Economics and Banking – James Tobin 1963 (Redux)

By JKH (Cross posted from Monetary Realism)


This essay (unlike the previous) is an unfiltered look at the famous 1963 James Tobin paper on banking:

Commercial Banks as Creators of “Money” (1963)

The Tobin paper has been the subject of recent vigorous blogosphere discussion. Steve Waldman has included a list of links at the conclusion of his own excellent post on the subject:

I have included links to several additional Tobin papers at the conclusion of this essay which may help round out the context. But the focus here is on the 1963 paper linked above.

Tobin’s essay revolves around the idea that banks are a type of financial intermediary. His generic description of bank balance sheet management holds up reasonably well today. Curiously, there has been some resistance from heterodox economics types to the idea that banks are financial intermediaries. Why has there been such resistance to this idea?

First, heterodox likes to focus on the notion that banks are “special”, as reflected in the idea that “loans create deposits”. This seems to be a pivot point for a kind of reflexive rebellion against mainstream economics and its analysis (or not) of banking. In fact, while Tobin focuses on the intermediary character of banking in his 1963 paper, he does not deny that banks are special. He does examine their special nature more intensively in a 1982 essay (noted further below), but the two perspectives are entirely complementary.

Given a set S = {e1, e2}, the element e1 is special in that it is not the same as the element e2, but general in that both e1 and e2 are members of S. Both characteristics are true, and the only question is the qualifying criterion for membership in the set S. And in this case, that qualification has to do with a Tobinesque portfolio management approach as it pertains to both bank and non-bank financial intermediaries, notwithstanding the special characteristics of banking within the intermediary group. The common intermediary quality is the focus of his 1963 essay.

Second, failure to see how banks are a type of intermediary flows from what might be described (somewhat tongue in cheek) as “deposit origination myopia” (DOM), an especially exuberant attachment to the mantra “loan create deposits”. Tobin’s 1963 essay is a 50 year old barometer of this syndrome. I think this was roughly Paul Krugman’s interpretation in the context of earlier blogosphere discussions.

(Krugman has faltered tactically in a number of exchanges with heterodox bloggers because of his apparent dismissal of the relevance of bank accounting. Perhaps there is a potential middle point between presbyopia and myopia in this regard – although positions in some cases appear to beentrenched.)

Neither mainstream nor heterodox economic schools offer a very refined analysis of banking operations. But Tobin’s essay, 50 years old, is insightful on the general framework of banking – certainly more so than “loans create deposits”.

Reserve management is one of many critical functions in banking. The role of bank reserves complements the “loans create deposits” perspective – in that reserves are used for interbank payments, not for deposit creation. That said, heterodox insight on the issue of bank reserves, while constituting a degree of forward progress, is hardly the decisive qualification for understanding what banking is about. This is inherent in the message of Tobin’s 1963 essay. His framing of banks as a type of financial intermediary is a more complete conceptual framework. In that context, the fact that loans create deposits is a minor operational detail that is of very little significance to the strategic management of banks.

Banks are financial intermediaries. So are insurance companies. Banks are special. So are insurance companies. Banks are special because they issue deposits. Insurance companies are special because they issue actuarial claims. Recent blogosphere discussion seems skewed toward the special part in the case of banks. And so the phrase “loans creates deposits” seems to have produced a fresh generation of corresponding banking experts.

Tobin focuses on the set of financial intermediaries and how its constituent members – banks, insurance companies, mutual fund companies, investment firms, etc. – are similar as a group. (Tobin also reveals a great deal in the 1963 paper about how banks are special, without dwelling on this aspect as the main theme.) These similarities have nothing to do with the way in which deposits are issued by banks or used by non-bank financial intermediaries.

Tobin starts his essay with a trifecta – dismissing the false money multiplier story of banking, acknowledging that loans create deposits, and recognizing that banks obtain required reserves after the fact of related deposit creation. These are standard heterodox claims today. Tobin pointed out these facts 50 years ago! But he then sets these observations aside, and moves on to describe banking as a type of financial intermediation.

Before moving further into Tobin’s paper, it may be worthwhile to explore the idea of media of exchange.

Media of Exchange

Commercial banks are “special” at the very least because they issue demand deposit liabilities. Demand deposits are a core medium of exchange for customers of banks. We can think of the medium of exchange for bank customers as a complex of related things: demand deposits, cheques, electronic transfers, banknotes and coins. This may be insufficiently precise for those with more refined sensitivities about definitions, but it suffices here. The instruction to effect an electronic transfer from one demand deposit to another blurs the distinction between deposit and the instruction itself as the medium of exchange. (Marshall McLuhan said the medium is the message.) Is the medium of exchange the demand deposit, the electronic transmission, or the instruction? Are these things in combination a substitute for the physical transportation of central bank money from one bank account to another? To simplify these issues, we’ll refer to bank demand deposits as the core medium of exchange for commercial bank customers, with central bank money, cheques, and electronic instructions affixed to that concept.

Central Banks are “special” as well, for at least the reason that they issue reserve deposits for use by commercial banks in making payments to each other. These reserve balances are the core medium of exchange for commercial banks. Central banks and government treasuries also issue banknotes and coins, which commercial banks hold as inventories in order to satisfy customer demand. Commercial bank customers hold a mix of government money and commercial bank money according to their liquidity preferences.

Commercial banks are in general the only institutions that both use and issue different media of exchange in stock form.

(One interesting exception to this rule occurs within the Euro system. The European Central Bank (ECB) issues TARGET2 balances to the national central banks (NCBs) of the Eurozone. The NCBs use TARGET2 balances as a medium of exchange and issue reserves to commercial banks. Commercial banks use reserve balances as a medium of exchange and issue demand deposits to their customers.)

There is an important parallel to be drawn between central banks and commercial banks as issuers of media of exchange. Both issue media of exchange to their respective clientele using asset swaps. For example, central banks can issue new reserve balances to commercial banks through asset repurchase agreements or by lending directly to banks. (In recent years, with quantitative easing, the US Federal Reserve has issued additional reserves by acquiring bonds originally held mostly by non-banks. Credits to commercial bank customer accounts are accompanied by an increase in bank reserve balances at the Fed.) Commercial banks typically issue new deposits by lending directly to customers or acquiring other market assets (usually liquid securities of third parties). While the process of medium of exchange creation is sometimes described as “ex nihilo”, both entities – central banks and commercial banks – expand their balance sheets by exchanging the medium of exchange for an asset of equivalent deemed value.

Customers of commercial banks can disburse funds from newly acquired deposits in order to pay for things by using cheques, electronic transfer, or conversion to central bank money. The creation, destruction, and transfer of bank deposits works according to the rules of accounting. Individual banks issue new deposits in conjunction with the transfer of funds from other banks. The deposit at the origin of the transfer is destroyed as the new one is created. The issuing bank receives central bank reserve credit as payment from the counterparty bank. System deposits remain unchanged in this type of transfer. This is perhaps the most basic of all banking transactions, but it is interesting because the medium of exchange (the source bank deposit) is being used to acquire the medium of exchange (the destination bank deposit). Indeed, it is reflected this way in macro flow of funds accounting.

A financial intermediary category is defined according to the kind of liability business it is in. Banks are allowed to create the medium of exchange as a liability. Other intermediaries for example offer insurance or pension policies or mutual fund shares.

The core asset business of a bank is lending and the core liability business is deposits. The core asset business of an insurance company is a financial asset portfolio of securities and the core liability business is the issuance of insurance policies and pension promises. The core asset business of a mutual fund company is investment in financial securities and the core liability business is mutual fund shares. And so on. (We use the term “investment” here in its vernacular meaning of financial investment in stocks, bonds, and money market securities.)

The banking system as a whole expands its balance sheet in a single step. The coincident swapping of a newly created demand deposit liability for a new loan asset is the iconic example of such system expansion. This produces new medium of exchange.

By contrast, the NBFI system (non-bank financial institutions) as a whole expands its balance sheet in two steps. An insurance company can swap an insurance policy liability for a demand deposit asset. Using the same demand deposit, it can subsequently acquire a corporate debt or equity security as an addition to its portfolio of assets, which in total “hedges” the actuarial claims it has issued. The net result is that the NBFI balance sheet will have expanded by the value of the insurance policy as a liability and the matching value of the financial asset. No new medium of exchange is produced in these two steps.

The net effect in the case of both banks and NBFIs is that balance sheets are constructed according to asset swaps. That is the essence of financial intermediation.

Financial Intermediation

In the case of both banks and NBFIs, the role played by the medium of exchange in balance sheet expansion is just the beginning of the story. It is the operational front end of a larger and more complicated management process. That process is inherent in Tobin’s portfolio management paradigm. It is the theme of his 1963 paper.

