February 16, 2022
Carolyn Sissoko, University of the West of England
In Chapter 13 of Raising Keynes Stephen Marglin seeks to augment his 21st century General Theory with a theory of money. In many ways he is successful. He explains convincingly how Keynes’ approach inverts the causality of the quantity theory of money: instead of the quantity of money determining prices, in Marglin’s model the central bank determines nominal rates and upon this basis the nominal ‘hurdle’ rate of investment – that is the rate (or more accurately rates) determined by asset market transactions that separate investments that will be undertaken from those that will not – is established. Output, employment, inflation, and ultimately both the real rate of interest and the quantity of money needed for transactions all result from this process.
In this chapter Marglin extends his view that liquidity preference is just a theory of spreads to incorporate endogenous money. He finds that the theory of liquidity preference as a theory of spreads is still valid but with the extension that the central bank sets the short-term rate in lieu of the General Theory’s exogenous fiat money paying a yield of zero.
Marglin, however, exhibits a failing that is common amongst mainstream macroeconomists who address the question of endogenous money. The goal is typically to tweak the existing models to take into account this endogeneity – effectively assuming ex ante that endogeneity is not that important. Thus, Marglin attempts to capture the endogeneity of money by allowing the central bank to set an interest rate that then determines the quantity of money. By limiting the implications of the endogeneity of money to this single phenomenon, his discussion vastly underestimates the implications of endogenous money (see Lavoie’s contribution for more on this).
In particular, Marglin makes some very interesting – and very questionable – adjustments to the model of money underlying liquidity preference in order to make his approach work. The dynamic underlying Keynes’ liquidity preference approach is driven by the relationship between the transactions demand for money as a means of exchange and the speculative – or portfolio – demand for money as an asset and store of value. Any decrease (increase) in output or prices will result in a decrease (increase) in the transactions demand for money, increasing (decreasing) the money available for portfolio demand and decreasing (increasing) the yield on bonds and therefore the hurdle rate for investment, which ultimately has the effect of increasing (decreasing) investment. In short, the dynamics of Keynes’ liquidity preference theory are entirely dependent on the fungibility of money across transactions in goods and services (transaction demand) and in asset markets (speculative demand).
When Marglin introduces banking and endogenous money into his 21st century general theory, he makes the very strong assumption that “it can hardly be argued that the logic of endogenous money creation for transaction purposes applies to the supply of money as a store of value” i.e. money available for portfolio investment (p. 507). This is a remarkable claim (that Laidler’s contribution also challenges). The endogenous money creation function of banks requires only (i) that bank deposits are accepted as a means of payment and (ii) that bank loans are funded by the creation of deposits. While the growth of endogenous money can be limited by regulation that limits the creation of deposits, it is far from evident how endogenous money can be kept from flowing into asset markets.[1]
When Marglin assumes outright that bank loans can only fund transactions in goods and services – and therefore that bank money creation can only affect the transaction demand for money – he posits that there is one form of money that is affected by endogeneity and a distinct form of money that is not. In short, he assumes away the fungibility of money itself – which is puzzling.
In fact, we know that endogenous money applies just as much to asset markets as to those in goods and services, because bank loans frequently fund activities that go beyond the purchase of goods and services. In particular, mortgages fund the purchase of houses and repurchase agreements fund the purchase of bonds. This fact settles definitively the question of whether endogenous money affects the speculative demand for money – of course it does. Money is fungible and the flow of bank-created money into asset markets is an extremely important source of the supply of money.
The reason mainstream economists avoid introducing this aspect of endogenous money into their models is because a key implication of these facts is that asset prices depend on bank regulation and on the flows of bank money creation into asset markets. While this was obvious to Schumpeter who concludes that economic stability is dependent on ‘the way – conscientious or otherwise – in which credit is handled in prosperity,’[2] it also undermines the basic premise of economic modeling – that one can safely abstract from legal and regulatory detail. And so, like Marglin, mainstream economists typically choose to introduce an assumption into their models that ensures that this real-world phenomenon simply cannot happen.
