Friday, February 24, 2012

What is Money?

Nick Rowe says we should not think of money as a store of wealth:
Money is what money does. There are two functions of money that define what is and what is not used as money: medium of exchange; and medium of account. That's it... We need to start worrying a lot more about how money works as a medium of exchange. We need to understand a lot better than we do how money works as a coordinating device in a decentralised economy. And we need to understand a lot better than we do how money can sometimes fail as a coordinating device. Because, outside a very simple economy, people can't barter their way back to full employment if the monetary exchange system fails. 
We need to stop thinking of money as a store of wealth, just like all the others. And let's start by changing the textbook definition of money, by deleting that bit about money being a store of wealth. 
I agree, but would add that we also need to start thinking about money at all levels of transactions. Most textbooks and many economists think of money assets at only the retail level (i.e. the M2 money supply).  This crisis has taught us that institutional money assets--those assets like treasuries, commercial paper, and repos that facilitate transactions in the financial system--matter too.  The bank run on the shadow banking system was a bank run using institutional money assets.  If we really want to understand money and its implications for the economy we need to be thinking about these money assets too.  Thanks to Gary Gorton, David Aldonfatto, and Stephen Williamson I have come to better appreciate this point.  And thanks to this perspective I have come to see the demand for safe assets and budget deficits in a different light


  1. It's not clear to me how much the assets you speak of were actually functioning as media of exchange. Were they bought and sold with money or were they used to make purchases directly? If the holders had to sell them before they could purchase something else, then they weren't media of exchange at all, but just highly liquid assets.

    Economists tend to confuse money and liquidity, in my view. Yeager had it right: while near monies may be close substitutes for money to an individual, since they can readily exchange such assets for money at short notice and at little cost, they are nothing like close substitutes for money with regard to their effect on the economy as a whole.

    An individual's liquidity position is a measure of how easily he can turn his wealth into all-purpose purchasing power, i.e. money balances. To realise such potential purchasing power, he must first find willing buyers of his assets. The trouble is that if everyone decides to realise his own potential purchasing power at the same time, then there are too many sellers and not enough buyers--prices quickly fall and everyone gets a nasty shock. The available liquidity in the economy appears to evaporate all of a sudden, but it's really, mostly, just an excess demand for money. (This is also, I think, why the whole people-aren't-as-wealthy-as-they-thought explanation of the recession has superficial appeal).

    What makes money special is that, unlike, near monies, you don't need to find a willing buyer, i.e. someone who just wants to hold that asset. People accept money in exchange all of the time because even though they have no desire to increase their average money holdings, because they know it can be readily used to buy what they actually want instead.

    Was the shadow banking system creating money or just near money? I don't know. I read some of Gorton's work, but I never found a satisfactory answer, though, admittedly, I did have a hard time understanding it in places.

    While it's plausible that a decrease in the supply of safe and liquid assets, instigated by the housing bubble, was responsible for the increasing money demand that caught the Fed off-guard in 2008, an excess demand for money would also tend to have that effect anyway.

    It's hard to disentangle these things, especially from where I am sitting.

  2. But isn't the dollar itself a safe asset? Doesn't that confirm it's store-of-value function?

    Excess reserves are $1.6 Trillion.

    That's a lot of stored value doing nothing for liquidity, nothing for AD, nothing for GDP growth.

    From your linked Feb 2 post:
    All this requires is running a budget deficit which may or may not imply higher government spending (i.e. it could also come from tax cuts).

    We have had tax cuts upon tax cuts and deficits for several decades now. This led to a collapse in Aggregate demand. GDP growth has been in decline four 4 decades.

    Isn't there some point at which abstract theory has to confront reality?

    Or am I missing something big?


  3. Lee,

    Repos are to institutional investors what checking accounts are to you and me. The both are both highly liquid assets that can be turned into instant purchasing power.