Tobin views those who acquire the liabilities of banks and other financial institutions as “would be lenders” in a counterfactual economy without financial intermediaries:

“…the essential function of financial intermediaries, including commercial banks, is to satisfy simultaneously the portfolio preferences of two types of individuals or firms. On one side are borrowers, who wish to expand their holdings of real assets… On the other side are lenders who wish to hold part or all of their net worth in assets of stable money value with negligible risk of default.”

Tobin’s general description of banking holds up well today. Commercial banks manage credit risk, liquidity risk, and interest rate risk, among other types of risk. They have systems for the measurement of risk and for the allocation of capital to those risks and to the individual banking businesses that incur them. There is an overarching oversight process that includes risk committees and asset-liability committees (ALCOs) undertaking regular systematic reviews of risk exposures. Strategy decisions are made about target asset-liability mixes and pricing parameters. There are internal funds transfer pricing systems to ensure that current market interest rates are reflected immediately in the interest margins that result for each separate banking business according to current asset and liability pricing. These various management components survey and steer the financial intermediation effect in its totality. Moreover, all types of financial intermediaries manage their balance sheets with such techniques of portfolio management, risk management, asset-liability management, and capital allocation.

Just as the bank lending officer ignores the reserve position in making lending decisions (a point made often by heterodox observers), risk and asset-liability managers ignore “loan create deposits”. The deposit origination process is simply not relevant to the portfolio management approach. It is an operational feature of minimal importance – according to Tobin’s strategic portfolio management description – and in fact. For example, the critical credit risk characteristic of an individual loan has nothing to do with the fact that loans create deposits. Banks are concerned with risk to net interest margins, not the fact that demand deposits are created endogenously.

The relationship between the strategic portfolio management approach and the operational endogenous money phenomenon can be explored by examining what happens to the medium of exchange after a commercial bank has issued it. What is its life cycle, so to speak? How is it used, and what becomes of it? Tobin explores this sort of thing by resorting to the analogy of the “widow’s cruse”.

Tobin’s “Widow’s Cruse” Analogy

Tobin refers to a “widow’s cruse” in describing commercial bank deposit creation. A somewhat archaic and awkward analogy, it suggests an unlimited supply of something. His point is that banks do not enjoy such a widow’s cruse capacity for deposit creation and balance sheet expansion. The widow’s cruse fails in the case of commercial banks.

Commercial banks are constrained in their deposit expansion both within their own banking intermediary category and against competing intermediary institutions such as insurance companies, pension funds, mutual funds, and investment firms. A number of factors impinge on the volume of bank deposits that can be produced and survive in final stock form. These factors operate at the original point of deposit creation, and subsequent to that point.

First, in order to create deposits, banks first must have suitable opportunities available to expand their assets. New asset originations must satisfy criteria for assuming the credit risk that is inherent in lending and asset acquisitions. Acceptable asset opportunities must be combined with a required return on equity capital. Loans are priced according to the cost of funding (including the cost of equity capital) and associated administrative expenses. The required margin translates to a capital constraint (quantity and price) rather than a reserve constraint. (Where required reserves are applicable, it is the cost rather than the availability of reserves that is a factor.) Moreover, the quantity of capital is sometimes a scarce resource. Banks can only take on new risk if they have excess capital. And this depends on their level of retained earnings and their access to new capital if required. And even in the presence of excess capital, banks may choose to use it in share buybacks rather than new asset acquisitions if the cost of equity capital is too high relative to available asset opportunities. A bank that decides on a share buyback is momentarily unconstrained by the quantity of capital but constrained in the supply of asset opportunities that meet required return on equity hurdle rates.

Second, the contractual nature of loan agreements require that they be repaid. This is a source of natural “reflux” of the medium of exchange, since demand deposits are extinguished when loans are repaid. The banking system balance sheet and medium of exchange expand on a net basis over time to the extent that asset acquisition and the corresponding “efflux” of the medium of exchange exceeds this contractual reflux.

Third, the ultimate fate of the demand deposits initially generated by asset expansion is generally uncertain even within the banking system. Banks compete within their own intermediary group for all of their desired liability forms, including demand deposits. Banks expect the deposits created by their own new lending activity to end up at other banks as the result of the borrower’s use of funds elsewhere. Individual banks are subject to continuing competitive pressures from other banks in retaining existing deposits. Demand deposits thus pose a risk in terms of both liquidity and interest rate exposure, because they can migrate easily to other banks and/or require higher interest rates to retain in the event of monetary policy tightening.

Moreover, the banking system as a whole requires demand deposits as a source of funds to convert to other liability forms. In order to avoid excessive risk, banks will attempt to convert demand deposits opportunistically to alternative funding forms such as time deposits, debt, and equity capital – so as to strengthen the risk profile of the overall funding structure for the balance sheet. (The cases of issuing new equity capital and generating equity internally through retained earnings both require demand deposit conversion.) The banking system is thus proactive in reshaping the risk profile of the original demand deposit flow created by lending and other types of asset acquisition. The default assumption that banks just sit on the fruits of their demand deposit creation is incorrect.

Fourth, the banking system competes with other financial intermediaries for desired liability forms. If an insurance company issues a new policy, it is competing with a bank that issues a demand deposit liability. This is fundamentally true in the sense that – in a counterfactual financial system without insurance companies – agents would be forced to construct their own insurance cash flows using bank liabilities. Tobin also illustrates the case where demand deposits are used to acquire insurance policies (for example). Insurers as a consequence then seek to acquire financial securities as assets and banks sell some of their liquid securities in response to that demand, resulting in the destruction (reflux) of demand deposits originally issued by the banking system. All intermediary institutions manage their balance sheets in such a competitive context, and they have considerable influence and control over their balance sheet compositions through product design, active pricing, and risk management.

Thus, Tobin essentially distinguishes between a strategic view of banking and an operational view. An operational or monetary view of banking focuses on micro transaction details – things like bank reserves and “hot potato” and “reflux” monetary behaviors. But such “hot potato” and “reflux” effects operate ubiquitously and continuously in financial markets. It’s called the flow of funds – involving the partial success and failure of the widow’s cruse in ongoing flux. As described above, the cruse can fail through limited expansion at origination, contraction at termination, internal conversion, or net sale of assets from the banking system.

Once deposits are created, the liability side of an individual bank balance sheet is in play. Funding is at risk, unless locked in from a liquidity management perspective. In the ongoing process of financial intermediation and balance sheet management, the maintenance of funding dominates the process of loan creation or extinction. Banks need to retain and attract a share of demand deposits (and other liability forms) once they have been created. They need them as an item in the ongoing liability mix and as a source for conversion into other liability forms or equity. No bank can afford to hemorrhage its share of the medium of exchange liability form as a stock item on the balance sheet.

All intermediaries are subject to such liquidity and other risks. In that sense, it is arbitrary to type a mutual fund or insurance company as a financial intermediary and a bank as not. If anything, the liquidity risk inherent in demand deposit liabilities emphasizes the importance of individual banks retaining and attracting them – notwithstanding their ability to create them. Their ability to create them is constrained by the economics of credit risk and capital management. Their ability to retain them is constrained by the effectiveness of their liquidity management. A time deposit provides liquidity and interest rate protection since it can’t be moved to a competing bank until it is converted to the medium of exchange at maturity. Debt and equity provide similar liquidity protection. Once new money is created, it becomes the object of competition within the banking system.

The Chicago Plan and its variations argued for the confiscation of demand deposit capacity from commercial banking and the ring fencing of demand deposits with a 100 per cent reserve requirement. These types of proposals tend to overlook the fact that banks have a self-interested motive to convert much of the demand deposit base that originates from lending to other liability forms for purposes of risk management. And in all of this, the operational characteristic of deposit creation is essentially irrelevant, given subsequent competition for desired liability forms.

The Interface of Government and Commercial Banking

Tobin distinguishes between the “fountain pen money” of commercial banks and the “printing press money” of government. The paper’s focus is the former. He makes the distinction between the natural reflux channels for commercial bank money, as described above, and the core “hot potato” characteristic of government money. But what he is really talking about in the latter case is the consolidated government financial position. This is the position that is popularly appreciated now in heterodox circles as the “net financial asset” position (NFA) held by the non-government sector. This position is created over time by government budget deficit efflux and surplus reflux. To the degree that the budget is exogenous, the private sector portfolio that results is a “hot potato”. But this may be debatable to the extent that there is an argument for the “deep endogeneity” of the government fiscal position.

In this context, Tobin does not illuminate the decomposition of NFA into its component parts – bank reserve balances, banknotes and coins, and government bonds. In particular, he does not discuss the scenario of open-ended structural reflux from commercial bank deposits to banknotes and coins. And that is understandable, since his paper is about an ongoing functional and viable commercial banking system.

