Access to Credit is the Key to a Win-Win Economy

Matt Klein directs our attention to an exchange between Jason Furman and Dani Rodrik that took place at the “Rethinking Macroeconomic Policy” Conference. Both argued that, while economists tend to focus on efficiency gains or “growing the pie”, most policy proposals have a small or tiny efficiency effect and a much much larger distributional effect. Matt Klein points out that in a world like this political competition for resources can get ugly fast.

I would like to propose that one of the reasons we are in this situation is that we have rolled back too much of a centuries-old legal structure that used to promote fairness — and therefore efficiency — in the financial sector.

Adam Tooze discusses 19th century macro in follow up to Klein’s post:

Right the way back to the birth of modern macroeconomics in the late 19th century, the promise of productivist national economic policy was that one could suspend debate about distribution in favor of “growing the pie”.

In Britain where this approach had its origins, access to bank credit was extremely widespread (at least for those with Y chromosomes). While the debt was typically short-term, it was also the case that typically even as one bill was paid off, another was originated. Such debt wasn’t just generally available, it was usually available at rates of 5% per annum or less. No collateral was required to access the system of bank credit, though newcomers to the system typically had to have 1 or 2 people vouch for them.

I’ve just completed a paper that argues that this kind of bank credit is essential to the efficiency of the economy. While it’s true that in the US discrimination has long prevented certain groups from having equal access to financial services — and that the consequences of this discrimination show up in current wealth statistics, it seems to me that one of the disparities that has become more exaggerated across classes over the past few decades is access to lines of credit.

The facts are a harder to establish than they should be, because as far as I can tell the collection of business lending data in the bank call reports has never carefully distinguished between loans secured by collateral other than real estate and loans that are unsecured. (Please let me know if I’m wrong and there is somewhere to find this data.) In the early years of the 20th century, the “commercial and industrial loans” category would I believe have comprised mostly unsecured loans. Today not only has the C&I category shrunk as a fraction of total bank loans, but given current bank practices it seems likely that the fraction of unsecured loans within the category has also shrunk.

This is just a long form way of stating that it appears that the availability of cheap unsecured credit to small and medium sized business has declined significantly from what it was back when early economists were arguing that we could focus on efficiency and not distribution. Today small business credit is far more collateral-dependent than it was in the past — with the exception of course of credit card debt. Charge cards, however, charge more than 19% per annum for a three-month loan which is about a 300% markup on what would have been charged to an unsecured business borrower in the 19th century. To the degree that it is collateralized credit that is easily available today, it will obviously favor the wealthy and aggravate distributional issues.

In my paper the banking system makes it possible for allocative efficiency to be achieved, because everybody has access to credit on the same terms. As I explained in an earlier post, in an economy with monetary frictions there is no good substitute for credit. For this reason it seems obvious that an economy with unequal access to short term bank credit will result in allocations that are bounded away from an efficient allocation. In short, in the models with monetary frictions that I’m used to working with equal access to credit is a prerequisite for efficiency.

If we want to return to a world where economics is win-win, we need a thorough restructuring of the financial sector, so that access to credit is much more equal than it is today.

Advertisements

Equity financed banking is inefficient

I see that Tyler Cowen and John Cochrane are having an exchange about banking. First, Cowen expresses a nuanced view of banking, then Cochrane takes the opportunity to promote his narrow (aka equity-financed) banking proposal, and Cowen questions how successful equity-financed is likely to be in practice.

With my latest paper, I have something different to contribute to the discussion: a model of how banking — and the leverage of banks — promotes efficiency. From a macro perspective the argument is really very simple: we all know from the intertemporal Euler equation that it is optimal for everyone to short a non-interest bearing safe asset. (The Friedman Rule is just an expression of this fact.) The point of my paper is that we should understand banking as the institutionalization of a naked short of the unit of account.

How is this efficiency-enhancing? A naked short position requires you to sell something that you do not have. It is a means of creating a temporary “phantom” supply of what is sold, until such time as the short position is closed out. The Euler equation tells us that a “phantom” supply that supports short positions is exactly what the economy needs to achieve intertemporal allocative efficiency.

Of course, the problem with a naked short position is that if a short squeeze (aka bank run) forces the closure of the positions too early, bankruptcy will be the result. The paper is a careful study of what is necessary to make this role of the banking system incentive feasible, and finds (alongside many other studies) that competitive banking is inherently unstable. Two means of stabilizing banking in the context of the model are (i) the natural monopoly approach: permit a non-competitive industry structure, but regulate what banks can charge; or (ii) the central bank approach: set a lower bound on the interest rate banks can charge.

So I don’t think that Cowen really captures what banks do when he presents “transforming otherwise somewhat illiquid activities into liquid deposits” as the primary liquidity function of banks. In my model banks promote allocative efficiency by creating “phantom” units of account. But I think Cowen does capture a lot of the regulatory complexity that is created by the liquidity function of banks.

Cochrane is the one, whom I really think is working from the wrong model. I’ll go through his points one by one.

1) We’re awash in government debt.

