realitybase - Journal

Paul Krugman says current high crude oil prices can't be a mere speculative bubble because he can't see any evidence of hoarding. While I don't think there necessarily has to be hoarding for speculators to drive up prices, it turns out there was a lot of hoarding a year ago and still some today.

The National Petroleum Council assessed in December 2004 that the "lower operational inventory" of crude oil in the US needed to avoid refinery disruptions as 260 to 270 million barrels. At II-5. Any domestic crude inventories above this level are "discretionary."

Companies may manage their individual level of discretionary inventories based on their assessment of future market conditions. There may also be incentives based on the futures market that influence the level of discretionary inventories. For example, if futures prices for delivery in coming months are higher than the price today, an incentive may exist to build inventory because it may be worth sufficiently more later to cover the carrying cost and financial risks. Under this scenario, the market is said to be in contango. Conversely, if the futures prices for delivery in coming months are lower than the price today, companies may have an incentive to draw discretionary stocks and keep them low while maximizing sales. Under this scenario, the market is said to be backwardated.

At II-3. In Figure II-2 (at II-7), the NPC compares its lower operational inventory assessment with the same Energy Information Agency inventory data plotted here, making clear that NPC, EIA, and my earlier post all use the same definition of "inventory."

Using the NPC assessment as a floor, it turns out that inventories have been well into the discretionary zone, reaching 355 millions barrels in June 2007 and then dropping sharply to 286 million barrels at the end of December. The sharp drop of 69 million barrels suggests that the herd was liquidating physical inventory in anticipation that wet barrels would not be significantly more costly or valuable in the coming months, and might be worth less. It seems the herd was wrong. In January and February there were modest additions to discretionary inventories, which then stood about 35 million barrels above "lower operational inventory," suggesting some sentiment that prices were more likely to go up than down.

It is possible that 355 million barrels in June 2007 was close to the maximum available crude oil storage capacity in the US. "Working storage capacity at operable refineries" in the US has been tracked by EIA as of January 1 each year since 1982. This storage at refineries has shrunk from 181 million barrels in 1982 to 158 million barrels in 2007 mostly, it seems, because of the closure of about 120 small refineries between 1981 and 1986, and continuing closures since. (Whereas there were once over 300 refineries, there are now fewer than 150.) NPC at I-7. I have not been able to find data on crude storage capacity at port and pipeline terminals and other upstream storage vessels and, therefore, can't say if the reduced refinery storage capacity has been offset by increases upstream. But I have compared the amounts by which total domestic inventories exceeded refinery storage capacity at three times when inventories peaked, and at the end of February 2008:

	April 1981	June 1990	June 2007	Feb. 2008
Inventory	397	384	355	302
Refinery tankage	181	174	158	158
Difference	216	210	197	158

If upstream storage capacity has not increased, domestic crude oil inventories were nearly as close to peak storage capacity in June 2007 as they were in April 1981.

Whether crude oil prices are or not a speculative bubble is currently a hot topic on Krugman's blog.