Do banks really create money out of nothing? Another empirical test of the three theories of banking Academic research paper on "Economics and business"

Accepted Manuscript

Do banks really create money out of nothing? Another empirical test of the three theories of banking

Richard A. Werner

PII: DOI:

Reference:

S1057-5219(15)00147-7 doi: 10.1016/j.irfa.2015.08.014 FINANA 895

To appear in:

International Review of Financial Analysis

Please cite this article as: Werner, R.A., Do banks really create money out of nothing? Another empirical test of the three theories of banking, International Review of Financial Analysis (2015), doi: 10.1016/j.irfa.2015.08.014

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Do banks really create money out of nothing? Another Empirical Test of the Three Theories of Banking

Richard A. Werner1

Abstract

The financial crisis has heightened interest in the question of how banks operate. This had been neglected by economists and finance researchers, but thanks to the crisis a wider debate on this issue seems to be the 'new normal'. Over the past century and a half, the role of banking in the economy has been described by three different theories, which are differentiated by their accounting implications. The currently dominant theory is the financial intermediation theory of banking, which says that banks collect deposits and then lend these out, without any power to create money, just like other non-bank financial intermediaries. The fractional reserve theory of banking says that each individual bank is a financial intermediary, gathering deposits and then lending these out without the power to create money, but the banking system collectively is able to create money through the process of 'multiple deposit expansion' (the 'money multiplier'). The credit creation theory of banking says that banks are not financial intermediaries and do not gather deposits to lend out, but instead each individual bank has the power to create credit and money out of nothing. Eminent scholars have been supporting each one of these different theories. Recently, the Bank of England (2014) argued that textbooks have been wrong in their presentation of banks, and that the credit creation theory is, after all, correct. Needless to mention, the question which of the three theories is correct has major implications for economics, finance, monetary and general government policy and bank regulation. The Basel approach has been predicated on the veracity of the financial intermediation theory. Yet, proponents of each of the theories — including the recent intervention by the Bank of England — merely assert their viewpoints. So far only one empirical test exists. It is the purpose of the present paper to present another empirical test of the three theories, using a different methodology. Analysing bank operations and bank accounting via the bank's annual accounts reporting software, which allows the control of all other factors, it is found that the financial intermediation and the fractional reserve theories of banking are rejected by the evidence. Policy implications are discussed, in particular for regulating bank capital adequacy as a tool to avoid banking crises.

JEL Classifications: E30, E40, E50, E60

Keywords: bank accounting; bank credit; credit creation; financial intermediation; fractional reserve banking; money creation

1 Professor of International Banking, University of Southampton Business School. Director, Centre for Banking, Finance and Sustainable Development, University of Southampton. Email: werner@soton.ac.uk; UK fax: 023 8059 3844

Do banks really create money out of nothing? An Empirical Test of the Three Theories of Banking

Abstract

The financial crisis has heightened interest in the question of how banks operate. This had been neglected by economists and finance researchers, but thanks to the crisis a wider debate on this issue seems to be the 'new normal'. Over the past century and a half, the role of banking in the economy has been described by three different theories, which are differentiated by their accounting implications. The currently dominant theory is the financial intermediation theory of banking, which says that banks collect deposits and then lend these out, without any power to create money, just like other non-bank financial intermediaries. The fractional reserve theory of banking says that each individual bank is a financial intermediary, gathering deposits and then lending these out without the power to create money, but the banking system collectively is able to create money through the process of 'multiple deposit expansion' (the 'money multiplier'). The credit creation theory of banking says that banks are not financial intermediaries and do not gather deposits to lend out, but instead each individual bank has the power to create credit and money out of nothing. Eminent scholars have been supporting each one of these different theories. Recently, the Bank of England (2014) argued that textbooks have been wrong in their presentation of banks, and that the credit creation theory is, after all, correct. Needless to mention, the question which of the three theories is correct has major implications for economics, finance, monetary and general government policy and bank regulation. The Basel approach has been predicated on the veracity of the financial intermediation theory. Yet, proponents of each of the theories — including the recent intervention by the Bank of England — merely assert their viewpoints. So far only one empirical test exists. It is the purpose of the present paper to present another empirical test of the three theories, using a different methodology. Analysing bank operations and bank accounting via the bank's annual accounts reporting software, which allows the control of all other factors, it is found that the financial intermediation and the fractional reserve theories of banking are rejected by the evidence. Policy implications are discussed, in particular for regulating bank capital adequacy as a tool to avoid banking crises.

JEL Classifications: E30, E40, E50, E60

Keywords: bank accounting; bank credit; credit creation; financial intermediation, fractional reserve banking, money creation

1. Introduction

Since the 2008 financial crisis, interest by researchers in finance and economics in the functioning of banks has increased significantly, and many observers have noted that prior to the crisis economics theories had left out banking entirely (Kohn, 2009; Werner, 2012), while in finance, the macro-economic feedback of banking activity had been neglected. This has led to the emergence of 'macro finance' as a new discipline, nested within the finance research agenda. Its emergence is another example of the post-crisis 'new normal', within the discipline of finance research.

The present paper is meant as a fundamental contribution to the 'new normal' research agenda of 'macro finance'. Specifically, it deploys empirical accounting research to settle a long-standing but unresolved central dispute concerning the role and function of banks, which has major implications for monetary and macroeconomics as well as finance and banking.

There are three distinct and mutually exclusive theories of banking, with quite different implications for economics, finance/banking and official policy (monetary, fiscal and bank regulation policy): The credit creation theory of banking maintains that each bank can individually create money 'out of nothing' through accounting operations. The fractional reserve theory states that only the banking system as a whole can collectively create money out of nothing, while each individual bank is a mere financial intermediary, gathering deposits and lending these out, without the power to create credit. The financial intermediation theory considers banks as financial intermediaries both individually and collectively, rendering them indistinguishable from other non-bank financial institutions in their behaviour, especially concerning the deposit and lending businesses and unable to create money individually or collectively.

All three theories have been dominant for several decades during the 20th century, with famous economists supporting them. It is surprising to find that the question which of the three theories is accurate still remains unsettled today. The question which of the three theories is correct has implications for research and policy. For progress to take place in the areas macroeconomics, monetary economics, finance, banking, development economics, but also central bank policy and bank regulation, it is necessary to resolve the question of which banking theory is correct. This is all the more so, as since the crisis there is a heightened interest in, and for modelling purposes, need for a better understanding of the details of bank operations. The first such empirical test was presented by Werner (2014) in this journal. However, in this live loan experiment, it was not possible to completely control the environment. This resulted in a more complex evaluation of the results.

The main contribution of the present paper is to provide another empirical test of the validity of the three theories, using a different methodology that allows the complete control of other factors. Thus the results are more clear-cut. The paper contributes to the literature in monetary and macroeconomics, as well as macro finance and bank regulation. The findings also help in understanding the mechanisms that lead to banking crises.

The paper is structured as follows: The following section will introduce and survey the literature on the three theories of banking, and draw out their differing accounting implications. Section three briefly summarises the results of qualitative empirical accounting research, namely the eye-witness testimonials from bankers. Due to a high

degree of specialisation and the dominant role that IT systems play in modern bank accounting, many bank staff are unable to observe directly the accounting operations that might help an observer distinguish between the three theories, and consequently bankers can be found supporting each of the three hypotheses, with the presently dominant financial intermediation theory boasting the largest number of supporters. It however emerges from these testimonies that the question can be settled by gaining access to banks' daily internal accounting records. Section four presents such a test. Section five analyses and interprets the results. An important conclusion is that the nature of banking in its economic, financial and legal implications is mainly manifested in the bank accounting records, and that a more careful examination of bank accounting practice could have prevented erroneous theories of banking and bank regulation from prevailing for so long.

2. A brief overview of the three main theories of banking and their accounting

The literature on the role of banking and its accounting is briefly surveyed below. It is confined to works by authors who are concerned with banks that cannot issue bank notes, thus concerning British authors on this topic confining itself to the era after the Bank Act of 1844.

(a) The financial intermediation theory of banking

The presently dominant financial intermediation theory holds that banks are merely financial intermediaries, not different from other non-bank financial intermediaries: they gather deposits and lend these out (Figure 1). Or, in the words of recent authors, "Banks create liquidity by borrowing short and lending long" (Dewatripont, Rochet and Tirole, 2010), meaning that banks borrow from depositors with short maturities and lend to borrowers at long maturities.

Figure 1 The Financial Intermediation Theory of Banking

('Financial Intermediaries7) = "indirect finance"

to CB as

reserve

Purchase of Newly Issued Debt/Equity = "'direct financing'Vdisinteraiediaiion

Investment

(Borrowers)

Proponents include, among others, Keynes (1936), Gurley and Shaw (1960), Tobin (1963, 1969), Sealy and Lindley (1977), Diamond and Dybvig (1983), Baltensperger (1980), Diamond (1984, 1991, 1997), Eatwell et al. (1989), Gorton and Pennacchi (1990), Bencivenga and Smith (1991), Bernanke and Gertler (1990), Rajan (1996), Myers and Rajan (1998), Allen and Gale (1994, 2004); Allen and Santomero (2001); Diamond and Rajan (2001), Kashyap et al. (2002), Matthew and Thompson (2005), Casu and Girardone

(2006), Dewatripont, Rochet and Tirole (2010), Gertler and Kiyotaki (2011) and Stein (2014). The theory is also supported by authors who purport to focus on bank accounting in their analysis, but in fact merely recite tertiary sources found in the textbooks (Wolfe, 1997).

Keynes' (1936) General Theory argues that for investments to take place, savings first need to be gathered. This view has also been reflected in the Keynesian growth models by Harrod (1939) and Domar (1947). They have had a significant influence on economic policy in the post-war era, as their work has been interpreted to the effect that developing countries could be helped by international banks who could substitute domestic savings with borrowing from abroad in order to fund economic growth.

Gurley and Shaw (1955, 1960) argue that banks and non-bank financial institutions largely share the function of being financial intermediaries, thus arguing that there was nothing special about banks. Tobin (1963) backed this view in his influential work. He argued:

"Neither individually nor collectively do commercial banks possess a widow's cruse" (p. 412).

"The distinction between commercial banks and other financial intermediaries has been too sharply drawn. The differences are of degree, not of kind... In particular, the differences which do exist have little intrinsically to do with the monetary nature of bank liabilities... The differences are more importantly related to the special reserve requirements and interest rate ceilings to which banks are subject. Any other financial industry subject to the same kind of regulations would behave in much the same way" (p. 418).

Sealy and Lindley (1977) develop a production theory for depository institutions:

"The transformation process for a financial firm involves the borrowing of funds from surplus spending units and lending those funds to deficit spending units, i.e. financial intermediation" (p. 1252).

".. .the production process of the financial firm, from the firm's viewpoint, is a multistage production process involving intermediate outputs, where loanable funds, borrowed from depositors and serviced by the firm with the use of capital, labor and material inputs, are used in the production of earning assets" (p. 1254).

Baltensperger (1980), a banking and monetary economics expert, also believes banks are merely financial intermediaries, unable to create money:

"The main economic functions of financial firms are those of consolidating and transforming risks on the one hand, and of serving as dealers or 'brokers' in the credit markets ... on the other hand" (p. 1).

Riordan (1993) holds that

"Banks serve as financial intermediaries between borrowers and lenders. More precisely, banks borrow from depositors and lend to investors. ... In a capitalist economy most investment projects are owned and managed by private

entrepreneurs and firms. Generally these investors lack enough equity fully to finance their projects and consequently seek loans to complete financing. Banks, on the other hand, aggregate deposits to make these loans" (p. 328).

Kashyap et al. (2002) believe that banks are pure financial intermediaries, presenting a model of banking in which the bank purchases assets with funds it has acquired in the form of deposits or the issuance of equity or bonds:

"The total assets to be financed at date 0 are L 1 S0. They are financed partly by demandable deposits. ... In addition to deposits, the bank can also issue claims in the public market... These claims mature at date 2, and can be thought of as either bonds or equity" (p. 41).

The large 'credit view' literature (such as Bernanke and Blinder, 1989, Bernanke and Gertler, 1995), the monitoring literature on financial intermediation (Diamond, 1984; Sheard, 1989), and the substantial literature on the various other theories of financial intermediation, which include banks, but do not distinguish them from other non-bank financial institutions (see Casu et al., 2006), all hold that banks are just another type of financial intermediary among many, without the power to create credit in any way.

