Is medicine as flawed as finance?

Events that took place this past holiday season have set me to thinking not just about the awful nexus that takes place when illness, addictive drugs, and the American medical system meet, but also about the nature of observation-based (as opposed to controlled-study-based) science and the relationship between the practice of this science and the giant corporations that have an interest in this practice. In short, I’ve been thinking about how failures in the world of medicine look very similar to failures in the world of finance.

What happens in an environment where data is important, but its interpretation is necessarily imprecise, and there are corporations whose goal is to profit off of any structural weaknesses in the methods used to interpret the data? The combination of weak antitrust enforcement that has placed immense power in the hands of a very small number of corporations and a corporate focus on shareholder value rather than stakeholder value means that there simply aren’t that many influential corporations left whose core business strategy is to serve those who buy their products to the best of the corporation’s ability.

In finance this means that clients are often treated as “the mark”, and client losses are justified by those who generate them on the Darwinian principle that good things will happen when dumb or uneducated people lose money. Financiers know that the nature of the data ensures that they can almost always come up for some kind of an explanation for why the product they use to garner some “dumb money” is in some way beneficial and should not be banned. (e.g. “in an efficient market, only people who need product X will buy product X, so we don’t need to worry about the losses of the “dumb money,” which exists to make the market more efficient.”) The tools of the academics are used, not for the purpose for which they were invented, but to make the world a worse place to live in.

Unfortunately I’m beginning to suspect that our drug companies function on the same principles as the financial industry. It seems to me that doctors have been trained not to listen too closely to patient complaints about side effects. Now there are probably good reasons for this: if the doctor is conservative about prescribing medicine so that you really need the medicine when you get it, then the side effects will need to be quite severe in order for them to outweigh the need for the medication. And it is true that doctors almost certainly receive many complaints about perceived side effects that are in fact due to other causes. In short, doctors have a very hard job.

It seems to me that pharmaceutical have turned the challenge of medicine into a profit opportunity through two mechanisms. First, they work aggressively to get doctors to prescribe their medications for minor ailments that could be addressed through over-the-counter or non-pharmaceutical means. When the pharmaceutical companies are successful, doctors end up prescribing drugs that are net “bads” for their patients, and frequently choose to address side effects not by taking the patient off the medication, but by prescribing another medication to address the side effect. A patient with a minor complaint can end up on a cocktail of drugs that causes far more damage to the patient’s health than the minor complaint itself. Who has not heard a doctor state when the patient questions whether her growing health problems are not in fact being caused by the cocktail of medication that “It’s not cause and effect,” pooh-poohing the patient’s concerns? While there are certainly very good doctors out there (and I recommend that you seek them out), the medical profession has done far too little to offset the nefarious influence of drug company incentives.

Secondly, it appears that drug companies have learned that addictive drugs are some of the most profitable. In my view this is likely to be due to the fact that these drugs often have the side effect of causing the malady they are prescribed to cure. That is, once you have become addicted to the drug, trying to get of the med will often cause you to experience the illness that you took it to address — but even worse than before you took it. It’s not unusual for patients to get into a pattern where the doctor keeps prescribing higher and higher doses of these medications and that the patient ends up facing very strong disincentives to go off the medicine. A profit-maximizing pharmaceutical company will likely prefer to develop this type of medicine than a medicine that can treat the ailment, but that is non-addictive. That is, the profit motive is very much adverse to what is in patients’ best interests. When you add to this dynamic the tendency of many doctors to pooh-pooh patient concerns about side-effects and in particular concerns that the medication may be worsening the condition (“It’s not cause and effect. Your symptoms are probably just the progression of your ailment.”), it hardly surprising that the way these medications are being used is often toxic.

Overall, when I hear complaints about how too much of the public doesn’t believe in science anymore, I can’t help wondering: Well, what is their experience of how science is applied in the modern world?

Evaluating heterodox monetary theory through a New Monetarist lens

I. Using New Monetarism to motivate a closer look at credit-based payments systems
a. The failure of economic theory to incorporate banking is as old as economic theory itself
II. Important implications of a credit-based payments model
a. Settlement by clearing is a substitute for settlement by cash
b. The credit-based payments model undermines the distinction between consumption and production credit
III. A final note

Money and the issues that the economy’s dependence on money raises are some of the most difficult topics to analyze in economics. All discussions of money are founded on some underlying framework explaining what money is and what money does. My preferred framework is based on “New Monetarist” models of the economy. A variety of competing approaches are typically described as “heterodox” in the economics literature. This post is a preliminary effort at organizing a comparison of these frameworks and was stimulated by Chapter 5 of Christine Desan’s Making Money which is a highly recommended and fascinating study of the historical development of money. Desan builds her argument based on the heterodox approaches.

My organizing structure will build on a verbal discussion of the implications of new monetarism. Three related, but distinct, heterodox approaches will be addressed in the context of this structure: the hierarchy of money developed by Stephanie Bell-Kelton (explained with extraordinary clarity in Perry Mehrling 2012), the cartalist approach to money, and Geoffrey Ingham’s focus on money as a unit of account and store of value as distinct from money as a means of payment. (I am aware that both Ingham and Kelton may fall into the cartalist category, but find it convenient to discuss each of these approaches independently.)

Using New Monetarism to motivate a closer look at credit-based payments systems

New monetarism is a sub-field of macroeconomics that has grown up over the past few decades and that uses formal economic modelling techniques to study money. While the principal innovation that this literature brings to economic theory comes from liquidity frictions that generate a role for a means of payment, as in the older monetary literature, money also serves as a store of value and via the study of relative and intertemporal prices as a unit of account.

The basic intuition that can be derived from the formal analysis of a means of payment is that at any point in time a given transaction in the economy may be viewed as taking one of the following forms:

  • Barter (requires double coincidence of wants)
  • Cash purchase-Cash sale
  • Credit purchase-Credit sale
  • No trade (takes place if there is either no coincidence, or the parties are unable to settle on terms)

Because economic agents are on both sides of any transaction, every cash purchase and every credit purchase must have a corresponding cash sale or credit sale. Because both credit and cash transactions can only take place if the seller anticipates that cash or credit will have value in the future, only barter transactions do not depend on future expectations about the value of some means of payment.

Some basic results that derive from this literature are:

i.  From a theoretic perspective cash payments and credit payments are alternative payments systems (Gu, Mattesini, & Wright 2016). Because conceptually a payments system can be based entirely on one or on the other, credit – or in the extreme case gift-giving – can be the basis for the payments system.

This result may be consistent with Ingham (2004)’s argument that the payments system is dependent on the sovereign for its unit of account, since the new monetarist literature can be viewed as assuming some institutional infrastructure arguably including coordination on a unit of account. On the other hand, this result contrasts with the basic premise of the hierarchy of money theory which states that bank money – or the modern credit-based payments system – is dependent on sovereign-authorized cash for settlement.

This theoretic result is a valuable insight, because it is far from clear that the banking system – abstracting from regulatory requirements that did not exist when the system was developed in the 19th century – is dependent on sovereign “cash” for settlement. Whatever remaining balances aren’t cleared by the banking system may be carried as debt by a member of the system to the next period, and it is unclear that sovereign “cash” is necessary to the system – except perhaps as a unit of account. (This is discussed in more detail below.)

 ii. Cash and credit transactions are similar because they can only take place if the seller anticipates that cash or credit will have value in the future.

This can be viewed as a formal realization of Ingham’s contention that all forms of money can be properly viewed as debt in the sense that the holder of money, whether cash-based or credit-based, has a claim with a future value determined by social relations (2004, Ch. 4). This is also related to the cartalist view (embraced by Ingham but also espoused by many others before him) that what is money is fundamentally determined by what the state accepts in payment of taxes. On the one hand, the theoretic result is a general finding that both cash and credit depend on future expectations, whereas the cartalist view is more of an empirical claim that the most important real-world source of those expectations comes from the state’s power to tax. The theoretic literature then raises two issues with respect to the cartalist view: first, are there any other means by which money enters into circulation in addition to sovereign tax policy; and, second, what makes it incentive compatible for a public or private issuer of money to accept/redeem it.

iii.  An ideal form of credit is strictly better than cash: the set of equilibria that can be attained using cash is a subset of the set of equilibria that can be attained using credit (cf. Kocherlakota 1998. See also Sissoko 2007).

Not just the heterodox hierarchy of money approach, but also most mainstream macroeconomics — and indeed a great deal of the New Monetarist literature  –emphasizes cash as the cornerstone of the payments system. The implication that I draw from this New Monetarist result is that this emphasis is misguided. We need to re-center the discussion so that the focus is on credit- or bank-based payments systems and then evaluate the role that cash may play in supporting these systems.

The failure of economic theory to incorporate banking is as old as economic theory itself

Treating the “cash” solution to the monetary problem as the reference point and then asking how credit expands the cash system has been the default approach in economic analysis dating back to Adam Smith. Indeed, from this point of view the heterodox hierarchy of money approach is similar to the monetarist treatment of banking: banking serves only to expand an existing monetary base. (If these arguments were limited to the claim that banking is dependent on an external unit of account, they might be sustainable, but these arguments always go well beyond such modest claims.)

Banking scholars have been attempting to debunk the economists’ approach to banking for more than 200 years. Henry Thornton, the banker who first explained the role of the lender of last resort and how it affected domestic and foreign trade differently, took Adam Smith to task for failing to understand the nature of Britain’s paper monetary system in 1802. Schumpeter was engaged in much the same debate with his contemporaries a hundred years later. And today we still find that economists have difficulty with the very concept of a credit-based payments system. One might hope that the formal modeling framework of New Monetarism will help the profession remove these intellectual blinders.

Important Implications of a Credit-Based Payments Model

Settlement by clearing is a substitute for settlement by cash

Hierarchy of money theorists tend to ask “in what will this transaction be settled” and to categorize credit based on the purported means of settlement.[1] This approach accurately models the relationship to bank money of near-monies that are settled in bank money, and thus it is a very useful approach. However, when this approach is extended by analogy to the relationship between bank money and cash (or claims on the state), the approach breaks down and it loses its usefulness. (Indeed, Knapp whose cartalist approach to money underlies the hierarchy of money theory did not draw this distinction.[2]) Not only is a cancelled check – or the transfer of bank money – legally proof of settlement of tax payments as well as civil obligations, but even in a financial crisis it is central bank liabilities that expand to support the system, not “cash.” While hierarchy of money theorists seek to reconcile this fact with the theory by explaining that these central bank liabilities are like “cash” in some sense sovereign obligations, it is more accurate to view these liabilities as a core part of the system of bank money itself.

