Every fuel price fell last month, so why did refining margins improve?

The 3:2:1 crack spread is 54.76, up 5.05 over the past 30 days. The spread is the refiner's rough gross margin: three barrels of crude in, two barrels of gasoline and one of diesel out, priced at the difference. A move that size in a month is worth tracking down. So I went looking for what moved, and where the extra margin is landing.

Crude is the first place to look, because the crack starts with the cost of the barrel going in. Brent is down 20.0% over the month. WTI is down 19.4%, sitting just above its 30-day low of 69.23. Both benchmarks lost about a fifth of their value in 30 days. ING's read on the wire is that prices overshot to the downside, with the war risk premium coming out of valuations as a U.S.-Iran peace deal calmed the market. Whatever the cause, the input cost for every refiner dropped hard.

The output side is where the spread is decided, so products are the next number. U.S. diesel is 4.832, down 12.5%. Gasoline is 4.048, down 12.1%. ULSD diesel futures are 3.1623, down 10.6%. Products fell too, but they fell less than crude did. Crude lost about 20%, products lost about 12%. The gap between those two declines is the whole story of the crack. The refiner's cost dropped faster than the price it sells at, and the difference between them widened by 5.05.

The 3:2:1 widens any time crude falls faster than the products made from it, even when every price on the board is going down. Nothing here required product prices to rise. Diesel and gasoline both got cheaper at the rack and the pump this month. The crack still widened, because the barrel going in got cheaper faster.

The harder question is where the wider margin actually lands, because a number at the refinery gate does not automatically reach anyone downstream. So I checked the retail-wholesale diesel spread, the cut the station keeps between what it pays at the rack and what it charges at the pump. It is 1.73, up only 0.095 over the month. The retailer's margin barely moved. The 5.05 of extra crack is not showing up in the retail diesel spread. Most of the gain is upstream, at the refiner buying crude and selling product, not at the marketer or the station moving gallons. The retailer kept roughly the same penny margin it had a month ago while the refiner's gross margin proxy widened sharply.

Whether that margin holds depends partly on capacity. EIA's annual Refinery Capacity Report puts U.S. operable atmospheric distillation capacity at 18.2 million barrels per calendar day on January 1, 2026, down over 250,000 b/cd, about 1%, from a year earlier. A smaller refining base makes it harder to flood the market with product as crude gets cheaper, which keeps product prices firm relative to crude and the crack wide. The 1% cut is small, but it leans the same way as the price data.

So the obvious read, that products got more expensive, is wrong. Every fuel price here is down. The spread widened because crude fell about 20% while products fell about 12%, and the eight-point gap between those declines is the refiner's gain. The margin is landing at the refinery gate, not at the pump, and the retail diesel spread at 1.73 says the station is not the one collecting it this month.

What I am sure of: the crack is 54.76, the move is plus 5.05, and it came from crude dropping faster than product, not from any fuel price rising. The refiner running cheap crude into product that held its value better had a good month, and a roughly 1% smaller national refining base gives that margin some room to persist. Right now that spread is generous, and the refiner is set up to capture it.

What I am not sure of is how long it lasts. Crude is close to its 30-day low of 69.23, with the war premium already out. If crude has finished falling, products will keep grinding down toward it and the crack will narrow back. The widening came from the speed of the crude drop, and that speed does not repeat once the price has reset. A peace deal that removed the premium is a one-time move, not a trend the refiner can bank on. Natural gas storage at 2835.0, up 10.0% on the month, is a side note here, but it says refiners running gas-heavy hydrogen and process units are not getting squeezed on energy input either, which helps net margin hold for now.

For an operator, the read is simple: this is a refiner's month, the gain is at the gate, and it is built on a crude drop that has probably mostly happened. Plan for the crack to narrow from 54.76, not widen further.

And that was just the data. See you next week.