Influential textbooks on money and banking also are proponents of the financial intermediation theory of banking, such as that by Cecchetti (2008), who does not consider banks able to create credit or money:

"...an institution like a bank stands between the lender and the borrower, borrowing from the lender and then providing the funds to the borrower" (p. 39).

...or the banking textbook by Casu et al. (2006):

"Banks, as other financial intermediaries, play a pivotal role in the economy, channelling funds from units in surplus to units in deficit. They reconcile the different needs of borrowers and lenders by transforming small-size, low-risk and highly liquid deposits into loans which are of larger size, higher risk and illiquid (transformation function)" (p. 18).

Matthew and Thompson (1995) state:

"Financial intermediation refers to borrowing by deficit units from financial institutions rather than directly from the surplus units themselves. Hence, financial intermediation is a process which involves surplus units depositing funds with financial institutions who in turn lend to deficit units" (p. 33).

"An exogenous increase in the demand for loans shifts the LL schedule up to LL' and increases the loan rate. The bank (or banking system in the case of a non-monopoly bank) will respond by supplying more loans and deposits. To attract more deposits, the bank (banking system) will bid for deposits by increasing the deposit rate" (p. 110).

As there is no clear distinction of banks from non-banks, economists also see no reason why banks need to be singled out for special treatment or indeed inclusion in their macroeconomic theories. Thus it came to pass that the semnal articles in leading journals and widely-used macroeconomics and monetary economics textbooks have long dropped out the banks: they do not feature in 'advanced macroeconomics' or 'advanced monetary economics', as is seen in the influential 785-pager by Woodford (2003), the 820 pages of Heijdra and Van der Ploeg (2002) or the 751 pages of Sorensen and Whitta-Jacobsen (2010).

(b) The fractional reserve theory of banking

The second theory of banking has a major similarity with the financial intermediation theory of banking: It also argues that each bank is a financial intermediary. However, it disagrees with the former theory concerning the collective, macroeconomic role of banks: it argues that together the banking system creates money, through the process of 'multiple deposit expansion'. Thus when Gurley and Shaw (1960) argued that banks and non-bank financial institutions are largely similar in that they were both financial intermediaries able to 'create financial claims', they were challenged during the 1950s and 1960s in influential journals by, among others, Culbertson (1958), Aschheim (1959), Warren Smith (1959), Solomon (1959) and Paul Smith (1966) and Guttentag and Lindsey (1968), many of whom were supporters of the fractional reserve theory.2 Phillips' citation of the credit or money multiplier rendered him one of the earlier and most influential economists to formulate the mechanics of fractional reserve banking.3 According to Phillips:

"What is true for the banking system as an aggregate is not true for an individual

bank that constitutes only one of many units in that aggregate" (Phillips, 1920, p.

Crick (1927) is another supporter of this theory. He argues that while each bank is a financial intermediary, the system as a whole can create money. At the same time, Crick recognises that many observers are uncomfortable with the idea that banks could be creating money:

Smith (1959), for instance, argued in the Quarterly Journal of Economics that banks 'can create money" and that "their credit-creating activities expand the supply of loanable funds available to finance expenditure" (p. 535).

"Commercial banks do have a special ability to expand credit for a reason that is simple but often overlooked.... What is truly unique... about commercial banks is... their distinctive role as issuers of means of payment [which] gives commercial banks a peculiar ability to expand credit."

Smith argues that banks are (presumably in aggregate) not financial intermediaries and their function is distinct from that of financial intermediaries (what in modern parlance is referred to as 'non-bank financial intermediaries'). According to Smith, the money creation by banks is due to a 'multiplier process' (which he also calls the "credit expansion multiplier' or "multiple credit creation").

"Commercial bank credit creation makes funds available to finance expenditures in excess of the funds arising out of the current income flow. Intermediaries, to the extent that their activities are as described so far, merely collect a portion of current voluntary saving and serve the function of making these funds available for the financing of current expenditures - i.e., they help to channel saving into investment in a broad sense. Thus, intermediaries are exactly what their name indicates. Commercial banks, on the other hand, are distinctly not intermediaries" (p. 538).

3 Earlier authors were Davenport (1913) and Marshall (1890), although Davenport, like many early authors on this issue, were ambiguous. Thus Davenport simultaneously supported the credit creation theory.

"To some minds the idea of "creating" anything is both objectionable and absurd, but disagreement on matters of terminology should not blind us to the relations, in sequence and amount, between the volume of bank credit outstanding and the quantity of bank cash held against it."

Like later Keynes in some contexts, as well as still later James Tobin, Crick adopted the habit of placing the concept of creation in inverted commas ('credit "creation"'), and while not entirely denying the potential for banks to create credit and money, succeeded in downplaying the significance and re-assuring the concerned public — or academia — by emphasising that while the system taken together has the ability to create money in some kind of diffuse way, individual banks are not able to do so. In any case, it all boiled down to the ratio of bank credit to cash reserves — a technical matter, in other words, of little direct consequence for the economic model builder.

"To say that a bank cannot in practice "create" deposits to an indefinite extent is one thing; to say it cannot "create" deposits at all is another. The first assertion is true because a bank cannot make an addition at will to its own cash reserves without reducing its earning assets, while it feels it must conform to a regular ratio. The second is untrue, because we know as a fact that an addition to bank cash is accompanied, or closely followed by, a multiple addition to deposits, which cannot be attributed to any other cause but action by the banks" (Crick, 1927, p. 201).

The last sentence, as well as the thrust of Crick's argument makes it clear that he believed, for all means and purposes, the system as a whole, not individual banks, can create money and credit.

Keynes (1930) is not a beacon of clarity on this important topic. Firstly, he supports the fractional reserve theory, citing both Phillips (1920) and Crick (1927) approvingly (p. 25). But he then discusses the concept of money 'creation' by referring to any increase in bank deposits as the 'creation' of deposits:

"There can be no doubt that, in the most convenient use of language, all deposits are 'created' by the bank holding them. It is certainly not the case that the banks are limited to that kind of deposit, for the creation of which it is necessary that depositors should come on their own initiative bringing cash or cheques" (p. 30).

Keynes may have been referring to bank transfers as the kind of deposit that allows a bank to 'create' a deposit, while remaining a mere financial intermediary, since Keynes (1930) deploys the expression 'creation of deposits' also for the instance of a cash deposit at a bank (p. 24), arguing that:

"only the bank itself can authorise the creation of a deposit in its books entitling the customer to draw cash or to transfer his claim to the order of someone else" (p. 24).

So deposit creation' "in the most convenient use of language" here is simply the act of recording a deposit in the bank's account, i.e. a bank accounting entry. If the adjustment of an account is termed the 'creation' of such an accounting record, by this definition banks are of course 'creating' entries whenever a transaction is made. However, by this

definition any non-bank corporation would equally be 'creating' assets and liabilities on its balance sheet, whenever a transaction is entered into the firm's accounts. Thus such terminology does not serve to clarify. In particular, as we also consider the third theory of banking, the credit creation theory, we are interested in the question whether an individual bank can create credit out of nothing, or whether it is only the banking system as a whole that is able to do so, as the fractional reserve theory of banking maintains.

A modern textbook favours the fractional reserve theory, but mirrors Keynes' ambiguous terminology:

"The process of multiple-deposit creation may seem somewhat like a magician pulling rabbits out of a hat: it seems to make something out of nothing. But it is, in fact a real physical process. Deposits are created by making entries in records; today electronic impulses create records on computer tapes. The rules of deposit creation are rules specifying when you may make certain entries in the books. It is these rules — in particular, the fractional reserve requirements — that give rise to the system's ability to expand deposits by a multiple of the original deposit increase" (Stiglitz, 1997, p. 737).

Again, the 'creation' of deposits and loans is defined by the creation of an accounting record. Such terminology distracts from the question whether banks can create new purchasing power out of nothing. What must be the most influential post-war textbook in economics, however, addresses this question head on, that by Samuelson (1948): The original first edition deals with the third theory of banking, the credit creation theory and dismisses it. Under the heading "Can banks really create money?", Samuelson argues against "false explanations still in wide circulation" (p. 324):

"According to these false explanations, the managers of an ordinary bank are able, by some use of their fountain pens, to lend several dollars for each dollar left on deposit with them. No wonder practical bankers see red when such behavior is attributed to them. They only wish they could do so. As every banker well knows, he cannot invest money that he does not have; and any money that he does invest in buying a security or making a loan will soon leave his bank" (p. 324).

Samuelson supports the fractional reserve theory of banking and holds that a bank needs to gather the funds first, before it can extend bank loans. At the same time he argues that, in aggregate, the banking system creates money. He illustrates his argument with the example of a 'small bank' that faces a 20% reserve requirement and considers the balance sheet accounts of the bank. If this bank receives a new cash deposit of $1,000, "What can the bank now do?", Samuelson asks (p. 325).

"Can it expand its loans and investments by $4,000 so that the change in its balance sheet looks as shown in Table 4b?"

Table 4 b. Impossible Situation for Single Small Bank

Assets

Liabilities

Cash reserves

+$1,000 Deposits

+$5,000

Loans and investments.... +$4,000

"The answer is definitely 'no'. Why not? Total assets equal total liabilities. Cash reserves meet the legal requirement of being 20 per cent of total deposits. True enough. But how does the bank pay for the investments or earning assets that it buys? Like everyone else it writes out a check — to the man who sells the bond or signs the promissory note. ... The borrower spends the money on labor, on materials, or perhaps on an automobile. The money will very soon, therefore, have to be paid out of the bank. ... A bank cannot eat its cake and have it too. Table 4b gives, therefore a completely false picture of what an individual bank can do" (p. 325f).

Samuelson argues that since all the money lent out will leave the bank, after loan extension the true balance sheet of this bank that has received a new deposit of $1,000 will look as follows:

Table 4c. Original Bank in Final Position

Assets Liabilities

Cash reserves............... +$ 200 Deposits......................+$1,000

Loans and investments.... +$ 800 _

Total..................... +$1,000 Total...................... +$1,000

(Samuelson, 1948, p. 326).

Samuelson argues that an individual bank cannot create credit out of nothing, while the banking system can do so:

"As far as this first bank is concerned, we are through. Its legal reserves are just enough to match its deposits. There is nothing more it can do until the public decides to bring in some more money on deposit" (p. 326).

"The banking system as a whole can do what each small bank cannot do!" (p. 324),

namely create money. This, Samuelson explains via the iterative process of one bank's loans (based on prior deposits) becoming another bank's deposits, and so forth. He shows "this chain of deposit creation" in a table, amounting to total deposits in the banking system of $5,000 (out of the $1,000, due to the reserve requirement of 20% implying a 'money multiplier' of 5 times (assuming no cash 'leakage'). As a result the consolidated balance sheet of the banking system will appear as follows:

Table 4i. Consolidated Balance Sheet Showing Final Positions of All Banks Together

Assets Liabilities

Cash reserves............... +$1,000 Deposits......................+$5,000

Loans and investments.... +$4,000 _

"If the reader will turn to Table 4b previously marked impossible, he will see that the whole banking system can do what no one bank can do by itself. Bank money has been created 5 for 1 — and all the while each bank has only invested and lent a fraction of what it has received as deposits!" (p. 329)

What Samuelson calls the "multiple deposit expansion" is described in the same way and with remarkable similarity in the fifteenth edition of his book (Samuelson, Nordhaus, 1995), only that the reserve requirement cited as example has been lowered to 10%: "All banks can do what one can't do alone" (p. 493). The table with the 'chain' of nth-generation banks to whom decreasing portions of deposits have moved is the same, as is the caption "All banks together do accomplish what no one small bank can do — multiple expansion of reserves..." (p. 492). Table 4i re-appears, with the same title ("Consolidated Balance Sheet Showing Final Positions of All Banks').

There are subtle though important differences. The overall space devoted to this topic is much smaller in 1995 compared to 1948. The modern textbook says that the central bank-created reserves are used by the banks "as an input" and then "transformed" "into a much larger amount of bank money" (p. 490). There no attempt to deal with the credit creation theory directly. There is no equivalent of Table 4b — the idea that an individual bank might create deposits is not mentioned at all. Instead, each bank is unambiguously represented as a pure financial intermediary, collecting deposits and lending out this money (minus the reserve requirement). The fractional reserve theory had become mainstream:

"Each small bank is limited in its ability to expand its loans and investments. It cannot lend or invest more than it has received from depositors" (p. 496).