Two related papers

Financial Intermediaries and the Effectiveness of Monetary Controls (with William Brainard, 1963)

The Commercial Banking Firm: A Simple Model (1982)

These papers along with the one discussed here constitute a consistent treatment of the banking system as a type of financial intermediary in competition with others, and with individual banks operating as active portfolio managers.


Got a comment or question about this post? Feel free to use the Ask Cullen section, leave a comment in the forum or send me a message on Twitter.
  • Tom Brown

    JKH, some questions:

    1. Does any of this analysis change when at the ZLB? Especially in regard to part of the motivation for banks to attract and retain deposits?

    2. It sounds like you’re saying that gov bonds contribute to the “hot potato” as well as currency and Fed deposits. True?

    3. Can you elaborate on this a bit?: “But this may be debatable to the extent that there is an argument for the “deep endogeneity” of the government fiscal position.”

  • LVG

    I am confused by the way you describe bank lending as an “asset swap”. When a bank makes an non-collateralized loan it is definitely not making an asset swap, right?

    What am I missing?

  • Tom Brown

    I think the asset acquired is the loan itself. I like to think of the aggregated banking sector as purchasing stuff from non-banks by crediting bank deposits. That “stuff” is everything non-banks have to offer: new loan agreements (which may require collateral), their time (salaries and contractors’ fees), dividends & interest payments, office supplies, rent, electric bill, etc. In a sense all those transactions are “asset swaps.”

  • Cullen Roche

    It might help to always think of a loan as having 4 accounting entries. When you take out a loan you take on an asset (the deposit) and a liability (the loan) and the bank takes on an asset (the loan) and a liability (the deposit). Maybe it’s better to say that a non-collateralized loan is a promise to deliver assets in the future. In this way, the bank is swapping its assets today for your assets tomorrow.

    Not sure if JKH is viewing it like that, but that might help conceptualize things better.

  • erestor

    Wow, great post. That resolves many of the issues that were causing me confusion.

  • Tom Brown

    An easy analogy is two people swapping IOUs. Both IOUs were created ex-nihilo, but they’re still swapped (as assets) to accomplish the loan/deposit. But like you say, one IOU provides purchasing power right now and is easily transferable, while the other is less liquid. Each IOU represents a financial asset and each ties a creditor (left hand side of BS) to a debtor (right hand side of BS) and thus four entries are required: 2 on the left and 2 on the right. It could eventually be anywhere from 2 to 4 entities having those BS entries (more if the IOUs are divided into pieces of course).

  • Tom Brown

    BTW, if anybody else can answer these that’s fine too!

  • Tom Brown

    The banks can also sell equity (stocks) in themselves and bonds. It seems to me that in doing so the role of banks and NBFIs is somewhat blurred. In other words, a bank which sells it’s own equity seems to encompass both the role of an NBFI and a bank to some extent.

    Looking at Ramanan’s post here as an example:

    The shares-issuing role of Ramanan’s NBFIs can be incorporated into the banks.

  • Tom Brown

    … into the model of the banks anyway.

  • Cullen Roche

    Yeah, and I think we’re all (even some Market Monetarists) moving towards a point of general agreement – banks are special because they create the primary means of payment….

  • Qcl

    Tom ,

    Can stocks be bought with reserves ?

    I am still trying to find a direct transmition mechanism between QE and rising indices .


  • Tom Brown

    Right, but some distinctions persist. For example in one extreme Sumner say only the MOA is important and that’s be cause it’s “paper gold.” This view seems odd to me for two reasons:

    1. Focus on MOA (“outside money” is roughly the same as MOA).
    2. Definition of MOA as “paper gold” seems circular, relying completely on an M M T style concept of pure fiat money given value only by the state.

    Then there’s Nick Rowe & Bill Woolsey: who claim that MOE dominates and that there’s NO law of reflux for MOE (or MOA)! I think JKH (and Tobin) is clearly saying something different here. Sumner would probably agree w/ Tobin on reflux applying to inside money, but not outside money.

    Then there’s Mike Sproul who is a pro-reflux extremist (reflux is what gives money it’s “backing” value.. even so called “fiat” money in his view). But Sumner & Sproul do not agree either: I think Sproul would claim both inside and outside money can reflux (at least eventually in the case of outside money).

    Now let’s add JKH’s view to this spectrum. I’m not sure it fits into any of those camps. Especially when he brings up ““deep endogeneity” of the government fiscal position.”

  • Tom Brown

    BTW, regarding the “Widow’s Cruse” (not curse!), I liked JP’s take on this as well:

  • Tom Brown

    I don’t see why not, but of course that implies another bank is the buyer. They’d be purchased with bank deposits (by letting them be debited) otherwise (i.e. by non-bank buyers).

  • Tom Brown

    Well not really the “dividends and interest payments [on bank deposits]“… those are more just like just plain payments… lust like interest payments back to the banks in the opposite direction.

  • Geoff

    Tom, this may may not be where you’re coming from, but I think JKH is saying that the core liability of a bank is (are) deposits. Equity is not a core liability, but the bank CAN fund itself by issuing equity under certain circumstances if it makes financial cents.

  • Geoff

    If banks trade, say, stocks amongst themselves in exchange for reserves, is that considered HPE?

  • Tom Brown

    I’m not getting your question. HPE really boils down to an excess supply of money (or perhaps other money-like assets), right? That’s how I think of it.

  • Tom Brown

    Stocks I don’t think can cause the HPE effect since their nominal price can shrink… whereas the nominal price of $1 worth of currency or bank deposits cannot (or a $1 worth of short term Tsy debt for that matter… which CAN shrink, but perhaps not much, depending on circumstances). There’s no limit to how far the nominal value of a share of stock can decrease.

  • Geoff

    JKH, your paper is extremely informative, but there is no conclusion. What is the bottom line? Would it be safe to draw this conclusion:

    Yes, loans create deposits, and, yes, banks are special because they create MOE, but let’s not get carried away and draw crazy conclusions from these facts. At the end of the day, banks are still financial intermediaries that are subject to many of the same constraints to which other financial intermediaries are subject?

  • Detroit Dan

    What does “HPE” stand for?

  • Tom Brown

    “hot potato effect” … see JKH’s article above. It’s also something that he MMists postulate.

  • Geoff

    Tom, I’m thinking about HPE as it relates to the banks as opposed to the general public. If the banks have excess money (aka excess reserves), and those reserves can’t leave the banking system in aggregate (for the most part) so the banks trade stocks for reserves among themselves, affecting stock market values in the process?

    Of course, I guess the banks don’t really need excess reserves to trade stocks in the first place, as long as they have the capital.

  • Tom Brown

    “financial cents” ha!

    Sure, but he’s also saying that the banks would like to manage their “liquidity risk” by getting depositors to exchange them for less liquid assets (liabilities to the bank), e.g. time deposits (savings deposits & CDs, etc). And I suppose if they can get them to purchase sub-ordinated debt or equity in the bank they can increase capital too.

    If you look at the BSs that Ramanan posted (in my link) with the NBFIs competing w/ the banks… ultimately by taking deposits away from the banks and issuing shares instead… another way to look at that might be just to view the banks issuing shares directly.

  • Tom Brown

    Ah, OK, similar to Sumner’s case 7 (cashless). I guess I could see banks stock prices rising in that case.

    “I guess the banks don’t really need excess reserves to trade stocks in the first place, as long as they have the capital.”

    That’s a very interesting question/statement. So let’s say you’re looking at my typical Example #1 type scenario: A overdrafts, B is credited, A borrows reserves back from B to repay overdraft. The end result is A borrows from B… by why can’t we just skip directly to that result? I think in practical terms right now at least some reserves come into play to make the inter-bank transactions happen. It kind of takes the pressure off Bank B to trust Bank A in the 1st place.

  • Detroit Dan


  • LXDR1F7

    I think the main reason banks create most of the broad money supply is because people cant directly deposit and transact in base money like banks.

  • Nick Rowe

    Suppose mutual fund companies in aggregate want to expand the quantity of mutual funds they create. They must first find customers who want to hold more mutual funds.

    Suppose banks in aggregate want to expand the quantity of money they create. It is not true that they must first find customers who want to hold more money. All they need to do is find people who want to sell them something, and whether that something is a computer or an IOU doesn’t matter. The person who sold the bank the computer doesn’t want to hold the money — he wants to spend it on something else. The person who sold the bank an IOU (borrowed money from the bank) doesn’t want to hold that money — he wants to spend it on something else.

    You can’t create more mutual funds if people don’t want to hold more mutual funds. You can create more money even if people don’t want to hold more money. Because people know they can always pass on that unwanted money to someone else in exchange for something they do want, who in turn will pass it on to someone else in exchange for something he wants,…etc. Because money is the medium of exchange.

    That’s my main beef with Tobin.

