State of muddlement in money creation affairs

“Commercial bank money – that is, people’s bank deposits – is created through the intermediation of credit.” — Bank of England (2021)

In the history of life on Earth, each mass extinction brings with it evolutionary opportunities to fill the niches that those extinctions have emptied (they are called “evolutionary radiations”). For example, the rise of mammals was made possible by a meteorite that hit the shallow waters near the Yucatan Peninsula 66 million years ago, which drove the dinosaurs extinct. Somewhat analogously,^[i] the 2008–9 financial crisis has upset some of the prevailing economic doctrines; to name one case, serious doubts have been cast on the usefulness of Dynamic Stochastic General Equilibrium (DSGE) models (Fagiolo and Roventini, 2017).^[ii] More generally, it has revealed limits in economists’ understanding of the economy (Yellen, 2016, p. 1) and somewhat tarnished their credibility in the public mind. While it is debatable the extent to which old idols have been shattered and new “research radiations” sprung (see Aigner et al., 2018),^[iii]some effects can definitely be detected. In my recent paper I zoom in on one of these cases: the way economists conceive banks.

Before the crisis, banks were regarded as non-money-creating financial institutions, whose basic role was to lend out the funds that depositors placed with them. In less than a decade, this view (known as the “financial intermediation theory,” sometimes also known as the “loanable funds model”) has been mercilessly thrown into the dustbin of history, in what is aptly described by Bezemer (2016, p. 1276) as a “striking turnaround in the communis opinio among researchers.”^[iv]

A similar fate has befallen on the “fractional reserve theory” of banking, which postulates an intermediation role for individual banks but a money-creating role for the banking system as a whole (Werner, 2015, p. 4). The little credibility this theory commanded at the end of the last century was shattered – starting in the early 2000s in Japan and post-2008 in the US and Euro area – by an unprecedented surge in excess reserves which should not have lasted, according to the most charitable interpretation (see Fig. 1). These events procured the final blow to an already convalescent theory (Goodhart, 2017, p. 2).^[v], [vi]

In their place, a different conception that we may call the “credit creation theory”^[i] of banks is currently, according to Goodhart (2017), “taking over as the consensus approach.” Figure 2 provides a snapshot of the upward trend in the number of papers I refer to collectively as the “post-crisis literature” on money creation.^[ii]

The details of this literature are rather intriguing, and my stance on it is of ambivalence, even slight puzzlement. On the one hand, I see in it a commendable shift that brings us an inch closer to economic truths and to re-establishing the deteriorated reputation of the science. On the other, I find reasons to remain sceptical, and I certainly do not feel comfortable joining with my colleagues in their proud, assertive, jubilant tones.^[i]

Think of how the new thesis has gained acceptance: swiftly, almost without resistance, and with very little to show in the way of empirical evidence and reasoned argument. One exception is Richard Werner,^[ii] who has provided both an empirical, corroborating test for the claim that banks “create” deposits when they extend loans (see Werner, 2014a, 2015), and a plausible legal explication of their capacity to do so (see ibid., 2014b).^[iii] But my charge does not lose force against the rest of authors, who rarely quote Werner (see Fig. 3 below)^[iv] or any other evidence-supplying source,^[v] and of whom it would be fair to say have changed course like a flock of birds.^[vi]

This begs the question: how well-anchored is this position, really? Do its exponents not seem superficially convinced? Perhaps the change has been perceived as a minor ad hoc adjustment, although the fervour with which the news has been divulged militates against this explanation. Perhaps the shift was a “no brainer” because the negated position was automatically refuted or shown to be unwarranted, in which case it would speak volumes about the discipline. A more encouraging possibility is that the significance and implications of the new view have not been fully grasped yet, e.g., due to lack of time for maturation of ideas.

Though not implausible, I believe a more convincing reason why this saltational change has that incomplete and mystifying character to it, is that it has not been accompanied by any piece of conceptual innovation, the most dangerous form of destitution in the history of ideas. If this diagnosis is correct, until the new view is subsumed into a solid set of new concepts, the “pendulum of ideas”^[vii] may at any time capriciously swing back, undermining and perhaps stultifying or even rolling back the post-crisis enterprise, for it may come to be regarded as a convoluted restatement of an older and simpler truth. This possibility, though not new in the history of monetary and banking thought (see Werner, 2015), is real enough and – if one believes, as I do, that there is value in the ideas contained in this literature – disquieting.

To show that this suspicion is not unfounded, one need only look at the details of what has been written, beyond the facade of pugilist self-confidence, to find the signs of a retreat “back to the drawing board,” which could be taken to adumbrate the incipient dissolution of the “consensus.” The notion of “banks as financial intermediaries,” an apparently pliable concept and around which the “financial intermediation theory” and the “fractional reserve theory” had succeeded in building an internally consistent structure, has revealed its recalcitrant, indomitable, mischievous side, and the post-crisis authors who had taken to the sea in tight formation are now beginning to drift apart and lose their way as if misled by its sirenic allure.