The modern credit-based monetary system is indeed anchored by a central bank that issues bank money in the form of reserves that can expand to support the system in crisis. It is also true that the central bank could not operate in this way without the sanction of the government. These facts do not, however, mean that the best way to understand this system of bank money is that it is dependent on the stock of government liabilities. The system of bank money was made effective only when it became dependent on an expandable source of central bank liabilities, and the concept of a central bank developed simultaneous with this system of bank money. Thus, bank money and the central bank liabilities that serve as the high-powered money of the banking system are part of a single bank money system such that the two parts of it cannot be disarticulated. It is misleading both to treat this relationship as analogous to the relationship between near-monies and bank money and to treat central bank liabilities as effectively sovereign liabilities.[3] Such an approach collapses the institutional detail of what is actually going on to such a degree that it risks misleading rather than clarifying the role played by money in the economy.

The processes of clearing and of interbank lending that are integral to every banking system ensure that only a small fraction of bank liabilities need to be settled at all and that many of these remaining balances are carried rather than settled. Thus, it is inaccurate to claim that when one pays by check (or otherwise by transfer of bank money) the transaction is settled by central bank reserves. The transaction is settled within the banking system itself. Furthermore, it is well established that central banks do not exercise quantity control over the stock of central bank reserves, but must issue reserves in a manner that is responsive to the demands of the banking system.[4] This is true both in normal times and in crises. Thus, the hierarchy of money framework when it claims that bank money is settled using central bank reserves creates an illusion of quantity-based central bank control over the money supply that is both erroneous and just one small step removed from the deterministic approach of the “old” monetarists.

For example, Perry Mehrling, whose theory of money relies heavily on the hierarchy, states: “At every level of the system, the availability of money from the level above serves as a disciplinary constraint that prevents expansion; credit is payable in money, but money is scarce.” Given that the specific hierarchy Mehrling is discussing explicitly places currency as a central bank liability above bank deposits, his implication is clearly that the quantity of central bank liabilities constrains the growth of bank deposits. My point is that this view misunderstands the basic function of the central bank: in order for a banking system to work central bank liabilities must expand in response to the demands of the banking system. It is true that the central bank constrains the growth of bank money, but it most definitely does not do so using quantitative controls on the supply of central bank liabilities. Instead the central bank uses interest rate policy, microprudential, and macroprudential supervision of the instruments on bank balance sheets (see Sissoko 2016).

Thus, by failing to understand the fundamentally credit-based nature of the modern payments system and by modeling the payments system as “inherently” cash-based with the banking system as simply an extension of the cash-based system, the heterodox literature perpetuates the error of the old monetarists. As Thornton argued in 1802 and many have tried to explain after him, the modern monetary system is a credit-based payments system that has properties very different from a cash payments system. Errors that can arise when scholars who are studying money fail to recognize either the possibility of or the characteristics of a credit-based payments system include:

  • The assumption that movements in a sovereign money supply determine movements in a contemporaneous credit-based money supply. This is of course a foundational principle of Old Monetarism. It is, however, also a fundamental principle of many heterodox approaches (see the discussion of Mehrling 2012 above). Desan too posits such a relationship and seeks support for it in the data on medieval English money. Unfortunately the data available from this era is sufficiently sparse that others find support for the view that credit actually expanded when the sovereign money supply shrank (pp. 225 ff).
  • The assumption that a deficiency of cash will constrain transactions. In an environment where there is clear evidence of credit-based payments, the possibility that credit fully substitutes for cash – at least within the group of individuals who participate in the system – must be considered. Even if cash is used sporadically (for example, seasonally during the harvest when it is suddenly available in a village economy), this does not necessarily imply that transactions are constrained by the absence of cash when it is unavailable.
  • The assumption that debts are paid in cash. In the absence of clear data indicating that this is the norm, the likelihood that some system of clearing or of transfer of claims is used to settle debts must be carefully evaluated. This possibility certainly cannot be rejected out of hand – and evidence that some currency is the unit of account is in no way evidence that this currency is also the means of payment. After all, almost every one of my transactions is denominated in dollars, but only a very small fraction of them are settled using dollar bills.
  • The assumption that debt that specifies a due date is actually expected to be paid off on that date. The rolling over of short-term debt obligations is such a fundamental part of historical payments systems that one must always evaluate whether or not a social norm was in place that would make it difficult to demand payment when due. For example, when studying credit in a 15th century English village, Elaine Clark establishes that 61% of the litigation over payments that arose involved debt that was incurred 4 or more years earlier and in more than half of these cases the debt was more than 7 years old. Thus, Clark concludes that in this small scale credit network the norm was for debt to be open-ended. Given examples such as this, one needs to consider the possibility that the due date for a debt is simply the date on which the terms of the credit line may be reset rather than assuming that it is a date on which payment is expected to be delivered.

The credit-based payments model undermines the distinction between consumption and production credit

The New Monetarist credit-based payments model points to an important fact about credit that is easily missed when the neoclassical model is your framework: the line that is sometimes drawn between consumption credit and productive credit begins to disappear in an environment where everybody needs to borrow in order to trade.

Desan has a very nice discussion of the implications of the ubiquitous use of what she calls “liquidity credit” (pp. 214 ff), and I prefer to call payments credit. She discusses the fact that payments credit does not sit easily within a dichotomy of (potentially exploitative) “exigency credit” and “investment credit”, but is best understood from the “revisionist” perspective that assigns a more affirmative role to credit.

I would add that payments credit lies squarely between the two poles of the dichotomy: it typically finances working capital – after all foodstuffs themselves can be considered inputs for someone whose main product is labor services and who needs payments credit in order to purchase food. While working capital has always been distinguished from fixed capital and it is the latter that is typically associated with investment, the productive role of working capital is indisputable.

Thus the credit-based payments model indicates that just as mercantile credit is viewed as being productive – precisely because it finances working capital (Desan pp. 228-29), the same reasoning justifies the view that payments credit in a rural village was also likely to have been productive.

A credit-based payments system may not be worse than a cash-based credit system

Desan discusses at length the ways in which the rural system of payments credit may have imposed costs on the poor. The main criticism I have of her analysis is that she does not spend enough time explaining how such costs might have been avoidable given a cash-based payments system. After all, the question is not whether the poor suffered in rural villages (given the heavy burden of taxation that Desan argues was a cornerstone of the monetary system, we know they did). The question is whether the fact that the payments system was credit-based made things worse.

For example, Desan argues that sellers on credit and those paying in advance could drive hard bargains with the poor (p. 223). But, as the New Monetarist framework makes clear, the presence of cash does not eliminate the bargaining problem. The poor are disadvantaged in a cash or a credit economy: if a buyer can only afford to pay cash for a quarter loaf of bread, the seller may well choose to charge that individual more than a quarter the price of a full loaf. In fact, in a world where everybody uses payments credit it may be easier to be poor, precisely because the need to borrow is unexceptional and can go unnoticed.

Desan also argues that enforcement of debt by litigation is costly, unpleasant, and can have the effect of souring relationships (pp. 218 ff). This is certainly true, but the relevant question is whether cash bargaining makes it possible for a community to avoid or at least reduce the measure of unpleasantness and soured relationships. The answer to this question is far from obvious. (While I pay my phone bills in full every month, I can assure you that the fact that I regularly have to call and have them corrected makes me feel that I know very well what a sour commercial relationship is – and it doesn’t require debt.) In short, disputes arise when people transact. When payments mostly take the form of credit, then debt litigation will frequently feature in the disputes. This is not, however, evidence that no dispute would have taken place without debt.

Desan remarks on how the credit-based payments system required vetting and long-term relationships that would be unnecessary in a cash economy (p. 218). This is certainly true, but cash is usually viewed as being most advantageous in stranger-trades where credit is impossible. Given the signs of almost universal access to credit in rural villages (Spufford, cited by Desan 209) – and the fact that these were the kind of close knit environments where credit is likely to be most effective – it’s not at all clear that these characteristics of credit were a handicap in rural villages. On the other hand, it seems reasonable to claim that the lack of cash and the reliance on a credit-based payments system may have slowed the growth of stranger trade and in this sense slowed the expansion of the economy.

Finally, Desan argues that the nonproductive nature of English credit exacerbated its dangers (p. 223), including “habituat[ing] people to living on prospect” (p. 224). As was discussed above it’s far from clear that this is the best way to view payments credit, which is indeed forward looking, but in a way that makes it possible for economic production to take place.

A final note

One of the advantages of the new monetarist approach is that it rejects the neoclassical model entirely in order to explore a variety of environments where trade is not easy. As a result, there are aspects of the heterodox monetary literature that are more reliant on the neoclassical framework than new monetarism. For example, Geoffrey Ingham appears to treat the idealized market of the neoclassical model as the basis for his general definition of a market. He writes: “a market is a system of multilateral exchanges in which bids and offers, priced in a money of account, can in principle produce a single price for a uniform good” (Ingham 2006). I find it surprising and unnecessarily restrictive to treat the idea of a market itself as referring to an environment where each good trades at a single price. Having spent time in parts of the world where the term “market” typically refers to an environment where many goods are sold and prices are reached through a process of one-on-one bargaining, the single price of the neoclassical model has always appeared to me a characteristic of the model, not of the “market.”

Precisely because discussions of money are inherently macroeconomic, requiring focus on the general operation of trade in the economy as a whole, close attention to the subtle ways in which simplifications drawn from the neoclassical model color the discourse is very important. For example, reading Desan I find that she periodically implies that if cash were in sufficient supply, prices would be neoclassical in character.[5] While this accords with Ingham’s claim that “a genuine market presupposes the existence of a money of account in which demand and supply can be expressed in prices,” it is far from obvious to me that Ingham’s market – or neoclassical prices – can be expected to exist anywhere but in the realm of purely abstract theory.[6] By introducing the reference point of idealized prices into the analysis of historical monetary phenomena and assuming that such idealized prices are specifically associated with the coin that served as a unit of account, Desan allows the neoclassical model to color her discussion of money and credit. In short, it seems to me that proponents of heterodoxy can sometimes end up taking the neoclassical pricing mechanism more seriously than many modern economists – who have after all been trained in all the shortcomings of the neoclassical model.