Meanwhile, bank deposit money is "supplied" by "the financial system" in an abstract process that each individual bank has little control over (p. 494).

Another supporter of the fractional reserve theory, published in a leading journal, is Whittlesey (1944), who stated that banks are "creating money" (p. 251), "exercising the sovereign function of issuing money" (p. 252), "administrators of the money supply" and engage in "deposit creation" (p. 247) — but only collectively, not individually, in line with

the fractional reserve theory:

"Despite the changes that have taken place, the mechanics of banking operations are essentially similar to what they were in the past. The process, whereby deposits are created — and may conceivably be destroyed — on the basis of fractional reserves and against changes in the volume of debts held by banks, is still fundamentally the same" (p. 247).

The author was aware that the policy conclusion that bank credit creation could be considered a mechanical process that did not need to be modelled explicitly in economic theories, was dependent on a number of assumptions:

4 Furthermore, unlike the original Samuelson (1948), the more recent textbook mentions nowhere that in terms of its operations an individual bank might also be able to 'create deposits' (even though it might then lose the money quickly), which can be said, somewhat contradictorily, to support the credit creation theory.

Moreover, the original Samuelson (1948: 331) offered an important (even though not prominently displayed) section headed 'Simultaneous expansion or contraction by all banks', which provided the caveat that each individual bank could, after all, create deposits, if only all banks did the same at the same rate (thus outflows being on balance cancelled by inflows, as Alhadeff, 1954, also mentioned). There is no such reference in the modern, 'up-to-date' textbook.

"The rise of a large and fluctuating volume of excess reserves is significant primarily because the assumption of a fixed reserve ratio underlies, to an extent that has not, I believe, received sufficient emphasis, the entire theory of commercial banking. The conventional description of the process of deposit expansion — with reserves overflowing from Bank 1 to Bank 2 and so on up to Bank 10, thereby generating a neatly descending series of deposit growth all along the line — rests on the assumption that reserves will be fully and promptly utilized" (p. 250).

Alhadeff (1954), a staff member of the US Federal Reserve system, also invokes Phillips

(1920) in supporting the fractional reserve theory of banking:

"One complication worth discussing concerns the alleged "creation" of money by bankers. It used to be claimed that bankers could create money by the simple device of opening deposit accounts for their business borrowers. It has since been amply demonstrated that under a fractional reserve system, only the totality of banks can expand deposits to the full reciprocal of the reserve ratio. [Footnote: 'Chester A. Phillips, Bank Credit (New York: Macmillan, 1921), chapter 3, for the classical refutation of this claim.'] The individual bank can normally expand to an amount about equal to its primary deposits" (p. 7).

(c) The credit creation theory of banking

The third theory of banking disagrees with the other two theories by arguing that banks are not financial intermediaries — neither in aggregate nor individually. Instead, each bank is said to have the ability to create credit and money out of nothing, when it executes bank loan contracts or purchases assets. Supporters include Macleod (1856), Withers (1909, 1916), Schumpeter (1912), Wicksell (1898), Cassel (1918), Hahn (1920), Hawtrey (1919) and others. There were more supporters of this theory in the era of widespread bank note issuance by commercial banks, but our concern is with writers that considered individual banks to be creators of credit and money even if they do not engage in note issuance.

No doubt the most authoritative writer supporting this theory in this context is Henry D. Macleod (1856), who was a banking expert and barrister at law. His influential work, published in many editions until well into the 20th century (the quote is from the 6th edition of 1905), illustrates the importance of considering accounting, legal and financial aspects of banking together. Based on an analysis of the legal nature of bank activity he concluded:

"Nothing can be more unfortunate or misleading than the expression which is so frequently used that banking is only the "Economy of Capital," and that the business of a banker is to borrow money from one set of persons and lend it to another set. Bankers, no doubt, do collect sums from a vast number of persons, but the peculiar essence of their business is, not to lend that money to other persons, but on the basis of this bullion to create a vast superstructure of Credit; to multiply their promises to pay many times: these Credits being payable on demand and performing all the functions of an equal amount of cash. Thus

banking is not an Economy of Capital, but an increase of Capital; the business of banking is not to lend money, but to create Credit and by means of the Clearing House these Credits are now transferred from one bank to another, just as easily as a Credit is transferred from one account to another in the same bank by means of a cheque. And all these Credits are in the ordinary language and practice of commerce exactly equal to so much cash or Currency (Macleod, 1905, vol. 2, p. 311, italics added)."6

However, the bank purchases this money not with other money, but instead with the bank's credit:

"the Credit [the banker] creates in his customer's favour is termed a Deposit. (p. 406).

"These banking Credits are, for all practical purposes, the same as Money. They cannot, of course, be exported like money: but for all internal purposes they produce the same effects as an equal amount of money. They are, in fact, Capital created out of Nothing" (Macleod, 1905, p. 408).

Macleod's message was spread far and wide by Withers (1909, 1916), who was a prolific writer about this topic and for many years editor of the Economist:

"In old times, when a customer went to a banker for a loan, the banker, if he agreed, handed him out so many of his own notes; now when a customer goes to a banker for a loan, the banker gives him a credit in his books, i.e. adds to the deposits on the liability side of the balance sheet (Withers, 1916, p. 42).7

6 See also: "We have seen that all Banking consists in creating and issuing Rights of action, Credit, or Debts, in exchange for Money, or Debts. When the Banker had created this Liability in his books, the customer might, if he pleased, have this Credit in the form of the Banker's notes. London bankers continued to give their notes till about the year 1793, when they discontinued this practice, and their customers could only transfer their Rights, or Credit, by means of cheques. But it is perfectly manifest that the Liabilities of the Bank are exactly the same whether they give their own notes or merely create a Deposit" (MacLeod, 1855-6, p. 338).

7 "It is true that the customer does not leave the deposit there but draws cheques against it, which he pays to people to whom he owes money. But these cheques, if paid to recipients who also bank at the bank which has made the advance, would simply be a transfer within the bank's own books, and the effect of the transaction upon its balance sheet would be that it would hold among its assets an increase — if the loan was for £100,000 — of this amount among its advances to customers; and on the liability side there would be a similar increase in the deposits. ... and if we could look at an aggregate balance sheet of the whole of the banks of the country we should see that any increase in loans and advances would have this effect of increasing the deposits as long as those who receive these banking credits make use of them by drawing cheques against them. In the comparatively rare cases where the borrower makes use of the credit by drawing out coin or notes from the bank, then the first effect would be that the bank in question would hold a smaller amount of cash among its assets and a larger amount of advances to customers. But even here the currency withdrawn would almost certainly come round again, either to this bank or another, from the shopkeepers or other people to whom the borrower had made payments. And so the cash resources of the banks as a whole would be restored to the original level, while the deposits, owing to the increase at the credit of the shopkeepers and others who had paid the money in, would be added to the amount of the advance originally made. (p. 42f)

"Exactly the same thing happens when, for example, in times of war the banks subscribe to loans issued by the Government, whether in the form of long-dated loans, such as the recent War Loan, or in the form of shorter securities, such as Exchequer Bonds, Treasury Bills or Ways and Means Advances. (p. 43).

"It follows that the common belief that a great increase in bank deposits means that the wealth of the community has grown rapidly, and that people are saving more money and depositing more with the banks is, to a certain extent, a fallacy. A rise in bank deposits, as a rule, means that the banks are making large advances to their customers or increasing their holding of securities, and so are granting a larger amount of book-keeping credit, which appears as a liability to the public in the shape of deposits. (p. 44)

"It may be objected that the deposits have to come first before the banks can make advances. Does this necessarily follow? (p. 44).

According to the credit creation theory then, banks create credit in the form of what bankers call 'deposits', and this credit is money.

"The function of the banker, the manufacturer of and dealer in credit, is to select from the gamut of plans offered by entrepreneurs, ... enabling one to implement their plans and deny this to another" (Schumpeter, 1912, p. 225).8

Later, Schumpeter (1954) would explain that

"this alters the analytic situation profoundly and makes it highly inadvisable to construe bank credit on the model of existing funds being withdrawn from previous uses by an entirely imaginary act of saving and then lent out by their owners. It is much more realistic to say that the banks 'create credit', that is, that they create deposits in their act of lending, than to say that they lend the deposits that have been entrusted to them. And the reason for insisting on this is that depositors should not be invested with the insignia of a role which they do not play. The theory to which economists clung so tenaciously makes them out to be savers when they neither save nor intend to do so; it attributes to them an influence on the 'supply of credit' which they do not have. The theory of 'credit creation' not only recognizes patent facts without obscuring them by artificial constructions; it also brings out the peculiar mechanism of saving and investment that is characteristic of fully fledged capitalist society and the true role of banks in capitalist evolution" (p. 1114).

US supporters of this theory included Davenport (1913) and Robert H. Howe (1915). Davenport (1913) argued:

".banks do not lend their deposits, but rather, by their own extensions of credit, create the deposits" (p. 263).

Robert H. Howe (1915):

"Banks do not loan money. They loan credit. They create this credit and charge interest for the use of it. It is universally admitted that the old State Banks that created credit in the form of bank notes, created currency — and our modern system of creating credit in the form of "Deposits" which circulate in the form of bank checks, is doing exactly the same thing — creating currency.

"All this in effect nullifies the National Banking Act, which provides for National Bank Currency based on U.S. Government Bonds, and also the act levying an annual tax of 10 per cent on all State Bank Currency. ... (p. 24)

"The public little realizes to what an extent Bank Credit, circulating in the form of bank checks, has supplanted all other circulating media. In 95 per cent of all the business done in the United States, the payments are made in bank checks and in only 5 per cent is any cash used; and of this 5 per cent an infinitesimal fraction only is gold. (p. 24f).

8 "Die Funtion des Bankiers, des Produzenten von und Händlers mit Kredit, ist in der Fülle der sich darbietenden Unternehmerpläne eine Auswahl zu treffen, die allen Lebensverhältnissen der Volkswirtschaft entspricht, dem einen die Durchführung zu ermöglichen, dem andern zu versagen" (Schumpeter, 1912, S. 225). Translated by author.

"The introduction of bank notes was useful in weaning the public from the use of gold and silver coins, and prepared the way for the introduction of Bank Credit as the means of payment for commodities. As a result of this evolutionary process, the checks drawn and paid in the United States amount to between two hundred billion and two hundred and fifty billion dollars a year. It is clear that it would be a physical impossibility to do this amount of business by the use of gold coin. There is only about eight billions of gold money in the world, of which amount less than two billions of dollars are in the United States. (p. 25)

"The banks have created fifteen billions of dollars of credit by discounting the notes of merchants and manufacturers, and crediting the proceeds to the borrower's account under the head of Deposits. As a result, the borrower is enabled to draw checks and pay his debts with them. (p. 25)

Swedish economist Gustav Cassel (1919, 1932) pointed out that

"In practice, deposits are also created and constantly fed by the bank's granting advances to its customers, either by discounting bills or by making loans and then crediting the clients with the amount in their accounts" (p. 414).

An important difference to the fractional reserve theory of banking is the use of singular in the above sentence: it is one bank that is able to create deposits. Hawtrey (1919), mirroring Macleod's (1856) exposition, also argued that banks create money out of nothing. Keynes was another prominent supporter of the credit creation theory, praising it enthusiastically in the early 1920s as an

"almost revolutionary improvement in our understanding of the mechanism of money and credit and of the analysis of the trace cycle, recently effected by the united efforts of many thinkers, and which may prove to be one of the most important advances in economic thought ever made" (Keynes, Moggridge, 1983, p. 419, as quoted by Tily, 2012).

Keynes gives the impression of a recent convert whose eyes had been opened. When Keynes credited the 'united efforts of many thinkers' for these 'revolutionary improvements in our understanding of money and credit' and 'most important advances in economic thought ever made' he referred to writings of authors backing the credit creation theory of banking.

In his Treatise on Monetary Reform (1924) Keynes was unambiguous about the ability of banks to expand or diminish "the volume of credit quoted" (p. 137). And, more clearly:

"The internal price level is mainly determined by the amount of credit created by the banks, chiefly the Big Five; ... The amount of credit, so created, is in its turn roughly measured by the volume of the banks' deposits — since variations in this total must correspond to variations in the total of their investments, bill-holdings, and advances" (p. 178).