    We should forget about understanding banks. We should concentrate on understanding money. If we understood money, understanding banks would be as easy (at least, no harder than understanding any financial intermediary).

  • jt26

    Hi Nick. Assume, we’re a credit economy and everyone has debt (i.e. a real economy without the top 2% ;-} ). You can’t create more money if people don’t want to hold more money. Because people know they can always pay off debt with it.

  • 1jump

    People always want more money.

  • Geoff

    People don’t always want more debt. I think you mean to say people always want more “free” money. In other words, they want more wealth. The Fed isn’t in that business, apart perhaps from capital gains.

  • Geoff

    GP, JT. The person who sold his IOU to the bank for a deposit will probably trade that deposit to someone else rather quickly, and around it will to. But the deposit should eventually get around to someone who uses it to pay off his loan, I would think.

  • 1jump

    Thats where the problem lies. Monetary policy needs to be separated from lending. People arent willing to enter further debt but are willing to have more money. The fact that people dont want more debt is resulting in lower AD, deteriorating balance sheets, lower credit worthiness etc…

    There is a point where people perceive they have too much debt then credit markets become disfunctional. The circuitbreaker is monetary policy that isnt debt based and is directly conducted with the public. Tne fed can easily create this mechanism.

  • Cullen Roche

    BUt Nick – Don’t you think it’s important that banks create the dominant form of payment today? If we lived in a world where there was just one bank that issued cash then I think you guys would argue that that bank matters, right? But we don’t. We live in a system designed around banks where banks issue the dominant form of payment. How can you say we should just “forget” that? It’s like living in a world where there is nothing but burritos to eat and saying that the burrito issuer doesn’t matter, but only the burritos matter. Well, don’t they go hand in hand! Shouldn’t we understand how both work???

  • Tom Brown

    I think I’ve got Nick’s argument down after reading several of his posts…. Ahem … so her goes, my Nick Rowe impersonation:

    If you have n goods one of which is “money” (bank deposits, cash, etc), then there are n-1 money markets. There’s just one apple market and gold market, but there are a LOT of money markets, so who’s to say that paying off loans is going to be the dominant money market? There are a lot of money markets out there than don’t involve anything a bank might sell.

    OK Nick, how’d I do? A reasonable Nick Rowe approximation?

  • Tom Brown

    I think he would agree with some of that Cullen:

    “Banks are special because money is special. If we used cows as money, then dairy farms would be special, because dairy farms would create money.” – N. Rowe

    Perhaps what he’s saying is not that banks aren’t important but that money (and it’s nature) is more important. Just a guess.

  • Tom Brown

    “Monetary policy needs to be separated from lending.”

    Some (perhaps most?) monetarists (think Scott Sumner) don’t spend any time discussing lending. Banks (and thus lending I suppose) are not important to them, but “monetary policy” sure is. Lol.

  • Nick Rowe

    Cullen: sure. But suppose the production of burritos was exactly the same as the production of burgers. But the marketing of burritos was totally different from the marketing of burgers, because people would buy burritos even if they didn’t want to eat a burrito, but planned to pass it on to someone else, who didn’t want to eat it either, like some chain letter (which is what money is). I would say economists need to focus a lot more on why burritos are different from burgers and all other foods, and a lot less on the production line.

    We need to focus a lot more on money, and a lot less on banks’ balance sheets.

    Someone could understand money without understanding banks. Nobody could understand banks without understanding money.

  • jt26

    Tom (or Nick or … argggh, you monetarist are always switching up! ;-} )
    Debt is not like other markets. Debt has a claim on my future income; iPhones do not.

  • Geoff

    Money is a chain letter? I’ve heard many different definitions of money, but that’s a new one. :)

  • Cullen Roche

    Hey Nick,

    Good thoughts. I agree there. But I do think it’s important that we understand how our monetary system is credit based. Loans create deposits and when someone wants money they usually have to take out a loan and work with a bank if they don’t already have access to the money. The overwhelming majority of transactions in our economy occur with credit. And understanding those transactions means we need to understand the entities that facilitate those transactions.

    I agree that it’s important to understand all forms of money. But I think some forms of money play a much more dominant role in our economy than other forms. For instance, Scott Sumner likes to focus on cash, but this is a flawed view of the world in my opinion. Cash exists almost entirely to facilitate the use of a bank depositor. The cash doesn’t even get out into the market unless someone draws down a bank account. Further, cash as a % of overall transactions is slowing substantially as we become an electronic economy dominated by credit transactions.

    So I agree we should understand all forms of money, but I think we need to better understand banks because they’re the issuers of the form of money that has come to dominate every day economic life.

    Do you disagree there?

  • Tom Brown

    I’m not a monetarist! I only play one on TV.

    Sorry, you’ll have to take further questions up with Nick… I’ve only memorized so many lines.

  • 1jump

    “But the marketing of burritos was totally different from the marketing of burgers, because people would buy burritos even if they didn’t want to eat a burrito, but planned to pass it on to someone else, who didn’t want to eat it either, like some chain letter”

    I think the idea of people not wanting money and passing it on is wrong. People want money but they may prefer other goods or financial assets over money or a balanced portfolio. Doesnt mean they dont want it.

  • jt26

    Lot’s of money-like things are special:
    (a) reserves are special because it is a safe asset that facilitates transactions in the banking system
    (b) bank MOE is special because it is a safe asset that facilitates transactions in the real economy
    (c) federal debt is special because it is a safe asset that facilitates:
    – shadow banking sector transactions
    – fluctuations of net savings desires of the private sector
    – international monetary system

  • Tom Brown

    I think that’s what Nick is getting at by saying that people are ALWAYS willing to accept money. He never said people aren’t willing to accept money.

    But people don’t usually take loans to do nothing with the money. They almost always take a loan to buys something else with. People always accept money because it’s so liquid.

    If somebody handed you $100k for free, what would you do with it? Keep it in your house? Deposit it in a checking account and leave it sit there earning 0.01% interest? I doubt it. You’d gladly accept the $100k, but then turn around and do something with it rather than just let it sit there. I think that’s all he’s getting at.

  • Cullen Roche

    Yeah, money is the only thing in the economy that gives you potential access to everything in the economy. You can’t always trade burritos for burgers. But if you have money you can always trade it for burritos OR burgers. That’s why people want money. And yes, I think Nick is obviously right that people want money for this purpose. But I don’t think that gets to the root of the discussion here as to why certain forms of money are more important than others and why some issuers of certain forms of money are important in understanding how the system works….

  • Tom Brown

    I.e. you probably don’t have a need for more than a few thousand $ to just sit there being liquid … to pay bills with, etc., or “just in case.” You probably have fewer $ still in physical cash stashed away in case of a natural disaster, etc. Unless you’re a drug dealer or tax evader… then you pretty much live off of $100 bills. :D

  • 1jump

    “We need to focus a lot more on money, and a lot less on banks’ balance sheets.”

    Most money is lent into circulation. Lending in heavily influenced by balance sheets of borrowers and lenders.

  • Tom Brown

    All these questions about the root nature of money are interesting to me… but I’d love to see an attempt to answer them using some kind of data or (maybe even better!) through some kind of “experiment.” That sounds like a really good question… How could an experiment be designed to test Nick’s hypothesis about the nature of money? Most of the arguments I come across seem to rely on thought experiments. Can’t we put some of those thought experiments to the test?

    Actually, I’m positive that somebody must have tried, right? I’m probably just ignorant of the literature.

  • Geoff

    Tom, haven’t you already done a type of experiment with your balance sheets? If all debt was repaid, you have proven that there would be no money left.

    I think this shows that money is determined by demand, endogenously.

    Money breathes.

  • jt26

    No. Nick wasn’t talking about accepting money, but rather “You can create more money even if people don’t want to hold more money.” For a single transaction that is true (e.g. 1 loan creates 1 deposit), but it might not be true for the whole private sector (e.g. deleveraging). Micro is not the same as macro … oh, now I sound like NIck! ;-}

  • Tom Brown

    My balance sheets don’t rise to level of experimental evidence I’m afraid… :D

    No, what I’m really suggesting be tested is this idea of reflux or lack of it. Rowe says that reflux doesn’t happen with MOE (bank deposits + cash + Fed deposits) to any great extent. I think Tobin in ’63 said it does happen with bank deposits.

    What do I mean by reflux? That money goes back to banks to buy bank products or repay loans. That’s what I *think* I mean! (Shoot, am I going to have to actually sit down and read Tobin now rather than wildly speculate based on what other people say. Oh well… I guess…) :)

    Check this paragraph out my JP Koning:

    “So what does it take to have a widow’s cruse? Two things. The liabilities of the issuer must be perpetual and non-convertible upon demand. Secondly, shops and markets must use those liabilities as a unit of account. Only when these two conditions will a widow’s cruse have emerged. Commercial banks pass the latter but fail the former. Stock-issuing non-financial corporations pass the former but fail the latter. Only modern central bank money is both. ”

    This ties into the “widow’s cruse” idea that JKH presents above. The “perpetual” part is one thing he says bank deposits fail at. I think this again has to do with this idea of “reflux.” Can we design a test to see exactly how true this assertion is?