[1] See, e.g., Gabor & Vestergaard 2016.

[2] See Bell 2001 at 159.

[3] Bell writes: “Although bank money is part of the ‘decisive’ money of the system, its acceptance at state pay offices really requires its conversion to state money (i.e. bank reserves). That is, bank money is converted to bank reserves so that (ultimately) the state actually accepts only its own liabilities in payment to itself.” (p. 160). First, most bank money is cleared, and only a tiny residual shows up in reserves, so there is no “require[ment of] conversion to state money.” Second, central bank reserves are not state liabilities, but central bank liabilities. The fact that the archetype of a central bank operated as a privately-owned institution for most of its history should make one think twice about casually asserting an equivalence between the central bank and the state.

[4] See Stigum & Crescenzi on the Fed’s so-called monetarist experiment, pp. 372 ff, 503. See also McLeavey et al. 2014 p. 21.

[5] “The process [of monetizing an economy with coin] shakes the items traded into particular relationships of value. Eventually, those relationships produce prices for goods and other resources in terms of pennies, the units of account. The pennies, as they are spent, traded, and taxed, ultimately create a set of equivalences: 5 pennies = a sword, 5 pennies = two cows. But there is nothing essential about the sword and the cows and their equivalence. To the contrary, they are interchangeable only in the world created by this community’s activity with [pennies]” (pp. 60-61). And later, the scarcity of small coin “effectively added a burden to those most handicapped by it, whether by increasing their vulnerability to [adversely-timed lawsuits] or affecting bargaining power over prices because they had credit not coin” (p. 223). Observe that the latter statements also implies that if the supply of coin is sufficient, sellers will be price-takers (which in economic theory is a prerequisite for neoclassical prices).

[6] Indeed, in the neoclassical model money is superfluous. So it is far from clear how neoclassical prices could possibly result from the introduction of money into an economy.

In search of financial stability II: Re-thinking money and banking

I.   Acceptance banking
II.  A simple model of money based on acceptance banking
III. What is money? An origin narrative
IV. So what do banks do?
V.  Conclusion: Banking as the fundamental source of liquidity

The challenge for any model of money and banking is to explain the two basic elements of the modern payments system.

  • First, the money supply is comprised mostly of private sector liabilities
  • Second, this money supply is backed on the asset side of bank balance sheets by a combination of (i) “cash” (e.g. central bank liabilities), (ii) private sector short-term debt, and (iii) private sector long-term debt (e.g. mortgages)

Why this is the structure of the payments system? I will simplify the question a little by observing that the backing of the money supply by long-term debt is a very recent phenomenon (see Jorda, Schularick, and Taylor 2014), and therefore by restricting my focus to the explanation of why the modern monetary system developed backed by private sector short-term debt. (Note that this post is designed to motivate in layman’s language a formal economic model of banking that is available here.)

There two common frameworks used to discuss the two sides of bank balance sheets in the bullet points above, a standard view and a heterodox view. The standard view is the “loanable funds” approach that assumes that cash is brought to the bank by depositors and then the bank takes some of that cash and lends it out. In this framework, no loan can be made unless a depositor first brings cash to the bank. (Classic economic models of banking such as Diamond-Dybvig rely on this framework.) The heterodox view claims that it is by making loans that a bank creates deposits. In this framework, a bank first underwrites a loan, and after the loan is approved the bank funds the loan by giving the borrower a deposit account with the value of the loan in it.

Charles Goodhart (2016) points out that both of these two frameworks are missing something very important about the relationship between the asset and the liability side of the bank balance sheet: the standard view implies that it is the depositor that drives the process, the heterodox view implies that it is the bank that drives the process, and both of these are wrong.[1] Goodhart explains that it is more accurate to think of banks as setting the parameters by which loans will be made – and in fact of typically offering borrowers credit lines on pre-specified terms – and then allowing the borrowers to determine whether or not to take out the loans that will cause the money supply to expand. In Goodhart’s framework banks are simply the intermediaries that allow the private sector to expand the money supply on an “as needed” basis. I will call this the “private money” model of banking.

Since the “loanable funds” approach models the “deposit taking” function of banks and is closely tied to the “goldsmith” story of bank origins, I think it is useful to connect Goodhart’s “private money” model with a specific banking activity, acceptance banking, and to present a corresponding origin narrative.

Acceptance banking

Banking in 19th century Britain largely took the form of acceptance banking. Whereas a bank that receives a deposit opens an account for a client, a bank that approves acceptance credit for a client opens for a client a discount – or an account that may go negative to the extent of the client’s credit line. The terms of the discount are set in advance, and the client draws down the credit line on an “as needed” basis. A bank discount is identical to a bank account in terms of the ability to deposit and to and withdraw funds. The only distinction between the two is that the discount is designed to carry a negative balance for an indefinite period of time.[2]

When a bank client draws a discount down from zero, the action simultaneously creates a bank asset and a bank liability. First, the draw automatically creates a bank loan on terms pre-specified at the time the discount was approved as noted above. But, secondly, because the draw is used to make payments using bank liabilities (that is, using either bank notes or bank acceptances), bank liabilities are also increased by the amount of the draw. Thus, acceptance banking – or for that matter any form of banking based on credit lines – doesn’t just have loans causing bank liabilities to be created, but also has the private sector driving the process by which bank liabilities are created.

A simple model of money based on acceptance banking

Before discussing an origin narrative that corresponds to the “private money” view of banking, let me lay out in abstract terms how we should think about this function of banking. The crucial point of this discussion is that in Goodhart’s “private money” model the money supply is expandable to meet the needs of the private sector, subject to the terms set out by the banks. Tying this view into the credit facilities with which we are familiar in the US, one may think of credit in the private money model as being as readily available as it is to businesses today through credit cards, but – because of careful underwriting and therefore the safety of the debt – bearing a low interest rate, such as 5% per annum.

The simplest model of money is a game with three people, each of whom produces something another seeks to consume: person 2 produces for person 1, person 3 produces for person 2, person 1 produces for person 3. Trade takes place over the course of three sequential pairwise matches: (1,2), (2,3), (3,1). Thus, in each match there is never a double coincidence of wants, but always a single coincidence of wants. We abstract from price by assuming that our three market participants can coordinate on an equilibrium price vector (cf. the Walrasian auctioneer). Thus, all these agents need is liquidity.

Let the liquidity be supplied by bank credit lines that are sufficiently large and are both drawn down by our participants on an “as needed” basis, and repaid at the earliest possible moment. Assume that these credit lines – like credit card balances that are promptly repaid – bear no interest. Then we observe, first, that after three periods trade has taken place and every participant’s bank balance is zero; and, second, that if the game is repeated foerever, the aggregate money supply is zero at the end of every three periods.

In this model the money supply expands only to meet the needs the trade, and automatically contracts in every third round because the buyer holds bank liabilities sufficient to meet his demand.

Consider the alternative of using a fiat money “token” to solve the infinitely repeated version of the game. Observe that in order for the allocation to be efficient, if there is only one token to allocate, we must know ex ante who to give that token to. If we give it to person 3, no trade will take place in the first two rounds, and if we give it to person 2 no trade will take place in the first round. While this might seem a minor loss, consider the possibility that people who don’t consume in the first stage of their life may have their productivity impaired for the rest of time. This indicates that the use of fiat money may require particularized knowledge about the nature of the economy that is not necessary if we solve the problem using credit lines.

Why don’t we just allocate one token to everybody so that we can be sure that the right person isn’t cash constrained in early life? This creates another problem. Person 2 and person 3 will both have 2 units of cash whenever they are making their purchases, but in order to reach the equilibrium allocation we need them to choose to spend only one unit of this cash in each period. In short, this solution would require people to hold onto money for eternity without ever intending to spend it. That clearly doesn’t make sense.

This simple discussion explains that there is a fundamental problem with fiat money that ensures that an incentive compatible credit system is never worse and in many environments is strictly better than fiat money. This is one of the most robust results to come out of the formal study of economic environments with liquidity frictions (see e.g. Kocherlakota 1998).

Now let’s continue our discussion of the payments function of credit lines by taking our simple model (the original one without fiat money), duplicating it twice (using ‘ and ‘’ to indicate the duplicates), assuming that preferences are such that participants do not wish to trade across duplicate groups, and offsetting the trading periods for our duplicate economies. In period 1 three pairwise matches take place: (1,2), (2’,3’) and (3’’, 1’’). Posit also that at the start of time there is a banking system that has loans outstanding to agents 1’ and 1’’, and deposits owed to agents 2’ and 3’’. (This is just the simplest way of creating a more complex, overlapping pattern of trade.)

Thus, we have an environment where there is always a stock of deposits and a stock of loans outstanding. Even so, every agent is regularly paying off his debt. The money supply still exists only to meet the needs of trade, and every participant’s account balance returns every third period to zero.

Now imagine that for every one of the participants in our triplicate economy there are n identical agents who have been excluded from the economy historically. If these agents are suddenly incorporated into the economy, then the money supply will increase by a factor of n. Because this increase in the money supply occurs only to meet the needs of trade, the increase in the money supply is entirely consistent with the existing price vector.

In short, because the debt created by the banking system is carefully constructed so that its only purpose is to provide liquidity to facilitate the operation of the payments system, the bank-based money supply is able to expand to meet the needs of trade, and will – in certain circumstances – expand without any tendency to affect the price level.[3]

Observe that this framework is the basis for the “real bills doctrine.” If the only debt in the economy finances the purchase of productive inputs, and if the banking system can enforce the requirement that this debt be paid off as soon as production takes place, then expansions of a money supply backed by this debt are not necessarily inflationary, but may reflect changes in the underlying real economy. (Note that, because we have assumed an equilibrium price vector, the question of how prices are anchored in this framework remains to be answered and is not addressed here.)