Two opposing factions can already be discerned, although their members do not seem aware of their membership. At one end stand those who think of the two notions as polar opposites (14.6% of the authors).^[viii] At the other end stand a more conciliatory group who do not see an inevitable tension between the two, ^[ix] and although they direct innuendos against the concept of “intermediary,” what transpires from their statements is that they perceive a complementarity and see banks as playing both roles, referring to them as not simply^[x] intermediaries,^[xi] or as both intermediaries and money creators, ^[xii] and tacitly believe a place ought to be reserved for both notions in the new pantheon (32.3%).^[xiii] The remainder 53.1% remain unpronounced or show ambivalence. Figure 3 gives a summary.

What this cacophony of views tells us is that the attitude of festivity and ostentation of the victor displayed by many of these authors is premature, and it conceals a theory of banking that is not entirely emancipated from the incumbent, for obvious reasons. It also reveals a literature that is “hanging in conceptual air,” so to speak, for not only have no new concepts been advanced, but neither have existing ones been amended nor, as has been shown, displaced with any tolerable degree of success. Quite the contrary, the search for new vistas has come at the price of a new, unresolved inconsistency – or, at best, a new uncomforting ambiguity – between the notions of “financial intermediation” and “money creation.” It is also very telling that, with the sole exception of Mehrling (2020, p. 22), this “analytical tension” remains unrecognised in this literature. To borrow imagery from Cape Fear, the post-crisis authors had happily packed their belongings and moved to their shiny new destination, unaware that the “financial intermediary,” a vengeful, unappeased, relentless concept, travelled silently along with them, tied to the underside of the car. 

In fairness, it must be said that these authors are not to be blamed for this unfortunate state of affairs – at least not fully – for they are only the victims of the irresponsible vagueness with which those preceding them made use of the term “financial intermediary,” which over the years has endowed it with the versatility of a Swiss knife and the environmental endurance of a cockroach.^[i] But the fact remains that the pendulum of economic ideas – at least those in the domain of money and banking – has swung in a new direction, and the hands that gave it the push have yet to find a way to secure that dense, inertial sphere to a fixed station. 

How this state of muddlement will be resolved is anyone’s guess …

Borja Clavero is a Banking Consultant at Local First (“Promoting local banks”), FinTech Consultant at PSP Lab, and PhD student at De Montfort University. 

This post is adapted from his paper, “State of muddlement in money creation affairs,” available on SSRN.

The views expressed in this post are those of the author and do not represent the views of the Global Financial Markets Center or Duke Law.

---

^[i] “One reason why the dispute still remains unsettled after such a long time is that discussions had been based on assertions, implying different accounting operations of banks. But the respective merit of the three theories cannot be settled in theoretical models designed from first principles: theoretical worlds might be conceivable in which each theory is plausible. Instead, the dispute 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 2014 – no scientific empirical test had been reported in the peer reviewed journals” (Werner, 2015, p. 10).

---

^[i] My position is similar to that of Goodhart (2017, p. 3), who writes: “no fully satisfactory analytical approach is currently available.” However, he is referring to money supply-determination theories, whereas my concern here is with theories on the role of banks with regards to money, which are not exactly the same. 

^[ii] Werner himself published many papers before his empirical test of 2014. But he did not adhere to the idea that banks create money blindly; he leaned on the existing (e.g., post-Keynesian) literature which, for a long time, has advanced the claim that banks create money when they lend. Also, the results from his work gave strong (in fact confirmatory) support to the idea, so he never saw reasons to drop it; as a working hypothesis, it worked really well. The same cannot be said of the post-crisis authors, with a few exceptions like Benes, Kumhof and Laxton (2014a, b), who incorporate money-creating banks in a DSGE model which, unlike those containing the financial intermediary (loanable funds) model of banks, successfully reproduces “the basic stylised facts of both the pre-crisis and crisis phases of financial cycles” (ibid., 2014b, p. 4).

^[iii] Kumhof does adduce his experience working for Barclays as evidence. See https://www.youtube.com/watch?v=OgsEyM82oCE

^[iv] Only 15 of the 96 papers surveyed (or 15.6%) directly cite Werner’s 2014-2015 papers. If we make the same calculation excluding the papers published before 2016, the percentage rises to a whopping 27.7%. They are: Abel, Lehmann and Tapaszti (2016), Bezemer (2016), Decker and Goodhart (2018), Didenko and Buckley (2019), Goodhart (2017, 2018), Gross and Siebenbrunner (2019), Jakab and Kumhof (2014, 2019), Kumhof and Jakab (2016), Ponomarenko (2016, 2017), Rendahl and Freund (2019), Storm (2017), Unger (2016). 

