Published on: 2026-08-12
Updated on: 2026-08-12
The 3-2-1 crack spread compares the cost of three barrels of crude oil with the value of two barrels of gasoline and one barrel of distillate, expressed per barrel. It is the standard proxy for gross refining margin: a widening spread means fuels are gaining value against crude, while a narrowing spread means they are losing it. That is why fuel prices can stay high while oil falls; the constraint sits in the refining system, not the crude market.
The 3-2-1 crack spread compares three barrels of crude oil with two barrels of gasoline and one barrel of distillate fuel, quoted per barrel of crude.
A widening spread means refined products are gaining value relative to crude; a narrowing spread means they are losing it.
The spread is a proxy for gross refining economics, not the net profit a refinery actually earns.
Crack spreads reveal downstream supply pressure that an outright crude oil chart cannot show, including why falling oil does not always mean cheaper fuel.
A crack spread is the price difference between crude oil and the petroleum products refined from it. The name comes from “cracking”, the refining process that breaks heavier hydrocarbons into lighter products such as gasoline and diesel.
The 3-2-1 structure uses a simplified product slate:
3 barrels of crude oil → 2 barrels of gasoline + 1 barrel of distillate fuel
The ratio is a benchmark, not a literal refinery yield. Real refineries also produce jet fuel, petrochemical feedstocks and other products, and their output mix varies with configuration and crude quality. The 3-2-1 simply standardises how the value of the two biggest transport fuels is tracked against their raw material.
Distillate covers fuels such as diesel and heating oil.
The common U.S. version prices crude per barrel and products per gallon, using NYMEX RBOB gasoline and ultra-low sulfur diesel (ULSD) as benchmarks. One barrel holds 42 U.S. gallons, so the product prices are converted into barrel equivalents first.
3-2-1 Crack Spread = [(2 × gasoline price × 42) + (1 × ULSD price × 42) − (3 × crude price)] ÷ 3
For example:
Crude trades at $70 per barrel, gasoline at $2.50 per gallon and ULSD at $2.80 per gallon.
Two barrels of gasoline: 2 × $2.50 × 42 = $210
One barrel of ULSD: 1 × $2.80 × 42 = $117.60
Three barrels of crude: 3 × $70 = $210
Spread: ($210 + $117.60 − $210) ÷ 3 = $39.20 per barrel
That $39.20 is the theoretical gap between the benchmark product basket and its crude input, before any refinery costs. The same method works with other regional crude and product benchmarks.
A widening spread means gasoline and distillate are strengthening relative to crude oil; a narrowing spread means they are weakening. Refined products have their own supply and demand, so they do not have to move in step with oil.
| The spread widens when | The spread narrows when |
| Fuel inventories tighten | Product inventories build |
| Gasoline or diesel demand strengthens | Fuel consumption softens |
| Refinery maintenance or outages cut fuel output | Refinery production increases |
| Crude falls faster than product prices | Crude rises faster than product prices |
Seasonality pulls the two product legs in different directions. Gasoline cracks often firm into the U.S. spring and summer driving season, while distillate margins tend to find support in colder months from heating demand.
The spread therefore reads differently from crude alone. A $10 fall in oil means one thing if gasoline drops $10 alongside it, and something else entirely if gasoline barely moves. And if crude is falling while the spread widens, the pressure has moved downstream rather than disappeared.
Crude is the raw material, gasoline and diesel are finished products, and refining capacity sits between the two. If crude supply improves and oil prices fall while refineries are offline or fuel inventories are thin, there is still not enough gasoline or diesel reaching the market. Product prices hold up even though the input became cheaper.
A refinery outage can even push the two markets in opposite directions at once. An offline refinery needs less crude, which weakens crude demand, while producing less fuel, which supports product prices. One disruption, two opposing price moves; which is why crack spreads are watched closely around hurricanes, maintenance seasons and unplanned shutdowns.
The 2020 demand shock showed the same separation in reverse. EIA weekly data recorded U.S. gasoline demand falling from about 9.7 million barrels per day in the week ending 13 March 2020 to 5.1 million by 3 April, and gasoline crack spreads briefly turned negative as gasoline fell faster than crude. Distillate demand declined less severely, and its firmer margin kept the overall 3-2-1 spread positive.
No. A common mistake is to treat a $30 or $50 crack spread as the amount a refinery earns on every barrel.
The spread is a gross refining-margin proxy. Actual profitability also depends on operating expenses, energy use, maintenance, transport, crude quality and the value of the rest of the product slate; the EIA notes that crack spreads exclude these other variable and fixed costs.
Nor is there one crack spread for every refinery. A Gulf Coast plant running one crude grade faces different input economics from a refinery elsewhere pricing off another benchmark. Treat the spread as a measure of relative value, not a refinery income statement, and not a standalone directional signal for crude, since a wide spread can close from either side.
There is no fixed level that always counts as high. Crack spreads vary with crude prices, seasonal fuel demand, refinery capacity and product inventories. Comparing the current spread with its own historical and seasonal range is more informative than any single permanent threshold.
The numbers describe the benchmark ratio: three barrels of crude oil are compared with two barrels of gasoline and one barrel of distillate. The mix approximates a common refinery product slate rather than the exact output of any individual refinery.
Not necessarily. A wider spread means gasoline and distillate are valuable relative to crude. It can widen because fuel prices rise, because crude falls, or both at once. The spread measures the relationship between the two markets rather than predicting the next move in either one.
Yes. If product prices fall far enough relative to crude, an individual product crack or the full spread can drop below zero. U.S. gasoline crack spreads briefly turned negative during the demand collapse of spring 2020, while distillate strength kept the 3-2-1 spread positive.
The 3-2-1 crack spread measures the refining layer that sits between crude and the pump: three barrels of oil in, two of gasoline and one of distillate out. Its trade-off is that it shows relative value, not refinery profit, and it says nothing on its own about where crude goes next. When oil and fuel prices stop moving together, it is the number that shows why.
