Streetwise Professor

September 14, 2008

Stuck on Stupid

Filed under: Commodities,Derivatives,Economics,Energy,Politics — The Professor @ 2:10 am

During the aftermath of Hurricane Katrina, General Russel Honore castigated a reporter for being “stuck on stupid.”

For some reason, that phrase has come to mind repeatedly in the past hours. Perhaps it is due to the fact that I am sitting on the 9th floor of the Hilton in College Station, Texas, watching the rains and winds of Hurricane Ike. But more likely, it is that I am passing the time reading several of the latest “studies” purporting to lay the blame for the runup of commodity prices (especially energy prices) in 2007-2008 at the feet of speculators, and more particularly, commodity index traders.

The most widely disseminated of these is the Masters-White “Accidental Hunt Brothers” piece. Another is “The Commodities Market Bubble: Money Manager Capitalism and the Financialization of Commodity Markets” by University of Missouri-KC professor L. Randall Wray.

In a nutshell–same old, same old BS.

I could go on and on deconstructing this bilge, but even though the opportunity cost of my time is only slightly above zero, being marooned as I am, it is still positive. So I will limit myself to hitting the low-lights.

The most fundamental problem with these papers is their failure to identify any channel whereby side-bets on commodity price outcomes affect the supply and demand for physical commodities. Masters and White are particularly slippery in this regard, equating in a conclusory way the purchase of physical oil by China to the purchase of forward and futures contracts in similar notional quantities by index funds, and then asserting that the side bets contribute just as much to demand as the Chinese purchases of physical oil, and hence are just as responsible for any price increases. Similarly, Masters-White state that index traders are holding “stockpiles” of oil. Uhm, no, they don’t. They hold contractual claims on oil that they liquidate without converting them into actual, bona fide stockpiles of physical oil. The Masters-White argument is akin to saying that the bettors at race tracks own the outcomes of horse races. No, they own betting tickets–contingent claims that have payoffs that depend on the outcome of a horse race. Absent some other mechanism (e.g., bribing the jockeys), these contingent claims don’t influence the outcome of the race.

Put differently, moneys invested in indices don’t actually flow into the market for the physical commodities. Nobody uses these dollars to buy oil or corn. Instead, these funds are used to place collateralized bets on future movements in prices. The actual money invested in these strategies (if the index investor follows a fully collateralized strategy) is used to buy T-bills in dollar amounts equal to the notional value of the commodity index. The actual moolah doesn’t flow into the commodity markets, the way that it does when Sinopec or PetroChina buys a supertanker of Arabian crude.

Masters-White and Wray also fail to address the fact that index investors liquidate their positions before they become claims on physical commodities, and hence they are typically sellers at the time that spot prices are determined.

In terms of “evidence” these authors merely point to the fact that index fund investments have risen at the same time that commodity prices have risen. They do not appear to separate out the effect of increases in the amount of commodity index fund assets that was caused by the effect of rising prices on the value of existing positions from the increases attributable to the flows of new moneys into the funds. There is no effort to carry out rigorous statistical tests (e.g., cointegration tests, Granger Causality tests) to establish causation or an equilibrium relation. (It is possible that the visual relationship between two strongly trending series like an oil price and the size of index positions could be a spurious one. The authors of these studies make no attempt to explore this possibility.)

Masters-White make a big deal of the fact that all commodities in the various indices have gone up, and say that this would be extremely unlikely unless the index investment was driving all of them. But commodities not in funds, such as MGE wheat, iron ore, minor metals, have been skyrocketing in price too.

This line of reasoning does suggest one potentially rigorous test of the hypothesis. Namely, if the increasing importance of index money is injecting a common element into prices of all index commodities, overwhelming individual commodity-specific fundamentals, one should observe increasing correlations across commodities in indices, and stable or declining correlations between indexed and non-indexed commodities. This effect should be especially pronounced in high frequency data (daily, or intra-day) because the index investments are by construction made simultaneously or nearly simultaneously across commodities in the indices. Of course, Masters-White-Wray don’t even come close to identifying, let alone testing, such a refutable hypothesis.

In terms of “theory,” Masters-White and Wray put great stock in the cost-of-carry relationship between spot prices and futures prices, asserting that by buying futures, index investors drive up futures prices, which equal spot prices plus carrying costs, thereby forcing spot prices up too. But, for instance, in copper and oil, the price rises in 2007-2008 occurred simultaneously with movements from contango into backwardation.

Masters-White assert that order flow moves prices, so the predominately long order flow of index funds therefore must move prices up. There are two fundamental problems with this “reasoning.” First, when index funds roll, their order flow is on the sell side–so why don’t they move nearby prices down when they roll?