    Here’s Nick’s comment and JP’s response to this:

  • Geoff

    Right. Money is primarily a medium of exchange. Therefore, most of us don’t need much. We store most of our wealth elsewhere, like in bonds, or stocks, or mint condition playboy magazines :)

  • Tom Brown

    Shoot… how many can possibly be in “mint condition?” … I’d think the vast majority would be well “used.” Ha! (I guess that’s what would make them so rare and valuable then, eh?)

  • Geoff

    Understood, Tom. Thx.

    BTW, the term “widow’s cruse” was killing me so I finally looked it up. I obviously don’t know my bible. It apparently refers to the widow’s cruse of oil that miraculously supplies Elijah during a famine (I Kings 17:8–16).

  • Tom Brown

    Check out the title and first couple sentences of this Rowe post:

    “Money is always and everywhere a hot potato [title]

    And the Law of Reflux is always and everywhere wrong. Unless the central bank decides to make it true. (The Law of Reflux says an excess supply of money will always revert to the issuing banks).”

    That second part:

    “And the Law of Reflux is always and everywhere wrong.”

    seems like it ought to be a testable hypothesis. Do we really have to rely on battling thought experiments to resolve this issue? That’s a hypothesis that the MMists will certainly not agree on, nor probably the PKEers nor anybody else.

  • Tom Brown

    Glasner and Koning discuss the painting JP used for an illustration and the various Biblical references:

  • Tom Brown

    That’s what I’ve found frustrating about pretty much the entirety of the long passionate MMist internecine battle over MOA vs MOE (and UOA) and “reflux” and “backing vs fiat” … it seems to all come down to these endless thought experiments. Well at least on the blogs. Why doesn’t someone like Sadowski come in with a pile of evidence to sort this out one way or the other? Maybe the data isn’t there… I don’t know.

  • Cullen Roche

    Can we call in a “full Sadowski” to resolve the safe asset shortage debate also? I’d like to know who’s right there and whether this thing is nearly as impactful as some people make it sound.

    How does one get a red phone for Sadowski? I’d like to be able to radio in a “full Sadowski” whenever possible. “Hey guys, we’ve got a rager in the comments section here. Can I get a full Sadowski over here ASAP?”

  • Tom Brown

    Lol, … I like to call it “threat level Sadowski” too… :D

    It’s kind of like putting up the bat signal night spotlight… just mentioning “Sadowski” tends to get a quick response.

  • Tom Brown

    Well he brings up the concept that banks (in aggregate) buy stuff by crediting bank deposits. He included computers and IOUs (loan agreements… i.e. making new loans) as examples. I agree… however, I do think there might be a difference between those two: Somebody will gladly sell banks computers by letting them credit their bank deposits… but it’s not clear to me that there’s always going to be *creditworthy* sellers of IOUs.

  • Cullen Roche

    Ha. Yeah, the guy’s so thorough I think I might just be bitter he’s on their side. Hence my prickish attitude with the accounting thing over the weekend. Plus, he was trying to Sadowski me in my own pad. Not cool, Robert Frost. Plus, we don’t have a Sadowski on our side. And that makes me feel very small. Oh well.

  • Tom Brown

    I hear ya! It’s be nice.

  • jt26

    Doesn’t San Diego have 3-4 military bases? There’s got to be a red phone somewhere!

  • JKH

    Most mutual funds are redeemable. There’s generally no requirement to hold a mutual fund share forever. And if somebody redeems a mutual fund share, there’s no distinction between that and somebody who wants to “get rid of” money – in the sense of exercising a choice for asset preference. More generally, lots of financial intermediary liabilities are tradable. Those who acquire them initially can get rid of them without reflux. Again, no distinction in the same way as above. And somebody who wants to get rid of money generally wants something for his money – not everything – which is a specific choice rather than a desperate exit from money at all costs – just like wanting money for your mutual fund share is a choice. The transaction is a swap of value in both cases. And the holding period is a continuum in the case of either money or anything else. So the holding period for money is not a binary distinction as compared to the holding period for anything else. And given that continuum, it’s definitely the case that you can’t create money without somebody wanting to hold it – just as it’s the case for other intermediary liabilities.

  • Tom Brown

    ” If all debt was repaid, you have proven that there would be no money left.”

    Not quite… I showed that the stock of private non-bank money (ignoring foreign & MBS etc) is:

    L + B + F

    So if L disappears you still have bank & Fed held Tsy debt (B+F).

  • Tom Brown

    JKH, was your comment here a response to Nick Rowe?

  • Tom Brown

    JKH, does what you are saying there overlap at all with Sumner’s statement here?:

    “However, in the short run asset prices adjust to prevent an excess supply of money.”

  • JKH

    It’s the swap of an asset of the borrower (the deposit) in exchange for an asset of lender (the loan).

    It’s a particular case of the more general idea of an asset swap.

    In this case, both assets are produced by the transaction itself.

    And each asset acquired in the exchange is the liability of the counterparty.

    But it’s still an asset swap.

  • JKH

    responded at monetary realism

  • JKH

    you’re right

    I didn’t clean it up with a proper conclusion – maybe because its a big picture topic with a conclusion still necessarily a work in progress – and/or because I ran over my time commitment on the piece

    but you’re statement combined with the following is heading in the right direction:

    “These papers along with the one discussed here constitute a consistent treatment of the banking system as a type of financial intermediary in competition with others, and with individual banks operating as active portfolio managers.”

    (final sentence of the post)

  • JKH


  • JKH

    “The person who sold the bank the computer doesn’t want to hold the money — he wants to spend it on something else. The person who sold the bank an IOU (borrowed money from the bank) doesn’t want to hold that money — he wants to spend it on something else.”

    I could say the same thing about a day trader at the Facebook IPO. He doesn’t want to hold the stock any more than the computer/IOU seller wants to hold the money.

    So nothing special about money insofar as that particular characteristic is concerned – i.e. the stuff about the holding period.

    Holding period is a specific outcome in a continuum of possibilities – for all assets.

    Money is special but not for that reason.

  • JKH

    I hope you’re being facetious.

    I’m OK with it if you are.

    I might become worried if you weren’t.


  • Mark A. Sadowski

    “Can we call in a “full Sadowski” to resolve the safe asset shortage debate also?…How does one get a red phone for Sadowski?”

    Historically the supply of U.S. private safe assets has been significantly larger than the stock of U.S. government safe assets.

    Consequently it is unlikely that the U.S. Treasury could create enough securities to fill the gap created by the shortage of private safe assets without undermining the safe asset status of Treasuries.

    Historically a sudden and permanent rise in the expected growth of NGDP leads to a rise in the supply of private safe assets and a decline of public safe assets.

    And in turn the current global shortage of safe assets is related to the shortfall in NGDP in much of the advanced world since 2008.

    The resolution to the safe asset shortage problem thus lies in monetary policy catalyzing the private sector into recovery.

  • Greg

    Hey Tom

    Actually I think the “handing you a 100k for free” is a terrible analogy for a loan. I would NOT necessarily spend a free 100,000$? Free meaning, no obligation to repay. I would only take a loan for 100,000 in oder to spend it however. No matter the interest rate. Why borrow 100,000 and have to pay back 110,000 in even 10 years if you are just going to leave it in cash. That would be stupid.

    Nicks stories often involve banks not being banks and borrowers not being borrowers.

  • Cullen Roche

    Ha. Good stuff.

    Do you actually have any historical data on safe assets and what would typically be considered a safe asset in the private sector? Do you use the SIFMA data to calculate something like that? If not, no big deal, but I would be curious if there’s a way to visualize how severe this problem is.

    Thanks Mark!

  • Tom Brown

    Sure, I didn’t mean it as an analogy for a loan. And I didn’t mean “spend it” necessarily: I just means you’d do *something* with it! Like buy bonds or stocks, etc. I presented it as a gift rather than a loan to emphasize that people generally just don’t leave $100k sitting there doing nothing. I share your skepticism about why that money wouldn’t “reflux” to pay back loans (had it originated from a loan at some point). I’m not convinced by the refluxers or the non-refluxers: both have thought experiment stories to back up their claims, but I don’t know how to tell which is true!

  • Mark A. Sadowski

    “Why doesn’t someone like Sadowski come in with a pile of evidence to sort this out one way or the other? Maybe the data isn’t there… I don’t know.”

    That’s the whole problem. It’s a theological question, kind of like “how many hot potatoes fit on a the head of a unit of account?”

    If I thought this question could be answered empirically you can be sure I’d search for the answer.

