I. The Early Theory of Banking
This is Part I of a lengthy critique of the paper, Bagehot was a Shadow Banker. In this section I explain first the nature of the 19th c. monetary system in Britain, then discuss the way that the system was understood and explained by contemporary theorists, and wind up by detailing some of the factual errors in the paper.
A. Money in the 18th-19th c.
1. Constraints on finance prior to the 19th century
As the middle ages were coming to a close Europe developed a financial instrument called the bill of exchange that was managed by a network of wealthy merchant bankers. In its initial form the bill of exchange was used to finance international trade between Europe’s leading cities. The bill was a short-term debt contract that was payable in a foreign country. Clearing mechanisms enabled trade in these bills to minimize the transport of gold and silver across Europe.
In the sixteenth century the bill of exchange evolved into a very different instrument. Endorsement allowed bills to circulate from hand to hand before being redeemed, and domestic bills became the norm in highly developed commercial economies. Local bankers managed local networks and stood ready to discount bills before they were due. Thus, a tradesman with a local bank account could write a bill in the name of a supplier, who could then choose to hold the bill, endorse the bill over to a creditor of his own or cash it – less a discount – at the bank. The bill was a form of commercial paper that was endorsable and effectively allowed banks to underwrite a system of trade credit for the local community. The result was that trade in urban economies began to take place on the basis of a paper monetary system that was supported by a network of banks.
In order for the system to work, standards had to be put into place to prevent the local tradesmen from writing too many bills. In practice a single principle was used to regulate this credit system: A bill was valid only if it was issued in exchange for goods. Bills that were written in the absence of a real exchange were described as “fictitious” or “accommodation paper.” Any tradesman who was caught issuing fictitious bills was considered a fraud and excluded from the financial network. Suspicion of such fraud could also derail a tradesman’s career.
The principle that bills were valid only when they were issued in exchange for real goods is now known as the Real Bills Doctrine. This doctrine was the standard our early modern ancestors put in place to ensure that finance served the needs of trade. It had the advantage of being applied at the individual level, creating a completely decentralized means by which the issue of financial paper could be controlled.
Nowadays the real bills doctrine is famous because it played an important part in the debate over monetary policy that took place in England in the early 19th century, and is associated with Adam Smith. The Currency School argued that the Bank of England should be constrained to issue bank notes in an amount that did not exceed the amount of gold it held in its vaults, while the Banking School argued that the real bills doctrine was sufficient to control the money supply and that the Bank needed to have the flexibility to issue an indeterminate quantity of bank notes when discounting real bills. It’s worth noting that the real bills doctrine was so fundamental to the 18th and early 19th century concept of financial stability, that no one questioned the necessity of adhering to the doctrine. The issue in the debate was whether or not the real bills doctrine alone was sufficient to ensure financial stability.
The denouement of this controversy took place when the Bank Charter Act of 1844 was passed. This was effectively a compromise. Only the Bank of England was allowed to issue bank notes and the Bank’s issue was fixed by the amount of gold in its vaults; however, the Act was subject to suspension by executive order. In practice, this meant that the Bank of England’s note issue was restricted – unless economic circumstances required a greater supply of notes. The Act was temporarily suspended in 1847, 1857 and 1866.
2. The 19th century evolution of the financial system
In the meanwhile, the British economy was steadily outgrowing the real bills doctrine itself. By the start of the 19th century in England the system of domestic bills had evolved into acceptance finance. A country tradesman who regularly shipped his wares to a London middleman for sale would draw on his account with the middleman when making purchases in his own local community. The tradesman would write a bill drawn on the London middleman to pay his local supplier. The supplier would go ahead and circulate the bill through endorsement. However, until the bill was discounted at the local bank, sent by the banker off to his London correspondent for settlement and formally accepted by the London middleman as an obligation, there was no certainty under the law that the middleman would pay.
