Tuesday, March 5, 2013

Line in the sand

Over at the Free Banking blog, Kurt Schuler has two good posts on where to draw the line between money and other assets. While Schuler likes to differentiate between the monetary base (deposits at the central bank + currency) and other assets, he points out that there are a number of other popular spots to scratch out a line. The monetarists, for instance, settled on M2. I seem to recall that in America's Great Depression, Murray Rothbard let the cash surrender value of life insurance policies slip over the line into money supply territory. There are a thousand-and-one places to draw the line.

Schuler notes that rather than drawing a sharp line between money and other assets, one can also recognize a spectrum of "moneyness." Anyone who's read this blog knows that I'm amenable to this idea. Before we can ask where do we draw the line? we need to ask how do we draw the line?. Either treat money as a set of distinct goods, or treat each good as more or less money-like. By money-like, I'm referring to a good's role as a medium-of-exchange, not its role as a store-of-value or unit-of-account.

I've tried to convey these two approaches in the graphic below. I don't think either approach is better than the other, but we should be consistent. Depending which one you choose, you'll probably be able to see the economy from a different perspective, and different perspectives can be helpful.

Rothbard and the monetarists took the first approach. Between which discrete goods should the line be drawn? Should bank deposits make it past the line or not? How about shares?

Rather than placing a line between discrete groups of goods, the second approach draws a unique line across each individual good. This line demarcates each good's monetary qualities from its non-monetary qualities. All goods are money, but some are more money-like than others. Cattle (i.e. commodities), for instance, are simultaneously capital/consumption goods while also having monetary properties. Even beer (consumer goods) has a degree of moneyness since specialized producers, wholesalers, and retailers hold bottles in inventory for the purposes of resale.

In his discussion of moneyness, Schuler invokes the same classic 1956 W.H. Hutt paper that I've mentioned before. According to Hutt, money throws off a constant stream of services, the essence of which is availability. Just as an unused fire extinguisher provides its owner with constant comfort, the availability of money in one's wallet provides a steady flow of relief. Hutt described this idea as the yield from money held. But Hutt's is still an expression of absolute money, not moneyness, for his choice of words implies that only money-proper yields availability services and all other goods be damned. I find it useful to convert Hutt's expression from one of absolutes to one of degrees by rephrasing it as the money-yield from goods held. Beer, cattle, houses, stocks, banknotes, and bank deposits all throw off availability services, though the size of this stream varies according to each good's marketability, or liquidity. Because this service is valuable, a premium gets built into the price of a given good, or a liquidity premium.

Continuing his discussion of moneyness, Schuler also brings up the Divisia index, a technique of aggregating monetary assets championed by William Barnett. Rather than simply summing up quantities of so-called money, a Divisia index is a weighted monetary index. A component's contribution to the index is determined (in part) by its degree of monetary usefulness. How are monetary services computed? A liquid asset's interest rate is compared to the rate yielded by an illiquid zero risk benchmark bond. The more interest that is foregone in holding the given liquid asset implies that larger monetary services are being offered to compensate the asset holder. In other words, the lower its interest rate, the more money-like the asset, and the larger a component's contribution to the Divisia index.

This is a fascinating approach. Theoretically, I'd go even further than Barnett in extending the continuum of moneyness beyond financial assets to stocks, houses, cows, and beer. Here the computations get difficult. Barnett uses market-determined interest rates from debt markets to determine each Divisia component's monetary services. But the return from a stock comes primarily in the form of expected capital appreciation, not interest, so teasing out a stock's monetary services by comparing it to some illiquid interest-yielding benchmark bond would probably prove to be difficult. The same goes for houses, and it gets harder to compute moneyness the further we wade into markets for commodities and goods.

Even if we arrive at some aggregate amount of money services, I'm not yet sure what it would be useful for, and for whom. In thinking about degrees of moneyness, my preferred application would be  liquidity spreads rather than liquidity aggregates. Being able to see the risk premium of a given asset via credit default swaps is certainly useful, at least to financial market participants. One would imagine that knowing an asset's liquidity premium—how that premium fluctuates over time and how it compares to other assets' premia—would be just as helpful as knowing its risk premium.


  1. Regarding the monetarists, let's be clear that they were interested in money as the unit of account, and saw the central bank's job as managing demand for the unit of account. This was confused by the fact that old monetarists thought V was stable in NGDP=M*V, so adjusting the quantity of M2 was (almost) equivalent to adjusting demand for the unit of account. More recently (and now that it's obvious that V is highly variable), market monetarists have been more clear on this issue, that demand for the unit of account is what matters (Scott Sumner post here: http://www.themoneyillusion.com/?p=17368). For market monetarists, the value of the unit of account is what's important because nearly all contracts are drawn up in it, so when it changes the terms of all contracts change (in real terms).