^[v] Though some authors do refer back to the post-Keynesian literature on money creation.

^[vi] One wonders whether there is a connection to the anemological expression economists often use when the economy is in trouble (“headwinds”).

^[vii] This is in reference to Werner (2014a, p. 12), where he writes that it “is possible that the pendulum is about to swing away from the financial intermediation theory to one of the other two [theories]. But how can we avoid that history will merely repeat itself and the profession will spend another century locked into a debate without firm conclusion?”. 

^[viii] See, e.g., Aldasoro and Unger (2017, p. 15); Ábel, Lehmann and Tapaszti (2016, p. 38); Castellano and Dubovec (2018, p. 9); ECB (2019); Gross and Siebenbrunner (2019, p. 4, 9); Jakab and Kumhof (2015, p. ii); Van Dixhoorn (2013, p. 11); Werner (2014a, p. 2, 2015, p. 1, 9).

^[ix] See, e.g., Awrey (2017, p. 957); Bê Duc and Le Breton (2009, p. 16); Borio and Disyatat (2011, p. 8); Faure and Gersbash (2017, p. 1); The Group of Thirty (2015, p. 7, footnote 11, p. 31); ING (2018, p. 1); Mehrling (2016); Pozsar (2014, p. 11); Rule (2015, p. 6); Turner (2011, p. 5, footnote 5); Urbschat (2018, p. 10).

^[x] One could interpret the assertion that “banks are not simply intermediaries” in two ways. It could mean that banks are intermediaries as well as something else, or it could mean that banks are something other and more complex than intermediaries. I interpret the phrase in the second sense. 

^[xi] See, e.g., Benes and Kumhof (2012, p. 5; Bezemer, 2016, p. 1276); Bundesbank (2017, p. 17); Decker and Goodhart (2018, p. 30); ECB (2011, p. 68); McLeay, Radia and Thomas (2014b); Ramanauskas, Matkėnaitė and Rutkauskas (2016, p. 21); Storm (2017, p. 2, 4); Turner (2015, p. 3).

^[xii] See, e.g., Bê Duc and Le Breton (2009, p. 16); Berry et al. (2007, p. 377); Bianchi and Bigio (2017, p. 32); Bluhm, Georg and Krahnen (2016, p. 15); Borio (2012, p. 11); Borio and Disyatat (2011, p. 8); Brunnermeier and Sannikov (2016, p. 2); Carney (2018); Choulet (2015, p. 14); Didenko and Buckley (2019, p. 1071); Disyatat (2010, p. 12); Doherty, Jackman and Perry (2014, p. 4–5); Reserve Bank of New Zealand (2019, p. 1),

^[xiii] A good example of this can be found in the Bank of England’s (2021) publication ‘New forms of money’, cited at the beginning of this paper:

“Commercial bank money – that is, people’s bank deposits – is created through the intermediation of credit”

An even more blatant example is the response by the Reserve Bank of New Zealand (2019, p. 1) to fourteen questions posed to it by an anonymous person. There we read: 

Question: “Are commercial banks in New Zealand intermediaries or do they create brand new money when they are Facilitating Credit? According to Working Paper No. 529 from Bank of England by Zoltan Jakab and Michael Kumhof titled “Banks are not intermediaries of loanable funds – and why this matters”.”

Response: “Banks are both intermediaries and money creators when they facilitate credit. When a bank lends money to a customer the money that is lent will be a mix of their own capital, customer deposits, money borrowed from other financial institutions, and newly created money.” (italics added)

---

^[i] I keep to the convention established by Werner, although I find the term ‘credit creation’ misleading. Werner (2014b, p. 74) writes that “banks do not just grant credit, they create credit, and simultaneously they create money.” I find the choice of words inappropriate. All lending and borrowing creates credit (an asset on the creditor’s balance sheet, a liability on the debtor’s), but only bank lending creates money. He elaborates:

“While other non-bank firms can also grant credit, in their case it would be misleading to speak of ‘credit creation’, since their granting of a loan results in a gross increase in credit (and temporary lengthening of their balance sheet); but the discharging of their accounts payable liability arising from the loan contract results in an equal reduction in another credit balance, resulting in a reduction of the overall balance sheet and thus no change in total net credit or money in the economy. There is no money creation in the case of firms that are not banks. The bank, on the other hand, creates gross credit, just like nonbanks, but this is not counter-balanced by an equal reduction in credit balances elsewhere, leaving a net positive addition to credit and deposit – hence money – balances: net credit creation. This credit creation is visible in the permanent expansion in the bank’s balance sheet, and is executed through the operation that makes banks unique” (ibid.) (italics added)