  4. Jazzbumpa,

    I don't think Nick is saying that money is not a store of value. Indeed, money has to be a store of value to be a medium of exchange at all, since there is always some elapse of time between receiving money and spending it. (Nick makes this same point by noting that money velocity cannot be infinite.) However, that money is a store of value is relatively unimportant. Technically, all goods are a store of value, and there are some goods which are much better stores of value than money. What makes money important, what makes it unique, what makes it special for the economy as a whole, is its unique roles as a medium of exchange and unit of account. These features make money qualitatively different than even the nearest of near monies, and they are why disequilibrium in the supply of and demand for money have such peculiar and sweeping macroeconomic consequences.

    In my opinion, lots of so-called monetary economists, seem to conceptualise as different from near monies only by degree. In this view, money is just the nearest of near monies, if you will, and pure money is just an abstraction which doesn't actually exist. However, by not accounting for money's qualitative differences, many monetary economists effectively leave money itself out of their models of the economy! That seems too stupid to be true, even to me, but I keep getting that feeling over and over again when reading what people have to say.

  5. Jazzbumpa,

    Short answer: money supply has yet to reach level of money demand. We still have excess money demand.

    Yes, the dollar is a safe asset that can be used as a store of value. But why? Because it is the most liquid asset.

  6. David,

    You say 'turned into instant purchasing power', but that's just the question: do they have to be sold for money first? Obviously, the answer is yes for lots of things, like if you wanted to go shopping at Wal-Mart.

    My guess would be that they really do serve as media of exchange in particular financial markets. In that case, they are, conceptually, a bit like banknotes in the United States before the civil war--they served as media of exchange in particular regions but were merely liquid assets far away from home.

    But how do you disentangle whether the collapse of the repo market, for example, was a cause or the effect of an excess demand for money? The causation could run either way, and that leads me to wonder how important that whole market really is to understanding the recession.

  7. Just wondering if you've seen Steve Waldman's latest at Interfluidity. (Incredibly cogent, as always.) Proposes a system that would greatly increase the supply of safe assets available to small savers, with darned interesting implications for investment incentives.

  8. If memory serves, JMK's 1937 summary journal article of the GT provides a quip about how foolish it is to talk about money as a store of value. The point being that money is desired for the purpose of liquidity, not as a store of value. In the JSTOR version of the article it was at the top of the page. About 7 pages in. (I went and looked. It's 10 pages in). The QJE article from 1937, p. 216 "Why would anyone outside of a lunatic asylum wish to use money as a store of wealth."

  9. You might call it "hyperinflation," or "improving liquidity." I would call it "counterfeiting."

    The Dreckgeld "issued" by the Financial Sector has done as much, heck, nearly infinitely more, to debase and damage the credibility and good-faith underpinnings of the REAL "money" the rest of us have to use, as all those funny-munny $100s printed and circulated by various economic rivals and enemies of the nation. Go look up what constitutes "counterfeiting" under the US Code, 18 USC Section 470 et seq. Sure looks like it to me, and to Matt Taibbi too.

    And us wage-earners, who EARN money, as opposed to "making" it by casino bets and flat-out inventions like CDOs and CDSs and the rest, have to now create REAL WEALTH to make up "value" to store up again, and try to plug the freakin' holes the hotshots have blown below the national waterline. And are continuing to do so, arrogantly, billions of counterfeit "notional" dollars of it, every freakin' day.

    Snicker, sneer, chuckle "IBG-YBG." Yeah, that's right. Uh-Huh.

  10. "What is money?" is a question that only invites circular answers like "It is the stuff we pay with." Let's try a question that has a meaningful answer: "Whose liability is that money?" Now we have a simple accounting problem with a real answer: Green paper dollars are the Fed's liability. Checking account dollars issued by B of A are B of A's liability. Credit card dollars are Mastercard's liability. Toys R Us dollars are Toys R Us' liability, etc. Once that question is clearly answered, we can move to the question on everyone's mind: "What determines money's value?", and the obvious answer is that money's value is determined by the assets and liabilities of its issuer, just like any other liability.