So what. Unless the government is going to start guaranteeing private sector naked short positions in government debt, it doesn’t matter how government debt we have, because it will do nothing to solve the monetary problem. We need banks because they do make possible for the private sector in aggregate to support a naked short position in the unit of account (that’s what bank deposits are) and this is necessary for intertemporal allocative efficiency.

2) Liquidity no longer requires run-prone assets. Floating value assets are now perfectly liquid

This view fundamentally misunderstands the settlement process in securities transactions. I responded to this view in a previous post and will simply quote it here:

Cochrane, because his theoretic framework is devoid of liquidity frictions, does not understand that the traditional settlement process whether for equity or for credit card purchases necessarily requires someone to hold unsecured short-term debt or in other words runnable securities. This is a simple consequence of the fact that the demand for balances cannot be netted instantaneously so that temporary imbalances must necessarily build up somewhere. The alternative is for each member to carry liquidity balances to meet gross, not net, demands. Thus, when you go to real-time gross settlement (RTGS) you increase the liquidity demands on each member of the system. RTGS in the US only functions because the Fed provides an expansive intraday liquidity line to banks (see Fed Funds p. 18). In short RTGS without abundant unsecured central bank support drains liquidity instead of providing it. (See Kaminska 2016 for liquidity problems related to collateralized central bank support.) In fact, arguably the banking system developed precisely in order to address the problem of providing unsecured credit to support netting as part of the settlement of payments.

Just as RTGS systems can inadvertently create liquidity droughts, so the system Cochrane envisions is more likely to be beset by liquidity problems, than “awash in liquidity” (p. 200) – unless of course the Fed is willing to take on significant intraday credit exposure to everybody participating in the RTGS system. (Here is an example of a liquidity frictions model that tackles these questions, Mills and Nesmith JME 2008). Overall the most important lesson to draw from Cochrane’s proposal is that we desperately need better models of banking and money, so we can do a better job of evaluating what it is that banks do.

3) Leverage of the banking system need not be leverage in the banking system.

Because the purpose of banking is to promote economic efficiency by providing society with “phantom” units of account, we need leverage in the banking system. What Cochrane calls “banking” cannot play the role of banks as I model them.

4) Inadequate funds for investment

My model of banking does not provide funds for investment — as least as a first order effect. My model of banking only provides funds for transactions. On the other hand, as a second order effect by promoting allocative efficiency, it seems likely that banks make investing more profitable than in an environment without banks. So an extension of the model that shows that banking promotes investment should not be difficult.

In short, both Tyler Cowen and John Cochrane are in desperate need of a better model relating the macroeconomy to banking. It’s right here.

 

Bank deposits as short positions: the details

So I’ve finally posted the paper I’ve been working on — a New Monetarist model of bank money — on SSRN. Warning for non-economists: lots of Greek  in this one.

Here’s the title and introduction.

The Nature of Money in a Convertible Currency World

This paper studies the nature of money in an environment where the means of payment is convertible at a fixed rate into the numeraire consumption good. By focusing on this environment we eliminate the possibility that the means of payment changes value over time, and deliberately construct a situation where the price level is disabled as a means of equilibrating the supply of money with the demand for it. To our knowledge no one else has studied such an environment in a Lagos-Wright-type framework. Our goal in this paper is to demonstrate that in this environment the first-best can still be attained – if the means of payment is effectively a naked short of the unit of account.

A naked short has the effect of creating a “phantom” supply of the shorted object that disappears when the short is closed out. We demonstrate here that banks can create this “phantom” supply of the unit of account in the form of acceptances of private debt.[1] This type of bank liability is issued when the bank stamps a private commercial bill “accepted,” and the bank obligation is put into circulation when the borrower makes purchases. Then, when the borrower pays off the loan, the phantom supply of the unit of account along with the outstanding, but contingent, bank liability that was used to create it is closed out.

Why do we model the means of payment as a naked short of the unit of account? We argue, first, that this is the best way to understand the nature of the banking system in its developmental stages. Second, by modelling the means of payment in this way our model demonstrates the efficiency gains that can be created through the introduction of a banking system. Third, by carefully evaluating the incentive feasibility conditions for our bank money equilibria, we are able to relate the monetary system to banking stability. We find that the implementation of central bank monetary policy via interest rates can be explained by the need to stabilize the banking system. Finally, we also find support for the use of usury laws as a means by which policymakers choose amongst multiple equilibria to favor the interests of non-banks over those of banks.

The monetary system modelled in this paper is based on the 18th century British monetary system as described in Henry Thornton (1802) An enquiry into the nature and effects of the paper credit of Great Britain. Privately issued bills function as a means of payment because they are “accepted” as liabilities by the banks that underwrite the monetary system. While these bills were denominated in a gold-based unit of account,[2] as a practical matter there was no expectation that they would be settled in gold. Instead, they were used as a means of transferring bank liabilities from one tradesman to another. Thus, bills that are simultaneously private IOUs and bank liabilities are used to make payment. The non-bank debtor pays off her debt by depositing someone else’s bank-certified liability into her account. (The 18th century monetary system was the precursor of the checking account system and operates just like a system of overdraft accounts.) The bank’s liability on a deposited bill is extinguished when funds are credited to the depositor’s account.