    In the meantime I just plan on watching the long twilight battle between Sumner and Rowe over the MOA vs. MOE question until the end of eternity, like the rest of you:

  • Tom Brown

    But aren’t there testable implications to each story? How about the reflux story? That’s a related concept… there’s really no way to tell if MOE is “no reflux” (Rowe) or “full reflux” (Sumner)?”

  • Tom Brown

    Well, I guess Sumner is probably pro-reflux for the bank deposit component of MOE, not necessarily ALL of MOE (i.e. cash too). That’s my interpretation in terms of the “reflux” debate, but I don’t know that I’ve ever seen Sumner address reflux head on.

  • Tom Brown

    … and to the extent they don’t just leave it sit there is the extent to which it’s an “excess supply” of money.

  • jt26

    Reply to Mark re:post below and links. There are several “safe asset shortage” problems mentioned in the MSM. One is the one your links describe, namely “shadow banks lend when they can issue assets perceived as safe by investors, benefiting the economy; thus shortage of collateral is a shortage of credit”. I don’t have a problem with that one. The one that bothers me and where there is little evidence is: “quality collateral is the high powered money for the shadow banking sector, and a shortage constrains lending like fractional banking”.

  • Nick Rowe

    Cars are an asset. There’s a demand for cars. There are two types of people who demand cars: drivers and dealers.

    Drivers buy cars because they want to drive them. Their holding periods may be short or long.

    Dealers buy cars only because they want to sell them again. They sell them to drivers. If a dealer could easily synchronise his purchases and sales perfectly, he would own a vanishingly small inventory of cars, for a vanishingly small holding period.

    If the drivers stopped demanding cars, the dealers would stop demanding cars too.

    Money is an asset. There’s a demand for money. But there are only dealers, and no drivers. Dealers buy money from other dealers, and sell money to other dealers. And we are all dealers in money.

    Unlike a car, (pure) money is totally useless if you can’t sell it again.

    The car dealer buys cars for money and sells cars for money, and hopes to get more money when he sells than when he buys.

    The money dealer buys money for apples and sells money for bananas, and hopes to enjoy eating the bananas more than he would have enjoyed eating the apples.

    Car dealers will only buy more cars from the car producers if they think they can persuade the car drivers to hold more cars. Money dealers know that if they buy more money from the money producers they can always sell it again to other money dealers.

  • Tom Brown

    JKH, you write:

    “Money is special but not for that reason.”

    Then what is the reason?

  • Tom Brown


    Doesn’t JKH’s point about stock fall outside the bounds of your car analogy? Nobody ultimately is an end user of the stock are they? Sure they possibly get a revenue stream from it… but not necessarily.

    Don’t we have to add a new player to your car analogy to encompass stock… like a shareholder in Avis perhaps… Lol.

  • Nick Rowe

    JKH: I can redeem a mutual fund share (open ended) at the shop that issued it. And I redeem it for money.

    I can redeem money in any shop in the country, not just the shop that issued it. And I can redeem it for anything. Only if I redeem it at the shop that issued it will it go out of existence. But why would i want to do that? And if the shop that issued it wanted to expand, they will immediately put it back into existence again.

  • Nick Rowe

    Tom: if a stock didn’t pay dividends, ever, and would never be redeemed by the company that issued it, would anybody ever buy it?

    Plus people own stocks that pay no dividends because they expect it to appreciate. People own Bank of Canada money despite expecting it to depreciate at 2% per year.

  • Tom Brown

    Nick, is there a testable implication of your view of the specialness of money (including bank money) which could be investigated to compare it with JKH’s view or Sumner’s view or Mike Sproul’s view?

  • Cullen Roche

    I think this gets way too messy. It’s easier just to stick to traditional definitions rather than trying to reinvent the wheel here. Stocks are savings vehicles much like bonds are. They’re securities issued by corporations that give its owner a claim on some cash flow. That cash flow may or may not protect you against inflation. This is a money-like instrument in much the same way that a savings account is. But it’s not money in the same sense that a cash bill or a bank deposit is because ultimately you can’t walk into a Wal-Mart and buy groceries or guns or whatever people buy at Wal-Mart. Nick, do you even know what a gun is? I don’t know if those things exist in Canada? :-)

    Anyhow, I think money is important, but I don’t know if I like thinking of money in terms that make it just merely another good like a car or a mutual fund share. They’re very different things that serve very different purposes. And they have very specific definitions in economics that I don’t think need to be totally reworked in order to understand their purpose or importance.

    But I am open to being convinced otherwise!

  • Tom Brown
  • Tom Brown
  • Tom Brown

    The difference being that woman is in the Canadian military… in the US she’d just be another satisfied Wal-Mart customer… pleased with a great deal on a back-to-school sale item!

  • Geoff

    Nice gun.

  • Mark A. Sadowski

    “Do you actually have any historical data on safe assets and what would typically be considered a safe asset in the private sector?”

    See the Appendix of the Gary Gorton et al. paper to which I linked. It lists what they included in their safe asset estimates by flow of funds identifier. I have not yet tried to construct their measures.

  • Mark A. Sadowski

    “Doesn’t San Diego have 3-4 military bases?”

    Wrong coast.

    “The one that bothers me and where there is little evidence is: “quality collateral is the high powered money for the shadow banking sector, and a shortage constrains lending like fractional banking”.”

    This is the first I have heard of this problem.

  • Geoff

    I’m going to guess for the same reason Cullen says. Money is special because it is what we all use as a means of final payment, at the point of sale. Good luck trying to buy groceries with t-bonds or mutual fund units.

  • Tom Brown

    Right I get that, but I’m trying to tie two worlds together here (if possible!).

    Nick’s view seems to be bank deposits are special because they are MOE, and MOE doesn’t reflux back to the bank necessarily: banks in aggregate can create an excess supply of MOE which then hot potatoes. Thus banks are important.

    Krugman, Sumner or “the 1st” Tobin (as Nick would say) claim that bank deposit MOE doesn’t hot potato because an excess supply CANNOT be created by banks because the law of reflux applies (i.e. excess supplies are just sent back to the bank), thus banks are NOT special. There cannot be an excess supply of bank deposit MOE.

    So now I’m trying to fit JKH’s position into the spectrum: he seems to endorse Tobin #1 on reflux, but… I don’t think he’s where Krugman or Sumner are either. Is he in between somewhere (partial reflux)… or is he saying that there’s a continuum of money like instruments (stocks, bonds, etc) thus there’s a number of assets with different holding period attributes… all of which perhaps have some limited reflux capacity. If the latter, how does that relate to the former. I’m having a hard time fitting it all together!

  • Mark A. Sadowski

    “That’s my interpretation in terms of the “reflux” debate, but I don’t know that I’ve ever seen Sumner address reflux head on.”

    I’m pretty sure Sumner’s never discussed The Law of Reflux in a post. But based on comments he’s written in Rowe’s posts addressing the Law of Reflux I was under the impression that Sumner largely agrees with Rowe (i.e. that the Law of Reflux is only true if the central bank lets it be true).

    In any case I’ve never seen the Law of Reflux brought up in the context of the MOA/MOE debate. Moreover I’m not sure why you think it is relevant.

  • Geoff

    Thanks for the clarification. Jeez, I can’t keep everyone’s views straight. Time to build a spreadsheet matrix.

    Or maybe a reflux capacitor. :)

  • Cullen Roche

    San Diego is actually loaded with bases. I have a lot of friends in the service out here. Most people think it’s just one big Navy base because it’s the homeport of the pacific fleet, but there’s a lot more here than just the base and the Seals. There’s the Marine Corps Air station ar Miramar where they do the big shows every year, the Marine base at camp pendleton up north, the recruit depot, the Coronado base & home to the Navy Seals, point loma base and then the Naval Base downtown which is the one most people know about.

    Still I wouldn’t call on them for help here. I’d much prefer the Full Sadowski. :-)

  • Cullen Roche

    Apparently cocaine also isn’t money. Who knew? Good thing to keep in mind.–225022842.html

  • Tom Brown

    Mark, I may very well not be understanding “reflux” correctly, but here’s where I think some daylight opens up between Nick and Scott: Nick thinks the law of reflux does not apply to bank deposits and Scott does. That’s my interpretation. I’ll point you to some Rowe articles which I think support this concept:

    “The quantity of medium of exchange is supply-determined in a way that is not true of other goods.”

    Notice the MOE (bank deposit version) is specifically mentioned here by Rowe, which Sumner does not think is important. But how could it be “supply-determined”… well I think it’s only if the law of reflux doesn’t apply to it.