In short, acceptance finance was a prototype for the checking account system that would develop decades later – just like a checking account system it required that (i) bad bills or checks be passed infrequently and (ii) middlemen or bankers could be relied on to honor their obligations. When one recognizes the sophistication of the financial system in Britain at the turn of the 19th century, one begins to understand why Henry Thornton considered the “science” of credit to be the fundamental source of British growth at the time.
Now here is the question: Is the bill drawn by the country tradesman on the London middleman a real bill or a fictitious bill? Assuming they have an ongoing relationship is there anything wrong with a middleman accepting the bill before he has received a delivery of goods? Is there anything wrong with a middleman extending an overdraft to a tradesman? It was probably inevitable that the practice of acceptance finance broke down the cultural barriers that had supported the real bills doctrine. Henry Thornton’s Paper Credit makes it clear that by the early years of the 19th century, some British bankers were beginning to realize that a “good bill” could be backed by nothing more than an individual’s personal credit.
Legal cases demonstrate that the use of accommodation paper was growing – and becoming more acceptable – through the first decades of the 19th century. The Banking Act of 1844 started a different trend: banks that were no longer allowed to issue bank notes found another way to create money, the checking account. These two trends combined to create a new financial system centered around banks as the arbiters of credit.
The 19th century witnessed a transition from a decentralized system of paper money that was controlled by adherence to the real bills doctrine to a more complex system in which short-term monetized credit was allocated by banks. A new approach had to be found to control the growth of the new monetary system based on checking accounts. The Banking Act of 1844 had been a first effort at direct control of the money supply. It was unsuccessful in many ways: In the first quarter century after it was passed, it had to be suspended three times in order to protect the economy from the ravages of liquidity crises. And the growth of checking accounts effectively neutered the Act.
In the meanwhile, however, the Bank of England had a discovered a new tool for controlling the money supply. In the 18th century the Bank’s discount rate had remained fixed at 5%. As a consequence in normal times competing banks took most of the trade, and the Bank’s discount business was relatively small. In a liquidity crisis, however, the Bank’s discounts would increase astronomically for a few days or even weeks only to fall back to normal when the panic had eased.
In the first half of the 19th century a major concern of the Bank was the maintenance of its gold reserves. Thus, the outflow of gold that was associated with crises and strong demand for discounts at the Bank caused concern. It didn’t take long for the Directors of the Bank to realize that by raising the discount rate, they could moderate the outflow of gold. In the 1820s Bank Rate, or the discount rate of the Bank of England, started to be used as a policy tool. By the middle of the century Bank Rate was the principal policy tool that the Bank used to control the flows of gold to and from the Bank and to moderate the growth of credit and of the money supply.
3. The genesis of fiat money
England developed a paper monetary system in the late 18th century. The monetary system was not uniform across the country. In commercial regions a large fraction of the circulating currency took the form of domestic bills. Local bank notes were often an important part of the currency too, especially in agricultural districts. Bank of England notes were issued in large denominations and were important for settling interbank accounts, but circulated very little in the countryside.
This state of affairs changed dramatically in 1797. The finance of the Napoleonic Wars had put an enormous strain on the financial system and the Bank of England risked running out of gold. The solution was the suspension of the convertibility into gold of the Bank of England note. This suspension lasted for almost a quarter of a century.
As the local banking networks had relied through their London correspondents on the gold reserves of the Bank England in order to meet the demands of their own customers, it was no longer possible for local banks to pay out their notes in gold upon request. To resolve the settlement problem in the countryside, the Bank of England began to issue notes in small denominations – making it possible for the local banks to pay out Bank of England notes instead of gold.
Thus, at the turn of the century the British economy shifted very smoothly from a gold standard to a fiat money standard. During the war the economy experienced a moderate level of inflation with the result that when the war finally ended in 1815 it was not immediately possible to resume convertibility of the Bank of England note into gold at the rate that prevailed in 1797. Policymakers, however, were committed to resumption at the original exchange rate. Thus, in the years following the Napoleonic Wars the British economy was put through a severe recession and in 1821 convertibility of the Bank of England note was restored.