  11. Mike,

    First, a definition cannot be circular; a definition is not an argument. Suppose someone defined 'money' as the stuff we use to pay for other stuff. That's not an argument: it's not like the word 'money' is a premise and 'stuff we use to pay for other stuff' is a conclusion. That just wouldn't make any sense.

    You may not like a definition. You may find it misleading, unimportant, or whatever, but whatever is unsatisfactory about a definition, it's not it's circularity. Goodness knows what you must think of dictionaries.

    Anyway, what you're saying about money is simply not true, or, at least, it's not true all of the time. Federal Reserve notes may be 'liabilities' in an accounting sense, but they're not really liabilities at all. You can't redeem them from anyone for anything. I suppose their kind of like liabilities for the whole United States economy, but no individual is forced to accept them in exchange.

    For money, supply creates its own demand in a way that ordinary liabilities of corporations or governments do not. An increase in the supply of money sets of a process of inflation that eventually increases the demand for money to match. This is not true for other kinds of liabilities. This is because people will not refuse the ordinary medium of exchange even if they do not desire more of it, since they know it can be readily exchanged for something they do want--that's what makes it a medium of exchange in the first place. This is not true for other liabilities, like those of Toys R Us. If Toys R Us wants to issue more bonds, then it needs to convince someone to hold more of them; supply doesn't create its own demand through a process analogous to inflation. Money really is special.

  12. By the way, this is exactly the point I was trying to make in an earlier comment. By conceptualising money as just another liability, distinguished only by its particular liquidity and safety, money, in the traditional sense, is being tacitly left out of the analysis. The things that make money special--it's general acceptance in exchange, and it's lack of a price of its own--are being assumed away in favour of, I suppose, simpler models.

    For the most part, one can only get rid of unwanted money balances by passing them onto someone else, and the next guy will accept them even if he doesn't want larger money balances, because he expects to be able to get rid of them in the same fashion. If there is nobody who wants larger cash balances, then the unwanted money doesn't just go out of existence, but rather it begins to drive up prices. Eventually, the previously unwanted money will be demanded at the higher level of prices.

    Contrast this to any ordinary bond or whatever. One can get rid of it by selling it to someone else, and if he doesn't want it he can then sell it again, and so on. The difference is that if there is nobody who wants the bond, then it's price eventually falls. If nobody wants it, even at the lower price, then it is just held until maturity and it goes out of existence.

    Bonds react to surpluses and shortages through quantity supplied and price changes. Money reacts to surpluses and shortages by shifting the general price level. The first is unlikely to be felt much outside of a particular market, while the second impacts all markets in the economy. The first could not alone be responsible for a general glut or fall in employment, while the second is almost guaranteed to have that effect if it is in short supply.

  13. Lee:

    1) I'll leave it to you to google "circular definition". It's not the main point of dispute anyway.
    2) I don't think you'd deny that money issued by private banks, toys r us, etc, is the liability of its issuer. The fed's dollars are a true liability of the fed. The fed will buy them back with bonds, and if the fed were ever 'unwound', then holders of the fed's dollars. Or think of when the fed was on the gold standard. Every weekend the fed closed, and you couldn't get gold for your dollar for 2 days. What difference does it make if that suspension is continued for 30 days or 30 years, when the fed's gold and bonds remain in its vault the whole time? You seem to think the dollar is unbacked, but it is actually backed but not convertible (into gold).
    3. Money is not special. If the fed issues another dollar bill, it must find someone who is willing to hand over a $1 bond in exchange. If a firm wants to issue a new share of stock, it has to find someone who is willing to hand over resources in exchange for the share.
    4. Unwanted money does go out of existence. This is called the Law of Reflux. If $10 billion is returned to the fed, the fed pays out $10 bil. of bonds, and $10 bil. in paper is either shredded or stored in the fed's vault.

  14. Unfinished sentence correction: "then holders of the fed's dollars will have a claim to the fed's bonds AND gold."