the standpoint of the country or agency granting the bailout. Not only are bailouts expensive, but the recipients would often do better simply by suspending convertibility or devaluing.
The real bills doctrine was discredited largely because of the writings of Henry Thornton (1801), David Ricardo (1810) and Lloyd Mints (1945). Each of these writers claimed that the real bills doctrine placed no effective limit on the amount of money that banks might create. While Mints, for example, was willing to admit that money issued in exchange for a given physical amount of assets will not cause inflation, he claimed that money that is issued for a given money's worth of assets presents the possibility that the new money will cause inflation, thus diminishing the real value of each borrower's debt, and allowing them to borrow still more. The result would be a self-perpetuating cycle of more loans, more money, and more inflation.
Mints, Thornton, and Ricardo erred by assuming what they were trying to prove. On real bills principles, a new issue of money, adequately backed by equally-valued assets, would cause no inflation, so Mints' "self-perpetuating cycle" would never get started. Only by assuming the validity of the quantity theory from the outset were real bills critics able to conclude that the real bills doctrine would lead to inflation.
In the early 1980s two papers were published that argued for a division of monetary theory into "quantity" and "real bills" schools of thought, and then used models based on Samuelson and Miller-Modigliani to assert that the quantity theory was not pareto optimal. Sargent and Wallace in their 1981 paper argued from a Samuelson model and concluded that it was impossible to argue that separating the market for money from the market for private credit was unquestionably better, and, in fact, the real bills model permitted pareto optimality where the quantity model did not. Thomas Cunningham in 1992 did an empirical survey which concluded: "The results provide clear evidence supporting the Real Bills doctrine, that the value of assets backing money determines its value, over the Quantity Theory."