In footnote 7, Marglin acknowledges that his assumption that ‘fractional reserve banking erects a fire wall between portfolios and transactions’ is not necessarily correct – and that the Federal Reserve Act in 1913 was in fact designed to create a formal legal structure that would establish this firewall.[3] Indeed, the generation of banking experts that drafted the Federal Reserve Act viewed the establishment of a regulatory firewall between bank money creation for the purpose of transacting in goods and services and bank money flows into markets for real estate and financial assets – or speculation – as the sine qua non of bank regulation: the real bills principle was designed to address the danger of asset price bubbles driven by such flows.[4] In short, Marglin assumes that endogenous money in the banking system can only flow into transactions in goods and services, instead of recognizing that when the banking system works in this way, it is in fact an outcome of the design of a regulatory system that deliberately favors bank finance of such transactions and deliberately discourages bank finance of the purchase of real estate and bonds. In short, in order for the banking system to operate in the way that Marglin assumes the banking system must operate, there must be regulation that strictly circumscribes mortgage and repo lending by banks.
Keynes, writing in the 1930s, may have been able to assume that contemporary regulation directed bank money creation into real activity, that the subsidy paid by depositors in the form of interest foregone could be conceptualized as returning to them as the employees of businesses that existed only because their deposits fund the banks, and that the businesses themselves participated in this system of funding banks via deposits (cf Galbraith’s contribution).
Of course, the world we live in is not a world where bank mortgage and repo lending are discouraged – on the contrary, these activities are often treated as fundamental to banking. Overall, in order for Marglin’s model to be applicable to the modern world, he needs to allow for bank loans and bank money creation to finance the purchase of bonds – since it is indisputable that this is an important aspect of modern repo markets.
Instead of presenting a realistic view of modern bank regulation, Marglin presents an idealized vision where regulation (in the form of reserve requirements) is actually binding and this limits banks’ ability to create money (even though he is apparently aware that banks have not been reserve-constrained for generations[5]). Marglin imagines that these theoretically constrained banks can then overcome their constraints by either securitizing bank loans that are then sold to portfolio investors as ‘intermediate’ risky assets (fitting easily into his version of liquidity preference as a theory of spreads) or by allowing money market funds to disintermediate banks by funding business transactions directly in the form of commercial paper. (Note that Marglin overlooks the role of off-balance sheet bank liabilities in making non-financial commercial paper eligible assets for MMFs.[6])
Marglin understands the contradiction with the latter claim: if MMFs pay interest but otherwise provide the same services as banks, then bank deposits will disappear, because everybody will prefer the higher interest-paying option – and the whole liquidity preference theory breaks down. His resolution of this conundrum is telling: Marglin effectively argues that deposits are for the unwashed masses and small firms, whereas it is the wealthy, institutional investors and mega-firms with savings (such as Apple) that will be interest sensitive and hold interest-bearing short-term debt.
To summarize, in the modern world regulation does not direct bank money creation into real activity, the subsidy paid by depositors in the form of interest foregone is not then paid back to them as the employees of businesses that exist only because of the financing that the mass of depositors is willing to provide, and the businesses themselves are not expected to participate in this system of mass finance of economic activity. Instead, we live in a world where there is a money for thee and a money for me.
The beneficiaries of the higher value money are what Zoltan Poszar calls institutional cash pools, i.e. institutional investors and the cash segments of the portfolios of wealthy individuals and wealthy firms.[7] Marglin makes the odd claim that firms like Apple, which ‘combine wealth holding with business pursuits in a single portfolio is exceptional.’[8] In fact, the developed world is suffering from an investment famine where the net portfolio of non-financial firms is no longer in debit reflecting the borrowing of funds to invest, but instead has a credit balance so that in aggregate non-financial firms are in fact lending money in bond and other financial asset markets.[9] In other words, firms in the aggregate are not borrowing for productive purposes, instead they are in fact speculating on financial markets. Again we find the mainstream economist’s habit of assuming that the world looks like his model, rather than running a robust check of his model against reality.
In short, Marglin imagines a world in which only those ‘who are strictly portfolio managers’ rely on short-term credit instruments as money.[10] In fact, we live in a world where access to higher value money is generally available to both the well-to-do (in the form of MMFs) and the wealthy, as well as all medium to large-sized businesses – and in a world where these wealthy investors have been disintermediating the banking system for almost half a century. For Marglin, this is not a problem as liquidity preference as a theory of spreads still holds – just with two classes of participants, those who are unsophisticated, use deposits, and subsidize the monetary system, and those who rely on interest-bearing forms of money.
