Far be it from me, some guy on the internet, to nitpick, but a slightly better way to draw the graph might be to equalize the units at some point, so as to make the scale on both Y axes the same.
It sort of illustrates the same point, the divergence between the oil price and the unleaded price, but everything is on the same scale and the axes are calibrated the same.
This graph might be even better:
which is the difference between WTI and unleaded for the same period. If you were to do this graph for the period between 1996 and 2000, the margin might be 10 cents or less for most of it.
Anything less than about 10 cents/gallon, the refiner is just about losing money on the variable manufacturing cost, which is the labor and utilities it takes to run the reactor. At about 20 cents, maybe they break even economically, when you consider capital costs and interest expense, and enough profit to justify being in the business. Obviously, different refiners have different break even points, so the cost structure at Murphy Meraux might be about 20 cents, per the above article, and some of the big, well run, efficient refiners, if any exist, might be less than that.
What you can see for sure is that since early in 2004, this business has been really variable. They fatten up for awhile, and for some short periods, they have even gone negative.
But, what we are saying is that at some point, evidently starting about now, if the margins are not high enough, the refiners will shut down for awhile and do maintenance until the pricing gets better.
Another point is that the refinery margins need to be high enough, for long enough, to justify reinvestment and capacity expansion. We hear all the time that the last greenfield refinery in the US was built in 1976 or something. The reason is: the refinery economics need to be consistently strong enough to justify risking multiple billions of dollars to construct the unit. As it is, even in this regime of higher prices, the pricing is still not consistent enough to do this, rather than much safer investments that can return more money, such as high-tech companies, or some factory in China.
So, you get no new refineries. People might expand existing plants if the economics justify it, but in general it puts us in the terrible situation we are now in, with little or no spare capacity, and a collection of 40-plus year old refineries which fall apart every spring, trying to provide the nation with the unleaded it needs.