    Here’s anther one:

    “Scott: “Unlike the Tobin view, I don’t think Glasner’s view differs in terms of any testable implications.” – SS

    This is something I plan to reflect on more, but I think it does have testable implications. For example, whether and under what conditions an increase in the money supply could be effective even if there’s no fall in bond yields.” – NR

    I’m pretty sure Glasner and Rowe don’t agree on HPE for MOE either.

  • Geoff

    A cocaine monetary system would probably involve a lot of reflux up the nose.

  • Cullen Roche

    I have no idea what you’re talking about. :-)

  • Tom Brown

    Here’s another example:

    “But I thought I understood David’s view to be that central bank money can be in excess supply or demand (it can hot potato) but that commercial bank money can’t. I’m saying the same thing can happen to both inside and outside money.” – NR

    But in that case Scott thinks he *might* agree with Rowe (in a later comment) against David.

    So I think there’s a tie in between “reflux,” MOE/MOA, and HPE.

    Also have a look at this Glasner post (in which he asserts Scott agrees with him) and in which he mentions the law of reflux, MOA, MOE, etc, starting with this paragraph in particular:

    “But not everyone agrees with this view of how the quantity of inside money is determined. There are those (like Scott and me)…”

  • Nick Rowe

    Cullen: (Scott focuses on central bank money because central bank money is the medium of account, and Scott thinks the MOA is important, because demand and supply of the MOA determine the price level.)

    If we want to understand banks in the real world, we need to imagine an unreal world in which (say) banks buy houses, and rent out those houses, instead of buying mortgages so other people can buy houses. Imagine a world in which banks have houses on the asset side of their balance sheets, and earn rental income, instead of interest income.

    In that imaginary world, banks would still create money. They would create money every time they bought a house, and destroy money every time they sold a house.

    That thought-experiment teaches us that banks’ ability to create money is totally independent of their ability to lend.

    And it lets us see clearly the difference between the demand and supply of money, and the demand and supply of whatever it is on the asset side of banks’ balance sheets. The demand and supply of money is obviously different from the demand and supply of houses. (Too many people get horribly confused between the demand and supply of money and the demand and supply of bank loans.)

    The fundamental/philosophical difference between you and me, Cullen, is that you want to think in terms of monetary realism and I don’t. I think we can only understand the real world if we ask counterfactual questions: what else would be different if one part of the real world were different?

  • Nick Rowe

    Plus, in that imaginary world, all the people whose slogan is “Loans create deposits!” would instead be saying “Buying houses creates deposits!” which would be rather amusing.

  • LVG

    Why do economists insist on thinking of the world in this abstract/fake way? How does that help us understand reality? Counterfactual questions are interesting for theoretical purposes, but at some point we have to tie it all back into the world as it actually exists. I think that’s what Cullen is very good at – staying grounded in reality and not making things up. That seems to be the biggest strength of all the MR guys. They just understand what IS and not what some people want there to be.

  • Tom Brown
  • Tom Brown


    I think JP Koning did a good job summarizing the HPE/Reflux debate here in terms of the position of Rowe, Glasner, and Sproul (all different). I’m guessing that Sumner is closest to Glasner on this from what I’ve read… although I admit it’s not totally clear.

  • JKH

    Leaving aside the car analogy, we’re all contingent dealers of financial assets in general.

    And financial assets can be characterized along a liquidity continuum for that purpose.

    (Not many of us buy a stock and lock in a note to self never to sell it under any circumstances.)

    So the holding period for just about any financial asset is variable and uncertain on that basis.

    Again, money is special, but I don’t think this is the characteristic that makes it special in a binary way.

  • Nick Rowe

    LVG: because we want to answer questions like: would the recent recession have been less severe than it was in the real world if monetary policy had been different than what it was in the real world? Real world people care about the answers to questions like that, because they want better policies for the real world future.

  • JKH

    Money is special because its the medium of exchange.

  • Tom Brown

    Nick a couple of questions:

    1. You write “central bank money is the medium of account.” Is that your view or just Scott’s? Given this statement:

    I thought perhaps you had adopted Koning’s view instead.

    2. Can you expand on this bit a little more?:

    “And it lets us see clearly the difference between the demand and supply of money, and the demand and supply of whatever it is on the asset side of banks’ balance sheets. The demand and supply of money is obviously different from the demand and supply of houses.”

    I’m not sure I see what you’re getting at there. I know you’ve written this before:

    “The quantity of medium of exchange is supply-determined in a way that is not true of other goods.”

    You’ve used that to argue that the banks in aggregate can create an excess supply of MOE I think. I still have a hard time seeing that since if the aggregate banks attempt to buy to many houses, won’t the price raise and make it uneconomical for them to do so? Or are you saying that all else being equal they can create their own rising inflation in prices (starting w/ houses and eventually spreading to all other goods)? So if one bank (for whatever reason) decides to overpay for a house, this creates a positive feedback loop which may induce other banks to start overpaying for houses? Something like that? But then of course rents would have to rise too right. Am I on the right track here? Thanks.

  • JKH

    I think the issue here is whether you can “get rid of” financial assets in a similar way that you can get rid of money.

    And it seems to me that most often you can.

    Whether that’s by redeeming it or by trading it is a secondary issue IMO relative to this idea of “getting rid” of it.

    And most mutual funds are just structured to be non-tradable, so redemption is the way to get rid of shares.

    Talking here about a continuum of liquidity really in the context of “getting rid of” – not other properties of money.

  • JKH

    And who says I need to get rid of my money any faster than the BMO stock I bought a few months ago?

    You should see my bank account. You’d be appalled.

    I’m a walking liquidity trap.

  • JKH

    I think the whole issue of “hot potato” and “reflux” needs to be treated with much more flexibility and on an empirical case by case basis. Asset “behavior” tends to be mixed rather than binary in this regard.

    I found it interesting that Tobin discussed reflux through government surpluses implicitly in the context of NFA rather than the medium of exchange – which happens to be how MMT converts the velocity equation.

  • Tom Brown

    Shoot… that was one episode I never saw.

  • jt26

    Hi Mark: “This is the first I have heard of this problem.”
    See e.g.

  • Mark A. Sadowski

    “See e.g”

    OK, I stand corrected. This is not the first time I have seen this mentioned.

    1) I agree with David Beckworth that the original “collateral shortage in repo markets” meme ignored the fact that QE would catalyze the creation of more safe assets than it would remove. So this was always a non-problem problem.

    2) I agree with Nick Edmonds that draining reserves through a reverse repo program available to non-banks is better than a bank only term deposit facility because, other things being equal, more liquidity is a good thing.But it’s a mistake to confuse the impact this will have on the repo market with the impact this will have on the real economy.

    3) Worrying about the exit-path implications for collateral chains is very likely a monumental waste of time since all indications are that the Fed plans on implementing a fixed rate, full allotment, reverse repo anyway.

  • Cullen Roche

    ” Worrying about the exit-path implications for collateral chains is very likely a monumental waste of time since all indications are that the Fed plans on implementing a fixed rate, full allotment, reverse repo anyway.”


  • Mark A. Sadowski

    Tom Brown,
    “I’m guessing that Sumner is closest to Glasner on this from what I’ve read… although I admit it’s not totally clear.”

    I looked at all your links. With the sole exception of the JP Koning post I’ve read all of these before. You’re to be commended for your efforts at summarizing the various debates.

    But it’s still not completely clear to me where everyone stands on all of these issues. More importantly I still don’t see any testable implications.

  • Tom Brown

    Mark, if there truly are not any testable implications, one reason might be that the differences in positions on these reflux/HPE/MOA-vs-MOE issues don’t actually matter. That’s not necessarily the case, but that’s one possibility. Is that your view?

  • Geoff

    This has been an excellent discussion, much of which frankly flew over my head. But this is my layman’s takeaway:

    1. Bank Deposits are special because they are the primary Medium of Exchange (MOE) in today’s system.
    2. Banks are special because they are the monopoly supplier of Bank Deposits.
    3. The amount of Bank Deposits in the system (i.e. the money supply) fluctuates mainly endogenously based on the interactions of lenders and borrowers.
    4. The price of money (i.e. interest rates) will influence the actions of both lenders and borrowers to some degree. To the extent that interest rates are determined exogenously, one could say that there is some exogenous influence on the money supply.
    5. So perhaps there is both Reflux and HPE, with the Reflux (aka endogeneity) dominating, IMO.

  • Mark A. Sadowski

    Tom Brown,
    “Is that your view?”

    Well, to be honest, I’ve never really understood why I should even care about the MOA/MOE fracus.

    Maybe if somebody involved in it stopped debating it long enough to explain why it really matters, then I would be more concerned about its resolution, and might even take a position.

  • salsabob

    Still looking for a definition/explanation of “deep endogeneity.” Could it have something to do with the private banking system having no choice but to create deposits when when federal deficit spending demands it? That customer makes them a little less special, no?