Yet, his later support for the other theories indicates that Keynes may not have been settled in his views on the credit creation theory of banking. Indeed, there is some evidence

9 This quote is from the English translation of the fifth German edition of the 1918 book, both published in 1932.

that he may have been open to the implication of the fractional reserve banking theory that high powered money is a key driving factor:

"Thus in one way or another the banks generally adjust their total creation of credit in one form or another (investments, bills, and advances) up to their capacity as measured by the above criterion; from which it follows that the volume of their 'cash' in the shape of Bank and Currency Notes and Deposits at the Bank of England closely determines the volume of credit which they create"

(p. 179).

A clearer statement coming from Keynes' pen can be obtained in the form of the final report of the Committee on Finance and Industry, commonly known as the Macmillan Committee (1931), after its chairman, Hugh Macmillan.10 The Committee gathered much evidence, mainly in the form of first-hand eye-witness accounts, and quickly identified bank credit creation as a central focus of their inquiry. It must be considered as one of the most thorough and wide-ranging investigations of banking and finance in the modern age conducted by such a broad group of stakeholders. The final report, submitted in June 1931, contained a number of statements on the question at hand. It is said to have been drafted and significantly influenced by Keynes, one of the committee members. The statement expressly refers to bank accounting of an individual bank:

"It is not unnatural to think of the deposits of a bank as being created by the public through the deposit of cash representing either savings or amounts which are not for the time being required to meet expenditure. But the bulk of the deposits arise out of the action of the banks themselves, for by granting loans, allowing money to be drawn on an overdraft or purchasing securities a bank creates a credit in its books, which is the equivalent of a deposit" (p. 34).

The last sentence uses the singular: a loan from one bank results in credit creation, which is the "equivalent" of deposit creation, amounting to the size of the loan. If the bank was a financial intermediary, it would not newly create the deposit of the borrower, but transfer the funds from another account, either inside or outside the bank. This is most clearly seen "If no additional in-payments were made by customers and there were no withdrawals in cash," because then

"the volume of deposits of a single banker would fluctuate only with the volume of the loans he himself made." (p. 12).

10 The committee was appointed by the Chancellor of the Exchequer in November 1929 to

"inquire into banking, finance and credit, paying regard to the factors both internal and international which govern their operation, and to make recommendations calculated to enable these agencies to promote the development of trade and commerce and the employment of labour' (p. 1). It consisted of leading experts, opinion-leaders and stakeholders of the day, including John Maynard Keynes and Professor T. Gregory, professor of Banking at the LSE, treasury and Bank of England representatives and senior executives of banks, but also a union representative, a representative of the cooperative movement and a politician. Over almost two years the Committee held 49 meetings and interviewed 57 witnesses, reflecting "a wide and varied range of representatives of banking and finance, both in this country and in the United States and Germany, as well as of industry and commerce from the point of view both of employers and of employed, while members of the Universities and the Civil Service and eminent economists of diverse schools have also lent their assistance" (p. 1). This included Mantagu Norman, the governor of the Bank of England, Professor A. Pigou of Cambridge University, as well as senior representatives from Barclays Bank, Midland Bank, Lloyds Bank, National Provincial Bank, Westminster Bank, the Scottish banks and the Treasury, and such long-standing and internationally active banking insiders as Otto Ernst Niemeyer and Henry Strakosch.

11 In his opening words to witness Josiah Stamp, chairman Lord Macmillan stated: "You appreciate that our main preoccupation is with the question of the basis of credit as affecting industry and employment." (Macmillan Committee, 1931, appendix, witness transcripts, p. 238, question 3710).

The credit creation theory of banking also featured prominently in textbooks, training a new generation of economists and policy makers: The US textbook on monetary economics by James (1930) was unambiguous and confident in the assessment that

"... the bank is enabled to make loans to an amount many times larger than the sum of cash which has been deposited with it, and it will already have become apparent that the greater part of the items appearing on the liabilities side of the balance sheet, under the heading of deposits, is created, not as a result of cash deposited with the bank by customers, but through the making of loans or discounts by the bank to those customers.

".the bank has monetized credit. It has created purchasing power which did not exist before, since it has supplied the borrower with a means of paying his debts, without in any way reducing the amount of money in the hands of the other members of the community. Each addition to the existing volume of bank loans, therefore, results in a net increase in the total supply of money in the community, and any diminution in that volume will decrease the total volume of money." (James, 1930, 194f, italics in original).

While the star of the credit creation theory was on the descent in the mid-1930s, despite the increasing use of the fractional reserve theory, a leading — if not the leading monetary economist of his day, Irving Fisher still insisted on the veracity of the credit creation theory:

"When a bank grants me a $1,000 loan, and so adds $1,000 to my checking deposit, that $1,000 of 'money that I have in the bank' is new. It was freshly manufactured by the bank out of my loan and written by pen and ink on the stub of my check book and on the books of the bank... Except for these pen and ink records, this 'money' has no real physical existence" (Fisher, 1935, p. 3).

(d) Assessment

From our review of the literature it is clear that despite today's dominance by the financial intermediation theory, whose rise started with Gurley and Shaw (1955), the question whether banks create money and are thus 'unique' "remains unsettled" (p. 992) — which is what Guttentag and Lindsey (1968) concluded almost half a century ago in their Journal of Political Economy article. The situation has not been helped by the fact that some influential authors, such as Keynes or the Bank of England, have been supporting all three mutually exclusive theories at one point or another.

Thanks to the financial crisis, interest in the question of how banks function has increased. Unlike prior to the crisis, now a number of authors have brought the previously dormant debate to the fore about which theory of banking is correct. In particular, they have disputed that the currently dominant financial intermediation theory of banking is correct and they have endorsed the credit creation theory. This includes most prominently Bank of England staff, who have expressed support for the credit creation theory of banking (Bank of England, 2014a, b), using an elaborate accounting visualisation to explain it. It also includes Ryan-Collins et al. (2011), cited by Bank of England (2014a), and Benes and Kumhof (2012) from the IMF.12

Triggered by an earlier financial boom, it had been used to predict imminent financial instability in Japan (Werner, 1992), to formulate policies to end the Japanese recession (Werner, 1992, 1997), and in the analysis of the European

However, such works have remained disputed by the majority of authors in economics and finance, who maintain that banks are merely financial intermediaries. This includes colleagues at the Bank of England. Its governor Mark Carney (2014) in his Mais Lecture at the Cass Business School cited the monetary theory of Brunnermeier and Sannikov (2013) in support of his arguments. In their paper they argue prominently, in the abstract (quite visibly to the governor) that banks are financial intermediaries that

"take deposits from .households to extend loans." and that "finance themselves by borrowing from households".

In their model, banks could not be anything else, because their model features no money creation whatsoever ("Assume there is a fixed supply of infinitely divisible money", p.

In late March 2014, external member of the Financial Policy Committee of the Bank of England, Dame Clara Furse, gave a speech to the Chartered Institute for Securities and Investment in Liverpool, in which she explains:

"The financial system performs vital functions for us all — it exists to intermediate savings and investment. Banks, non-banks and markets all contribute to this.." (Bank of England, 2013d).

Further, she argues that for economic growth to take place, bank activity can be substituted by 'direct finance', and she recommends, as one of the lessons of the crisis, to enhance 'market based finance', i.e. funding via channels other than banks. Other economists at the Bank of England also seem supporters of the financial intermediation theory of banking, as can be seen from the Bank's forecasting models, which do not include banks (Bank of England, 2014a).

Such conflicting views on the question of just how banks operate — are they individually able to create credit out of nothing, are they mere financial intermediaries, but able to create credit collectively, or are they intermediaries without any individual or collective power to create credit — are not new: this issue has been disputed vigorously over the past one hundred and fifty years, with the accounting treatment by banks being at the heart of the difference of the three theories. The choice of theory shapes different responses to banking, in particular concerning monetary policy and bank regulation.

One reason for the ongoing and still unsettled dispute is that so far discussions have taken place based on theories and assertions concerning the accounting operations of banks. But the respective merit of the three theories cannot be settled in theoretical models designed from first principles: worlds can be conceived theoretically in which each theory is plausible. Instead, it can be settled through empirical evidence on the actual operations and accounting practices of banking. Surprisingly, in the observation period — from the mid-19th century until this day - hitherto no scientific empirical test has been reported in the literature that is able to settle the issue.

and US financial crises since 2007 and the question of how to organise banking systems to enhance social welfare (see Werner, 2003, 2005, 2012).

As a result, in their model banks are pure intermediaries: "Intermediaries can take deposits from unproductive households to extend loans to entrepreneurs" (p. 6). Whether such a model is appropriate for a central bank engaged in 'quantitative easing' is an interesting question.

What all the relevant works in the past century and beyond have therefore in common — irrespective of which of these three theories they support — is that they have been content to make diverging assertions and counter-assertions on bank accounting, without attempting to resolve it on the basis of an empirical examination. This can be done by gathering empirical evidence of a kind that is able to differentiate between the three hypotheses, rejecting two of them, and failing to reject the third.

The following section briefly presents results of a testimonial-style inquiry, similar to the one conducted by the Macmillan Committee, followed by a more formal empirical test of the various banking theories.

3. A Brief Comment on Empirical Field Work Results

About twenty semi-structured interviews were arranged with bankers at various levels of seniority, and working for a variety of banks. They included very senior bankers, such as a former global CEO of Deutsche Bank, as well as loan officers and ordinary bank staff of small, local banks. Many of the meetings were tape-recorded. The findings, to be reported in more detail separately, can be summarised as follows: The majority of experienced bankers argued that either the financial intermediation theory or the fractional reserve theory was true. A minority supported the credit creation theory. The views were scattered without discernable pattern with respect to seniority, size of bank, type of bank or country.

When asked for evidence concerning the accounting procedures to be followed when a bank loan is extended, all witnesses testified that nowadays such details are embedded in the electronic accounting systems and IT of the banks, and therefore bank staff no longer see directly what happens in terms of precise booking procedures. As a result they could not provide firm evidence. When asked for permission to 'see into the IT', the response was that this was not possible, due to the high requirements concerning security, as well as client confidentiality, regulatory restrictions, and the need to keep commercial secrets safe. So staff were unable to back up whichever theory they supported with easily accessible evidence from the bank's internal processes, because the relevant accounting operations are embedded in secure IT systems to which they are not privy.

The field work thus found (a) that many bank staff are not aware of any potential power of one bank to create money, and (b) in today's computerised banking world, the answer to the crucial questions are tied up in 'black box' bank accounting software and cannot be conclusively settled by observation of human actions, such as accounting entries, manual transfers or other explicit commands. As a result, it is concluded that other empirical techniques are needed.

Meanwhile, the first empirical test of the three theories (Werner, 2014) came in supportive of the credit creation theory. However, since the environment in this live experiment could not be 100% controlled, the evaluation of the results was somewhat more complex. Thus there was a need for a fully controlled test.

4. A Controlled Empirical Test

Below, a controlled empirical test of the question whether an individual bank can create credit and money out of nothing is presented.

4.1. Predictions of the three theories

Before the theories are tested, it is necessary to distinguish the accounting implications and differences of the three theories of banking concerning the case of a newly extended bank loan, in order to identify the expectations according to each theory, if the researcher borrowed E200,000 from a bank.

(a) Accounting implications of the financial intermediation theory

According to this theory, banks are, as far as payments and accounts are concerned, not different from non-bank financial institutions, such as stock brokers or asset management companies. The latter are required by Client Money rules (see CASS in FCA, PRA, 2014) to hold deposits in custody for customers (likened to warehousing or bailments). In this case client funds are kept separately from the bank's own funds, so that customer deposits are not shown on the balance sheet as liabilities. All funds are central bank money that can be held in reserve at the central bank or deposited with other banks or financial intermediaries (where they are held off-balance sheet).

Table 4 Account Changes due to a E200k Bank Loan (Financial Intermediation Theory)

Assets Excess Reserves. Loans and investments. Total..................

Liabilities

- E + E

200 200

Total.

According to this theory, the bank balance sheet does not lengthen as a result of the bank loan. To test the veracity of this theory, the bank balance sheet needs to be examined before and after the extension of a bank loan. If it is found that a bank loan lengthens the balance sheet, then this theory is rejected.

(b) Accounting implications of the fractional reserve theory

According to this theory each individual bank is a financial intermediary that cannot create credit out of nothing. Funds are being treated as equivalent to cash or precious metals in the sense that they are thought to have the ability to flow between banks and the central bank. A bank can only lend money, if it has previously received the same amount either in cash or in excess reserves from another bank, whose own reserve balances will have declined after the transfer, or from the central bank.

