A couple of points that are often overlooked. The US oil industry in years past, outside the five year window, has carried significantly more gasoline and crude oil inventories, both in terms of Days of Supply and in absolute numbers.

Following are the Days of Supply and absolute numbers, for late April this year and for late April for the first available EIA data for gasoline and crude oil:

Gasoline:

Late April, 1991: 28.9 days (206 mb)

Late April, 2007: 20.8 days (193 mb)

Crude Oil:

Late April, 1983: 31.6 days (365 mb)

Late April, 2007: 22.1 (336 mb)

IMO, the industry has basically gone to a just in time inventory system, especially for crude oil--presumably because of the SPR--but we don't have a SPR for gasoline.

Also, the latest EIA data, for January, show a one mbpd drop in world crude + condensate production (relative to 5/05), which presumably translates to around 1.1 mbpd or so less refined product on the markets, relative to 5/05.

Fundamentally, the problem we have in the US is the expectation of a continued exponential increase in our crude + product imports, while the new emerging reality is an exponential decline in crude + product exports.

This is our expectation (Khebab's chart): http://www.theoildrum.com/uploads/28/Data_4weeks.png

Re: Stock coverage.

The number of days for crude oil stocks is still close to the average value (22 days) observed for the period 1991-2006:



but gasoline stocks are close to an all time low:



IMO, the key difference for crude oil is the SPR. Prior to the SPR, it only makes sense that refiners had to keep more oil on hand. With hundreds of millions of barrels of crude oil sitting in salt domes on the Gulf Coast, why tie up all of that capital?

WT, what's your feeling on the use of the "SPR" as a "PPR", or political petroleum reserve. instead of strategic uses , increasingly it will be a "tide over" facility in the near future, to fill gaps in oil supply until it's depleted. then, TSHTF.

In effect, the release of emergency reserves is the new "swing producer."

As you suggested, the key problem is the release of oil from the SPR based on near-cornucopian assumptions that we won't peak for decades to come.

DOE is refusing to buy oil because it is too expensive for the SPR. So we will expect even higher price swings?