What is money? An origin narrative

We have laid conceptual underpinnings that explain: first, the relationship between the use of bank liabilities as money and the fact that these liabilities are backed by short-term private sector debt; and second, the fact that a system of “private money” has the advantage that it can very naturally expand to meet the needs of trade. We now demonstrate that there are also historical foundations for the model of money presented here. Before expanding upon the historical details, we discuss in more general terms the implications of a private monetary system that is not anchored by any sovereign unit of account.

One of the great inventions at the dawn of the early modern era in Europe was that of monetary systems that existed in the abstract without any physical embodiment of the unit of account. Specifically, by the 1530s the process of clearing and settling European trade was taking place using the ecu de marc, a unit of account that was stable precisely because it was not tied to the coin issued by a sovereign.[4]

When such an abstract unit of account is combined with a sophisticated system of clearing and settlement, a monetary system is established that is purely abstract. Such a monetary system exists only in account books. While net balances will be convertible into real goods or traditional financial assets, the monetary system itself has an existence that is independent of the real economy. For example, even if the monetary system typically pays off net balances in francs, the sovereign government that issues the francs can go bankrupt and the monetary system will simply shift to paying off net balances in the next best instrument, whether it be dollars or euros or drachmae.

While a monetary system can exist purely in the abstract, the danger that the value of the abstract unit of account will be devalued is every bit as much a risk as the danger that a more traditional sovereign unit of account will be devalued. On the other hand, the abstract unit of account typically develops because a sovereign instrument is being devalued and the bankers seek to maintain the value of their own interactions. The decision to treat the “old” value of the sovereign instrument as the “true” value for the purposes of the monetary system has the effect of converting the monetary system into one that is purely abstract – but convertible into real instruments. Thus, just as a complex web of institutions (e.g. independent central bank, separate public Treasury, democratic polity) supports the value of sovereign money, so an equally complex institutional structure (e.g. personal liability for debt, shunning of bankrupts, hierarchical structure that exploits reputation effects at every level from the international to the local) is required to protect the value of an abstract unit of account.

The first observation of this phenomenon of a monetary system that existed only in the abstract took place in the mid-16th century at the Lyons fairs where the “imaginary” ecu de marc was the unit of account for the money market.[5] By the turn of 17th century the European money market had moved to Venice where the Banco della Piazza di Rialto and its bank ducat became the next “imaginary” currency of account for European trade. In 1609 Amsterdam founded the Wisselbank which deliberately copied the model of the Venetian bank and its bank ducat. By the end of the 1620s the European money market, along with its international trade had shifted to Holland.[6]

During the same time period Amsterdam adopted the techniques of decentralized clearing that had been developed in Antwerp (during a period when banking was a prohibited activity). Clearing was decentralized by formalizing legally the rules for endorsement and circulation of bills of exchange. Thus, by the time the Bank of England was established in 1694 (with the advice of Dutch financiers), (i) the intellectual foundations for a stable and imaginary bank-based unit of account – that is for fiat money – had been firmly established by a century and a half of practice in Europe; and (ii) Europe’s system of clearing and settlement had been so thoroughly established in international trade that bills drawn on European banks could circulate among merchants in Russia, India, and the Americas.

The brilliant innovation of the founders of the Bank of England was to address a political problem: sovereign authorities understood very well the challenge to their authority posed by an autonomous abstract monetary system, [7] and sometimes deliberately took action to weaken it. Thus, by combining the issue of Bank of England notes with an important role in the finance of government debt, the bankers successfully aligned the interests of the sovereign with those of the issuer of the bank-based unit of account.[8] In short, the founders of the Bank of England deliberately laid the foundations of a fiat money that was backed, not as it had been in the past by gold and private debt, but by a combination of gold, sovereign and private debt. The effectiveness of the institutional structure established in 1694 was proven a century later as the Bank of England note enabled the British economy to shift very smoothly to a Bank-based monetary standard and the Bank was thus able to play a crucial role in the finance of the Napoleonic Wars.

The point of my brief review of monetary history is this: clearing and settlement is money. There is no need for some sovereign token to serve as a final means of payment. In short, the theory of the essential role of government in the monetary system is, just that, a theory. It is true in the sense that monetary systems that develop without the consent of the governments within the boundaries of which they function rarely last more than a few decades because they compete with governments which are therefore incentivized to undermine their stability. It is not true, however, in the sense that monetary systems cannot function without being tied to some government unit of account. The history of Europe in the modern era is proof of this latter statement.

By contrast to the consensus view, there is a strong argument that the inverse of the conventional view – that is, of the view that monetary systems are dependent on sovereigns that are institutionally capable of issuing debt without defaulting and base money without inflating – is equally true: modern sovereigns are only able to issue sound debt and money, because of their close ties to banking systems that support robust economic activity by underwriting unsecured, but safe (and therefore low-cost), debt that allows the payments system to operate smoothly and facilitates access to the payments system for a broad spectrum of society. After all, historically bank-based units of account and payments systems were established centuries before British government debt became a safe asset, and the role played by the Bank of England in establishing the safety of British debt ensured that this debt was inextricably tied to the performance of both the banking system and the British economy.

So what do banks do?

So what do banks do? Banks operate the payments system. This entails not just mechanistically processing customer payment orders, but also the design and maintenance of a safe system of short-term lending to support the payments system.

Economic efficiency is fundamentally dependent on the banking system to manage and alleviate the fundamental problem that for each market participant the flow of funds is not synchronized. In the absence of unsecured credit to support payments, many market participants will face prices that are determined by the fact that they are liquidity constrained and that result in an inefficient allocation relative to an economy where these liquidity constraints are obviated by short-term credit. (In our toy model the economy is autarkic if there is no monetary instrument.)

Precisely because payments system credit addresses only the inherent timing problem in payments, these systems can be designed so that they are extremely safe. Thus, from the 17th through the 19th centuries the interest rates paid by businessmen on such credit were typically in the range of 2 to 6% per annum.[9] Then, when we say that banks operate the payments system, we need to include in that description the business of setting the terms of credit lines and monitoring borrowers’ behavior, so that borrowing for purposes of transacting is an activity that can be done at very low cost.

So what are the most important functions of the banks? They:

  • set parameters for credit lines including the credit limit and the interest to be paid
  • monitor borrowers’ use of credit lines and financial positions more generally, adjusting credit terms as needed, and
  • impose penalties on (or withdraw the credit line from) borrowers who violate the terms of the credit line

When these activities take place in an environment where there is interbank competition, the interest rates charged to businessmen with no history of default for such a credit line should be in the low single digits. If we don’t see this kind of unsecured credit readily available to almost all businessmen,[10] then we can assume that something is going very wrong with our banking system and that it is failing in its most important function.

Conclusion: Banking as the fundamental source of liquidity

The modern payments system should be understood as the modern evolution of an abstract monetary system that dates back to the 16th century and one of the earliest money markets established in Europe. At the heart of the payments system lies a system of unsecured credit in which banks set the terms of credit lines and individual market participants draw down those credit lines on an “as needed” basis. This clearing and settlement process together with the short-term credit lines that are intrinsic to its functioning comprise the fundamental source of liquidity in a modern economy.

This analysis indicates that there’s another way to define liquidity. Liquidity is created by the unsecured credit lines that are extended by the banking system in order to make the payments system function smoothly. Thus, one can define liquidity itself as the unsecured credit lines that facilitate the settlement of asset trades and other obligations. This definition is almost the same as that of “funding liquidity” or “the ability to settle obligations with immediacy,”[11] but focuses attention not on settlement, but on the extension by banks of unsecured credit lines that facilitate settlement.

Market liquidity, by contrast, is the ease with which an asset can be bought or sold and is determined by the difficulty of finding a counterparty for your trade (see Harris 2003, p. 394). Clearly when buyers have access to unsecured credit lines, this plays an important role in making it easy for sellers to find buyers, to trade in large size, and to get a good price for the asset. Thus, liquidity, as we have defined it, is also likely to be an important determinant of market liquidity. By contrast, temporary fluctuations in market prices (driven for example by bargaining dynamics in an over-the-counter market) are unlikely to have a significant effect on liquidity. After all, credit lines are generally committed, so temporary fluctuations will frequently disappear before the bank has the opportunity to change the terms of the credit line. This structure makes economic sense, because such temporary fluctuations are unlikely to affect a borrower’s capacity to repay the loan over time. Thus, while we might expect the structure of market liquidity (e.g. whether most assets can be traded on an exchange vs. over-the-counter) to affect the willingness of banks to extend unsecured credit lines and therefore to affect liquidity, as we have defined it, there is little reason to expect that day-to-day changes in market prices should affect liquidity.

Overall, by defining liquidity as the unsecured credit lines that facilitate the settlement of asset trades and other obligations we have a single definition of liquidity that determines both funding liquidity and to a large degree market liquidity. Contrast this approach to liquidity with that of Brunnermeier & Pedersen 2008 (BP). BP define funding liquidity as the fraction of an asset that a trader can finance. That is, BP assume that what is defined as liquidity here – that is, the extension of unsecured credit lines – is necessarily nonexistent. Thus, given our definition of liquidity, BP can be reinterpreted as stating that when liquidity is unavailable, adverse dynamics are easily generated by the interaction between market price fluctuations and collateralized financing constraints.

This analysis raises a host of questions: If the unsecured credit lines that make the payments system function smoothly are liquidity, then are these credit lines also money? Should they be money? If these credit lines that are so important to the operation of the payments system are not money, then what is the point of defining money at all? I am still puzzling over these questions so I only ask them and don’t pretend to answer them here.

[1] Paper presented at the Macro, Money and Finance Conference, Bath University, UK, September 8, 2016.

[2] For accounting purposes, a discount (like a derivative) creates complications. When the discount has a positive balance, it is equivalent to a deposit account and is therefore a bank liability. However, when the discount has a negative balance it is a loan made by the bank and therefore is a bank asset. For analytic simplicity the text here assumes that discounts always have negative balances and therefore are always bank assets.