  • Cullen Roche

    The banking system does not create deposits when deficit spending demands it. Govt spending is a redistribution of existing bank deposits, not the creation of new bank deposits..

  • JKH
  • Geoff

    Bob, I think it was in the context of NFA’s, not deposits. Are NFAs endogenous? In other words, is the govt deficit endogenous? As JKH said, the part of the deficit that ebbs and flows in line with the economy (like the automatic stabilizers) could be called endogenous.

  • jt26

    re:Mark A. Sadowski says:
    09/25/2013 at 2:19 PM

    Beckworth is saying the ‘high power money’ collateral is important, but the shortage is probably overblown. I’m still interested in whether the first assertion is true. Heh, I’m getting only a Quarter Sadowski here. ;-}
    I just want to know how the monetary system works!

    BTW related to FRFSRRepoF: with the shortage meme, I’ve often wondered why the Fed’s securities lending facility has not been strained, but maybe that’s more of a problem with the Primary Dealer-to-Shadow-Banking interface.

  • Cullen Roche

    I see you’re dropping a big bag of full Sadowski on Phillippe over at TMI. He’s been known to troll these parts also. You’re better off arguing with a wall though. He won’t budge.

    Plus, MR doesn’t say deficits are permanently stimulative. You have to view it in the context of the entire economy and the other sectors. Just so you know….For instance, right now, I don’t think the deficit is really austere or stimulative. In fact, I think it’s more representative of real private sector growth in that its mostly due to tax receipts increasing. So there’s a net financial asset contribution, but as you’ve noted, there’s a cyclical slowing in spending. I wouldn’t call it austere, but I also wouldn’t call it hugely stimulative.

    If you ever get really really bored you might read my MMT critique. That way you won’t confuse us with them, which is common….

  • Tom Brown

    Cullen, you might find this interesting:

    I’d never really understood that before. I’d heard the pieces: monetary preferable to fiscal (because it doesn’t involve more debt), monetary offset cancels out fiscal (unless CB incompetent… then fiscal might work), etc. But I’d never put it together into a coherent whole. Again, not saying I agree, but at least I can see the logic of it now in a rough sketch anyway. Maybe you’d already understood the reasoning there.

  • Cullen Roche

    Two things I don’t get:

    1) Why does monetary policy have to change in order for fiscal to work?

    2) The MM guys don’t seem to understand that fiscal policy is really just a form of monetary base expansion via the US Tsy. If the US Tsy stopped paying interest on bonds and said all us govt issued notes now have overnight duration then they’d be issuing something is a near perfect equivalent of the monetary base. It would be clean money printing as someone like Friedman might think of it. So it’s illogical to claim that QE does something and fiscal policy doesn’t when all fiscal policy really is is the issuance of a net financial asset, often in the form of very short duration bills. The main difference of course, is that the T-bond issuance actually changes the net worth of the pvt sector while the QE issuance just swaps assets. So I don’t get it. Is the Chuck Norris effect really more powerful than the asset printing effect????

  • Tom Brown

    Cullen, I’m certainly not the one to explain it… however I came to the conclusion I did about MM ideas after reading this Sadowski comment in the same Sumner post:

    Especially the Tyler Cowen chart he provides in the link, and his description of what’s going on there. I think that’s the justification. That’s their answer for you question 1.

    Regarding your point 2, I tend to agree, but the MMists look at it different. Keep in mind that I don’t think any of them really like QE. Again w/ my MM hat on, perhaps they’d liken it to a really loud unintelligible yell from the CB: it’s gets people’s attention, and it perhaps conveys some information, but it doesn’t do a great job of it. Ideally QE would not be necessary and clear messaging from the Fed could do most of the job better: so instead of a loud unintelligible yell, the Fed simply would speak in a clear unambiguous way what exactly their intentions are. I have to say I do find this part of their argument somewhat compelling: I don’t think the Fed is doing a great job communicating and perhaps it’s trying to compensate for lack of clarity by how loud it’s yelling (to torture the metaphor). I’m going to anticipate your response: They’re saying Chuck Norris, and you say Bruce Lee… and I think there’s a good argument to be made there, but they clearly think Chuck Norris wins this fight and he doesn’t have to yell: Sumner has said himself that QE is a mess and ideally, even at the ZLB, base money is endogenous. It all comes down to the message and the credibility of the message, I think they’d say.

    So, like I say, given that the MMs see Chuck dominating Bruce, and everything is really about expectations… I can see the logic of this argument. Whether it’s true or not is a separate issue!

    Also, they don’t like fiscal in comparison to credible MM because fiscal increases the debt which they see as unnecessary.

    So yes, I agree there’s a lot of assumptions about what’s true there… I agree! But it has an internal consistency anyway. And if it WAS true, that would be neat trick! No doubt about that! If Chuck could just look menacing and steer NGDPLT w/o even having to resort to (much) QE and w/o any fiscal stimulus… well that would be great I think! It’s a nice theory, but is it true? :D

  • Tom Brown

    Overall I think the Fed could do a better job of communicating its intentions better. Whether that’s sufficient… I’m skeptical of that. Also, I’m not convinced that fiscal can’t be a useful tool in the overall bag of tricks… but it probably is most effective with what I’d call a coordinated, complementary and clear monetary policy (if legally possible). I’m also attracted to the idea of getting monetary to do as much of the heavy lifting as possible, again for the debt reason that they point out.

    So do I see a case where monetary can do a large part of the heavy lifting w/o actually doing much lifting (i.e. w/o massive monthly asset purchases) but instead with clearly expressed and defined goals in terms of expected targets? (either inflation or NGDP)? Yes that seems like it might be plausible… but I’d also say it’s plausible that the optimal way to approach this task is with a complementary and coordinated fiscal policy. That seems plausible to me too. I’m sure Mark can present of lot of evidence to favor (what appear to me at least) one plausible concept over another. But at least I think I’m a little closer to getting inside the head of an MMist w/o thinking they are crazy. :D

    … however I remain a skeptic. It’d be nice if some country were to adopt the MMist ideas in their entirety (no half measures!) and put them to the test so we could see how they work. When I asked Sumner what nation on Earth comes closest, he said Australia (but perhaps that was a cherry picked example… I don’t know!… if someone were to ask me which country were following my advice I’d be tempted to pick the one that’s doing the best… not that I even know how Australia stacks up in that regard… because I don’t!).

  • Cullen Roche

    SOOO many assumptions in part 1. Neutral money, general equilibrium, etc. sigh.

    As for part 2 – who wants Chuck OR Bruce when you could have both? They seem totally closed minded to using both. I don’t get it. Like you, I find parts of the thinking compelling. But they don’t seem to see anything compelling on our side which just stinks of politics and bias. That doesn’t sit right with me.

  • Tom Brown

    I wonder if the Fed is afraid to be too clear about what they are attempting to accomplish for fear of failing and undermining their credibility. What do you think? Because it’s not terribly clear is it? They aren’t hitting their inflation target (from what I understand) and they haven’t raised the inflation target. They have established some objectives regarding the unemployment rate… but they don’t have a clear schedule for any of this really do they? They seem to be leaving themselves lots of outs… in case they fail. It’s understandable. It’s like telling the guy who’s going to dive off the cliffs in Mexico for the 1st time “It’s all about the power of positive thinking! Believe you’re going to time the tide correctly and you will!! Don’t let negative thoughts creep into your head! (i.e. the surf washing over your broken & mutilated body on the rocks below)”

    (as you can see, I’d be terrible at it!)

  • salsabob

    It would seem then that “deep endogeneity” is the lucky coincident that growth in bank deposits (from endogenous bank activity) and portfolio preference covers the approximately $1T in annual deficit spending. Do we have any evidence of that? I assume there’s some preference to hold some of those bank deposits so the growth in bank deposits would have to exceed the annual federal deficit, no?

    If that is not the case, would the banks have to establish new bank deposits (in the Treasury’s account), hold the Ts themselves until a depositor came along or the Fed. Reserve? That would seem more analogous to what they do for all other customers. That would also seem to broaden what “deep endogeneity” is, no?

  • Tom Brown

    salsabob, I’m not sure I understand your question. I don’t think that’s what JKH meant by “deep endogeneity”… I think he just means that in contrast to an “exogenous” government budget, and budget that is has “deep endogeneity” may have built in responses to economic conditions, for example, in a recession less taxes come in as people lose their jobs, but at the same time more foodstamps and unemployment checks go out. It’s an “endogenous” change in the budget due to “automatic stabilizers.”

    That’s one possibility… but perhaps he was getting at something else. Really though I think he means that the budget isn’t determined in a vacuum… decisions are made or automatic budgets adjustments happen because it’s all in a big feedback loop: budgets are determined by what’s happening in the rest of the economy. That’s my take.

    I don’t follow your logic with the $1T. You might take a look at this:

    It might help to clear up some misconceptions.