Research Note
The Byproduct
A quarter of Lower 48 gas supply is drilled for oil, not for gas.
Brent fell from $96.92 a barrel on 21 August to $88.24 on 25 August, after Iranian and Omani officials were reported to have agreed a temporary joint maritime corridor through the Strait of Hormuz. The Henry Hub gas price fell over the same days, from $2.82 to $2.70 per million British thermal units. Two commodities, one piece of news, both lower.
The second move is the one worth stopping on. Close to a quarter of Lower 48 natural gas supply now comes out of the Permian basin, where almost nobody drills for gas. It arrives because somebody drilled for oil and the gas came up with it. News that lowers the oil price lowers the return on the drilling that produces that gas, which raises a question the gas price did not appear to be asking: how much of American gas supply is a byproduct of the oil business, and what is the forecast for that supply assuming about the oil price?
What the Gas Market Is Reading
The number the gas market is judged on is production, and it keeps going up. The Energy Information Administration's August forecast has US dry gas production at 107.6 billion cubic feet per day in 2025, rising to 111.2 in 2026 and 116.0 in 2027. Gross LNG exports rise over the same period from 15.1 to 18.6 billion cubic feet per day, which is a record export book by some distance, and the Henry Hub price still falls: $3.53, then $3.44, then $3.31.
That is a coherent picture and it is the one most readings of the gas market start from. Supply is growing faster than even a record export programme can absorb, so gas stays cheap. What the picture leaves out is where the supply comes from, and the composition has changed enough over ten years to matter more than the total.
A Quarter of the Supply Is Drilled for Oil
Appalachia is still the largest gas region in the country and has been for a decade. It is no longer where the growth is.
A Quarter of Lower 48 Gas Now Comes Out of an Oil Basin
Marketed natural gas production by region, billion cubic feet per day, 2015 to 2025
The Permian went from 8.1 percent of Lower 48 marketed production in 2015 to 23.9 percent in 2025. Appalachia fell from 32.9 percent in 2018 to 31.7 percent while still growing in absolute terms.
Source: US Energy Information Administration, Short-Term Energy Outlook.
Permian marketed gas production went from 6.0 billion cubic feet per day in 2015 to 27.6 in 2025, which took it from 8.1 percent of Lower 48 supply to 23.9 percent and past the Haynesville on the way. Appalachia grew too, from 33.1 in 2020 to 36.7 in 2025, but that is 11 percent over five years against the Permian's 73 percent.
The Permian is not a gas basin. It produced 6.60 million barrels of crude a day in 2025, close to half of all US output of 13.59 million. The wells are drilled on oil economics, permitted on oil economics and paid for out of oil revenue. The gas is a coproduct that has to be sold, flared or reinjected, and in a basin with pipeline capacity it gets sold at whatever the market offers.
The Year That Tested It
That is an assertion until something separates the basins that respond to the gas price from the basins that do not. 2024 did. The Henry Hub price fell 13.5 percent between 2023 and 2024, from $2.54 to $2.19, while Brent was close to flat at $82.41 and then $80.56. A gas price signal, more or less on its own.
When the Gas Price Fell, Only the Gas Basin Cut
Change in marketed natural gas production by region, 2023 to 2024, percent. The Henry Hub spot price fell 13.5 percent over the same two years
Rigs fell in both places: the Permian averaged 334 in 2023 and 308 in 2024, the Haynesville 58 and 37. Only the Haynesville's production followed them down.
Source: US Energy Information Administration; Seven Measures calculations.
The Haynesville, which is drilled for gas and nothing else, cut production 12.3 percent. The Permian raised production 9.4 percent and the Eagle Ford 4.9 percent into the same falling price, and Appalachia was flat. Rig counts came down in the oil basins as well as the gas one, which is rather the point: what separated them was not how much drilling happened but what the drilling was for.
This is worth being precise about, because it is the load-bearing evidence in the note. It does not show that Permian operators are indifferent to the gas price; they would rather have a higher one, and gas revenue is part of the well economics. It shows that the gas price is not what decides whether the well gets drilled. When the only thing that moved was the gas price, only the gas basin moved with it.