Werner makes a distinction between “gross credit creation” and “net credit creation”. He points out, correctly, that the balance sheet of the lender expands only in the case of bank lending (a loan asset matched by a deposit liability), making banks “net credit creators”. Loan-granting by non-banks, on the other hand, changes the composition of their assets (a loan asset replaces a deposit asset) but not its size, making them “gross credit creators”. However, the balance sheet of borrowers expands in both cases, because a debt liability is created on their balance sheet when borrowing from banks and non-banks alike. If we consider the balance sheets of the lender and the borrower in conjunction (as we should), then, by Werner’s account, this would imply that “net credit” is created independently of who is the lender, and the distinction between “net” and “gross” credit creation is rendered meaningless. – On the positive side, the word ‘credit’ provides a hint of where the power to create money resides (i.e., on the liability nature of money). It also has a better ring to it.

^[ii] In the period from 2007 until today, I could find 96 papers subscribing to the new thesis that banks create money when they “lend” it, and stating so fairly unambiguously. Many of these are either official publications by central banks or they are published in reputed outlets, like the central banks’ own series of working papers. The list is given in the Appendix.

---

^[i] The comparison is not meant to suggest that both events are comparable in relevance; that would be petty and arrogant. Nevertheless, it works as an analogy. 

^[ii] Disapprovingly, a professor of mine once said of me that “I don’t believe economics can be a science”. What he seems to have meant is that only DSGE models have a legitimate claim to having a scientific status, other approaches having no such claim. I do not agree, of course – and I would go as far as to invert the proposition.

^[iii] Aigner et al. (2018) conduct a bibliometric study and of more than 440,000 economics articles before the crisis (1956–2007) and after the crisis (2008–2016). They “observe a slight shift away from the idea that financial markets are efficient by default and prices only follow random walks” (ibid., p. 23), an increased emphasis on the concept of liquidity, and “only a temporary increase of interest in classic contributions dealing with financial and economic instability” (ibid.). But overall, they find that, “unlike the Great Depression of the 1930s, the current financial crisis did not lead to any major theoretical or methodological changes in contemporary economics, although the topic of financial instability received increased attention after the crisis” (ibid., p. 1).

^[iv] See Werner (2015) for an overview of the three main theories of banking and their historical rise to and fall from prominence.

^[v] The ‘money multiplier’ continues to be a target of scorn and derision, and distrust in it is probably widely shared among the post-crisis authors. It has been variously described as slight in information content (Goodhart, 1989, p. 136), “misleading” (Disyatat, 2008, p. 14), “flawed and uninformative” (Disyatat, 2010, p. 2), “totally divorced from reality” (Feroli, 2010, p. 11), “misleading, atheoretical and […] without predictive value” (Goodhart, 2010), “mechanistic” (ECB, 2010, p. 292), “not a particularly useful framework either for understanding changes in monetary aggregates or for designing appropriate monetary policy responses” (ECB, 2011, p. 67), “not an accurate description of how money is created in reality” (McLeay, Radia and Thomas, 2014b, p. 15), an account that “significantly oversimplifies the role of commercial banks in money creation and confuses the causality between narrow and broad monetary aggregates” (Rule, 2015, p. 8), “wrong” (Werner, 2015, p. 13), “rejected” (Werner, 2014a, p. 15), an unhelpful way of thinking about money creation (Kent, 2018, p. 7), “fundamentally mistaken” (Jakab and Kumhof, 2015, p. 5), with “serious drawbacks” (Ábel, Lehmann and Tapaszti, 2016, p. 41). According to Goodhart (2017, p. 8), the “money base multiplier almost never operated in practice; now it is defunct even in theory and in principle”. Carpenter and Demiralp (2010, p. 28) provide the most devastating diagnosis: “While the institutional facts alone provide compelling support for our view, we also demonstrate empirically that the relationships implied by the money multiplier do not exist in the data […] Changes in reserves are unrelated to changes in lending, and open market operations do not have a direct impact on lending. We conclude that the textbook treatment of money in the transmission mechanism can be rejected.” The Bundesbank (2017, p. 24, footnote 35), on the other hand, though it warns that the multiplier “should not be broadly interpreted as a causal relationship between reserves and the money supply”, and that “it makes sense to look at the driving forces behind the multiplier”, it concedes that for “certain analytical purposes, the simplification involved here may be useful”. 

^[vi] Tobin (1963, p. 13), adducing the “redundant reserves of the [1930s]”, argued similarly: “the textbook description of multiple expansion of credit and deposits on a given reserve base is misleading even for a regime of reserve requirements. There is more to the determination of the volume of bank deposits than the arithmetic of reserve supplies and reserve ratios”. Tobin’s 1963 article contributed substantially to dethroning the money multiplier theory in the 1960s (Werner, 2014a, p. 9).