After my 1700 word rant the other night, I thought I’d tell you the same story with two figures.
The first figure shows you two things – the difference between crude oil prices in the Midwest US (WTI in PADD 2) and the US Gulf Coast (Brent or LLS in PADD 3) is shown in blue, with the widening Brent-WTI spread as the rising portion on the right. In red, you see the difference in gasoline prices, in $/gallon, scaled on the right axis. Through 2011, averaged weekly, crude oil in the US midwest was $15.75/bbl LESS expensive than on the Gulf Coast, while gasoline was, on average, 11.1c/gallon MORE expensive. The historic depreciation in crude oil in the midwest over the past year actually coincided (correlation, not causation) with higher gasoline prices. The real story, however, is that gasoline differentials do not track local crude differentials.
The second image gives some indication as to why this may be the case – the market for refined products is a national one, and so local refiners and not local consumers are able to profit from the depressed local crude prices. This figure shows the 3:2:1 crack spread, or the difference between the value of a hypothetical refinery output (2/3 of a barrel of gasoline and 1/3 of a barrel of distillate fuel oil) and a barrel of crude.
This isn’t a matter of collusion or market power, it’s simply a matter of a broad market for refined products not affected in the same way by pipeline shortages and transportation bottlenecks which affect crude oil prices in one region but not another. Since refiners can sell their product in higher priced markets, they won’t sell locally at a discount, and so margins increase. Obviously, over time, gasoline prices track world crude oil prices very closely – they just don’t necessarily track local ones. Think about this before you decide that blocking export pipelines and stranding crude in Canada, which will lead to a discounted Canadian crude oil price, will translate to savings at the pump in Canada.