[3] If we take this simple model as a metaphor for a much more complex monetary system with the same properties, we can consider the kind of monetary expansion that drove Schumpeter’s process of creative destruction (see Schumpeter 1939 which I discuss here). Thus, imagine a Walrasian economy with trading frictions where the banking system operates as described above to eliminate the frictions and make competitive equilibrium attainable. Now assume that someone has an idea for a better production method. Using the credit-based transactions system that individual can buy inputs and convert the production method into an operating business easily. Assuming the new production method is genuinely more efficient than the older one, the newcomer will sell his product more cheaply, demand for the product will slowly (due to information transmission costs and the costs of building up production capacity) shift to the new production method, and the payment system will shift very smoothly to financing the better method.

[4] The inevitable and steady devaluation of all the late medieval coinage systems due simply to the use and circulation of the coins is documented in great detail by Lane and Mueller (1985).

[5] Boyer-Xambeu, Deleplace & Gillard (1994, p. 78).

[6] This history is recounted in detail in a chapter of my dissertation, “The Political Economy of Private Paper Money,” Sissoko (2003).

[7] Montesquieu, The Spirit of Laws, XXII(13).

[8] Martin (2013) calls this monetary innovation the Great Monetary Settlement.

[9] Barbour (1963), p. 85.

[10] Those who are just starting out may be expected to have very low credit limits and/or be required to have a co-signor.

[11] Matthias Drehmann & Kleopatra Nikolaou, Funding Liquidity Risk, BIS Working Paper No. 316 (July 2010).

In search of financial stability: A comparison of proposals for reform

I. The liquidity view
a. Solution: Expansive LOLR
b. Solution: Narrow banking

II. The solvency view
a. Solution: PFAS – the dealer of last resort meets narrow banking
b. Solution: Controls on credit

The vast literature on the financial crisis includes a segment comprised of books that propose reforms to the financial system that are designed to promote financial stability. The initial goal of this post was to evaluate and compare some of the more recent contributions to this literature: Morgan Ricks’ The Money Problem (2015), Adair Turner’s Between Debt and the Devil (2015), and Mervyn King’s The End of Alchemy (2016). In order to help balance the discussion, I am also including Perry Mehrling’s The New Lombard Street (2011), Hal Scott’s Interconnectedness and Contagion (2012), and John Cochrane’s Toward a Run-Free Financial System (2014).

A first basic organizing principle for comparing these proposals is to separate the works by their view of the essential problem to be solved: some argue that we should focus on panics or on avoiding liquidity droughts, whereas others see the fundamental problem as one of solvency or too much private sector debt. Those who take the liquidity view make proposals that fall into two broad categories: the establishment of an expansive lender of last resort, and narrow banking proposals where the government backstops short-term debt. While some proponents of the solvency view also put forth narrow banking proposals, their proposals typically attempt to address the potential danger of too much government support for short-term debt and therefore are distinguished from the liquidity-based narrow banking proposals. Finally some advocates of the solvency view argue that financial stability necessitates controls that limit the private sector’s ability to originate debt.

This post addresses each of these arguments in turn.

The liquidity view

The list of authors who argue that the key to addressing financial stability is to focus on liquidity crises and their prevention is long. Here we will discuss the proposals put forth by John Cochrane, Perry Mehrling, Morgan Ricks, and Hal Scott.

Each of these authors is explicit that in his view the key to financial stability is the prevention of liquidity crises. For example, Morgan Ricks writes: “when it comes to financial stability policy, panics— widespread redemptions of the financial sector’s short- term debt— should be viewed as ‘the problem’ (the main one, anyway). More to the point: panic-proofing, as opposed to, say, asset bubble prevention or ‘systemic risk’ mitigation, should be the central objective of financial stability policy” (p. 3). This view is echoed by both John Cochrane: “At its core, our financial crisis was a systemic run. … The central task for a regulatory response, then, should be to eliminate runs” (p. 197); and Hal Scott: “Contagion occurs when short-term creditors run on solvent institutions, or institutions that would be solvent but for the fire sale of assets that are necessary to fund withdrawals” (CNBC comment) and “contagion, rather than asset or liability interconnectedness, was the primary driver of systemic risk in the recent financial crisis” (p. 293). Perry Mehrling also frames the crisis as fundamentally a matter of liquidity, acknowledging first that it was catalyzed by the decline in collateral valuations, but then explaining: “from a money view perspective, price is first of all a matter of market liquidity, and this perspective focuses attention on the dealer system that translated funding liquidity into market liquidity.” (p. 125).

All four of these authors focus on the fact that the financial system that faced crisis in 2007-09 was constructed upon a foundation of short-term liabilities of non-banks. They differ, however, on the question of whether central bank policy was a cause or a consequence of this financial structure. Both Mehrling and Scott focus on what the Federal Reserve did to address the 2007-09 crisis, whereas Cochrane and Ricks argue that lender of last resort support played an important role in moral hazard and the deterioration of financial institution balance sheets in the decades leading up to the crisis (Cochrane pp. 231-32; Ricks p. 195). Indeed Ricks argues against not just the implementation of last resort lending in the lead-up to the crisis, but even against the traditional lender of last resort, because, first, in his view it functions as a distortionary subsidy to financial institutions and, second, it will fail if these institutions do not have enough of the right sort of collateral (pp. 186-87).

Threading a path between these views I would argue that during the decades preceding and fostering the growth of this financial system built on the short-term liabilities of non-banks, a naïve view of the lender of last resort was promoted by Federal Reserve officials. Alan Greenspan declared that: “The management of systemic risk is properly the job of the central banks. Individual banks should not be required to hold capital against the possibility of overall financial breakdown. Indeed, central banks, by their existence, appropriately offer a form of catastrophe insurance to banks against such events” (speech 1998). And through these formative decades Timothy Geithner, who would be President of the Federal Reserve Bank of New York and then Treasury Secretary during the crisis, was learning to ignore moral hazard concerns when dealing with crises (Geithner 2014).

Before exploring the details of the “panic-proofing” proposals, let’s briefly preview the contrary view that the crisis was a solvency crisis, and the critiques that the solvency proponents have to offer of the liquidity view. Mervyn King references Keynes’ exposition of uncertainty, animal spirits, and the fact that “a market economy is not self-stabilizing” to explain that sometimes an interim period of disequilibrium may be part of a necessary adjustment process as it becomes clear that the current pattern of behavior is no longer sustainable and that “the debts and credits that have built up … will eventually have to be cancelled” (pp. 294-323). In short, due to radical uncertainty, liquidity neither is nor should be “a permanent feature of financial markets” (p. 151). He remarks that: “Political pressures will always favor the provision of liquidity: lasting solutions require a willingness to tackle the solvency issues” (p. 368).

Adair Turner is more direct in his critique. His view that modern economies are reliant on too much private sector debt is supported by extensive empirical research (Jorda, Schularick & Taylor 2014, Mian & Sufi 2014), and he argues that those who deny that too much private sector debt has been originated are misled by a “presumption in favor of … as many financial contracts as possible as widely traded as possible [that] was an accepted article of faith” prior to the crisis (p. 29). Thus, from the perspective of Between Debt and the Devil, proponents of the liquidity view are likely to be captive minds who simply cannot conceive of the possibility that the debt that was originated prior to the crisis was in fact unsustainable and will at some time in the future end up in default.[1]

Only Ricks directly addresses and rejects the solvency view. His discussion does not, however, reach the question of whether a systemic panic is a necessary consequence of an environment with an unstable build-up of debt. Instead he focuses on how damaging the panic itself was. Thus, while one can read Ricks as arguing that the problem can be addressed either at the level of the debt bubble or at the level of the panic, the fact that he chooses to address the problem at the latter stage because it is only then that the problem becomes acute indicates that he considers “too much debt” to be a distinctly secondary concern.[2] This approach lends credence to Turner’s view that current modes of thought about finance preclude serious discussion of the problem of too much debt.

Unsurprisingly, neither King nor Turner supports the broad government guarantees that underlie all of the solutions proposed by the liquidity view proponents. Despite the common reliance of all four liquidity view authors on government guarantees to prevent crises, the form that these guarantees take is very different. Perry Mehrling and Hal Scott would implement these guarantees through expansive access to the lender of last resort without requiring major structural reform to the financial system. John Cochrane and Morgan Ricks, by contrast, propose complete transformation of the financial system before they would advocate government liquidity support.

Solution: Expand the role of the lender of last resort

Perry Mehrling’s argument in support of an expansive role for the lender of last resort is premised on the assumption that complete transformation of the financial system is not a practical solution. He writes: the “capital-market-based credit system … is now a more important source of credit than the traditional banking system. I take it as given that this brave new world is here to stay.” (p. 113). Similarly, even though Hal Scott does discuss proposals that place a cap on short-term funding for banks (p. 160 ff), he does not clearly address the possibility that such caps could be applied to non-banks as an alternative to lender of last resort support. In short, Scott implicitly, though not explicitly, adopts Mehrling’s approach: financial stability is a problem of stabilizing a financial system constructed upon a foundation of short-term liabilities of non-banks. (As we will see below, Cochrane and Ricks do not share this view.)

The most famous proponent of the lender of last resort as a form of “panic-proofing” is probably Timothy Geithner, who views 2007-09 as fundamentally a liquidity crisis and argues that the right way to deal with such a crisis is by providing government support to the financial institutions involved until such time as their balance sheets are repaired and they can function without government support.[3] This naïve view of the lender of last resort treats the moral hazard concerns of this central bank function as something that must be ignored during a crisis.[4]

Mehrling and Scott seek to lay analytic foundations for an expansive lender of last resort as a solution to panics. Scott in his book recounts the aggressive actions that did indeed have the effect of saving the financial system from contagion (though many have observe that economic performance subsequent to this bailout of dysfunctional finance has left much to be desired, e.g. Mian and Sufi 2014) and argues that: “History has taught us that contagion is an unavoidable risk of financial intermediation and that a strong lender of last resort is necessary to prevent it” (CNBC). In fact, Scott views the Lehman bankruptcy as a lesson that “to be effective, a central bank lender-of-last-resort must be unlimited and non-discretionary. The current [post Dodd-Frank] regime leaves open the risk that lender-of-last-resort assistance will be withheld from a distressed financial institution at a critical moment, and thus short-term creditors remain incentivized to withdraw in the face of such distress. An explicit guarantee, as opposed to the implied guarantee that existed before Lehman’s failure, assures short-term creditors that they will recover all of their funds, thus removing their incentive to run in anticipation of large losses” (p. 292). He makes clear in a later article that “the ability to lend to non-banks in a crisis is a crucial matter, and will become even more important, as over regulation of banks fuels the further growth of the shadow banking sector” (CNBC).

