Friday, October 14, 2011

Brent is a Better Reference Price than WTI

During the discussion of Brent vs WTI following my latest monthly global oil supply post, commenters raised a couple of questions: how did Brent or WTI contrast with the prices received by the swing producer (Saudi Arabia), and how did they contrast with US gas prices?

The answers are reasonably clear in both cases.  Above, I show the price of Saudi Arabian Light grade from the beginning of 2000 (on the y-axis) versus either Brent or WTI on the x-axis.  Historically, Arab Light has always sold at a slight and only moderately variable discount to Brent/WTI (because Arab Light is more sour).  The black line above represents exactly equal prices.  However, in recent months, the historical relationship with WTI has completely broken down and Arab Light is selling for much more than WTI.  The relationship to Brent is unchanged.

Similarly if we look at the average weekly retail price of US gasoline versus WTI (top panel) and Brent (lower panel):

you can see that the correlation with Brent is a bit tighter and in particular the current anomaly in the WTI graph where gas prices are too high relative to WTI (the "loose end of the string" at about ($80, $3.50)) is non-existent in the Brent graph where current prices are pretty much normal.

One could do a more thorough analysis looking at all oil blends around the globe but the data above is enough to convince me: WTI is now reflecting an anomalous situation in the US midwest and no longer even explains US gas prices well, let alone global trends.  Therefore I will redo all my monthly supply graphs using Brent as the marker and henceforth use mainly Brent in analysis.


jhm said...

This was in today's FT, ICYMI:

Oil traders unravel Cushing mystery

Peter Vesborg said...

Having established that Brent is the better metric, are you going to make a Brent-based version of the (excellent) oilproduction + spot price combined graph in future posts?

Stuart Staniford said...

Peter - yes, that's the plan.