The Crack Spread: How Refiners Actually Make (and Lose) Money Turning Crude Into Fuel

In October 2022, the diesel side of the refining business printed a number that veterans still talk about. The distillate crack spread hit roughly $1.72 a gallon, more than four times the prior five-year average, and the headline 3-2-1 spread that summer had already blown past $60 a barrel from its usual $10-to-$20 home. Refiners were not smarter that year. They were scarcer. That is the whole story of the crack spread in one sentence: it does not measure how good a refinery is, it measures how badly the world needs what a refinery makes.

What the 3-2-1 actually is

Strip away the jargon and the crack spread is a subtraction problem. You take the price of the fuels a refinery sells, subtract the price of the crude it buys, and the difference is your gross margin per barrel. The reason it is called a "crack" is literal: refineries crack long crude molecules into lighter product molecules.

The 3-2-1 version, which the U.S. Energy Information Administration treats as the industry proxy, uses a ratio: three barrels of crude in, two barrels of gasoline out, one barrel of distillate out. That is not an arbitrary choice. It roughly matches the real yield slate of a typical U.S. refinery, which produces something close to two barrels of gasoline for every barrel of distillate. Match the math to the plant and the number means something. Refineries built for a heavier diesel slate get modeled with other ratios, and you will see 5-3-2 and 2-1-1 spreads quoted for exactly that reason.

Doing the arithmetic

The trap for newcomers is units. Crude trades in dollars per barrel. Gasoline and distillate futures trade in cents per gallon, and there are 42 gallons in a barrel. So you cannot subtract them directly. You convert the products to a per-barrel basis first, using the standard formula taught in Penn State's energy markets course:

• (2 barrels x 42 gallons x gasoline price per gallon) plus (1 barrel x 42 gallons x distillate price per gallon), minus (3 barrels x crude price per barrel), then divide the whole thing by 3.

Run it with the course's worked example. WTI crude at $84.54, RBOB gasoline at $2.57 a gallon, heating oil at $2.79 a gallon. Two barrels of gasoline and one of distillate come to $79.44 more than three barrels of crude. Divide by three and you get a crack spread of about $26.48 a barrel. That is the gross line.

The word gross is doing heavy lifting. The crack spread ignores every refining cost except the crude itself, including labor, catalysts, natural gas to run the units, and the price of compliance credits. Those variable costs can run on the order of $20 a barrel. A $26 crack can be a comfortable business or a marginal one depending on the plant, and it never captures fixed costs at all. Treat the spread as a revenue proxy, not a profit statement.

Why it blows out, and why it collapses

Crack spreads move on the gap between fuel demand and refining capacity, not on the price of oil. This is the part outsiders get backwards. Expensive crude does not help refiners. What helps refiners is when there is not enough refining to go around.

2022 was the textbook blowout. Roughly a million barrels a day of U.S. refining capacity had closed or converted during the pandemic. Europe was scrambling to replace Russian diesel after the invasion of Ukraine. Demand had snapped back. Crude was expensive, but refining was the actual bottleneck, so the spread on the diesel side exploded to that $1.72-a-gallon record. When the constraint is the plant, the plant owner gets paid.

The reverse happens when capacity catches up. By September 2024 the EIA reported global refining margins had fallen to multiyear seasonal lows, running below their 2019-to-2023 averages since that spring. The drivers were the mirror image of 2022: softer product demand, U.S. distillate consumption down about 6 percent year on year over the summer, weaker economic activity in China and Europe, and a wave of new capacity from Kuwait's Al-Zour and Nigeria's Dangote refinery coming online. Add supply, subtract demand, and the crack compresses.

How refiners lock it in

A refinery is long a bet it did not choose to make. It buys crude today and sells fuel weeks later, exposed to both prices moving against it in between. The crack spread is also the tool to neutralize that.

The trade is to "crack the spread" in the futures market: buy crude futures and simultaneously sell gasoline and distillate futures in the same 3-2-1 proportion. That locks in the margin between input and output regardless of which way flat prices run. If crude spikes, the long crude leg pays; if fuel prices sag, the short product legs pay. Exchanges list crack spread products directly so a refiner can put on the whole structure without legging into three separate contracts. It is a margin hedge, not a directional bet, and a refiner that has hedged forward will keep running hard even when the cash market looks ugly, because its margin was banked months ago.

The same instrument works for speculators and for the physical merchants who move product between regions. When they buy the crack they are betting refining stays tight; when they sell it they are betting the squeeze eases.

Why the Gulf Coast, Rotterdam and Singapore never match

There is no single global crack spread, because there is no single crude and no single fuel demand. Each hub prices off its own crude and its own product barrel.

The U.S. Gulf Coast benchmarks against Light Louisiana Sweet, models a gasoline-heavy slate to match American driving demand, and exports aggressively out of enormous capacity. Rotterdam, the trading heart of the Amsterdam-Rotterdam-Antwerp region, prices off Brent and skews toward distillate because Europe runs on diesel. Singapore benchmarks against Dubai, the region's medium-sour crude, and hosts a mix of simpler plants serving Asian demand. Product mix, crude quality and logistics all differ, so the spreads differ.

That divergence is the trade. When the European distillate crack runs richer than the U.S. one, cargoes move across the Atlantic to chase it, and the flow itself eventually pulls the two spreads back toward each other. Watch three regional cracks at once and you are watching where the world's marginal barrel of fuel wants to go.

The number to actually watch

The crack spread is a quick-and-dirty gauge, and honest analysts call it exactly that. It leaves out most of a refinery's costs, flattens a complex product slate into three numbers, and tells you nothing about a specific plant's efficiency. What it does, better than any other single figure, is tell you in real time whether the refining step is scarce or abundant. When the 3-2-1 is sitting in the teens, refining is a normal, grinding, cost-controlled business. When it rips toward $60, the market is telling you the barrels of crude are fine and the shortage is everything downstream of the crude. In 2022, the whole world learned that lesson at the pump.