In our model productivity is stochastic, and as a result the demand for money is stochastic. We show that the bank-based money described in our model can accommodate this stochastic money demand so that a first best is attained. Thus, our model can be viewed as a model of the “banking school” view where money is issued on an “as needed” basis at the demand of non-banks.

We argue that the convertible currency environment forces a reconsideration of the nature of money. Typically the monetary literature views money as “an object that does not enter utility or production functions, and is available in fixed supply” (Kocherlakota 1998). Shifts in the price of money equilibrate the economy in these environments. Historically, however, stabilization of the price of money by tying it to a fixed quantity of gold was a foundation of economic success in the early modern period (van Dillen; Bayoumi & Eichengreen 1995). Thus, we consider how money functions in an environment where its price is “anchored”. We show that a solution is for the means of payment to be a debt instrument that is denominated in the anchored unit of account and is certified by a bank. This solution is based on actual market practice in the early modern period.

This approach allows us to reinterpret general results such as Gu, Mattesini, and Wright (2014)’s finding that when credit is easy, money is useless, and when money is essential, credit is irrelevant. While their conclusion is correct given their definitions of money and credit, we argue that this standard definition of money is not the correct definition to apply to an environment with banks. We argue that the means of payment in an environment with banks is a naked short of the unit of account, which would be categorized in GMW’s lexicon as “credit”.

This paper employs the methods of new monetarism. Our model combines an environment based on Berentsen, Camera, and Waller (2007) with an approach to banking that is more closely related to Gu, Mattesini, Monnet, and Wright (2013) and Cavalcanti and Wallace (1999a,b). Our model of banking is distinguished from GMMW because non-bank borrowing is supported not by collateral, but by an incentive constraint alone, and from Cavalcanti and Wallace because our banks don’t issue bank notes, but instead certify privately issued IOUs. We find that for values of the discount rate that accord with empirical evidence, such a payments system can be operated with no risk of default simply by setting borrowing constraints.[3] We start by finding the full range of incentive feasible equilibria of the model, and then discuss how, when there are multiple equilibria, a policymaker may choose between these equilibria.

In this environment competitive banking is incentive feasible only when enforcement is exogenous. In the case of endogenous enforcement, competition in banking typically drives the returns to banking below what is incentive feasible and the only equilibrium will be autarky. This result is consistent with many other papers that have found that the welfare of non-banks is improved when there is a franchise value to banking (Martin and Schreft 2005, Monnet and Sanches 2015, Huang 2017. See also Demsetz et al. 1996).

Thus, the challenge for a policymaker is how to regulate competition in the banking sector so that banking is both incentive compatible – and therefore stable – and also meets the policymaker’s goals in terms of serving non-banks. One solution is to treat banking as a natural monopoly, allowing an anti-competitive structure while at the same time imposing a cap on the fees that can be charged by banks. This solution explains usury laws, which by capping interest rates at a level such as 5%, the rate in 18th century Britain, is able to generate both a robust franchise value for the banks that provide payments system credit and at the same time to ensure that a significant fraction of the gains created by the existence of an efficient means of payment accrue to non-banks. An alternate solution is to impose a competitive structure on the banking industry, but also to set a minimum interest rate as a floor below which competition cannot drive the price. We argue that this is the practice of modern central banks and thus that monetary policy should be viewed as playing an important role in preventing competition from destabilizing the banking sector.

Section I introduces the model of a convertible currency. Section II describes the equilibria of the model. Section III presents the equilibria using diagrams. Section IV discusses the means by which policymakers choose between the difference equilibria of the bank-based monetary system. Section V concludes.

[1] While it would be easy to reconfigure the means of payment to be deposits or bank notes, we believe the monetary function of bank liabilities in this paper is sufficiently different from the existing literature that it useful to present it using an unfamiliar instrument.

[2] For the purposes of keeping the exposition simple, assume that we model the monetary system prior to 1797 (when gold convertibility was suspended).

[3] Indeed, we argue elsewhere that the credit based on precisely such constraints constituted the “safe assets” of the monetary system through the developmental years of banking (Sissoko 2016). Treasury bills, the modern financial world’s safe assets, were introduced in 1877 and modeled on the private money market instruments of 19th century Britain (Roberts 1995: 155).

Bank deposits as short positions

A quick point about monetary theory and banking.

Monetary economics has a basic result: nobody wants to hold non-interest bearing fiat money over time unless the price level is falling, so that the value of money is increasing over time. Many, if not most, theoretic discussions of money are premised on the assumption that fiat money is an object and that therefore one can hold no money or positive quantities of money, but one can’t hold a short position in fiat money.

Maybe this is one of macroeconomics greatest errors. Perhaps the whole point of the banking system is to allow the economy as whole to hold a short position in fiat money. After all, from the perspective of a bank what is a bank deposit if not a naked short position in cash? And by lending to businesses and consumers banks allow the rest of us to be short cash, too. This makes sense, because the basic principles of intertemporal economic efficiency state that we should all be short cash.