Despite the fact that gold was now readily available, country banks continued to settle their obligations in Bank of England notes with frequency for the simple reason that Bank notes were accepted by almost everyone. Bank of England notes displaced gold as a means of settling trades, because they were in practice “good as gold”.
Thus, the foundations of a modern banking system were laid in 19th century Britain. Paper bank notes were universally accepted in final settlement of debt. The banking system offered checking accounts to the general public and short-term credit to those that met the criteria of the bankers. And finally the whole system was moderated by the Bank of England’s control over the short term interest rate on bills discounted at the Bank. The great 20th century innovation would be the shift to a true fiat money standard with no convertibility of bank notes into any kind of real asset.
B. Early Monetary Theorists’ View of the Relationship Between Banking and Growth
Almost a century before Bagehot, Adam Smith and Henry Thornton discussed the role played by the banking system in the economic growth of Britain. Chapter 2 of Book II of the Wealth of Nations is devoted to explaining how banks contribute to the recent increase in trade and industry of Britain. In addition, Smith remarks on the importance of the Bank of England and the role it played in “support[ing] the credit” of the largest banks in England, Germany, and Holland, including an anecdote about how in 1763 the Bank advanced in a single week £1.6 million (equal at the time to almost a quarter of the Bank’s total liabilities) “a great part of it in bullion” (at II.2.85). For Smith, the mechanism by which banking contributed to growth was by expanding the metallic money supply and conserving on the circulation of gold and silver. In short, it was because bank money was not collateralized by “safe assets” like gold or silver that it could play the role that it did in increasing the money supply.
As noted above, Henry Thornton was one of the early proponents of the acceptance finance system that developed in the first decades of the 19th century, and of the concept that bankers’ lending should be tied to the individual credit of the borrower rather than to specific transactions (Paper Credit at 89). Thornton, too, emphasized the importance of the expansion of the money supply that the use of bills entailed. He criticized Smith for failing to recognize that bills didn’t just expand the metallic money supply, but also the supply of bank notes, and emphasized that bills improved the money supply, because they were an interest bearing form of money. As a result, Thornton found that bills were a preferred form of money for those engaged in commerce (at 92).
Thornton, too, had a fine understanding of the role of a lender of last resort in a crisis: “That a state of distrust causes a slowness in the circulation of guineas, and that at such a time a greater quantity of money will be wanted in order to effect only the same money payments, is a position which scarcely needs to be proved” (at 99). He continues to establish the difference between a central bank, which “is completely subjected to the interests [of the public]” and a “private house” which “may be in general considered as having in the bank [of England] a sure resource” (at 126 -27). Thornton defends the 1797 suspension by the Bank of England of payments in gold and argues that, if the Bank of England erred in this time period, it erred “on the side of too much restricting its notes in the late seasons of alarm” (at 127). Thornton makes it abundantly clear that a central bank which can expand the money supply is needed to support the banking system through a liquidity crisis.
Finally, Thornton attributes “the flourishing state of our internal commerce” to the role played by the banking system in providing a credit-based monetary system (at 175-76). He also explains very clearly that transactions based on credit are essential to the operation of the British economy (at 100-02).
In short, both Adam Smith and Henry Thornton understood the importance of banking to the economic growth that they were witnessing in Britain on the eve of the 19th century. They also understood the role that the Bank of England played during crises in supporting that growth and in supporting the existence of a credit-based economy. Thus, it is literally painful to read statements in Bagehot was a Shadow Banker such as “At the time, that Bagehot was writing, this backstop function [of the central bank] was not yet fully understood, much less accepted” or “the central banks of Bagehot’s time . . . employed their balance sheets to stem the downturn . . . [but] did so without much prior theory about why it would work, and with hardly any thought about possible implications for more normal times” (both at 1).