    Units of account are really dichotomous: IBM stock is *in no way* used as a unit of account, outside perhaps employees discussing stock options. And of course, in the United States legal tender laws enforce the roll of USD as the only contractually enforceable unit of account.

    However, almost anything can be used as a means of exchange. Some things are more useful than others, which is where thinks acquire different degrees of money-ness. People can and do trade beers, cows, and musical performances without any involvement of federal reserve notes.

    1. Scott Sumner's definitions are always and everywhere sloppy. Firstly, Scott doesn't put the focus on unit of account, he puts it on medium of account. Secondly, he's using the wrong definition for MOA.

      See these two discussions.

      A Federal Reserve $100 bill is a medium of exchange, not a unit of account. The unit of the account is just a word or symbol; dollar/$. The medium of account defines that word; today it is CPI. In times past it was grains of gold. You can't "demand" the unit of account, just like you can't demand "metres" or "yards".

      Sticker prices are set in terms of some unit of account, not the medium of account. No one posts prices in terms of CPI, they set them in $. Payments are made using various media-of-exchange, one of which is dollar bills. Another is IBM shares. There is a demand for the various media of exchange, not a demand for the unit of account.

      "Regarding the monetarists, let's be clear that they were interested in money as the unit of account, and saw the central bank's job as managing demand for the unit of account."

      You shouldn't take Sumner as the be-all end-all of monetarism. David Laidler was a strong advocate of M2, for instance. See Essays on Money and Inflation.

      "And of course, in the United States legal tender laws enforce the roll of USD as the only contractually enforceable unit of account."

      Legal tender laws are largely meaningless. Firstly, they don't enforce a unit-of-account. Americans can invoice in any currency they wish. Rather, legal tender laws enforce what sort of media of exchange can be legally used to discharge any debt, unless debtor and creditor have agreed differently. See my posts on legal tender.

      "IBM stock is *in no way* used as a unit of account, outside perhaps employees discussing stock options."

      Don't forget corporate acquisitions. The quantity of stock used as media of exchange is far more than most people think.

    2. This medium of account idea is certainly an interesting way of looking at it. Nevertheless I think if you rephrase my post as arguing that the medium of account is fundamentally different than other media of exchange, the point stands. If the vast majority of contracts in a country are made out in a particular unit of account, with payment expected in a particular medium of account, it would seem that that medium of account is qualitatively different from other media of exchange.

      In some sense the biggest way in which it's fundamentally different from other assets is simply that it is the last part of the economy that remains centrally planned. Its price doesn't get magically set properly by markets like all the other assets do: somebody has to actually decide what its value will be.

    3. Let me see if I can dehomogenize myself from Scott. I'm going to avoid all use of the word "money" to make things easier.

      Scott: Fed liabilities (currency + deposits) are different from private banking liabilities because latter are basically an option to provide the former. Private banking liabilities are derivatives, so to say, and therefore their market value is derived from the market value of Fed liabilities. Thus there is something that is "primary" about Fed cash and deposits. How the Fed chooses to set the value of cash and deposits is very important since it affects the value of all derivatives of Fed cash/deposits ie. banking deposits.

      I've got no problem with the above. Note that I've just rewritten this quote from Sumner:

      "I don’t care if currency is only 1% of all financial assets. Give me control of the stock of currency, and can drive the nominal economy and also impact the business cycle."

      While that is interesting, it doesn't deal with liquidity/media of exchange/moneyness. The latter is the subject that interests me. How do Fed liabilities compare to private banking liabilities on the liquidity axis? How would we measure their different liquidities? What benefits accrue to those who issue liabilities that are more liquid than their competitors? What would happen if the moneyness of Fed liabilities was to decline?

      Scott wants to drive a line between what is the unit-of-account and what isn't. Fine. On the entirely different subject of media of exchange, I'm saying we don't have to draw a line between what is "the" medium of exchange and what isn't.

  2. Any thoughts on the massive surge in BTC lately? Obv it's related to the big news of late (Silicon Valley Bank, Namecheap), but do you think it's sustainable?

    1. Perhaps it's reached the point that Selgin used to describe fiat money: once it reaches escape velocity, it can circulate on its own w/o any fundamental value.

