Krugman on Calvo on Commodities
In his discussion of the cause of skyrocketing commodity prices, which he attributes in large part to central bank behavior creating excessive liquidity, Guillermo Calvo said:
Absence of a substantial increase in physical commodity inventories has been mentioned as evidence of absence of speculative activity (by Martin Wolf and, more guardedly, Paul Krugman). But that is not valid.
Paul Krugman replies:
Calvo on commodities: Guillermo Calvo is one of my favorite economists. But - you know there had to be a but - I just don't buy his latest missive. Still, it's important that we have this debate: something awesome is happening to oil and other commodities, and figuring out what it means is crucial.
So, a couple of points.
First, Calvo dismisses the argument that the absence of physical hoarding is evidence against a speculation/liquidity source of high commodity prices. "Suppose, for the sake of the argument, that the demand for commodities for current consumption or production is completely inelastic..." Well, that's assuming your conclusion.
When I think about speculation, I always start from Paul Samuelson's classic analysis in terms of intertemporal price equilibrium (a 1957 paper - and not available, as far as I can tell, online. Why isn't Weltwirtschaftliches Archiv on JSTOR?). Speculation can affect spot prices because it takes physical stuff off the market. Argue, if you like, that the inventory data are unreliable, or that stuff is being held in the ground; but don't tell me that physical quantities are irrelevant.
Second, Calvo argues that inflation risks stem mainly from excess liquidity. He's in good company there, but I won't join in that chorus. In general, I don't trust hydraulic metaphors for monetary economics. And we are in a world where central banks target interest rates, not monetary aggregates; the risk of inflation, if there is one, would come from Bernanke and Trichet keeping rates too low too long, not on what's happening to M2 or some broader M-something.
On a happier note, it's great to see top-flight economists weighing in on the crucial issues of the day. It's kind of like the Asian financial crisis of 1997-1998, which was bad for the world but a sort of golden age for policy-relevant theory.
I'm not sure I understand the statement that "the risk of inflation, if there is one, would come from Bernanke and Trichet keeping rates too low too long, not on what's happening to M2 or some broader M-something." The way to keep interest rates low is to increase the quantity of bank reserves and hence to pump up the money supply. So keeping the interest rate too low, too long is the same as pumping too much money into the system. Isn't it? What am I missing?
I thought the problem was measurement. We aren't sure how to measure money properly, exactly where to draw the line between assets that are liquid enough to count as money and those that do not have sufficient liquidity, so tracking monetary aggregates is difficult. But that doesn't mean the quantity of money, properly defined, isn't a source of inflation, just that we might have a hard time uncovering a stable relationship between money growht and inflation due to measurement difficulties. I don't meant to imply that targeting money and targeting interest rates are identical monetary policies if there are no measurement issues, they're not (and this argument is independent of the main debate, the source of the commodity price boom), I'm only saying that the quantity theory still holds in the long-run, MV=PY, and that as M grows, P grows.
Update: Maybe I should elaborate. The liquidity of an asset is not fixed, it varies with market conditions. When financial markets get in trouble as they have recently and markets seize up, the liquidity of some assets - those with lots of revealed uncertainty about their true value - will see their liquidity drop. As this happens, it is possible that assets that were "near money" and liquid enough to count in a broad definition of money, are no longer liquid enough to be part of the money supply and M drops.
It is also possible that as the Fed intervenes and backs some of these assets, the liquidity of the assets will increase once again as the markets come back to life. This would increase the volume of near money and, with a broad definition of M, cause on increase in the money supply and a reduction in interest rates (and, with respect to recent history, financial innovation can increase liquidity and thus increase the supply of money independent of the quantity of financial assets). Thus, it's possible for the quantity of money measured as the quantity of some fixed set of financial assets to be relatively constant, yet have the underlying money supply vary. Therefore, even though we haven't seen the traditional measures of reserves and M increasing substantially, that doesn't necessarily mean that money, broadly defined, hasn't varied substantially as the liquidity properties of financial assets have changed with financial market conditions.
Update: A bit more on this - I think I get it now. Krugman's point is that Calvo is, in effect, saying that the SWFs and others are pumping liquidity into the system, which will lead to inflation. SWFs might be able to add liquidity to the system if the Fed was, say, holding M1 constant, i.e. had a monetary target and didn't act to offset any movement in the interest rate (that's one thing I had in mind above in noting that a monetary target differs from an interest rate target). But what's actually happening is that the Fed has an interest rate target that is based upon its reading of economic conditions; basically that means that private actions that increase liquidity will automatically be offset by the Fed as it stabilizes the target interest rate. It's not that excess liquidity cannot create inflation, it can and will if Bernanke/Trichet ease too long (though the admonition that this is different from the 1970s comes into play here), it's a question about whether SWFs are actually adding liquidity to the system. Given that the Fed will neutralize any change in the interest rate, they won't be able to do that.
Posted by Mark Thoma on Saturday, June 21, 2008 at 09:00 AM in Economics, Inflation, Monetary Policy |
You can follow this conversation by subscribing to the comment feed for this post.