Perry Mehrling does not advocate for an “unlimited and non-discretionary” lender of last resort. Instead he argues that the Federal Reserve should convert into a regular facility the Primary Dealer Credit Facility, which was a program the Federal Reserve put into place during the crisis to support the value of private sector assets that were used as collateral in the tri-party repo market. (At its peak this facility held more than $60 billion of equities. See PDCF data .) Mehrling argues that the modern capital-market-based financial system needs such a “dealer of last resort” to set a price floor on private sector assets and that any moral hazard concerns created by this proposal can be addressed by careful pricing (pp. 134, 137-38).[5] Mervyn King doubts that central banks can implement such a policy successfully: “one of the most difficult issues in monetary policy today is the extent to which central banks should intervene in these asset markets – either to prevent an ‘excessive’ rise in asset prices in the first place or to support prices when they fall sharply. … I am not sure that their track record justifies an optimistic judgment of the ability of central banks [to do this]” (p. 265).

Overall, proponents of an expansive lender of last resort as a solution to the problem of liquidity crises generally start with the assumption that the existing financial structure cannot change and do not address the argument the existing financial structure is in fact a product of the expansion of central bank guarantees in the 1980s and 1990s. Adair Turner (likely with substantial agreement from Charles Goodhart, Mervyn King, and Martin Wolf) would probably argue that proponents of this view are captivated by pseudo-economic delusions and mistaken ideas that forestall an understanding of the fundamental problem of “too much debt.” In short, critics of the expansive lender of last resort proposal argue that far from stabilizing the financial system, the policy has a history of being destabilizing.

Solution: Narrow banking

John Cochrane and Morgan Ricks are united in their view that, even though excessive origination of debt is a predictable consequence of misguided government support for the financial system, the correct way to address this problem is to focus on run-prone (or short-term) financial claims and to design a monetary system backed by government obligations that will put an end to runs. While both authors favor structural financial reform that would effectively end – or at least severely restrict – private short-term debt, the monetary frameworks that the two authors adopt as they formulate their solutions are very different: Cochrane’s view of money is a fairly direct distillation of Milton Friedman’s approach, whereas Ricks develops more of a practitioner’s view that owes as much to Marcia Stigum and Diamond-Dybvig-type coordination problems as to any particular monetary theorist. The only common ground in the two views of money is that both treat money issued by the government as the anchor of their systems (Cochrane p. 224, Ricks p. 146).[6]

Both Cochrane and Ricks would transform the financial system by aggressively restricting the ability of both banks and non-banks to issue short-term, run-prone debt. In Cochrane’s proposal “demand deposits, fixed-value money-market funds, or overnight debt must be backed entirely by short-term Treasuries”(p. 198). Cochrane would restrict the degree to which any other short-term debt (except for trade credit) could be used to finance intermediaries by imposing a tax on such liabilities (p. 199). The result would be that “Intermediaries must raise the vast bulk of their funds for risky investments from run-proof securities [i.e. equity]” (p. 198). Ricks’ plan is more comprehensive because it would entirely prohibit nonbank issue of short-term debt, but somewhat less restrictive because it relies on government guarantees of bank liabilities rather than a mandate that banks hold government debt. Specifically, Ricks restricts the issue of short-term debt (except for trade credit) via “unauthorized banking provisions” that only permit banks to issue such debt, and requires that all short-term bank liabilities be explicitly guaranteed by the government (pp. 201, 235). Ricks’ proposal also imposes bank regulation similar to, but more strict than, what we have today including portfolio restrictions and capital requirements (p. 211). Ricks indicates that this proposal can be viewed as making explicit government guarantees that were formerly implicit (p. 25).

Both Cochrane and Ricks argue that government backing of short-term debt will eliminate the danger of runs (with of course the caveat that we are talking about the right sort of government). Whereas Ricks focuses in some detail on the structure of the monetary system, Cochrane’s emphasis is on the value of ensuring that most financial assets are backed by equity: “For the purpose of stopping runs, what really matters is that the value of investors’ claims floats freely and the investors have no claim on the company which could send it into bankruptcy” (p. 215). Ricks’ critique of Cochrane’s proposal is that he underestimates the demand for money-claims on banks and thus ties the supply of money to the quantity of short-term Treasuries available to back them. The advantage of Ricks’ sovereign guarantees of bank liabilities is that it allows the money supply to be backed in part by private sector assets and thus makes it possible for monetary policy to operate independent of fiscal policy (p. 182).

This significant difference in the two proposals is a consequence of the different monetary frameworks that the two authors employ. As noted above, Cochrane’s approach derives directly from Friedman’s and thus bank money, when it exists, is simply a function of government constraints. Ricks, by contrast, views banks as creating money and thus as playing an important part in determining the money supply. It is this latter approach that motivates Ricks to design a “narrow banking” system that nevertheless can allow for expansion of the money supply independent of government debt. Ricks observes that proposals like Cochrane’s (and Friedman 1960’s) envision a monetary system without a significant role for banks (p. 171).

In short, when Cochrane argues that the costs of his transformational plan are not too large, he does so without first modeling why money claims issued by banks are backed by private sector assets. Not only Ricks, but also Adair Turner, Martin Wolf and Charles Goodhart have argued that there are “positive benefits to private rather than public creation of purchasing power” and indeed, that this structure may play a role in “investment mobilization and thus economic growth” (Turner 188-89; see also Wolf 212-13).

Given that Cochrane – and all those who rely on Friedman’s monetary framework – have not thought through why we have the monetary and banking system that we have, his assertions appear “mystical and axiomatic” to use his own words (p. 223). For example, Cochrane writes that by limiting finance to equity finance “we can simply ensure that inevitable booms and busts, losses and failures, transfer seamlessly to final investors without producing runs” (p. 202). “Liquidity is now provided by the liquid markets for these securities, not by banks’ runprone redemption promises.” 226 This Friedman-esque vision of markets plus government as providing all the liquidity that an economy needs is combined with the remarkable claim that we no longer have a transactions need for bank liabilities.[7] Cochrane asseverates that “technology renders this ‘need’ [for short-term bank debt in transactions] obsolete. … We can now know exactly the prices of floating-value securities. Index funds, money market funds, mutual funds, exchange-traded funds, and long-term securitized debt have created floating-value securities that are nonetheless information-insensitive and thus extremely liquid. Consumers already routinely make most transactions via credit cards and debit cards linked to interest-paying accounts, which are in the end largely netted without anyone needing to hold inventories of runnable securities” (p. 222).

In short, 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.[8]

Another aspect of money that Ricks takes into account, but Cochrane with his simple Friedman-based monetary framework barely addresses is that banks are able to expose themselves to runnable, short-term debt even when they aren’t financing their balance sheets. Ricks argues: “Our monetary theory of banking … suggests that derivatives dealing is properly the domain of nonbank financial firms,” because “the amount of cash exchanged upfront [and therefore the money provided] is almost always very small in relation to the risk taken” (p. 208). Cochrane would not restrict such off-balance-sheet activities, and argues that “a few regulators” will be able to detect any dangerous behavior since leverage ratios will be very low (p. 216). Of course, one of the lessons of the crisis is that off-balance-sheet bank liabilities can be very large: Citibank (as well as UBS and Merrill Lynch) had to recognize upwards of $50 billion of derivatives exposures in the form of super-senior CDOs when its “liquidity puts” were drawn down (FCIC  Report, p. 260).

When we combine Cochrane’s casual approach to the danger of off-balance-sheet bank exposures with the view that “invoices [and] trade credit … are not runprone contracts” (p. 202), we find that his formulation of narrow banking leaves open the possibility that after his reforms the financial system could regenerate a very old – but not necessarily very stable – form of banking, acceptance banking. Whatever is classified as trade credit in Cochrane’s regime may be accepted or guaranteed by banks or unacknowledged shadow banks – and these acceptances may circulate as money just as they did in the 19th century with destabilizing effect. In fact, Ricks’ proposal is also permissive of trade credit and therefore is subject to a similar critique: nothing prevents nonbanks from guaranteeing trade credit obligations and this is an avenue through which a new, unstable banking system can develop. This analysis points to another common criticism of narrow banking proposals: they may be impossible to design due to the “remarkable ability of innovative financial systems to replicate banklike maturity transformation” (Turner p. 189).

Overall, narrow banking proposals raise very important questions about whether our monetary system can be better designed to avoid liquidity crises, but (i) will be very hard to formulate in a way that precludes their circumvention, and (ii) are probably best read as evidence that we need much better models of money and banking, so that we can actually understand what the connections are between money, bank liabilities and private sector bank assets, before pursuing transformative change.

The solvency view

Transformational reform is also proposed by scholars who believe that the essential problem that must be addressed in modern financial systems is not liquidity, but solvency. “The fundamental problem is that modern financial systems left to themselves inevitably create debt in excessive quantities, and in particular debt that does not fund new capital investment, but rather the purchase of already existing assets” (Turner p. 3-4). Turner argues that when banks expand the money supply by creating debt that is used to purchase existing assets, the result is an increase in the prices of the assets thus purchased, which then justifies an increase in the debt collateralized by the asset – and thus an expansion of the money supply. The ultimate consequence of this “self-reinforcing credit and asset price cycle” is an asset price bubble (p. 6). When the bubble bursts, as eventually it must, the problem is not liquidity, but solvency. The economy is then burdened with an overhang of debt that is either bad in the sense that repayment is not feasible or uneconomic, because the debtor is servicing debt that is greater than the value of the asset. This basic critique of modern finance – and in particular of the finance of real estate – is advocated not just by Turner, but also by Martin Wolf 2014 and Charles Goodhart & Enrico Perotti 2015.

