How Big a Contribution Comes from Oil Speculation?

Includes: DBO, OIL, USO
by: James Hamilton

A key reason why oil prices have been going up is that Asia and the oil producing countries are consuming more while global oil production has stagnated. That means Europe and America had to consume less, and a very high price proved necessary to accomplish that.

I do believe that speculation has been another factor that contributed to recent high oil prices. However, a key element of the bubble story is that there needs to be a very limited response of quantity demanded to the price increases, which the most recent data persuade me is no longer the case. Some of the estimates I've been hearing of the size of the contribution speculation is currently making to the price are therefore difficult to defend. Here I explain why, essentially elaborating on Paul Krugman's theme.

Senator Barack Obama's (D-IL) recent proposal to "crack down on excessive energy speculation" collects some of the claims recently being made on this issue:

Akira Yanagisawa, Senior Economist at Japan's Energy Data and Modeling Center: "In the most recent terms (the third and fourth quarter in 2007), the fundamental prices [of oil] are estimated around 50 to 60 dollars. On the other hand, it is estimated the premium has risen up to around 40 dollars at maximum." [Institute of Energy Economics, 3/08, p. 13]...

Larry Chorn, Chief Economist of Platts: "says the actual costs incurred in producing the most expensive oil is only around $70 or $80 a barrel, meaning that about $50 of the current price represents 'the market's risk premium plus speculation.'" [BusinessWeek, 5/13/08]...

J. Stephen Simon, Executive Vice President, Exxon-Mobil: the price of oil should be $50 to $55 per barrel based on supply and demand fundamentals. [House Testimony, 4/1/08].

John Hofmeister, President, Shell Oil Co: The proper range of crude oil is "somewhere between $35 and $65 a barrel." [Financial Post, 5/22/08].

The average price of West Texas Intermediate during 2007 was $72/barrel. If you compare that with the numbers that these authorities are suggesting the price of oil should be, they appear to be claiming that, based on fundamentals, the price of oil should have fallen rather than risen in 2008.

If anyone claims that the price of oil today should be no higher than it was last year, then I think it's reasonable to ask them also to provide us with the following detail underlying their assertion-- If oil were selling today for the same price as last year, what would be the quantity demanded?

Here I offer some calculations to illustrate why I think it's necessary to ask this question. I'll use data for the first quarter of this year, because that's the most reliable currently available. U.S. real GDP was 2.5% higher in 2008:Q1 than in 2007:Q1. With an income elasticity of 0.5, I would therefore have anticipated something like a 1.25% increase in quantity demanded if the price had stayed where it was.

The U.S. consumed 20.8 million barrels/day of crude oil and petroleum products in 2007:Q1, so with constant prices I would have expected consumption in excess of 21 mb/d for 2008:Q1. In fact, we consumed not 21 mb/d in the first quarter of this year but instead 19.9 mb/d. That 1.1 mb/d net drop in quantity demanded relative to baseline growth is surely related to the fact that the price of oil did not remain at its value of the previous year, but instead averaged $98/barrel during the first quarter of this year, 36% above the average value during 2007, consistent with a short-run price elasticity of

(21 - 19.9)/21 x 72/(98 - 72) = 0.15

If the price had stayed the same, the quantity demanded would have been significantly higher than what we observed. That invites the follow-up question-- Where is the physical product to satisfy this extra demand supposed to have come from?

If oil producers had been selling us 21 mb/d but we were only consuming 19.9 mb/d, that would imply that in excess of 1 mb/d would have been added to inventory during the first quarter. An extra million barrels going into inventory each day would have increased inventories by 90 million barrels by the end of March 2008. But EIA estimates that U.S. stocks of crude oil and petroleum products in March were 1,653 million barrels at the end of March 2008, compared with 1,662 at the end of December 2007 or 1,677 at the end of March 2007. Inventories did not increase, they fell, over this period.

We were only able to buy 19.9 mb/d in the first quarter when we offered a price near $100. So why would it have been possible to secure the 21 mb/d that consumers would likely have wanted at a price of $72?

Given these data, I think it is impossible to argue that the volume of futures market purchases alone could be the reason why oil prices went up this year. A key and necessary element of any speculation-based interpretation must be some explanation for the factors governing the physical quantity of oil being supplied to the market.

We're hearing from a number of experts asserting that there's no reason why the oil price should have gone up. I wish one of them would tell me where an extra million barrels per day in supply is supposed to come from.