So what does our modern financial system look like after this discussion? Instead of the world that Keynes inhabited where:
- Businesses, small and large, and individuals, wealthy or not, must all rely on the same money: bank deposits that pay little or no interest;
- Regulatory constraints force those bank deposits to fund commercial activity and business investment; and
- Regulatory firewalls prevent the flow of bank money creation into asset and real estate markets,
we live in a world where
- The wealthy and large businesses have access to a more valuable form of money, than those who are not well-off
- Bank money creation is not directed by regulation into commercial activity and business investment, and in fact there is an investment famine, as non-financial businesses are no longer net borrowers in the economy — instead they are in aggregate portfolio investors too. And finally:
- Bank money creation flows largely into real estate and asset markets,[11] where there is every reason to believe that this money creation drives up prices.[12]
In short, just as the 19th century real bills approach predicted, and as recent research confirms,[13] when bank money creation flows largely into financial asset and real estate markets it has an effect on asset prices. In this environment, it’s not surprising that even businesses find that investment in real activity isn’t worth it, because central bank supported asset investment is a much safer alternative. Finally, the whole modern system is designed to make the wealthy wealthier, because it relies on the existence of an underclass that continues to hold non-interest-bearing deposits as a subsidy to this asset-oriented financial system.
Overall, Marglin’s claim that Keynes’ liquidity preference theory still applies as a theory of spreads to the modern financial world is not convincing. A careful review of the existing data makes it clear that deposits no longer play the role in financing transactions and business activity that they did in Keynes’ time. Instead, bank money creation now funds asset price speculation – and the returns on this speculation are actively backstopped by the central banks. We live in a world where the monetary tail is wagging the real economy dog – and any discussion of the role played by money and bank money creation in the real economy needs to take these realities into account. As Keynes put it:
Speculators may do no harm as bubbles on a steady stream of enterprise. But the position is serious when enterprise becomes the bubble on a whirlpool of speculation.[14]
[1] Indeed, it is an open question whether modern regulators are typically successful in implementing constraints on the creation of deposits and their equivalents. BACK TO POST
[2] Joseph Schumpeter, Business Cycles: A Theoretical, Historical and Statistical Analysis of the Capitalist Process (New York: McGraw Hill Book Company, 1939), 156. BACK TO POST
[3] Stephen A. Marglin, Raising Keynes: A Twenty-First Century General Theory (Cambridge: Harvard University Press, 2020), 515. BACK TO POST
[4] The real bills principle is closely related to the real bills doctrine, but allows for more activity in the ‘grey’ areas between real and financial transactions. Carolyn Sissoko, “How to Stabilize the Banking System: Lessons from the Pre-1914 London Money Market,” Financial History Review 23, no. 1 (2016): 1-20. BACK TO POST
[5] Marglin, Raising Keynes, 515. BACK TO POST
[6] Carolyn Sissoko, “Is Financial Regulation Structurally Biased to Favor Deregulation?” Southern California Law Review 86 (2013): 365-420. BACK TO POST
[7] Zoltan Pozsar, “A macro view of shadow banking: Levered betas and wholesale funding in the context of secular stagnation,”INET Working Paper (2015). BACK TO POST
[8] Marglin, Raising Keynes, 513. BACK TO POST
[9] Carolyn Sissoko, “Repurchase Agreements and the (De)construction of Financial Markets,” Economy and Society48(3) (2019): 315-41; Joseph Gruber & Steven Kamin, “The Corporate Saving Glut in the Aftermath of the Global Financial Crisis,” International Finance Discussion Papers 1150 (2015). Jan Loeys, David Mackie, Paul Meggyesi, & Nikolaos Panigirtzoglou, “Corporates are driving the global saving glut,” JP Morgan Research (June 2005); Richard Portes, “Comment on ‘A Global Perspective on External Positions,’” in G7 Current Account Imbalances, ed. Richard Clarida (Chicago: University of Chicago Press, 2007); Martin Wolf, “Why the future looks sluggish,” Financial Times(Nov. 19, 2013), available at http://www.ft.com/intl/cms/s/0/a2422ba6-5073-11e3-befe-00144feabdc0.html. BACK TO POST
[10] Marglin, Raising Keynes, 513. BACK TO POST
[11] Oscar Jorda, Moritz Schularick & Alan M. Taylor, “The great mortgaging: housing finance, crises and business cycles,” Economic Policy 31 (Jan. 2016): 107-152. BACK TO POST
[12] Markus M. Brunnermeier & Lasse Heje Pedersen, “Market liquidity and funding liquidity,” Review of Financial Studies 22(6) (June 2009): 2201-2238. BACK TO POST
[13] Ibid; Tobias Adrian & Hyun Song Shin, “Liquidity and Leverage,” Journal of Financial Intermediation 19(3) (2010): 418 – 437. BACK TO POST
[14] John Maynard Keynes, The General Theory of Employment, Interest, and Money (London: Macmillan, 1936), chapter 12. BACK TO POST