    2. John S, my thoughts on bitcoin remain the same. See here. When I sell my bitcoins, I'll be sure to write a short post.

      I should note that we call fiat money - central bank notes and deposits - does have a "fundamental value". See here.

  3. "Before we can ask where do we draw the line? we need to ask how do we draw the line"

    Related question: WHY draw the line? Rather than asking "What is money?" we might ask "What is a bond?" You'd get another broad spectrum of things with various degrees of "bondness". By then, hopefully, you'd realize that it doesn't matter what we call a bond. What matters is: Whose liability is that bond? There are very clear lines between bonds issued by the US, by Canada, by GM, by New York, etc. The value of all those bonds is clearly determined by the assets/liabilities of their issuers, and not by how many things out there can qualify for some arbitrary definition of bonds.

    1. Mike, that's a good question.

      Isn't the point of monetary economics to inquire into the monetary nature of things?

      I don't disagree with anything you've said about the nature of assets and liabilities. In fact, you're the one who alerted me to the non-fiat nature of so-called fiat paper. That was certainly an eye-opener.

      But your approach reduces everything to pure Miller/Modigliani. I'd go so far as to say that it destroys monetary economics as a discipline. Who knows, maybe that's a good thing.

      You make the point that money:money what bond:bondness, your point being that bondness is a silly category and therefore so is moneyness. Yes, and goldness or pizzaness would also be odd, I agree. The reason I always draw attention to moneyness (let's just call it liquidity) is because liquidity is a universal property of all tradeable things. I see it all around me, both at work and at play.

      I like to think about how to go about creating a Miller-Modigliani monetary world rather than just a Miller-Modigliani non-monetary world. I'm pretty sure the best way to go about it is via moneyness and liquidity premia, not via absolute money and not-money.

    2. When people start holding silver as money, silver gets a premium. You could call it a moneyness premium. That premium is limited by competition from other money-like goods like gold, copper, tobacco, etc, and it is also limited by competition from paper and electronic IOU's denominated in (but usually not backed by) silver. Those IOU's can potentially reduce peoples' monetary holdings of silver to zero, eliminating silver's moneyness premium and leaving us with what you call pure M/M, and what I call the real bills doctrine.

      What might keep that moneyness premium from falling to zero? Well, things like being unable to trust IOU's, restrictions on IOUs' issuance, or IOUs' unsuitability for trade. So the moneyness premium can be positive, and your interest in moneyness is justified. But a moneyness premium creates an arbitrage opportunity for anyone who can issue IOU's that displace silver as money. That's why I expect moneyness premia to be small, and that's why I think the value of modern paper money is 99% backing and 1% moneyness.

      The rest of the economics world thinks that it's 1% backing and 99% moneyness. The words 'real bills doctrine' do not appear in ANY introductory macro textbook. They all talk about fiat money in a way that you and I know is wrong. The real bills doctrine is opposed by guys like Scott Sumner and Nick Rowe, and even by its natural allies over at Freebanking.org and Mises.org. Meanwhile, everyone is friendly to the moneyness view.

      So while you are inquiring into the monetary nature of things, keep in mind that the moneyness view has been hashed over a million times, while the real bills doctrine nearly untouched territory.

    3. I agree that the moneyness premium is small. Do you remember this post? Assuming my logic and data were right, the premium of one illiquid share over a more liquid share is usually non-existent, and when a premium does open up, it's never big.

      "The rest of the economics world thinks that it's 1% backing and 99% moneyness."

      I always read the rest of the economics world as saying 0% backing. And they don't use moneyness, since they tend not to think in degrees of liquidity, but absolute categories of "money" or "not-money". In some respects this is what drives the whole problem that you face. In devising an absolute category called "money", economists have decided that "money" objects abide by a unique set of forces, and that the value of "money" objects isn't determined by the same set of financial forces to which all other objects are beholden... like backing. The moneyness view restores the importance of financial forces (like backing) while keeping the monetary element alive, although it confines this monetary element to a sliver on top of fundamental value.

      The real bills/pure MM idea reminds me of an Arrow Debreu world. Monetary phenomena could never find a place in a pure Walrasian economy because as long as exchange is effortless, then pure barter will be the result. We need something like real bills/pure MM + moneyness to understand the real patterns of monetary exchange we see around us.

  4. I credited the rest of the economics world with a 1% backing view because guys like Nick Rowe and David Andolfatto sometimes write about the central bank's assets as if they matter.

