Saturday, April 30, 2011

The fundamental question for whether the market can fix itself: "can it be arbitraged?"

Nick Rowe responds in the comment section of this post:

"Daniel: There is no "money market". When people talk about the "money market" they really mean the market in loans. Which is the bond market.

There is a market in which bonds are traded for money. That's the "bond market". The demand for bonds depends (inter alia) on the rate of interest.

There is a market in which goods are traded for money. That's the "goods market". The demand for goods depends (inter alia) on the rate of interest.

Liquidity preference theory says the rate of interest adjusts to clear the bond market.

Loanable funds theory says the rate of interest adjusts to clear the goods market.

Clearly it can't do both at once (except by fluke). Or rather, the only way it can do both at once is if some of those other things in the "inter alia" adjust too. And that would include the stock of money and the price of goods
."

I think I may be closer to Nick than I had originally thought. To start with, though, I am somewhat confused about why he says that hte loanable funds theory says interest rates clear the goods market and liquidity preference theory says they clear the loanable funds market. I would have thought loanable funds theory says the interest rate clears the loanable funds market and liquidity preference says it clears the demand for and supply of liquidity - the decision whether to hold cash or not. I'm not sure what the goods market has to do with it (is this an intertemporal goods market point?), but I'm not sure that matters all that much.

The point is, there's no reason to think the interest rate satisfies both conditions simultaneously - and that is something I've always maintained. But other things like the stock of money, prices, and output can adjust so that it does satisfy both simultaneously (Hicks essentially forces this outcome - he doesn't allow any non-clearance in the loanable funds market or in the supply and demand for money). The point is, what is the nature of these "other adjustments"?

Nick makes the fair point on his blog that prices are probably rigid. We usually assume that the money supply is exogenously determined, so that leaves us with output. So let's say output takes the brunt of the adjustment and now the conditions of the loanable fund theory and the liquidity preference theory are satisfied. Fine. Now what about the depression in output?

This is where the contrast with the Walrasian system is important. In a Walrasian system excess supply of output in one corner of the system is offset by excess demand in some other part of the system (Brad DeLong would say excess demand for bonds, secure assets, etc.). That's the whack-a-mole world I was talking about earlier. Unless there are institutional frictions (zero lower bound, disrupted credit channels) that situation actually isn't all that worrisome. Why? Because we think market actors are good at arbitrage. Excess supply in one corner and excess demand in another corner is an opportunity for arbitrage.

But is that the case in this system that Nick and I have been describing? No, it's not. Before the dual constraints of the loanable funds theory and the liquidity preference theory weren't both satisfied. Taking after Keynes, let's say the supply and demand for cash/liquidity cleared but there was excess supply in the market for loanable funds. Income is then reduced to satisfy the loanable funds theory (so that both the loanable funds and liquidity preference theory are now satisfied). So now we have reduced income (deficient demand/excess supply in the goods market) but the loanable funds market is now cleared by the interest rate, as is supply and demand for cash. Unlike in the Walrasian system, we now have excess supply (deficient demand) in one corner of the system without any excess demand in another corner of the system.

Above, I said the Walrasian co-existence of excess supply and excess demand offered the opportunity for arbitrage and that we think people are good at arbitrage. In this situation, though (unlike in DeLong's original whack-a-mole glut with frictions) there is no opportunity to arbitrage because there is no excess demand to balance out the excess supply.

That's why the market can't fix itself. The market is an institution designed to arbitrage. If you can't arbitrage your way out of the situation, then you can't get out of it with a market, and that's why we say that you need government to increase demand (either through monetary or fiscal policy).

Steve Horwitz would counter that if you don't assume money supply is exogenous you get a new arbitrage opportunity... I have not sufficiently explored those arguments yet to comment on that one.

4 comments:

  1. Daniel:

    I take a similar view to Nick in that general gluts can only arise because of an excess demand for money. A spike in demand for safe assets can only do the same if that demand spills over to money, since money is a safe asset too.

    Here is why I think this way. Money is the only asset on every other market--goods, services, other financial assets--and thus a disruption to money via a money demand shock will disrupt all these other markets directly if prices are sticky. And sticky they are.

    Other safe assets like tbills are not on every other market--they are on their own market. If there were a sudden surge in demand for tbills it could cause a glut if there are not enough safe non-money assets. At such a point, people would turn to money which would cause a disruption to all the other markets.

    ReplyDelete
  2. Daniel: "To start with, though, I am somewhat confused about why [Nick] says that hte loanable funds theory says interest rates clear the goods market and liquidity preference theory says they clear the loanable funds market. I would have thought loanable funds theory says the interest rate clears the loanable funds market and liquidity preference says it clears the demand for and supply of liquidity - the decision whether to hold cash or not. I'm not sure what the goods market has to do with it (is this an intertemporal goods market point?), but I'm not sure that matters all that much."

    That's a good confusion to have. Nobody (including me) is clear on this. But it does matter; it matters a lot.

    The price of apples is set in the apple market. The price of apples adjusts (if it's not sticky) to equilibrate the demand and supply of apples.

    Loanable funds says that r adjusts to equilibrate desired I and S. Throw in a bit of National Income Accounting (closed economy, no government), and I=S can be re-written as C+I=Y. Ergo, r adjusts to equilibrate the output market. Which doesn't make sense, because r is (the reciprocal of) the price of bonds, and we pay for output with money, not bonds.

    Liquidity preference says that r adjusts to equilibrate the demand and supply of money. Which doesn't make sense, because there is no (single) market where the demand for money meets the supply of money. Plus, just as the price of apples is set in the apple market, the price of bonds (1/r) should be set in the bond market.

    ReplyDelete
  3. Warning to anyone else trying to comment here: copy your comment onto the clipboard first, because Google tends to eat it on the first attempt.

    ReplyDelete
  4. Ya - sorry to everyone - blogger has been acting weird for the last couple weeks.

    Let me know if you think you posted but lost it too, because some are identified as spam and if that's the problem I am able to rescue those.

    Thanks guys - I'm very much a novice trying to get a sense of how to think about these things, and I appreciate all the thoughts. Not every novice/student is lucky enough to get this kind of feedback and I'm grateful - I'm digesting all this now.

    ReplyDelete

All anonymous comments will be deleted. Consistent pseudonyms are fine.