Second, and more importantly, they are very confused on the relationship between order flow and price discovery. Although Masters-White are correct that orders in futures markets are anonymous, they draw the wrong conclusion from this. The impact of any particular individual, anonymous order on prices depends on the average information content across orders. Holding the intensity of informed trading constant, increasing the intensity of index trading reduces the average informativeness of order flow. Thus, more orders are entered, but their individual impact is smaller, leading to little or no net impact on prices. (Index trades are likely to be deemed uninformative on average for reasons identified by current colleague Praveen Kumar and fellow Chicago GSB PhD classmate Duane Seppi. It is unlikely that a particular trader has private information on all components of an index, and if he has private index on any individual components, it is more effective to trade on that information in the markets for the individual components. This is why markets for index products, e.g., SP500 futures, are typically much more liquid than the markets for the individual components.) Moreover, informed traders adjust the intensity of their trading in response to changes in uninformed order flow, and generally trade to offset the effect of uninformed orders on prices. Thus, even if “dumb money” moves prices away from where they “should be,” this is an opportunity for smart money to trade against that, thereby forcing prices back towards the level justified by fundamentals.

Masters-White also make a big deal out of the fact that many market participants quote spot prices basis a futures to conclude that futures prices are driving spot prices, and hence noise injected into futures prices by index trading must therefore contaminate spot prices. In other words, futures prices are the dog, spot prices are the tail.

This is to confuse mechanics with substance. The key point is still that when it comes to determining futures prices as the market moves to expiry, participants recognize that if they hold positions to term they must make or take delivery of physical product. Thus, futures prices at expiration reflect supply and demand fundamentals in the physical market. I will not take delivery of something at a price if I do not believe that I can sell it at that price or better. I cannot sell it at that price or better unless there is physical demand for the product at that price. Thus, even though it is common to price some physical trades basis a futures price, that basis is determined in the market. Moreover, it adjusts to reflect physical supply and demand fundamentals, just as the delivery mechanism connects fundamentals to futures prices.

Put differently, basis pricing does not mean that spot prices do not reflect physical market realities. The best illustration of this (discussed in a SWP post over the summer) is the heavy/sour crude basis, which widened dramatically over the summer. This reflected the high demand for low sulfur diesel, the low yield of such fuel from heavy crude, and disruptions in the markets for sweet crude (driven primarily by problems in Nigeria.) (All of these are nuances that Masters et al, and Michael Greenberger, miss completely–even though they are salient factors in understanding the summer’s market dynamics.) Physical crude traders did not mindlessly pass on increases in WTI prices to the prices of heavy crudes. Differentials adjusted to reflect physical market supply and demand realities.

This gets back to the fundamental problem with all these analyses, one that is highlighted by Masters’ and White’s choice of a title for their opus–”Accidental Hunt Brothers.” I have said it a zillion times, but maybe the zillionth-and-first repetition will make it sink into heretofore impervious skulls: PRICES SERVE TO ALLOCATE REAL RESOURCES. IF PRICES ARE SUBSTANTIALLY DISTORTED, THAT MUST MANIFEST ITSELF IN PRONOUNCED DISTORTIONS IN THE ALLOCATION OF REAL RESOURCES. In the case of the real Hunt brothers, to keep prices high, they had to accumulate huge quantities of physical silver–real physical inventories, not phantom paper inventories. When the government propped up the price of wheat and milk, it had to buy and hold (or destroy) huge quantities of physical wheat and milk. As another example, when a manipulation takes place, the manipulation causes manifest distortions in commodity flows. For instance, in 1910 famous (or infamous) speculator James A. Patten (benefactor of the Patten Gymnasium at Northwestern University as well as the Chicago Art Institute) cornered the New York cotton market. Prices became so distorted that English spinners shipped cotton back to the US to take advantage of the price distortions–the economic equivalent of water flowing up hill.

If speculators are distorting prices, that necessarily causes distortions in stocks or flows of physical commodities. If index investors, or other speculators, were Hunt brothers, accidental or otherwise, they would necessarily find themselves in the same circumstances–paying for, holding, and financing large positions in physical commodities. Positions so large–if the price distortions are as large as asserted–that could not escape notice. Neither Masters-White, Wray, nor any of the other critics of index investment have provided any evidence of such distortions. Indeed, they haven’t even tried to do so.

