Why has the 3:2:1 crack spread narrowed to 59.31 this month?

The 3:2:1 crack spread is 59.31, down 11.06 over the past month. That number stands in for refiner margin on paper, three barrels of crude turned into two of gasoline and one of diesel. The question is what pulled it down, and where the margin went instead.

The obvious read is that the products got cheaper. They did not. Gasoline is 4.61, up 9.3% over 30 days, off a low of 2.907. Diesel is 6.529, up 15.5%, off a low of 3.459. Both legs of the output rose, and diesel rose hard. So the crack did not narrow because the finished fuel lost value.

Crude is the other side of the spread. WTI is 92.41, up 12.4% over the month, with a low of 68.25. Brent is 97.44, up 10.9%. Crude climbed too, and it climbed faster than gasoline. Gasoline is two of the three barrels in the 3:2:1, the heaviest leg. When crude rises more than gasoline, the gap between them shrinks, and that gap is most of this move.

Diesel breaks the pattern. It rose 15.5%, more than crude and more than any other benchmark on the desk this month. One of the three barrels is diesel, so it pushes the crack the other way, but two barrels of lagging gasoline outweigh it. The blended number came down anyway.

There is a reason diesel ran while gasoline trailed. Ukraine struck Russia's Ilsky refinery this month and set it on fire, per the wire. A refinery offline pulls barrels out of the global pool, and it pulls distillate first. At home the harvest is starting, and a North Carolina report has Triad farmers absorbing higher diesel as fall harvest begins. Tractors and freight both run on diesel, so the demand lands right as the supply side tightens. Diesel is the leg where a real event moved a real price.

Natural gas belongs in this too, even though the crack ignores it. Henry Hub is 3.225, up 13.5% over the month. Storage is 3351, up 5.2% and near its 30-day high of 3364. The 3:2:1 counts crude in and fuel out and nothing else, but a refinery burns gas for process heat and buys hydrogen made from it. Gas up 13.5% means the paper crack overstates what a refiner actually kept. Real margin is thinner than 59.31 by an amount the spread alone does not show.

Now to where the margin landed. The diesel retail-wholesale spread is 1.639, up 0.255 over the month. It measures the gap between what a station pays for diesel and what it charges, and it widened while the refiner crack shrank. Margin moved down the chain. The paper margin at the refinery gate came off; the retail cut on diesel went up.

A refiner making more diesel than gasoline fared better this month than one weighted toward gasoline, since diesel held while gasoline lagged crude. The farmer bringing in the harvest and the hauler moving freight carry the cost, paying 6.529 for diesel that traded as low as 3.459 inside the same 30 days.

So what am I sure of. The crack narrowed mainly because crude outran gasoline, and gasoline is the heaviest leg of the 3:2:1. Diesel was the exception, held up by a refinery fire abroad and harvest demand at home, but one strong barrel could not lift a spread that two weaker ones were dragging down. On those points the numbers are clear.

What I am less sure of is whether any particular refiner actually earned less this month. The 3:2:1 is a blended benchmark, not any one company's book. A plant that buys a cheaper crude than WTI, or makes more diesel than the recipe assumes, could have held its margin or grown it while the headline crack fell. Rising natural gas cuts the other way and is not in the spread at all. The number tells you the direction of the average and very little about the edges.

Where the margin is landing looks clearer than why. It is shifting toward diesel and toward the retail end, away from the gasoline-heavy refiner. The Ilsky fire and the harvest could keep diesel firm for a while, though I would hedge that; outages get repaired and harvest demand passes. If diesel softens and crude holds, the crack could widen back out on its own.

And that was just the data. See you tomorrow.