Mervyn King in The End of Alchemy takes a slightly different approach. He argues that “the most serious fault line in the management of money in our societies today” is “the alchemy of banking” or the system by which money “is created by private sector institutions” and then used to finance illiquid and risky investments (pp. 86, 104). In his view, however, it is important to emphasize that the causal force generating “too much debt” was not the banks themselves, but the demand for borrowing to finance real estate investment due to the savings generated by the structural current account surpluses of Asian countries and Germany together with the decline in real interest rates that resulted from deficit countries’ efforts to keep their economies growing when faced by these surpluses (p. 319, 325).[9] In short, while King agrees that we are currently faced with a state of disequilibrium characterized by too much debt, he explains this outcome via a change in our understanding of the state of the world, not via an inherently unsustainable asset price bubble (pp. 356-57).

Proponents of the solvency view believe that not only does financial stability require that our financial structure be transformed, but also that the only path forward will require debt forgiveness of some sort (King p. 346, Turner p. 225ff). Because the focus of this essay is on proposals for transformational reform of the financial system, devices to deal with the debt overhang will not be discussed. Instead we evaluate King’s proposal for a pawnbroker for all seasons and Turner’s argument that direct controls on the financial sector’s origination of debt instruments are necessary.

Solution: PFAS – the dealer of last resort meets narrow banking

Like John Cochrane and Morgan Ricks, Mervyn King focuses his attention on the design of a more stable monetary system. His proposal for a pawnbroker for all seasons (PFAS) combines aspects of the dealer of last resort and narrow banking proposals. In particular, he would allow the central bank to lend against risky collateral, but only upon terms that are specified well in advance, and he would combine this policy with a restriction that all short-term unsecured liabilities of a bank must be backed by a combination of cash, central bank reserves, and the committed central bank credit line.

King motivates his proposal as an improvement over the traditional lender of last resort, which he (like Ricks) views as suffering from a time inconsistency problem: “The essential problem with the traditional LOLR is that, in the presence of alchemy, the only way to provide sufficient liquidity in a crisis is to lend against bad collateral – at inadequate haircuts and low or zero penalty rates. Announcing in advance that it will follow Bagehot’s rule … will not prevent a central bank from wanting to deviate from it once a crisis hits. Anticipating that, banks have every incentive to run down their holdings of liquid assets” (p. 269). But in contrast to some proponents of an expansive lender of last resort, King argues that moral hazard concerns must be addressed ex ante: “It is not enough to respond to the crisis by throwing money at the system … ensuring that banks face incentives to prepare in normal times for access to liquidity in bad times matters just as much” (p. 270).

Specifically, under King’s proposal, as under the dealer of last resort, the central bank provides liquidity against risky assets and does so subject to a haircut, but importantly the PFAS would not just specify the haircut in advance, but would specify it with the expectation of not changing it for years (p. 277). Thus, the first step of the PFAS proposal is that assets must be pre-positioned as collateral for a specific loan amount. The second step of the proposal caps the short-term unsecured debt of the bank by the sum of the cash, the central bank reserves held by the bank, and the amount that the bank can draw from the central bank on the basis of pre-positioned collateral (p. 272). “The scheme would apply to all financial intermediaries, banks and shadow banks, which issued unsecured debt with a  maturity of less than one year above a de minimis proportion of the balance sheet” (p. 274).

King’s proposal addresses two important design concerns. First, even though banks can create money, “only the central bank can create liquidity” or “the ultimate form of money” (pp. 190, 259). For this reason, King finds that “liquidity regulation has to be seamlessly integrated with a central bank’s function as the lender of last resort” (p. 259).[10] This is achieved by using the credit line commitment of the central bank as a determinant of the cap on a bank’s runnable assets. Second, when a central bank increases its collateralized lending to a bank, the bank’s unsecured lenders are disadvantaged and this form of central bank liquidity support can have the effect of reducing the availability of – or even generating a run on – unsecured market-based lending to the bank. For this reason what is needed is a “single integrated framework within which to analyze the provision of money by central banks in both good time and bad times” (p. 208). Because unsecured lenders will know in advance that the pre-positioned collateral will be used to draw from the central bank, they will not expect it to be available to support their own claims and will demand to be paid a rate on the unsecured debt that compensates them for this fact.

This proposal achieves stability in much the same way that narrow banking does: “all deposits are backed by either actual cash or a guaranteed contingent claim on reserves at the central bank” (p. 271). Unlike Cochrane’s narrow banking, however, only indirect control is exercised over the bank’s asset portfolio. In comparison with Morgan Ricks’ proposal, the public guarantee is provided not with respect to the liabilities of a bank but instead with reference to its assets, and it is the central bank – or the ultimate provider of liquidity – not bank regulators who will make the decisions that affect the bank’s asset portfolio.

The issue of the degree of control exercised by the PFAS is, in fact, an interesting question. One of King’s goals is to “design a system which in effect imposes a tax on the degree of alchemy in our financial system” (p. 271). While each bank nominally is left to determine how to allocate its asset portfolio, the central bank has almost total control over how the tax is structured and, in particular, over which assets will be highly taxed and which will not. According to King the central bank “should be conservative when setting haircuts and, if in doubt, err on the high side. … on some assets they may well be 100%. … It is not the role of central banks to subsidize the existence of markets that would not otherwise exist” (p. 277-78). At least to the degree that a financial intermediary finances itself with deposits and other forms of unsecured short-term debt, it would appear that the PFAS will exercise a great deal of control over the assets that are thus financed.

Unsurprisingly End of Alchemy includes a robust defense of central bank discretion (p. 167). Thus, whether or not this proposal is subject to Ricks’ criticism of narrow banking as serving as an excessive constraint on the money supply will depend on the decisions of central bankers and how they exercise the control over the banking system granted to them by the PFAS proposal.

Solution: Controls on credit

Control over the types of assets that are financed by bank credit creation is also the solution that Adair Turner proposes. It is Turner who advocates most strongly for the view that “too much debt” explains the increasing instability of modern economies. Thus, for Turner “the amount of credit created and its allocation is too important to be left to the bankers; nor can it be left to free markets in securitized credit” (p. 104); instead it is necessary for bank regulators to control the growth of credit. Turner argues more specifically that the most important driving force behind instability was the “interaction between the potentially limitless supply of bank credit and the highly inelastic supply of real estate and locationally specific land. … Credit and real estate price cycles … are close to the whole story [of financial instability in advanced economies]” (p. 175).

Thus, Turner proposes that bank regulation should directly constrain certain types of finance including lending against real estate and shadow banking (p. 195). He would also constrain borrowers’ access to credit and slow international capital flows, which when they took the form of short-term debt simply increased the excess of funds flowing into “hot” real estate markets (p. 196).

Constraints on shadow banking are necessary because in the run-up to the recent crisis it had the effect of “turbocharg[ing] the [credit] cycle, [and] increasing the danger of the wrong sort of debt” (p. 90). Like Ricks and King (p. 94), Turner emphasizes that it was shadow banks that caused bank funding markets to seize up when “wholesale secured funding markets went into a meltdown driven by the very risk management tools that were supposed to make them safe” (p. 103).

While Adair Turner does not promote any version of narrow banking, he draws inspiration from narrow banking’s vision of a system where financial assets are financed by equity. Because “in principle the more that contracts take an equity and not a debt form, the more stable the economy will be,” “implicit taxes on credit creation can be a good thing” (p. 192) and “free market approaches to [credit markets] are simply not valid” (p. 190).

Turner’s focus is, however, very different. Whereas John Cochrane argues that there is no need to differentiate between the different types of credit markets (p. 213), Turner emphasizes the importance of the real estate market: Nowadays “most bank lending … finances the purchase of real estate. … [This] also reflects a bias for banks to prefer to lend against the security of real estate assets … [which] seems to simplify risk assessment” (p. 71). As “banks, unless constrained by policy, have an infinite capacity to create credit, money, and purchasing power … [this combination results in] credit and asset price cycles [that] are not just part of the story of financial instability in modern economies, they are its very essence” (p. 73).

Overall, Turner’s bottom line is that “we should not intervene in the allocation of credit to specific individuals or businesses, but we must constrain the overall quantity of credit and lean against the free market’s potentially harmful bias toward the ‘speculative’ finance of existing assets.” This policy “does not mean less growth, since a large proportion of credit is not essential to economic growth” (p. 208).


Discussions of financial stability and how to achieve it are characterized by a remarkable breadth of views. At one extreme are those who believe that modern finance is here to stay and that its stabilization requires a lender of last resort which plays a much expansive role than in the past. Critics of this approach argue that on the contrary, the expansion of the lender of last resort’s responsibilities over the course of the last three or four decades is what generated the modern financial system which is so very unstable.

Some of these critics of the modern financial system emphasize the liquidity problems it generates and others the solvency problems. All, however, are in agreement that, if financial stability is the goal, substantial reform of the modern financial system is necessary.

Proponents of the solvency view explain that the design of the modern financial system is so flawed that the origination of too much debt is a structural problem. As a result proponents of the solvency view find that either regulators or the central bank must constrain the capacity of all financial intermediaries to finance certain forms of debt – and real estate loans, in particular – using short-term instruments.

The proponents of the liquidity view who propose transformational reform of the financial system argue that only government backing of short-term liabilities can stabilize them. They differ on the degree to which banks have a role to play in a reformed financial system, however. And the comparison of these proposals leads me to conclude that we are in desperate need of better – formal, economic – models of money and banking in order to evaluate these questions.

So what’s my bottom line? I’ve been working on a model of money, bank liabilities, and private sector debt that speaks to all these issues. This model demonstrates that banks’ economic function is to underwrite the unsecured debt that makes the payments system work. By doing so banks bring agents who would otherwise be anonymous and autarkic into the economy. In effect, banks are paid enforcers of intertemporal budget constraints – and it is only because they provide this service that you and I can participate in the payments system and therefore in a modern economy. In short, I think we need a “banking school” model to help us tackle these problems. (Warning to Friedmanites: banking school is the devil that it was Friedman’s agenda to exterminate.) The details will, however, have to wait for another day.

[1] While Hal Scott’s opus has been described as showing “that none of the banks that fell or were rescued were important enough to another big institution to cause its failure” (Authers 2016), this fails to address the question of whether the whole system was beset by too much debt. The danger to the financial system of a “bad equilibrium” in which every participant underwrites too much debt has been recognized for decades (Goodhart 1988 p. 48).