"A bank will not lend more than its excess reserves because, by law, it must hold a certain amount of required reserves. ... Each depository institution can create loans (and deposits) only to the extent that it has excess reserves." (Miller, VanHoose, 1993, p. 331).

Following the exposition in Miller, VanHoose (1993, 330-331), the balance sheet evolution is as follows:

Table 5 Account Changes due to a E200k Bank Loan (Fractional Reserve Theory)

Step 1 Precondition for the Bank Loan

Assets Liabilities

Excess Reserves. +E 200 Deposits......................+E 200

Total..................... +E 200 Total...................... +E 200

Step 2 The Bank Loan

Assets Liabilities

Excess Reserves. — E 200

Loans and investments.... +E 200 _0

Total..........................................0 Total............................................0

For the bank to be able to extend a loan, it has to check its excess reserves, as this is, according to this theory, a strictly binding requirement and limitation, as well as its distinguishing feature. The bank cannot at any moment lend more money than its excess reserves, and it will have to draw down the reserve balance to lend. Thus, the balance sheet expansion is driven by the prior increase in a deposit that boosted excess reserves, not by the granting of a loan, which actually merely draws down a prior increase in reserve balances.

To test the veracity of this theory, it needs to be verified whether a bank extending a loan first confirms the precise amount of its available excess reserves before entering into the loan contract or before paying out the loan funds to the customer, so as not to exceed the reserve figure. If the bank is found not to have checked or not to be concerned about its reserve balances then this constitutes a rejection of the fractional reserve theory.

(c) Accounting implications of the credit creation theory

According to this theory, banks do not separate customer funds from own funds. Thus when lending, banks credit the borrower's account with the borrowed amount, although no new deposit has taken place (credit creation out of nothing). The balance sheet lengthens. Cash, central bank reserves or balances with other banks are not immediately needed, as reserve and capital requirements only need to be met at particular measurement intervals and are not a direct precondition of granting a loan. In other words, a bank can extend a new loan, even though it has not received any new deposit money or reserves. Bank loans create new deposits, not the other way round.

Table 6 Account Changes due to a E200k Bank Loan (Credit Creation Theory)

Assets Liabilities

Deposits (borrower's A/C). +E 200

Loans and investments____ +E 200 _

Total..................... +E 200 Total...................... +E 200

To test the veracity of this theory, the balance sheet of a bank needs to be examined before and after the extension of a bank loan. If the bank loan increased its balance sheet, without any reserve movements taking place, then this theory would be consistent with the evidence.

4.2. The Test

In Werner (2014), a bank was identified that would cooperate in the implementation of an empirical test able to identify which of the three theories is rejected by empirical observation and which is consistent with the facts. Raiffeisenbank Wildenberg e.G., a small cooperative bank in Lower Bavaria, headed by director Marco Rebl, was willing to cooperate. The researcher took out an actual bank loan from the bank, after the usual due diligence, and internal records were examined closely. The evidence seems only consistent with the credit creation theory (reported elsewhere). It was also found that the bank neither checked the level of its reserve holdings nor made any fund transfers from reserves or other sources before the loan was granted and paid out.

However, a complication with this test is that due to 24-hours electronic banking, there was no way for the researcher to control the test entirely so that no other transactions took place. Therefore director Rebl suggested a method of testing the banking hypotheses, which would allow the researcher to control for all other transactions without fail. Applying this method and reporting the results are a main contribution of this paper.

Mr Rebl explained that all bank accounting takes place within the IT system that is used on a daily basis by bank staff. Although the code of the software would directly show the commands following the entry of a bank loan, gaining access to the internal software code is difficult, given the high security requirements of bank IT systems. However, Mr Rebl then pointed out that there are in fact two parallel IT systems in operation at all cooperative banks, and both maintain all the accounting information of each bank. The daily balance sheets are from the software called 'BAP agree' (Bankarbeitsplatz — 'bank work place' — agree). This software is however not used for the compilation of the formal annual accounts, which are submitted to bank auditors and the regulatory authorities. For these formal accounts, a second, parallel system is utilised, called Hersbrucker Jahresabschlussprogramm (below 'HJAP'; literally: Hersbruck annual accounts programme, named after the town where the Raiffeisen cooperative bank is located whose director, Mr Weidinger, originally developed this programme). HJAP is also a full-blown bank IT system, but with the specialisation of meeting the more stringent annual reporting requirements and having features useful for their compilation, checking and submission of these accounts. Mr Rebl pointed out that the system contains all the bank accounting rules and it conforms with all bank supervisory,

prudential and legal requirements, regulations and procedures (which may not necessarily be relevant or enforceable on a daily basis as visible in BAP agree).

All transactions are aggregated in HJAP for the annual accounts at the end of the calendar year. While transactions booked in BAP feed into HJAP, sometimes transactions take place late in December that, possibly due to the holidays, were not properly recorded or reflected in the BAP agree system. In this case, the bank directors have the opportunity to ensure that these are booked in the HJAP system as appropriate for the formal annual accounts by manual entry.

Thus Mr Rebl explained that the empirical test can also be conducted by using the latest annual accounts (as of writing, the 2013 annual accounts), and using the latest HJAP software (as of writing, 2.0.2013/5), in order to book the test bank loan of E200,000 as if it was a missed trade that had to be booked after 31 December 2013, to be added to the official accounts for reporting purposes. Since in this case only one transaction will be booked — the bank loan from the researcher — there is no noise due to other autonomous transactions being undertaken by other bank customers. In other words, all other factors are controlled for. Since the bank loan could be entered into by the researcher after the end of 2013 in such a way that it needed manual input in HJAP, as indeed happens on occasion with standard loans, this does constitute a realistic and actual empirical test. This was implemented as suggested by Director Rebl, using the audited accounts of 2013.

Appendix 1 shows the audited and formally submitted accounts of Raiffeisenbank Wildenberg for the year 2013. Appendix 2 shows the same accounts after our empirical test bank loan of E200,000 has been transacted in the annual accounts bank IT software (HJAP). The summary accounts are shown below:

Table 7 Raiffeisenbank Wildenberg eG: Audited Annual Accounts 2013

ASSETS

1 Cash

2 Bills of exchange

3 Claims on financial. inst.

6,123,707.01 24,066,899.94 19,655,934.00

31 Dec. 2013

227,072.87

Post-Test

227,072.87

6,123,707.01 24,266,899.94 19,655,934.00

Difference

4 Claims on customers 2

5 Bonds, bills, debt instr. 1

6 Stocks and shares

7 Stake holdings

8 Stakes in related firms

9 Trust assets

10 Compensation claims on the

397,768.68

4,713.81

397,768.68

4,713.81

200,000.00

public sector

11 Immaterial assets

12 Fixed assets

13 Other assets

14 Balancing item

15 Difference from asset

188,977.92 335,969.95 2,126.22 46,334.50

221,549.46 335,969.95

2,126.22 46,334.50

valuations

16 SUM OF ASSETS

51,049,504.92

51,249,504.92

200,000.00

The only two items that are affected are the claims on customers — the bank loan as a claim by the bank on the borrower due to the borrower's obligation to repay the loan, and the total balance of assets. Both increased by the loan amount of E200,000.

Table 8 Raiffeisenbank Wildenberg eG: Audited Annual Accounts 2013

LIABILITIES 31 Dec. 2013 Post-Test Difference

1 Claims by financial inst. 5,265,491.16 5,265,491.16

2 Claims by customers 41,462,424.00 41,662,424.00 200,000.00

2A Savings accounts 10,494,856.16 10,494,856.16

2B Other liabilities 30,967,567.84 31,167,567.84 200,000.00

- BA daily 14,069,056.09 14,269,056.09 200,000.00

- BB with agreed maturity 16,898,511.75 16,898,511.75

4 Trust liabilities 4,713.82 4,713.82

5 Other liabilities 33,812.09 33,812.09

6 Balancing item 12,787.37 12,787.37

7 Reserves 682,874.80 682,874.80

11 Fund for Bank Risk 420,000.00 420,000.00

12 Own Capital 3,167,401.68 3,167,401.68

13 SUM LIABILITIES 51,049,504.92 51,249,504.92 200,000.00

As can be seen, the only difference in the accounts before and after the loan has been extended lies in the items expected a priori by the credit creation theory. This test design is in line with the assessment by the Macmillan Committee (1931), which said that a feature of

the credit creation theory is:

"If no additional in-payments were made by customers and there were no withdrawals in cash, the volume of deposits of a single banker would fluctuate only with the volume of the loans he himself made." (p. 12).

Written confirmation was received by the director of Raiffeisenbank Wildenberg that the reserve balance of the bank was neither checked nor in any way manipulated in order to grant the new loan. The bank also confirmed in writing that no other transaction or operation was necessary to credit the borrower's account and that, in particular, no money was transferred away from anywhere else to the borrowers account. We thus conclude that the fractional reserve and financial intermediation hypotheses are rejected.

5. Conclusion

There are three theories of banking, with differing claims about how bank accounting operates. The first theory argues that banks are financial intermediaries gathering deposits and then lending these out, just like non-bank financial institutions such as brokers or asset management companies. In this case, the extension of a loan does not

lengthen the lender's balance sheet. The second theory argues that while each bank is a mere financial intermediary, like a stock broker or asset manager, the banking system as a whole creates credit and money, as reserves are lent on many times (multiple deposit expansion of fractional reserve banking). In this case, reserves with the central bank are lent on, also not lengthening the bank's balance sheet. The third theory argues that banks are not financial intermediaries that gather deposits and lend these out, and that do not use central bank reserves to lend them on; instead, they create credit and money individually out of nothing when they extend what is called a bank loan (credit creation).

While influential economists and thinkers can be found supporting each of these theories throughout the past century, surprisingly, no empirical test has been reported in the very long history of banking (modern banking being over 300 years old, and banking in general likely going back as far as 5000 years in Mesopotamia) which attempts to refute the theories in order to identify which one if any is not rejected by the evidence.

In this paper the results of an empirical test were presented, whereby the researcher took out a loan from a bank and this was booked in the bank's accounting and settlement IT systems under controlled conditions that excluded unrelated transactions. It was found that the credit creation theory of banking is consistent with the empirical observations, while the other two theories are not.

5.1 Lack of accounting rigour as a cause of confusion

The core activity of banking, what is commonly called 'receiving deposits' and 'lending', are in actual fact the creation and maintenance of accounting records and thus can be considered a form of applied accounting. However, this central feature of banking has been unduly neglected in the treatment of banks and their impact on the economy by academic authors, whether in journal articles, books or text books. It seems that a greater emphasis on bank accounting and a more careful consideration of its implications should have raised serious doubts about the theoretical viability and consistency of both the fractional reserve and the financial intermediation theories much earlier, even without our conclusive empirical test.

Accounting problems of the fractional reserve theory:

Customer deposits become the property of the bank and are fully shown on the balance sheet. Legally, they are not a deposit (bailment) but a loan to the bank, of which the bank keeps a record — the account statements. Thus it is a priori not clear why reserves should be the constraint on bank lending as claimed by the fractional reserve theory. This theory neglects, despite its rhetorical awareness of the 'creation of accounting records', the very transaction of booking a loan on the bank's balance sheet: the borrower's account is not shown, as it is simply assumed that the money 'leaves the bank immediately', effectively on the assumption that the loan is paid out in cash. Let us assume therefore that the borrower, due to delays, initially does not spend the money and it stays in their account, or alternatively, that Step 1 and Step 2 of fractional reserve banking (Table 5) happen on the same day, as may easily be possible (Table 9):

Table 9 Account Changes due to a E200k Bank Loan (Fractional Reserve Theory)

Steps 1 and 2 all in one

Assets

Excess Reserves. +E 200

Excess Reserves —E 200

Loans and investments________+E 200

Total..........................................+E 200

Liabilities Deposits......................+E 200

Total......................+E 200

The customer deposit, according to this theory, is the primary step, and it takes the form of a cash deposit by a customer. This produces an excess cash reserve. The borrower is then assumed to receive the loan in the form of cash, drawing down the excess cash balance. This scenario seems an unusual special case. Let us consider the more common case that a deposit is made by bank transfer from another account, assumed here with the same bank, to keep out the complication of interbank transfers. Moreover, as is usually the case in practice, the borrower receives the loan as credit to the borrower's cheque account at the bank. The customer deposit balance does not change due to the transfer of a deposit from one customer to another at the same bank. There is no change in excess reserves. Now it is time to book the new loan. Table 10 shows the result of this,

more common, scenario:

Table 10 Account Changes due to a E200k Bank Loan

(Fractional Reserve Theory, but borrower receives non-cash)

Steps 1 and 2 all in one

Assets Excess Reserves. Excess Reserves

Loans and investments____ +E 200

Total..................... +E 200

Liabilities Deposits......................+E 200

Total......................+E 200

As can be seen, by dropping the unrealistic assumption that deposits and loans are transacted in cash, we are right back at the credit creation theory: the asset side expands by the amount of the loan (reflecting the loan contract) and so does the liability side, as the borrower's account is credited, as shown by Table 6. In that case, the bank is not restricted to lending only as much as its excess reserves, because what is called a 'deposit' is actually just a record of the bank's debt to others. Despite Samuelson's (1948) protestation that "A bank cannot eat its cake and have it too" (p. 325f), the bank still has its reserves at the moment it has granted the bank loan and credited the borrower's account. In other words, in Table 9, instead of being a necessary requirement as claimed by the theory, the movement of reserves is redundant and there is no reason why such a movement is necessary in order for the bank to extend the loan. Equally, the bank is able to lend more than its excess reserves, since, as seen, the amount of reserves does not fall when the loan is granted. As was found in the empirical test, banks do not actually check for excess reserves before they grant credit. A careful examination of the accounting entries involved should have made this clear to authors of the fractional reserve theory. The argument that the newly created deposit entry of the borrower will 'soon leave the bank' also does not change the results: in this case, in practice, the bank simply swaps a liability to the borrower (the newly created deposit) with a liability to a bank (the bank of the

receiver of the payment made by the borrower from their newly created deposit). The balance sheet total remains unchanged, in its lengthened form.