To his (mild) credit, Wray recognizes the problem, anyways. He notes that Krugman has argued that speculative distortion should lead to accumulations of large and growing inventories in the hands of speculators. He attempts to avoid confronting this implication by stating that Krugman’s argument presumes perfect competition, and that commodity markets are not perfectly competitive, therefore Krugman’s argument doesn’t hold.

It is Wray’s argument that doesn’t hold. Perfect competition is not the essential linchpin of the analysis. Even if there is some market power on the producer side of the market, sellers will respond to higher prices by increasing output. Similarly, consumers will respond by reducing consumption. If somebody–a speculator–injects additional demand for the real commodity into the market, and drives prices up, producers will increase output and consumers will reduce consumption. This is sustainable only if the speculators willing to pay the (putatively) artificially high price put their money where their mouths are, and hold inventories of the commodity. (Wray tries to finesse the analysis by stating that maybe large oil producers are keeping oil in the ground rather than producing it. Maybe. But that is something that cannot be laid at the feet of the speculators. Moreover, it is hard to reconcile with the movement of the market into backwardation.)

Wray makes some other basic economic errors. For instance, he notes: “Americans have responded to rising gasoline prices in the manner economists expect, with consumption falling sufficiently to offset China’s increased use of crude oil—yet crude prices barely responded.” Catch the freshman mistake? If the reduction of consumption is an equilibrium response to higher prices, you wouldn’t expect the drop in consumption to cause a fall in prices. The reduction in consumption is a movement along a demand curve, and absent any shift of the demand curve, holding all else equal, the price should not fall. Wray’s is essentially a yo yo theory of prices. Rising prices cause falling consumption cause falling prices cause rising consumption cause rising prices . . . Maybe Wray should get rich by inventing a perpetual motion machine.

In brief, there is nothing new here. Indeed, these papers–trumpeted to the skies by Senators Lieberman, Cantwell, and others–are just a repetition of the same blarney that’s been flying about for months now. Can we please get unstuck from stupid, and start having an intelligent debate on these issues, one that reflects at least a modicum of comprehension of basic economics? Alas, I seriously doubt it.

Print Friendly


  1. Maybe Masters-White should update their framework and study exchange traded funds that actually take supply off the market such as GLD?

    Comment by Teresa Lo — September 14, 2008 @ 2:43 am

  2. GLD, however, was specifically setup to increase demand and the price of gold.

    On futures surely it depends on the actions of counterparties. If I make a bet that oil is going to go up with you, for say a notional 100m barrels tehn there’s no direct impact on the price. However if you don’t want to be short 100m barrels of oil, and decide to offset this, one way in which you might is to purchase physical oil (or someone else down the chain might do so). This however gets back to rising stocks, of which there isn’t obviously much evidence. Could it be that the stocks are being held in the supply chain at some position?

    Comment by James — September 15, 2008 @ 4:12 pm

  3. Craig,

    I have not been able to find, much less read, Randall Wray’s paper, yet I’m sure your ultimate conclusions about White, Masters, Wray et al. are correct. They are wrong on the critical points.

    You are undoubtedly correct that the data on inventories show that the allocation of real, physical resources has not been distorted (i.e. there are not massive stockpiles anywhere).

    Nevertheless, one of your remarks got me thinking about how changes in the term structure of crude and refined product futures might affect current production:

    “Wray tries to finesse the analysis by stating that maybe large oil producers are keeping oil in the ground rather than producing it. Maybe. But that is something that cannot be laid at the feet of the speculators. Moreover, it is hard to reconcile with the movement of the market into backwardation.”

    I agree that long-futures positions do not directly affect production of actual oil companies.

    Is it possible though, that long-futures positions have an indirect effect by changing the term structure of futures prices?

    Litzenberger and Rabinowitz wrote a paper []
    on oil futures backwardation back in 1995 and concluded that contango reduced current production and backwardation (especially strong backwardation) increased current production.

    If institutions are continually rolling futures positions forward, thereby reducing or eliminating backwardation, might that have some influence over production decisions of producers (especially government-owned oil companies)?

    Comment by John McCormack — September 17, 2008 @ 2:56 am

  4. John–

    The question is whether continually rolling futures positions forward by index speculators reduced backwardation.

    On a factual level, it is interesting to note that the rise in oil prices in 07-08 was accompanied by an increase in backwardation, not a move to contago. (Prices were in contango until July ’07, then moved into backwardation–that’s also when the spike began. Similar events in copper.)

    On a theoretical level, this post argues that the anticipated rolling of positions will not affect the shape of the forward curve. As long as funds roll, and it is believed that they will always roll, their deferred purchases will not affect the expected spot price.

    The ProfessorComment by The Professor — September 18, 2008 @ 3:51 am

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress