Self-Fulfilling Prophecy?
Crude oil prices have trended up for the last several weeks, but retail gasoline prices haven't kept pace. Several commentators, including one in USA Today, have predicted that pump prices could soon jump 24 cents per gallon to catch up with crude. While it's possible that pump prices have simply lagged crude prices by that much, it's equally likely that this story could create a self-fulfilling prophecy. To see why, you have to understand the relationship between inventory and demand.

The Energy Information Agency (EIA) of the US Department of Energy tracks week-to-week statistics for crude oil and petroleum product production, imports, exports and inventory. For example, the EIA shows current stocks of gasoline at 224 million barrels, 10% higher than a year ago and well above the typical range for this time of year. Inventories aren't this high because refiners produced unusual amounts of gasoline; they mirror the relatively weak demand that is typical in the winter months, currently running at 8.9 million barrels per day. (This may explain why prices haven't gone up in line with crude.)

But getting a handle on true demand is difficult, because no one is measuring how much gasoline we actually burn every day. Instead, the EIA looks at how much was produced or imported each week, and how much inventories went up or down. This still doesn't get at end-user demand, because the volumes leaving refineries and large distribution terminals reflect deliveries to local distributors, who watch sales at the gas stations they serve and adjust their own inventories in anticipation of rising or falling demand and prices.

Even the demand at service stations is distorted by a final level of inventory: the amount of gasoline sitting in our cars every day. It gets little attention, because it rarely changes in aggregate. Assuming every car in America has a fourteen gallon tank, usually half full, this hidden inventory amounts to 40 million barrels of gasoline, equal to 20% of the primary stocks the EIA measures. If consumers all decided to keep an extra 3 or 4 gallons in their tanks, apparent demand would increase by 10-20% for a few weeks and send a false demand signal to distributors.

So if we all go out and tank up because we think gas prices are about to jump 24 cents, are we making such an increase more or less likely?