[2] He writes: “this chapter offers reason to doubt that debt-fueled bubbles and the like pose a grave threat to the real economy in the absence of a panic” (p. 106) and “my claim is not that debt-fueled bubbles are insignificant … Rather, my claim is that panics appear to pose a far graver threat to the broader economy” (p. 141). This certainly seems to imply that is possible to have debt-fueled bubbles without also having a panic.

[3] In an interview Geithner states: “What’s unique about panics, and most dangerous, is the amount of collateral damage they do to the innocent, to people who had borrowed responsibly, who weren’t overexposed. The banking system is the lifeblood of the economy. It’s like the power grid. You have to make sure the lights stay on, because if the lights go out, then you face the damage like what you saw in the Great Depression … That requires doing things that are terribly unfair and look deeply offensive. It looks like you are rewarding the arsonist or protecting people from their mistakes, but there is no alternative. We didn’t do it for the banks. We did it to protect people from the failures of banks” (Wessel 2014).

[4] For a view of the lender of last resort which is more nuanced see Sissoko 2016. In fact, the origins of the term “lender of last resort” itself indicates that the central bank is rightly the “court of last appeal” which makes the ultimate determination of whether a financial firm is solvent or not. Implicit in the moniker is the idea that central banks should sometimes uphold the market’s death sentence for a financial firm – just as courts must sometimes uphold real-life death sentences (Sissoko 2014).

[5] Cochrane’s dry comment on the expansion of policy to the regulation of prices is: “What did the old lady eat after the horse?” (p. 238).

[6] This is unsurprising given that almost all modern academic analyses of money, including the heterodox literature, also emphasize the role of government in the money supply. Whether or not this consensus is well-founded is a topic for a different post.

[7] Perhaps Cochrane’s view of the capacity of markets to provide liquidity has changed in recent years. He writes in an October 2016 essay titled Volume and Information: “Information seems to need trades to percolate into prices. We just don’t understand why.” which would seem to imply that markets both demand liquidity and provide it.

[8] Indeed, this is clearly Morgan Ricks agenda (see p. 210). The weakness of Ricks’ approach is that he is a legal scholar and the agenda calls for formal economic analysis.

[9] Note that Turner and Wolf both agree that current account imbalances played an important role in generating the asset price bubbles.

[10] Here King is apparently questioning whether the liquidity coverage ratio specified by the Basel III accords makes sense.

Re-imagining Money and Banking

I’ve written a new paper motivated by my belief that the recent financial crisis was in no small part a failure of economic theory and therefore of economic thinking. In particular, there is a missing model of banking that was well understood a century ago, but is completely unfamiliar to modern scholars and practitioners. The goal of this paper is to introduce modern students of money and banking to the model of money that shaped the 19th century development of a financial infrastructure that both supported modern economic growth for more than 100 years and was passed down to us as our heritage before we in our hubris tore that infrastructure apart.

Another goal is to illustrate what I believe is a fundamental property of environments with (i) liquidity frictions and (ii) a large population with no public visibility but a discount factor greater than zero: in such an environment anyone with a notepad, some arithmetic skills, and some measure of public visibility can offer – and profit from – the account-keeping services that make incentive feasible a much better allocation than autarky for the general populace. Importantly collateral is completely unnecessary in a bank-based payments system.

This model has two key components. First, banks transform non-bank debt into monetary debt. Thus, the transformative function of banking is not principally a matter of maturity, but instead of the nature of the debt itself, that is, of its acceptability as a means of exchange. Second, monetary debt is money (contra Kocherlakota 1998). There is no hierarchy of moneys where some assets have more monetary characteristics than others. Instead there is only monetary debt and non-monetary debt. When we study this very simple model of money in an environment with liquidity frictions using the tools of mechanism design, we see that the economic function of the banking system is to underwrite a payments system based on unsecured debt and thereby to make intertemporal budget constraints enforceable or equivalently to make it possible for the non-banks in our economy to monetize the value of the weight that they place on the future in the form of a discount factor. Banking transforms an autarkic economy into one that flourishes because credit is abundantly available. In this model, constraints on the economy’s capacity to support debt are not determined by “deposits” or by “collateral”, but instead by the incentive constraints associated with banking.

In this environment, banking provides the extraordinary liquidity that is only possible when the payments system is based on unsecured debt. Underlying this form of liquidity is the banks’ profound understanding of the incentive structures faced by non-banks, as it is this understanding that makes it possible for banks to structure the system of monetary debt so that it is to all intents and purposes default-free. (This is actually a fairly accurate description of 19th century British banking. The only people who lost money were the bank owners who guaranteed the payments system. See Sissoko 2014.) Although this concept of price stable liquidity is unfamiliar to many modern scholars, Bengt Holmstrom (2015) has given it a name: money market liquidity.[1] In such a system the distinctions between funding liquidity and market liquidity collapse, because the whole point of the banking system is to ensure that default occurs with negligible probability. Thus, the term money market liquidity references the idea that in money markets, the process by which assets are originated must be close to faultless or instability will be the result, because the relationship between money – when it takes the form of monetary debt – and prices is not inherently stable (cf. Smith 1776, Sargent & Wallace 1982).

This paper employs the tools of New Monetarism, mechanism design, and more particularly the model of Gu, Mattesini, Monnet, and Wright (2013) to explain the extraordinary economic importance of the simplest and most ancient function of a bank: in this paper banks are account-keepers, whose services support a payment system based on unsecured credit. Unsecured credit is incentive feasible, because banks provide account-keeping services and can use the threat of withdrawing access to account-keeping services to make the non-bank budget constraint enforceable.

The basic elements of the argument are this: an environment with anonymity, liquidity frictions and somewhat patient agents is an environment that begs for an innovation that both remedies the problem of anonymity and realizes the value of the unsecured credit that the patience of the agents in the economy supports. I argue that the standard way in which economies from ancient Rome to medieval Europe to modern America address this problem is by introducing banking – or fee-based account-keepers – in order to alleviate the problem of anonymity that prevents agents from realizing the value inherent in the weight they place on the future. I demonstrate that in this environment, the introduction of a bank improves welfare. The improvement in welfare can be dramatic when the discount factor is not close to zero.

This paper uses the environment of Gu, Mattesini, Monnet, and Wright (2013) but is distinguished from that model, because here the focus is on a different aspect of banking. We study how the account-keeping function of banks serves to support unsecured credit, whereas GMMW studies how the deposit-taking function of banks is able to support fully collateralized credit.

The model of banking in this paper has implications that are very different from much of the existing literature on banking. This literature typically assumes the anonymity of agents and then argues – contrary to real-world experience – that unsecured non-bank credit is unimaginable (see, e.g., Gorton & Ordonez 2014, Monnet & Sanches 2015). In other words, the existing literature takes the position that in the presence of anonymity, no paid account-keeper will arise who will make it possible for agents in the economy to realize the value of unsecured credit that their discount factor supports. In the absence of unsecured credit, lending is generally constrained as much by the available collateral or deposits, as by incentive constraints themselves. This paper argues that standard assumptions such as loans must equal deposits (see, e.g. Berentsen, Camera & Waller 2007) or debt must be supported by collateral (see e.g. Gu, Mattesini, Monnet, and Wright (2013), Gorton & Ordonez 2014) are properly viewed as ad hoc assumptions that should be justified by some explanation for why banking has not arisen and made unsecured credit available to anonymous agents.

[1] While Holmstrom (2015) and this paper agree on the principle that money market liquidity is characterized by price stability, the mechanism by which that price stability is achieved is very different in the two papers: for Holmstrom it is the opacity of collateral that makes price stability possible.

Collateral and Monetary Policy: A Puzzle

A stylized fact about post-crisis economies is that asset markets have become segmented with “safe assets” trading differently from assets more generally. I have argued elsewhere that the collateralization of financial sector liabilities has played an important role in this segmentation of markets.

I believe that this creates a puzzle for the implementation of monetary policy that provides at least a partial explanation for why we are stuck at the zero lower bound. Consider the consequences of an increase in the policy rate by 25 bps. This has the effect of lowering the price of ultra-short-term Treasury debt, and particularly when combined with a general policy of raising the policy rate over a period of months or years this policy should have the effect of lowering the price of longer term Treasuries as well (due to the fact that long-term yields can be arbitraged by rolling over short-term debt).

A decline in the price of long-term Treasuries will have the effect of reducing the dollar value of the stock of outstanding Treasuries (as long as the Treasury does not have a policy of responding to the price effects of monetary policy by issuing more Treasuries). But now consider what happens in the –segmented — market for Treasury debt. Assuming that demand for Treasuries is downward sloping, then the fact that contractionary monetary policy tends to shrink the stock of Treasuries itself puts upward pressure on the price of Treasuries that, particularly when demand for Treasuries is inelastic, will tend to offset and may even entirely counteract the tendency for the yield on long-term Treasuries to rise. (Presumably in a world where markets aren’t segmented demand for Treasuries is fairly elastic and shifts into other financial assets quash this effect.)

In short, a world where safe assets trade in segmented markets may be one where implementing monetary policy using the interest rate as a policy tool is particularly difficult. Can short-term and long-term safe assets become segmented markets as well? Given arbitrage, it’s hard to imagine how this is possible.

These thoughts are, of course, motivated by the behavior of Treasury yields following the Federal Reserves 25 bp rate hike in December 2015.fredgraph


Finance and Economics: A Research Agenda

My research agenda employs deconstructive method to motivate a reconsideration of the meaning of neoclassical economics. Thus, an economic theory paper introduces a liquidity friction into a competitive model to study how standard models are constructed on the assumption of perfect intermediation (Sissoko 2007), an economic history paper demonstrates that in fact the markets of industrializing Britain relied on a carefully calibrated banking system that successfully stabilized money growth (Sissoko 2016a), an economic theory paper uses new monetarist methods to model banking and how it stabilizes the relationship between unsecured debt and the money supply (Sissoko 2016b), and another paper analyzes modern finance and explains how the growth of market-based lending has disrupted market liquidity by circumventing the stabilizing force of banks (Sissoko 2016c). Together these papers invert the mainstream view by arguing that starting in late 18th c Britain the banking system effectively stabilized the money supply, allowing it to be treated as a stable background condition: this made modern capitalism possible and neoclassical economics itself imaginable. This post explains the “big picture” of the role played by innovations in banking on European industrialization, and this chapter of my dissertation gives even more detail.