    I'm all for that "sliver on top of fundamental value" view. The biggest remaining question is what goes on between the world of effortless exchange and the world of paper/credit money. I'm starting to spend a lot of time looking at the currency of the American colonial period, since it was the period where government paper money really got off the ground. Lots of interesting histories of coin shortages, booms triggered by the introduction of paper money, inflations and recessions, competing currencies, etc.

    1. I think Nick is slowly working from a 0% backing view to 1% and maybe even a little more than that, probably due to the exchanges you two have had on his blog.

      Mike, I don't know much about the American colonial period but I've read some papers by Dror Goldberg on the subject that you may find interesting. Adam Smith also has some interesting commentary on that period. If you find anything interesting, do keep me in the loop.

    2. In one of our early arguments, Nick said that the central bank should keep about 30% assets in order to buy back its money in case of a decrease in demand for money. Baby steps.

      So far, the best resources on the colonial period are
      1) Andrew McFarland Davis, Colonial Currency Reprints
      2) Leslie Brock, Currency of the American Colonies
      3) John McKusker, Handbook of Colonial Currency
      4) Curtis Nettels, Currency of the American Colonies Before 1720

      I'll probably make Nettels' book into a pdf next week. I'll send you a copy.

  5. Personally, I believe We can eliminate the need to differentiate. What would happen if We did not NEED exchange (trade, barter, money, metal, etc.) to survive richly? I ask because there IS a way to eliminate this need. Such exchange is merely the accounting for meaningful energy expended in an energy-scarce society.

    Given this, it is clear that adding abundant free energy – and programmed machines for any necessary work no One WANTS to do – will remove the necessity of accounting for energy, freeing every Human to choose where, when and how Their energy is applied. It will free Them from HAVING to add Their energy to a system to survive.

    This will also pull the money rug out from under the banksters and other “elite.”

    It will not take materially from anyOne in terms of home, comforts, and so on, but it will remove power over Others in favor of autonomous power over Self, while allowing the abundance of this planet to flow to ALL of Us without the impedance of a money system.

    For more and for details, please see:

    Analysis: http://www.thelivingmoon.com/forum1/index.php?topic=657.0

    PLAN: http://www.thelivingmoon.com/forum1/index.php?topic=2759.0

    Governance: http://www.thelivingmoon.com/forum1/index.php?topic=2103.0

    The LOVE of money is the root of all evil; remove the soil in which the root grows…

    On Twitter @AmatreasuSolar

  6. I have a lot of respect for backing theory. I also think it's a tautology.

    I've found backing theory immensely useful in trying to explain what a fiscal deficit does (and does not do), what monetization vs bond financing is really about (duration and rate risk, not a fundamental qualitative difference) and getting people to focus on the assets of the sovereign complex rather than its liabilities.

    The trouble is - the main asset on a sovereign-CB balance sheet is unmonetized, and oftentimes, the sovereign-CB complex seems to be able to increase its assets at will. Whether or not, and to what extent, and in what conditions and through what channels does this degree of freedom to affect the valuation of your own balance sheet by fiat exists and operates are the central questions of monetary policy/ macroeconomic management.

    Thus, while backing theory helps re-frame the standard questions and concepts in a different, perhaps more analytically useful, manner, it does not invalidate the usual theories, debates, points of difference that monetary theorists have.

    To use an engineering/physics analogy, I see backing theory not as a competing description of the world, but as a Laplace transform. Sometiems it may be more helpful to look at things in a different dimension, but the central problem to be solved has not been changed/voided/invalidated.

  7. "Whether or not, and to what extent, and in what conditions and through what channels does this degree of freedom to affect the valuation of your own balance sheet by fiat exists"

    Role of expectations, monetary 'transmission' channels, interest rate bounds, liquidity premiums et al. The usual stuff.

    1. Everyone probably sees there own "central problem". To me, the central problem of monetary economics is that it has never been properly integrated with microeconomics, or value theory. Efforts have been made, but nothing satisfying has ever definitively emerged. We can still talk about empirical phenomena and monetary policy, but it would be far easier if we'd resolved the core problem to begin with. Money doesn't fit into micro (and micro doesn't fit into money).

      Backing theory, or the Miller Modigliani approach, helps move us there.

  8. http://koyakei.blogspot.com/2014/08/how-to-devide-all-information-to.html
    From this reason. I think all information can be currency.
    I guess you insist only adjunctive information can become money. Is it true?
    And I define money and currency is diffrent.

    Do you think, is there any possibility to qualitative approach of currency transaction?
    In addition, Do you think equality for chance to choose way of transaction is necessary for just competition?