The authors’ source for the first claim is Forrest Capie, but I believe they misread him (and no source is given for the second claim). Capie writes that the lender of last resort role was addressed “comprehensively” by Thornton in 1802 and that Bagehot wrote about it in the 1840s. In the context of this theory of a lender of last resort, well-developed more than a quarter century before the publication of Lombard Street, Capie writes: “But the Bank of England learned its role as lender of last resort slowly. It resisted for a long time the advocacy of theorists.” The continuation of Capie’s discussion indicates that the Bank had fully mastered the lender of last resort role by 1866. In 1866 the Bank allowed Overend Gurney & Co. to fail – in an event similar to Lehman’s failure, but without requiring any subsequent bailouts – and this was according to Capie the apparent apogee of the Bank of England’s learning process, resulting in “over 100 years of financial stability” (at 16). Lombard Street was published in 1873, six year later. Bagehot agreed that Overend’s failure caused “a crisis for which the Bank of England can[not] be blamed” (Lombard Street at VII.37).
Not only was the theory of the lender of last resort well developed by the mid-1800s, but the problems in normal times of moral hazard created by the actions of the lender of last resort were well understood. A model of moral hazard is the best explain for the contradictory stance of the Bank of England described by Bagehot, who found that by the 1850s the lender of last resort role was both understood and performed by the Bank, while at the same time the role was unacknowledged and sometimes even denied by the Bank (at VII.33). Further evidence of a conscious effort to manage the moral hazard of the lender of last resort’s role is that the Bank of England carefully laid the groundwork for the failure of Overend in 1858, eight years before the Bank refused to save it. In fact, Overend was a bill-broker, which like modern shadow banks was a type of money market lender that held less capital than traditional banks and therefore put extreme stress on the central bank in times of crisis (Lombard Street at XI.21 ff). Because the bill-brokers had borrowed more than the banks in the 1857 crisis, in 1858 the Bank of England, put in place a policy of not standing ready to support the shadow banks in a crisis, but of lending on a “special” basis only. Thus, according to both Bagehot and Capie, the great success of the Bank of England as a lender of last resort was the decision to let the largest of the shadow banks fail – and it was this decision that led to “over 100 years of financial stability.”
The authors of Bagehot was a Shadow Banker may be correct that the central bankers of our own time acted during the crisis without any theory of why their actions would work and “with hardly thought about possible implications for more normal times,” but they are not correct that the Bank of England acted in the 19th c. without a theoretic framework or concern for the moral hazard created by the Bank’s actions.
C. Additional Errors of Fact in Bagehot was a Shadow Banker
“Reading Bagehot, we enter a world where securities issued by sovereign states are not yet the focal point of trading and prices …” (at 4). On the contrary, at the very heart of the 19th c. English money market lay “government stock,” that is, a government bond called a consol that paid 3% per annum, and Exchequer (i.e. Treasury) bills. 19th century manuals on the business of banking explain “the proper investment of the surplus funds of the bank,” and the relative merits of consols, Exchequer bills, and commercial bills. Such texts make clear that the analysis of liquidity, interest rate, and credit risk was well understood in 19th c. money markets. (The advice is, however, extraordinarily conservative by today’s standards.) Furthermore, in a seminal article written a quarter century ago, Douglas North and Barry Weingast argued that the development of a stable market for public debt likely contributed significantly to the market for private debt that the authors of Bagehot was a Shadow Banker are discussing.
The authors model of the 19th c. banking system in Figure 2 appears to indicate that when non-financial firms had a surplus and wished to lend money on the money market, they held accounts at the Bank of England. In fact, the deposits at the Bank of England were composed largely of government deposits and bankers’ deposits – i.e. bank reserves (Lombard Street XII.5 ff). The Bank of England’s role was, instead, to backstop a circular system of lending between firms. That is, most firms had significant liabilities to their suppliers, while simultaneously holding as assets the liabilities of the purchasers of their products. Thus, a decentralized system of credit was supported by local banks which stood ready to discount and provide cash against these short-term business debts on an as needed basis. When a local bank needed cash it could, in turn, discount the bill with a London banking house or bill broker. For the larger local banks their signature on the bill qualified as an acceptance at the Bank of England, so the London discounter could use the bill to draw cash from the Bank of England. In normal times, the need for cash was not severe and the local firm was likely to just hold on to the interest-bearing liabilities of his purchasers. Of course, the willingness of the local firm to hold the purchaser’s debt was predicated on its liquidity and the ability to cash it out on a moment’s notice. In addition, to the degree that there were asset-rich firms — particularly in agricultural regions — they often held their deposits in local banks which gave them access to money-market returns via London correspondent bankers who could lend the funds out on the government debt and public and private bill markets. The Bank of England was an essential backstop to the English banking system, but it was not a key intermediary in the system as Figure 2 implies.
Note that Figure 2 also errs in indicating that bills discounted at the Bank of England were accompanied only by the guarantee of the acceptor. In fact, the discounter also guaranteed payment on the bill. Thus, every bill held by the Bank was supported by the credit of at least three parties, the issuer, the acceptor, and the discounter. Each of these parties was liable for the full value of the bill.
The authors describe the private bill market of Bagehot’s time as “a market in short-term private debt, typically collateralized by tradable goods.” (at 5) The private bills that circulated in Bagehot’s era were not collateralized. Thornton explained the situation clearly: “it may be observed, first, that the notes given in consequence of a real sale of goods cannot be considered as, on that account, certainly representing any actual property. Suppose that A sells one hundred pounds worth of goods to B at six months credit, and takes a bill at six months for it; and that B, within a month after sells the same goods, at a like credit, to C, taking a like bill; and again that C, after another month, sells them to D, taking a like bill, and so on. There may then, at the end of six months, be six bills of £100 each existing at the same time; and every one of these may possibly have been discounted. Of all these bills, then, one only represents any actual property.” (at 86). Bagehot’s discussion in Chapter II of Lombard Street makes it clear that he too is discussing the same type of bill, because there is no discussion of collateral or the possibility that a creditor forecloses on collateral.
The uncollateralized nature of the British money supply in the 19th c. is unsurprising given that both Adam Smith and Henry Thornton were convinced that the banking system promoted growth by making it easy for the money supply to expand. Under this theory, as Thornton clearly understood, limiting the expansion of the money supply to the supply of collateral would have functioned as a constraint on economic growth itself. In short, the whole point of the 19th c. system of private bills in Britain is that they were not collateralized, instead they were typically generated as part of the process of trade (that is, even under the acceptance credit system, goods were regularly transferred, there was just no requirement that every bill be created in the process of such a transfer).
 James Rogers, The Early History of the Law of Bills and Notes, Cambridge University Press, 1995, pp. 113, 171-173, 188-189.
 Henry Thornton, An Enquiry into the Nature and Effects of the Paper Credit of Great Britain, 1802, pp. 175 – 176.
 James Rogers, The Early History of the Law of Bills and Notes, Cambridge University Press, 1995, chapter 10.
 In fact, this obvious possibility had been raised in 1802 by Thornton who makes it clear that usury laws interfered with the operation of this mechanism. Paper Credit, p. 254.
 Forrest Capie, 200 year of financial crises: lessons learned and forgotten 12 (2012).
 Capie, at 8; Bagehot, Lombard Street, XI.30.
 Flandreau and Ugolini argue that the fact that the Bank lent freely to the remaining bill-brokers after allowing by far the largest of them to fail is evidence that the 1858 policy was simply rhetoric. Flandreau and Ugolini, at 14. However, if one views the Bank’s purpose as being one of saving the money market, not destroying it, then one can easily conclude that by allowing Overend to fail, the Bank’s goal was achieved and the subsequent aid to other bill-brokers was entirely consistent with that goal.
 See, e.g. James William Gilbert, The Logic of Banking 196-98 (1856).
 Constitutions and Commitment, 49 J. Econ. Hist. 803, 825 (1989).
 See also, Flandreau and Ugolini, at 21-22.