Why has the 3:2:1 crack spread widened to 54.48 this month?

The 3:2:1 crack spread is 54.48, up 8.94 over the past 30 days. The crack is the refiner's rough gross margin on a barrel: buy three barrels of crude, sell two of gasoline and one of distillate, keep the difference. A nine-dollar move in a month means the gap between what crude costs and what fuel sells for has opened up, and the question is which side did the moving.

Crude did most of it. WTI is down 28.4% over 30 days. Brent is down 26.3% over the same stretch. WTI is now sitting a nickel above its 30-day low of 68.58. Whatever else happened this month, the price of the thing a refiner buys fell by more than a quarter.

The products held up far better. ULSD diesel futures are down 15.4% over the month and RBOB gasoline is down 11.4%. Both fell. Neither fell anywhere close to as fast as the crude they are made from. Since the crack is product prices minus crude, product prices dropping at roughly half the speed of crude is the whole nine dollars. The margin widened because the cost side collapsed while the sell side only sagged.

So this is not a story about fuel demand surging and pulling product prices up. Gasoline and diesel did not rally. They came down. They just came down slower than the barrel underneath them, and the space that opened between the two lines is what a refiner captures.

The reason crude fell is sitting in the wire this week. OPEC production jumped in June, with the 11 members in Reuters' survey pumping 19.43 million barrels per day, up 3.3 million bpd from May. May was the lowest reading that survey has recorded since at least 2000, because the Strait of Hormuz disruption had knocked Gulf flows sideways. Those barrels are coming back online. More crude reaching the market with the war premium bleeding out is exactly the kind of move that drops a price 28% in a month.

Products came down too, and I looked for why they did not come down as hard. Part of the answer may be on the refining side rather than the crude side. Drones hit Lukoil's Nizhegorodnefteorgsintez refinery at Kstovo, which suspended operations on July 2. A refinery that is offline is not turning crude into diesel and gasoline. When crude supply is loosening but refining capacity is taking hits, the barrel gets cheaper faster than the fuel does, and that split is the refiner's margin. I want to be careful here. Kstovo is a Russian plant, and its diesel does not land at a U.S. rack. The effect on the products in our numbers is indirect, through global distillate balances, not a direct pipeline. It fits the shape of what the spread is doing, but I would not hang the whole nine dollars on one refinery.

So where is the margin landing. The crack lands at the refinery gate, and 54.48 is a strong month for a refiner. A plant buying crude near a 30-day low and selling gasoline and distillate that only gave back a third to a half as much is earning a wide margin on every barrel it runs. Refiners took real exposure through the Hormuz scare, and this is the kind of spread that pays them back for running hard while crude was volatile. Good month, earned.

The retail side tells a different part of it. The diesel retail-wholesale spread is 1.336 now, down 0.375 over the past 30 days. It is the marketer's cut between what diesel costs at the rack and what it sells for at the pump, and it narrowed. U.S. retail diesel is down 12.7% over the month while wholesale diesel dropped 15.4%. Retail is passing the crude drop through to the buyer, and the marketer's own margin on that gallon got thinner while it did. The refiner's margin widened by nine dollars a barrel over the same month the diesel retailer's margin shrank by about 38 cents. The wide crack this month is a refining number, not a retailing one.

I do not have a retail spread for gasoline, only for diesel. Retail gasoline is down 10.7% and RBOB is down 11.4%, which are close enough that the pump and the rack fell together and the marketer's gasoline margin probably held roughly flat. That is a guess from two percentages, not a spread I can read directly, so I will leave it there.

What I am sure of: the crack widened to 54.48 because crude fell about twice as fast as product prices, OPEC barrels coming back off the Hormuz disruption drove the crude leg, and the margin is landing at the refiner. What I am less sure of: how long it lasts. WTI is a nickel off its 30-day low right now. If OPEC keeps returning barrels the crude leg could keep falling, which would hold the crack wide, but if products finally follow crude down or the refinery outages clear, the gap could close as fast as it opened. A crack this wide is a signal to run crude while it lasts, and it may not last.

And that was just the data. See you tomorrow.