The Gas Comes Up Regardless
There is a second mechanism underneath this, and it is geological rather than commercial.
Each Permian Barrel Brings More Gas Than It Used To
Permian marketed natural gas production divided by Permian crude oil production, thousand cubic feet per barrel. 2026 and 2027 are EIA forecasts and are drawn open
The ratio has risen in every year since 2018, from 2.92 to 4.17 thousand cubic feet per barrel, an increase of 43 percent. It was flat to falling in the three years before that.
EIA's forecast, not a Seven Measures forecast.
Source: US Energy Information Administration; Seven Measures calculations.
Every barrel the Permian lifts brings more gas up with it than it used to, and by 2025 it was bringing 43 percent more than in 2018. Reservoir pressure falls as a field is produced and more of the hydrocarbon arrives at surface as gas; it is a characteristic of a maturing basin rather than a decision anybody took.
The consequence is that Permian gas grows faster than Permian oil, and would grow somewhat even if oil output were flat. That is the part of American gas supply which is furthest of all from responding to the gas price, and EIA has it continuing to 4.49 by 2027.
What the Supply Forecast Is Assuming
Put the two together and the forecast for US gas supply turns out to be, in substantial part, a forecast about oil drilling.
Half the Forecast Supply Growth Comes From Basins Drilled for Oil
Forecast increase in Lower 48 marketed natural gas production by region, 2025 to 2027, billion cubic feet per day
Differences on EIA's forecast, not a Seven Measures forecast. The Permian, Eagle Ford and Bakken are drilled for oil and account for 4.89 of the 9.46 Bcf/d increase, or 51.7 percent. Regional figures sum to 0.08 Bcf/d less than the published Lower 48 total in 2025 through EIA's own rounding.
Source: US Energy Information Administration Short-Term Energy Outlook; Seven Measures calculations.
Lower 48 marketed production is forecast to rise 9.46 billion cubic feet per day between 2025 and 2027, and a little over half of that comes from the three basins on the exhibit that are drilled for oil. The Haynesville, at 2.78, is the largest genuinely gas-directed contributor and the one piece of the growth a gas price actually commands. Appalachia adds 1.17 across the two years, which is the pipeline constraint speaking rather than a price response.
The same forecast has Brent averaging $86.81 in 2026 and $69.39 in 2027, a decline of $17.42. EIA is explicit about why the near-term number is high. It raised its estimates of Middle East shut-in production, in its words, "due to continued severe constraints on Strait of Hormuz transits, which we assume persist through August", and expects regional production to return "to near pre-conflict averages in early 2027". The high oil price is a disruption premium and the agency has written down the date it comes off.
So the largest single-year increase in Permian gas anywhere in the forecast, 2.38 billion cubic feet per day in 2027 against 1.61 in 2026, is scheduled to happen in the year the oil price that pays for it falls by seventeen dollars. And the $3.31 Henry Hub price at the end of the forecast is what that supply arriving produces. The gas price forecast and the oil price forecast are resting on each other, and they are quoted as though they were independent.
Investment Implications
The firm reads three things from this. First, that exposure to US gas supply is now substantially exposure to the oil price, and the two are underwritten in different books by different people. A long position in gas is in part a short position in Permian drilling activity, whether or not it was put on that way, and an oil price falling on a Hormuz reopening is not the neutral event for gas that 25 August treated it as.
Second, that the basins which respond to the gas price are where a gas price recovery is actually expressed. The Haynesville cut when the price fell in 2024 and came back when it rose, with rigs going from an average of 37 in 2024 to 59 in the second quarter of 2026. Appalachia, whose constraint is pipeline takeaway rather than price, does not do this and should not be underwritten as though it did.
Third, that the forecast is falsifiable and worth watching rather than arguing with. If Brent holds near $87 the associated gas keeps coming and the supply path is roughly right. If the Strait reopens and oil moves toward $69, the interesting question is not whether Permian drilling stops but how much of the 2027 supply increase was resting on it.
The work this note has not done, and which would turn a caution into a position, is the decline curve: how quickly associated gas volumes fall when Permian drilling slows, rather than whether they fall. That is answerable from well-level production data and it is the thing that would put a date on any of this.