One Piece at a Time - The Numbers Behind U.S. Crude Oil Balances and Inventories

Last week, crude oil prices dropped below $50/bbl, in part due to continued increases in U.S. crude oil inventories, and fell further over the next few days. Then yesterday, prices perked up by $1.14 to $48.86/bbl; again one of the factors was the weekly inventory number from the Energy Information Administration which showed inventories down by a fraction of a percentage point for the week. The market seems to react spontaneously to changes in that crude-stocks statistic. Up is bearish, down is bullish. These days even a very modest decline in inventories is bullish. But serious analysis requires a more detailed, more nuanced understanding of why crude oil inventories behave as they do. Were inventories driven up by higher production or lower refinery runs? By higher imports? By lower exports? The reasons behind the inventory change are more important than the change itself. Today we continue our series on the modeling of U.S. crude oil supply and demand, and the sourcing of input data used in those calculations.

The latest sharp drop in crude oil prices, which was blamed in part on unexpected gains in already record-high U.S. inventories, is a stark reminder of the importance of understanding and routinely calculating estimates of the oil supply/demand balance. Only by keeping up with the ever-changing relationship between crude availability and crude consumption—and by anticipating shifts in that relationship—can oil traders and others whose daily success or failure depends on crude pricing trends make informed decisions. Today we begin a blog series on the modeling of U.S. crude oil supply and demand, and the sourcing of input data.

The numbers don’t add up. Literally. The most closely watched energy statistics in the world have a problem, and it’s been getting worse over the past two years. We’re talking about EIA’s U.S. crude oil supply, demand and inventory balances, which are published each week and then trued up about 60 days later in monthly data. The problem is that the balances don’t balance. EIA uses a plug number alternatively called “adjustment” or “unaccounted for” to force supply and demand to equate. That would not be an issue if the plug number was small and flipped frequently from positive to negative, likely due to timing inconsistencies with the input data. But that’s not the case. The number is mostly positive, meaning more demand than supply. And the difference can be mammoth: last week it was 2.3 MMb/d, or 18.4% of U.S. crude production. It seems like barrels are somehow materializing out of nowhere. But now we know where, because EIA just finished a 90-day study of the crude imbalance that reveals the sources of the problem and what it is going to take to fix it. In today’s RBN blog, we will delve into what has been causing the problem, what it means for interpreting EIA statistics, and what EIA is doing to address the issues.