A Washington Post article claiming that one firm, Vitrol, "at one point in July,... held 11 percent of all the oil contracts on the regulated New York Mercantile Exchange" reinforced the beliefs of those who claim that unregulated speculation is behind the recent swings in commodity prices.
Jim Hamilton throws cold water on this idea. He has other objections to the article, but here's the one relating to the 11 percent figure that has received so much attention:
More speculation about those oil speculators, econbrowser: I normally leave it to folks like Dean Baker to beat up on the press. But I can't resist shining a bright light on today's story about oil speculators in the Washington Post, which has also been discussed by Mark Thoma and Tyler Cowen. ...
What ... does [David Cho] dig up? The article continues:
Even more surprising to the commodities markets was the massive size of Vitol's portfolio-- at one point in July, the firm held 11 percent of all the oil contracts on the regulated New York Mercantile Exchange.
That does sound like a lot, though enough details are left out to make me wonder what is actually being claimed here. Surely Cho doesn't literally mean "all the oil contracts," i.e., light sweet, Brent, heating oil, gasoline, and so on. If light sweet alone, are we talking about just futures, or futures plus options? Or is Cho possibly referring just to one very specific contract, such as the August CL futures contract? And were these positions held outright by Vitol or purchased on behalf of its clients?
Cho gets more quantitative a few paragraphs down:
By June 6, for instance, Vitol had acquired a huge holding in oil contracts, betting prices would rise. The contracts were equal to 57.7 million barrels of oil-- about three times the amount the United States consumes daily.
Again I'd like to know if we're including options somehow in this number. But more importantly, the claim that you can compare the number of notional barrels of oil implied by a futures contract if it were held to settlement with the number of physical barrels that the U.S. consumes repeats the egregious error committed by Michael Masters in his Senate testimony this May. You can't compare the outstanding NYMEX open interest with U.S. daily petroleum consumption numbers directly because they are measured in different units. Open interest is a stock-- it is measured in number of outstanding contracts at a particular point in time. Consumption is a flow-- it is measured in barrels per unit of time. You can't measure how many barrels of oil the U.S. consumes without specifying a time unit. We consume about 20 million barrels per day, or 14,000 barrels per minute, or 7.3 billion barrels per year. With which of these 3 numbers are we supposed to compare 57 million? Fifty-seven million sounds like a lot more than 14,000, and a lot less than 7.3 billion. You can make 57 million sound as big or as small compared with U.S. consumption as you want, because there's an arbitrary time interval associated with consumption that is not associated with open interest. If you want the futures volume to sound big, you compare open interest with daily consumption, as Masters and Cho both do.
Cho was trying the best he could to convince us that unregulated speculation was the cause of this summer's spike in oil prices.
But instead he convinces me that he really couldn't find much of a case.