Accounting problems of the financial intermediation theory:

The financial intermediation theory maintains that banks are, like non-banks, mere financial intermediaries. The accounts of non-bank financial intermediaries are characterised by a required separation of client money and the firm's own funds. Thus a careful study of accounting quickly reveals that the financial intermediation theory is impossible to reconcile with observed bank balance sheet information: the latter show customer deposits on their liabilities side. Stock brokers do not show their clients' assets, even if invested by them on a discretionary basis, as part of their own balance sheets. Mutual fund management firms and the assets of their fund investor clients are kept completely separately. Stock brokers' assets are boosted by their own investments, but not those of their clients. Thus an insolvency of a stock broker leaves client funds unencumbered: they are fully owned by the clients. But bank 'deposits' are owned by the banks and bank insolvency means that the client funds are part of the assets of the bankrupt firm. Depositors are merely general creditors, ranking ahead of shareholders (although smaller amounts may be covered by deposit insurance schemes, which are a separate issue). Thus a comparative analysis of stock brokers (as representative examples of non-bank financial intermediaries) and banks reveals that banks are different from non-banks, because they do not segregate client assets. A careful analysis of bank balance sheets and their accounting information would have provided ample notice to supporters of the financial intermediation theory, so dominant over the past half-century, that it always was a nonstarter, since banks could not possibly be financial intermediaries: how else could the rapid growth and massive scale of their own balance sheets be explained? Alas, it seems researchers in banking, finance and economics have woefully neglected accounting realities. Meanwhile, Arnold (2009) argued that, likewise, accounting research has not considered the economic implications of accounting facts. Thus more recognition of accounting facts is needed in economics, while accountants should not be shy to include an analysis of economic implications of accounting practice in their research remit.

5.2 Implications for Bank Regulation

The implications of our empirical findings are far-reaching for bank regulation and the design of official policies. Since bank regulation is based on the prevailing understanding of the role of banks, it is not surprising that we find that the prevailing theory of banking has also shaped the approach to bank regulation. Thus during the past forty years when the financial intermediation theory of banking has been dominant, bank regulation has focused on capital adequacy. During the earlier thirty years or so, when the fractional reserve theory of banking was dominant, bank regulation mainly took the form of reserve requirements. Neither has been very successful. This is likely connected to their underlying theories of banking, as will be briefly reviewed below.

Regulation via reserve requirements and the fractional reserve theory of banking

Bank regulation used to focus on reserve requirements. This regulatory treatment was based on, and theoretically supported by the fractional reserve theory of banking. It was found, however, that this regulatory policy was impracticable for central banks to operate (Goodhart, 1989). This article has found out why this should have been the case: the fractional reserve theory of banking is wrong. As careful analysis of bank accounting shows, banks' reserves with the central bank — the central feature of this theory and also the reserve requirement type of bank regulation based on it — never leave the accounts of the

central bank: like 'deposits' of the public with banks, which in reality are simply records of units of accounting money owed by banks to the public, 'reserves' by banks at the central bank are simply accounting records of money units owed by the central bank to the banks. Such indebtedness cannot result in money circulating in the economy, with the exception of the small amount of legal tender cash involved (Ryan-Collins et al., 2011). This explains why reserve requirements have gradually come to be de-emphasised, and some central banks, such as the Bank of England and the Swedish Riksbank, have even abolished reserve requirements altogether (playing havoc with the fractional reserve theory of banking by abductio ad absurdum, since the 'money multiplier' in this case is infinity).

Capital adequacy regulation and the financial intermediation theory of banking

In parallel with the policy to de-emphasise reserve requirements to regulate banks, central banks, via their influence over the Basel Committee on Banking Supervision, have shifted to the policy to regulate banks via capital ratios. This approach is predicated on the veracity of the financial intermediation theory. If banks are only financial intermediaries, then as agents their actions cannot have a macroeconomic feedback. But since the money supply is "created by banks as a byproduct of often irresponsible lending", as Martin Wolf called it (Wolf, 2013), the ability of capital adequacy ratios to reign in expansive bank credit policies is limited: imposing higher capital requirements on banks will not necessarily enable the prevention of boom-bust cycles and banking crises, since even with higher capital requirements, banks could still continue to expand the money supply, thereby fuelling asset prices, since some of this newly created money can be used to increase bank capital.

The link between bank credit creation and bank capital was most graphically illustrated by the actions of Barclays Bank during the financial crisis. This case study demonstrates that the money creators can effectively produce any level of capital, whether directly or indirectly: Unwilling to accept public money to shore up its capital, as several other major UK banks had done, it arranged for a Gulf investor to purchase several billion pounds worth of its newly issued preference shares, thus raising the amount of its capital and avoiding bankruptcy. The Gulf investor did not need to take the trouble of making liquid assets available for this investment: Barclays Bank generously offered to lend it to the Gulf investor — or, in other words, Barclays invented the money through credit creation. The preference shares of Barclays served as collateral. Table 10 illustrates such bootstrapping. ^

Table 10 Capital Out of Nothing: Boot-strapping Barclays Bank during the Crisis

Step 1: Loan to Gulf Investor Assets

Liabilities

Loans and investments____ + 5.8

Total..................... + 5.8

Deposits Capital

Step 2: Capital Raising

Assets

Liabilities

Loans and investments____ + 5.8

Total..................... + 5.8

Deposits Capital

Since it is now an established fact that banks newly invent the money that is 'loaned' by creating it out of nothing, the loan to the Gulf investor created (in step 1) a simultaneous asset and liability on the bank's balance sheet, whereby the customer's borrowed money is the fictitious customer deposit on the liability side. In step 2, the newly issued preference shares boost equity, and are paid for with this fictitious customer deposit, simply by liability swap from item 'customer deposit' to item 'equity'. Barclays was then able to report a significant rise in its equity capital, and hence in its capital/asset ratio.

We learn from this that it is possible for banks under the right circumstances to show almost any amount of capital to regulators, irrespective of the true state of their books. For instance, the attempt to improve on Basel I through the Basel II framework focused on measures that were known to reduce the capital that needed to be reported by large banks, which was at the time thought to lower risks and deliver net benefits, including a 'slight' reduction in the frequency of banking crises.

Despite such egregious examples such as Barclays Bank, it will be pointed out that in practise banks' ability to create money meets some limitations. This is especially true for small banks not acting in collusion with others. However, if banks of roughly equal size and customer base create credit at about the same pace, they may in practice not face any limitations on their ability to create money and credit, as they may find that their mutual liabilities due to transfers between them cancel out. Even higher required capital can be supplied from the very money created by the banks.

In this paper the reason why bank regulation based on the fractional reserve and on the financial intermediation theories of banking have not been successful could be identified. On the other hand, having no bank regulation is also not likely to be successful, as the 2008 financial crisis has shown: Bank credit derivatives had been entirely unregulated on the advice of Alan Greenspan and other supporters of unregulated markets. They have since concurred with their critics that regulation would have been better. But what type of bank regulation is likely to be more successful?

In the era when the credit creation theory of banking was dominant, its proponents pointed out that bank credit creation and growth in economic activity are connected, and credit for different types of transactions has a diverging effect on the economy. They have thus

14 A PriceWaterhouseCoopers Risk Management study commissioned by the European Commission on the effects of Basel II (PWC, 2004) argued that "A key objective of banking regulation is to ensure the efficient and safe operation of banks through the economic cycle, and any assessment of the impact of Base II needs therefore to take account of its macroeconomic effects. These effects can be divided into two kinds: the short- to medium-term effects on the current economic cycle and the longer-term effects. .. .our analysis of the impact of Basel II on the balance sheet and capital structure of banks. revealed that Basel II could, on balance, reduce the amount of regulatory capital in a number of EU countries, in turn potentially leading to lower lending rates. So the trade-off outlined above between possible short-term loss and logner term gain does not really occur. There is little or no short-term loss at a [sic] EU level. On the other hand, the National Institute's historical analysis of past banking crises suggests that Basel II will only have a moderate effect on the frequency of banking crises. The effect of changes in regulatory capital rules are likely to be helpful in this regard, but they are small relative to the other causes of banking crises. They are likely, at best, to reduce only slightly the frequency of such crises."

favoured bank regulation that directly targets bank credit, both its quantity and its quality (i.e. the type of transaction that gets funded by bank credit), whereby economically desirable bank credit is encouraged, and economically harmful credit creation is forbidden or restricted quantitatively. The relationship between disaggregated bank credit creation on the one hand and nominal GDP growth, real GDP growth and asset prices on the other was identified by the Quantity Theory of Credit (Werner, 1992, 1997, 2005, 2012), which can serve to guide the direction of credit. In particular, guidance would be used to restrict credit for transactions that do not contribute to nominal GDP: such credit for financial transactions creates asset boom-bust cycles and instability in the banking system. Before the use of reserve requirements, capital adequacy or interest rate targeting became dominant in the second half of the 20th century, central banks focused more on controlling bank credit directly. This policy was pioneered by the Reichsbank in 1912, but has been tried and tested by most central banks sometime between the 1920s through to the 1960s (with some continuing the practice until the 1980s, such as the Bank of Japan and the Banque de France with their 'window guidance' and encadrement du credit techniques). Credit guidance has an excellent track record in achieving the targeted credit growth and sectoral allocation (Werner, 2005).

The findings also have broader implications for policies to ensure economic growth and minimise unemployment, as well as policies by developing countries concerning the question of how to maximise sustainable growth. It is notable that the highly successful achiever economies of East Asia, which demonstrated rapid economic growth in the postwar era, all used mechanisms to guide bank credit. Some authors argue that their 'widnow guidance' bank credit control mechanisms were the real secret to their success (Werner, 2003, 2005), and our findings provide fundamental support for such claims.

Overall it can be said that one of the implications of this study is that it does not make much sense to build economic theories of the financial sector, if these are not based on accounting realities. The role of accounting in economics should thus be increased, both in research and in the teaching of economics. This includes the role of national income accounting and flow of funds information (see Winkler et al., 2013a, b), which have to be reconciled with those records by the banks, who are not only the "central settlement bureau, .a kind of clearing house or bookkeeping centre for the economic system" (Schumpeter, 1934, p. 124), but also the creators and allocators of the money supply. The reflection of empirical bank reality within theories and textbooks surely must become the 'new normal' in finance.

Bibliography

Ahmed, A. S., C. Takeda and S. Thomas (1999), Bank loan loss provisions: a reexamination of capital management, earnings management and signalling effects, Journal of Accounting and Economics, 28, 1-25

Arnold, P. J. (2009). Global financial crisis: The challenge to accounting research, Accounting, Organizations and Society, 34(6—7), 803—809.

Alhadeff, David A. (1954), The Rise of Commercial Banking, Berkeley: University of California Press; reprinted in 1980 by Arno Press as: Monopoly and Competition in Banking

Allen, Franklin & Carletti, Elena & Gale, Douglas, 2014. "Money, financial stability and efficiency," Journal of Economic Theory, Elsevier, vol. 149(C), pages 100-127

Allen, F. and D. Gale (2000), Financial Contagion, Journal of Political Economy, 108, 1-33.

Allen, F. and D. Gale (2004a), Financial intermediaries and markets, Econometrica 72 1023—1061.

Allen, F. and D. Gale (2004b), Competition and financial stability, J. Money, Credit, Banking 36 453-480

Allen, F. and Santomero (2001), What do financial intermediaries do?, Journal of Banking & Finance 25 (2001) 271-294

Aschheim, Joseph (1959). Commercial Banks and Financial Intermediaries: Fallacies and Policy Implications," Journal of Political Economy, LXVII (February, 1959), 59-71.

Ball, Ryan, Robert M. Bushman and Florin P. Vasvari (2008), The Debt-Contracting Value of Accounting Information and Loan Syndicate Structure, Journal of Accounting Research, Volume 46, Issue 2, May 2008, Pages: 247-287

Baltensperger, Ernst (1980), Alternative approaches to the theory of the banking firm, Journal of Monetary Economics, vol. 6, pp. 1-37

Baltensperger, Ernst (1989), Credit, in: Eatwell, John, Murray Milgate and Peter Newman (eds.) (1989), The New Palgrave Money, Basingstoke: Macmillan

Bank of England (2014a), The Bank of England's forecasting platform: COMPASS, MAPS, EASE and the suite of models, Bank of England Working Paper No. 471

Bank of England (2014b), Money in the modern economy: an introduction, by Michael McLeay, Amar Radia and Ryland Thomas of the Bank's Monetary Analysis Directorate, Quarterly Bulletin, Q1 2014.

Bank of England (2014c), Money creation in the modern economy, by Michael McLeay, Amar Radia and Ryland Thomas of the Bank's Monetary Analysis Directorate, Quarterly Bulletin, Q1 2014.

Bank of England (2014d), Taking the long view: how market-based finance can support stability, Speech given by Dame Clara Furse, External Member of the Financial Policy Committee, Bank of England at the Chartered Institute for Securities and Investment, Liverpool, 28 March 2014

Barth, Mary E, Wayne R. Landsman, Mark Lang, Christopher Williams (2012), Are IFRS-based and US GAAP-based accounting amounts comparable?, Journal of Accounting and Economics, Volume 54, Issue 1, August 2012, Pages 68-93

Barth, Mary E., Greg Clinch and Toshi Shibano (1999), International accounting harmonisation and global equity markets, Journal of Accounting and Economics, 26, (1—3), January 1999, 201-235

BBC (2013), Negative interest rates idea floated by Bank's Paul Tucker, 26 February 2013, accessed online

BCBS (1999), Capital requirements and bank behaviour: the impact of the Basel Accord, BCBS Working Paper No 1, April, Basel: BIS

BCBS (2009), Strengthening the Resilience of the Banking Sector, Consultative Document, December, Basel: BIS

Bellerby, John Rotherford (1923) (ed.), Control of Credit as a Remedy for Unemployment, International Association on Unemployment, London: P. S. King,

Bencivenga, V.R. and B. Smith (1991), "Financial Intermediation and Endogenous Growth", Review of economic Studies, 58, 195-209.

Benes, Jaromir, and Michael Kumhof (2012), The Chicago plan revisited, WP/12/202, Washington DC: IMF,

Bernanke, Ben (1993), Credit in the macroeconomy, FRBNYQuarterly Review, Spring (1992-93).

Bernanke, Ben and Mark Gertler (1995), Inside the black box: the credi tchannel of moentary policy transmission, Journal of Economic Perspectives, 9 (4), 27-48

Bernanke, Ben and Alan S. Blinder (1989), Credit, money and aggregate demand, American Economic Review, Proceedings, 78, 435-39

Bezemer, Dirk (2009), Understanding financial crisis through accounting models, Accounting, Organisations and Society, 26, 676-688

Bischof, Jannis (2014), Identifying Disclosure Incentives of Bank Borrowers During a Banking Crisis, Journal of Accounting Research, Volume 52, Issue 2, May 2014, Pages: 583—598

Branson, William (1968), Financial Flows in the US Balance of Payments, Amerstam: North-Holland

Brunnermeier, Markus and Yuliy Sannikov, The i theory of money, mimeo

Bushman, Robert M., Abbie J. Smith and Regina Wittenberg-Moerman (2010), Price Discovery and Dissemination of Private Information by Loan Syndicate Participants, Journal of Accounting Research, Volume 48, Issue 5, December 2010, Pages: 921—972

Carney, Mark (2014), One Mission. One Bank. Promoting the good of the people of the United Kingdom, Speech given by Mark Carney, Governor of the Bank of England, Mais Lecture at Cass Business School, City University, London, available at

www.bankofengland.co.uk/publications/Documents/speeches/2014/speech715.pdf

Cassel, Gustav (1918), Theoretische Sozialökonomie, Leipzig: C. F. Wintersche Varlagshandlung

Casu, Barbara and Claudia Girardone (2006), Bank competition, concentration and efficiency in the single European market, The Manchester School Vol 74 No. 4 Special Issue, 441—468

Casu, Barbara, Claudia Girardone and Philip Molyneux (2006), Introduction to Banking, Harlow: FT Prentice Hall, Pearson Education Ltd.

Cecchetti, Stephen G. (2008), Money, Banking and Financial Markets, 2nd edition, New York: McGraw-Hill Irwin

Crick, W. F. (1927), The genesis of bank deposits, Economica, 191-202 says the same.

Culbertson, John. "Intermediaries and Monetary Theory: A Criticism of the Gurley-Shaw Theory," AER, XLVIII (March, 1958), 119-31

Davenport Herbert J. (1913), The Economics of Enterprise, New York: Augustus M. Kelley.

Dewatripont, Matthias; Jean-Charles Rochet and Jean Tirole (2010), Balancing the Banks: Global Lessons from the Financial Crisis, Princeton: Princeton University Press

Diamond, Douglas W. (1984): Financial Intermediation as Delegated Monitoring, Review of Economic Studies, 51: 393-414.

Diamond, Douglas W. and Dybvig, P. H. (1983): Bank Runs, Deposit Insurance, and Liquidity:

Journal of Political Economy, 91(3): 401-419.

Diamond, Douglas W. & Rajan, Raghuram G., (2001). "Banks, short-term debt and financial crises: theory, policy implications and applications," Carnegie-Rochester Conference Series on Public Policy, Elsevier, vol. 54(1), pages 37-71, June

Domar, E. (1947), Expansion and employment, American Economic Review, March Douglas, C. H. (1924), Social Credit, London: Eyre & Spottiswoode

Dye, A. Ronald and Sri S. Sridhar (2008), A positive theory of flexibility in accounting standards, Journal of Accounting and Economics, 46, (2-3), December 2008, 312-333

Eatwell, John, Murray Milgate and Peter Newman (eds.) (1989), The New Palgrave Money, Basingstoke: Macmillan

Eccher, E. A., K. Ramesh and S. R. Thiagarajan (1996), Fair value disclosures by bank holding companies, Journal of Accounting and Economics, 22, 79-117

Fama, Eugene (1985), What's different about banks? Journal of Monetary Economics 15 (1985) 29-39.

FCA, PRA (2014), Client Assets (CASS), FCA Handbook, London: Bank of England, online at fshandbook.info

Fields, Thomas D., Thomas Z. Lys, and Linda Vincent (2001), Empirical research on accounting choice, Journal of Accounting and Economics, 31, 255-307

Fisher, Irving (1935), 100% Money, New York: The Adelphi Company

Gertler, M. and Kiyotaki, N. (2011), "Financial Intermediation and Credit Policy in Business Cycle Analysis", in: Friedman, B. and Woodford, M. (eds.), Handbook of Monetary Economics, Volume 3A, North Holland: Elsevie

Goodfriend, Marvin (1991), Money, credit, banking, and payments system policy, Economic Review, Federal Reserve Bank of Richmond, Jan/Feb 1991, pp 7-23

Goodhart, C. A. E. (1989), 'Disintermediation', in Eatwell, John, Murray Milgate and Peter Newman (eds.) (1989), The New Palgrave Money, Basingstoke: Macmillan

Goodhart, C. A. E. (1989), Money, Information and Uncertainty, 2nd ed., London: Macmillan Gormley, Todd A., Bong Hwan Kim and Xiumin Martin (2012), Do Firms Adjust Their Timely Loss Recognition in Response to Changes in the Banking Industry?, Journal of Accounting Research, Volume 50, Issue 1, March 2012, Pages: 159-196,

Gorton, Gary and George Pennacchi (1990), Financial Intermediaries and Liquidity Creation. The Journal of Finance, Vol. 45, No. 1. (Mar., 1990), pp. 49-71

Granja, Joao (2013), The Relation Between Bank Resolutions and Information Environment: Evidence from the Auctions for Failed Banks, Journal of Accounting Research, Volume 51, Issue 5, December 2013, Pages: 1031-1070

Gurley, John G., and Shaw, E. S. "Financial Aspects of Economic Development," AER, XLV (September, 1955), 515-28.

Guttentag, Jack M. and Robert Lindsay (1968), The Uniqueness of Commercial Banks, Journal of PoiticalEconomy, Vol. 76, No. 5 (Sep. - Oct., 1968), pp. 991-1014

Hahn, Albert C. (1920), Volkswirtschaftliche Theorie des Bankkredits, Tübingen: J. C. B. Mohr

Harrod, R. (1939), An essay in dynamic theory, Economic Journal, March

Hawtrey, Ralph George, (1919). Currency and Credit. Longmans, Green and Co

Heffernan, Shelagh (1996), Modern Banking in Theory and Practice, Chichester: John Wiley and Sons.

Heijdra, Ben J. and Frederick Van der Ploeg (2002), Foundations of Modern Macroeconomics, Oxford: Oxford University Press

Hellwig, Martin (1977), A model of borrowing and lending with bankruptcy, Econometrica, vol. 45, pp. 1879-1906

Hellwig, Martin (1991), Banking, financial intermediation and corporate finance, in: Alberto Giovannini and Colin Mayer (eds.), European Financial Integration, Cambridge: Cambridge University Press

Hellwig, Martin F. (2000), Financial intermediation with risk aversion, Review of Economic Studies, 67(4):719—742

Houston, Joel F., Liangliang Jiang, Chen Lin, Yue Ma (2014), Political Connections and the Cost of Bank Loans, Journal of Accounting Research, Volume 52, Issue 1, March 2014, Pages: 193-243

Howe, Robert Harrison (1915), The Evolution of Banking; A study of the development of the credit system, Chicago: C. H. Kerr & Company

James, F. Cyril (1930), The Economics of Money, Credit and Banking, New York: Ronald Press Co.

Kashyap A., R. Rajan, J. Stein (2002), Banks as liquidity providers: An explanation for the coexistence of lending and deposit-taking, Journal of Finance 57, 33-73.

Keynes, John Maynard (1924), Tract on Monetary Reform, London: Macmillan

Keynes, John Maynard (1930), A Treatise on Money, London: Macmillan

Keynes, John Maynard (1936), The General Theory of Employment, Interest and Money, London: Macmillan

Keynes, John Maynard and D. E. Moggridge (1983), Collected Writings XI, Economic Articles and Correspondence: Academic, Collected Works of Keynes, p. 419

Kiyotaki, Nobuhiro and John Moore (1997), Credit cycles, Journal of Political Economy, 105, 211-48

Klein, Michael A. (1971), A theory of the banking firm, Journal of Money, Credit and Banking, vol. 3, May, pp. 205-218

Knapp, Georg Friedrich (1905), Staatliche Theorie des Geldes, Leipzig: Duncker & Humblot

Kohn, Donald (2009), Monetary policy research and the financial crisis: strength and shortcomings. Speech at the Federal Reserve Conference on key developments in monetary policy, Washington, D.C., October 9, 2009.

Koo, Richard and Shigeru Fujita (1997), Koka asai ryoteki kin'yu kanwa (Quantitative Monetary Easing is Hardly Effective), Keizai Kyoshitsu, Nikkei, 26 March 1997, p. 31

Konchitchki, Yaniv and Panos N. Patatoukas (2014), Accounting earnings and gross domestic product, Journal of Accounting and Economics, Volume 57, Issue 1, February 2014, Pages 76-88

Krugman Paul R. and Maurice Obstfeld (2000), International Economics, Theory and Policy, Reading, MA: Addison-Wesley Publishing Company

Law, John (1705), Money and Trade Consider'd with a Proposal for Supplying the Nation with Money, London.

Lo, Alvis K. (2014), Do Declines in Bank Health Affect Borrowers' Voluntary Disclosures? Evidence from International Propagation of Banking Shocks, Journal of Accounting Research, Volume 52, Issue 2, May 2014, Pages: 541-581

Lutz, F. A. (1939), Velocity Analysis and the Theory of the Creation of Deposits, Economica, Vol. 6, No. 22 (May, 1939), pp. 156-169

MacLeod, Henry Dunning (1855-6), The Theory and Practice of Banking, in 2 volumes; citations from the 6th edition, London: Longman, Greens and Co., 1906.

MacLeod, Henry Dunning (1891), Theory of Credit, vol. 2

Macmillan Committee (1931). British Parliamentary Reports on International Finance: The Report of the Macmilbn Committee. H.M.Stationery Office, London, 1931.

Marshall, Alfred (1888), Report by the Gold and Silver Commission of 1887, Command 5512 Marshall, Alfred (1890), Principles of Economics

Matthews, Kent and John Thompson (2005), The Economics of Banking, Chichester: John Wiley and Sons

Mayer, Colin (19888), New issues in corporate finance, European Economic Review, vol. 32, pp. 1167-88

Miller, Roger L. and David D. VanHoose (1993), Modern Money and Banking, International Editions, 3rd ed., New York: McGraw-Hill

Mints, Lloyd (1945), A History of Banking Theory, Chicago: University of Chicago Press

Moeller, Hero, (1925), Die Lehr vom Oelde, Leipzig: Quelle und Meyer.

Monti, Mario (1972), Deposit, credit and interest rates determination under alternative bank objective funcitons, in: G. P. Szegö and K. Shell (eds.), Mathematical Methods in Investment Finance, Amsterdam: North-Holland

Müller, Adam Heinrich (1816), Versuch einer neuen Theorie des Geldes, mit besonderer Rücksicht auf Großbritannien. Leipzig

Myers, S. C. and Rajan, R. G. (1998), The paradox of liquidity, The Quarterly Journal of Economics, 113 (3), 733-771

(OECD) H.J. Blommestein, (1995), "Structural Changes in Financial Markets: Overview of Trends and Pros-pects", in: H.J. Blommestein and K. Biltoft, eds., The New Financial Landscape (Forces Shaping the Revolution in Banking, Risk Management and Capital Markets), Paris: OECD.

Phillips, Chester A. (1920), Bank Credit, New York: Macmillan Pigou, Arthur C. (1927), Industrial Fluctuations

PriceWaterhouseCoopers (2004), Study on the financial and macroeconomic consequences of the draft proposed new capital requirements for banks and investment firms in the EU, MARKT/2003/02/F

Ramanna, Karthik (2008), The implications of unverifiable fair-value accounting: Evidence from the political economy of goodwilll accounting, Journal of Accounting and Economics, Volume 45, (2—3), August 2008,253-281

Riordan, Michael H. (1993), Chapter 11: Competition and bank performance: a theoretical perspective, in: Mayer, Colin and Xavier Vives (eds.), Capital Markets and Financial Intermediation, Cambridge: Cambridge University Press, pp. 328-343

Robertson, Dennis (1926), Banking Policy and the Price Samuelson, Paul (1948), Economics, New York: McGraw-Hill

Romer David (2006), Advanced Macroeconomics, 3rd edition.

Ryan-Collins, Josh, Tony Greenham, Richard A. Werner and Andrew Jackson (2011), Where Does Money Come From? London: new economics foundation

Sadka, Ronnie (2011), Liquidity risk and accounting information, Journal of Accounting and Economics, Volume 52, Issues 2—3, November 2011, Pages 144-152

Samuelson, Paul and William Nordhaus (1995), Economics, New York: McGraw-Hill

Saving, Thomas R. (1977), A theory of the money supply with compettive banking, Journal of Monetary Economics, vol. 3, pp. 289-303

Schumpeter, Joseph Alois (1912), Die Theorie der wirtschaftlichen Entwicklung, Leipzig, Duncker und Humblot

Schumpeter, Joseph Alois (1917/1918), Das Sozialproduct und die Rechenpfennige. Glossen und Beiträge zur Geldtheorie von heute. Archivfür Sozialwissenschaft und Sozialpolitik, 44.

Schumpeter, J.A. (1934). The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Business Cycle. Cambridge (Mass.): Harvard University Press.

Schumpeter, Joseph Alois (1954), History of Economic Analysis, New York: Oxford University Press

Schumpeter, J.A. (1970). Das Wesen des Geldes, edited by F.K. Mann. Göttingen: Vandenhoeck & Ruprecht.

Sealey, C. and Lindley, J. T. (1977). 'Inputs, Outputs and a Theory of Production and Cost at Depositary Financial Institutions', Journal of Finance, Vol. 32, pp.

1251-1266.

Sheard, Paul (1989), The mainbank system and corporate monitoring and control in Japan, Journal of Economic Behaviour and Organisation, vol. 11, pp. 399-422

Smith, Adam (1776), An Inquiry into the Nature and Causes of the Wealth of Nations

Smith, Paul F (1966). "Concepts of Money and Commercial Banks," Journal of Finance, XXI (December, 1966), 635-48.

Smith, Warren (1959), "Financial Intermediaries and Monetary Controls," Quarterly Journal of Economics, LXXIII (November, 1959), 533-53.

Soddy, Frederick (1934), The Role of Money, London: Elkin Mathews & Marrot

Solomon, Ezra (1959), "Financial Institutions in the Savings-Investment Process," Conference on Savings and Residential Financing. Chicago: U.S. Savings and Loan League, 1959 Proceedings, pp. 10-55

Sorensen, Peter Birch and Hans Jorgen Whitta-Jacobsen (2010), Introducting Advanced Macroeconomics, 2nd Ed., Maidenhead: McGraw-Hill

Stamp, Josiah (1927), Industrial Fluctuations. by A. C. Pigou, Review by: J. C. Stamp, The Economic Journal, Vol. 37, No. 147 (Sep., 1927), pp. 418-424

Stein, Jeremy C. (2014), Banks as patient debt investors, American Economic Association/American Finance Association Joint Luncheon, Philadelphia, Pennsylvania, 3 January 2014

Steuart, James (1767), An Inquiry into the Principles of Political Economy

Stiglitz, Joseph (1997), Economics, 2nd Edition, New York: W. W. Norton

Stiglitz, Joseph E. and Andrew M. Weiss (1989), Banks as Social Accountants and Screening Devices for the Allocation of Credit, in: A. Couralies and C. Goodhart (eds.), Esscys in Monetary Economics in Honor of Sir John Hicks, 1989 ed. Macmillan Publishers

Thornton Henry (1802), An Enquiry into the Nature and Effects of the Paper Credit of Great Britain

Tily, Geoff (2012), Keynes's monetary theory of interest, BIS Papers No. 65, Basel: BIS

Tobin, James (1963), Commercial banks as creators of ,money', Cowles Foundation Paper 205 reprinted in D. Carson (ed.), Banking and Monetary Studies, Irwin: Homewood

Tobin, James (1969), A general equilibirum aprpoach to moentary theory, Journal of Money, Credit and Banking, vol. 1, 15-29

Tobin, James and William C. Brainard (1963), Financial intermediaries and the effectiveness of monetary controls, American Economic Review, vol. 53, 383-400

Tooke, Thomas (1838), An Inquiry into the Currency Principle

Thornton, Henry (1802), An Enquiry into the Nature and Effects of the Paper Credit of Great Britain, London: Hatchard

Tooke, William (1838-57), History of Prices and of the State of the Circulation during the Years 1703—1856 (6 vols)

Vyas, Dushyantkumar (2011), The Timeliness of Accounting Write-Downs by U.S. Financial Institutions During the Financial Crisis of 2007—2008, Journal of Accounting Research, Volume 49, Issue 3, June 2011, Pages: 823-860

Wagenhofer, Alfred (2011), Towards a theory of accounting regulation: A discussion of the politics of disclosure regulation along the economic cycle, Journal of Accounting and Economics, 52, (2-3), November 2011, 228-234

Walsh, Carl (2003), Monetary Theory and Policy, Cambridge: MIT Press

Werner Richard A. (1992), 'Towards a quantity theory of disaggregated credit and international capital flows', Paper presented at the Royal Economic Society Annual Conference, York, April 1993 and at the 5th Annual PACAP Conference on Pacific-Asian Capital Markets in Kuala Lumpur, June 1993, reviewed by the Economist on 19 June 1993 in the 'Economics Focus'

Werner Richard A. (1997). 'Towards a New Monetary Paradigm: A Quantity Theorem of Disaggregated Credit, with Evidence from Japan', Kredit und Kapital, vol. 30, no. 2, July 1997, pp. 276-309.

Werner, Richard A. (2005), New Paradigm in Macroeconomics, Basingstoke: Palgrave Macmillan

Werner Richard A. (2003), Princes of the Yen, Japan's Central Bankers and the Transformation of the Economy, New York: M. E. Sharpe, 2003

Werner Richard A. (2010), Comment on Strengthening the Resilience of the Banking Sector, official submission to public call for comments on 'Strengthening the Resilience of the Banking Sector, Consultative Document' by the Basel Committee on Banking Supervision, September 2009 (http://www.bis.org/publ/bcbs164.htm). Submitted 16 April 2010; published by the Bank for International Settlements, Basel.

Werner Richard A. (2012). Towards a New Research Programme on 'Banking and the Economy' — Implications of the Quantity Theory of Credit for the Prevention and Resolution of Banking and Debt Crises, International Review of Financial Analysis, 25, 94-105, doi: 10.1016/j.irfa.2012.06.002

Wicksell, Knut (1898), Geldzins und Güterpreise, Jena: Fischer

Wicksell, Knut (1907), The influence of the rate of interest on prices, Economic Journal, 17, 213-20

Wicksell, Knut (1922), Vorlesungen über Nationalökonomie auf Grundlage des Marginalprin%ipes / Bd. 2: Geld und Kredit, Jena

Wicksell, Knut (1935), Lectures on Political Economy, London: Routledge

Whittlesey, C. R. (1944), Problems of our Domestic Money and Banking System, The American Economic Review, Vol. 34, No. 1, Part 2, Supplement, Papers and Proceedings of the Fifty-sixth Annual Meeting of the American Economic Association (Mar.,1944), pp. 245-259

Wicksell, Knut (1907), The influence of the rate of interest on prices, Economic Journal, 17, 213220

Winkler, Bernhard, Ad van Riet and Peter Bull (2013a) (eds.), A Flow-of-Funds Perspective on the Financial Crisis, Vol. I, Money, Credit and Sectoral Balance Sheets, Studies in Economics and Banking, Basingstoke: Palgrave Macmillan

Winkler, Bernhard, Ad van Riet and Peter Bull (2013b) (eds.), A Flow-of-Funds Perspective on the Financial Crisis, Vol. II, Macroeconomic Imbalances and Risks to Financial Stability, Studies in Economics and Banking, Basingstoke: Palgrave Macmillan

Withers, Hartley (1909), The Meaning of Money,

Withers, Hartley (1916), The Business of Finance,

Wolf, Martin (2013), The case for helicopter money, Financial Times, 12 February 2013

Wolf, Martin (2014), Strip private banks of their power to create money, Financial Times, 24 Apr 2014

Wolfe, Simon (1997), An Economic Analysis of Financial Institutions' Accounting Practice, Dissertation submitted in the Department of Management, Faculty of Social Sciences, University of Southampton

Woodford, Michael (2003), Interest and Prices, Princeton: Princeton University Yohe, William P. (1995), Interactive Money and Banking, Oxford: Blackwell

Zimmerman, Jerold L. (2001), Conjectures regarding empirical managerial accounting research, Journal of Accounting and Economics, Volume 32, Issues 1—3, December 2001, Pages 411-427

Highlights

This paper identifies and reviews the three main theories of how banks function, concerning their ability to create money.

It presents a new empirical test of the three theories

Unlike the only other existing empirical test, the test enables the researcher to control for all transactions, delivering clear-cut results.

It is found that the fractional reserve and financial intermediation theories of banking are rejected by the empirical evidence.