Oil and gas producer concentration: when a company sample misses regional supply

A company watchlist is not a regional production census

Can a handful of listed oil and gas companies tell you whether an entire producing region is growing? Only after you establish how much of the relevant output the sample covers and what happened outside it. Counting the companies in a watchlist does not answer that question. A useful supply note begins with the population being measured, then separates changes in physical production from changes in who reports it.

EIA's September 22, 2026 analysis provides a historical starting point. For 2025, public companies represented 2% of about 12,000 producers while accounting for 68% of combined Lower 48 oil and natural gas output, measured in barrels of oil equivalent. The page carries a correction notice concerning production per well. We use its current version and do not reconstruct the earlier figure.

This is a guide to interpreting company samples, not a new October production release. The worksheet below uses invented numbers to show why a sample can grow even when the surrounding region contracts.

Keep operators, wells and output in separate columns

An operator count gives each operator one place in the count. A well count counts wells. An output share weights observations by the quantity produced. None of these denominators can stand in for the others without additional data. Start a worksheet with separate headings for company coverage, physical coverage and the reporting period; leave a cell blank when its source is missing.

A separate EIA report, released December 8, 2025, covers wells through 2024. It reports 918,481 producing U.S. wells in 2024 and says wells producing 15 barrels of oil equivalent per day or less represented 78% of wells in 2023–2024. Those observations do not measure 2025 Lower 48 operator ownership.

The wells report also describes varying state data completeness and the use of earlier observations for late-reporting states. Both EIA production documents draw on Enverus data, so they are not independent confirmations of the same estimate. Do not divide the national well count by the later operator count to manufacture an average: the dates, geographical coverage and counted entities differ.

Measure coverage in the region and fuel you actually follow

EIA's ownership analysis identifies a regional counterexample to the Lower 48 pattern: private companies accounted for 55% of combined oil and natural gas production in the Haynesville in 2025. Applying the overall public-company share to every producing region would therefore obscure a material difference.

For your own sample, define the boundary before computing coverage. Is the question about one region or all domestic production? Is it about crude oil alone, natural gas alone, or combined output? Do the company figures cover the same interval? Keep the numerator restricted to the chosen boundary instead of importing an entire company's output because its name appears on the regional watchlist.

EIA's units explainer lists barrels of oil equivalent among the common units used to compare different fuels. Such conversion supplies an energy basis, not a revenue or earnings weight. A production sample designed to examine gas supply should retain a gas-volume view; a combined energy share cannot by itself establish the sample's exposure to gas prices.

Calculate the missing sample before projecting growth

Consider a fictional region producing 100 units over one quarter, with every unit measured on the same basis. Three listed companies in your sample produce 60 units, while all producers outside the sample produce 40. Assume the boundary and sample membership stay fixed. These figures are a teaching example, not estimates for any actual basin.

Next quarter, sample output rises to 66 units, a 10% increase. Output outside the sample falls to 32, a 20% decline. Total regional production becomes 98, down 2%. Extending the sample's 10% growth to the full region would instead predict 110 units, overstating the actual illustrative total by 12 units.

The weighted calculation reaches the same answer: the sample's initial 60% share times 10% growth contributes 6 percentage points; the remaining 40% times negative 20% contributes negative 8 points. Together they produce negative 2%. Use initial-period weights for this two-period calculation, rather than the sample's larger share of the ending total.

If output outside the sample is unknown, do not silently assign it the sample growth rate. Record that assumption explicitly and test alternatives. In this example, outside output of 34 would keep the regional total at 100. That is a 15% decline from its starting 40. This threshold shows exactly how much missing evidence could reverse the conclusion.

Remove ownership transfers from the physical growth story

Now reset the fictional region to its original 100 units: the sample accounts for 60 and producers outside it account for 40. Suppose a sample company buys assets producing 10 units from an outside producer. Assume production at every asset stays unchanged and use a full-period comparable allocation to isolate the effect.

After the transfer, the sample accounts for 70 units and the outside group for 30. The sample has grown by about 16.7%, while the region remains at 100. No additional production is required to explain the sample increase. Treating those transferred units as new regional supply would count an ownership change as a physical expansion.

For an actual transaction, ask for the effective date, reporting boundary and a comparable-period bridge. A partial-quarter contribution cannot be compared mechanically with a full-quarter run rate. Keep acquired output, disposed output and changes from the continuing asset set on separate worksheet lines. If disclosures do not support that separation, mark organic growth as unresolved rather than inventing a precise adjustment.

Write a conclusion that matches the coverage

A concise research note should name the region, fuel, period, sample output, total output and coverage source. Add a membership-change column and identify any unresolved outside production. This gives the next reviewer a specific task: check the missing population, reconcile an acquisition, or replace an incompatible denominator.

The strongest counterevidence to a sample-based supply thesis is often another part of the same population. In the first hypothetical case, the omitted producers change the direction of the regional result. In the second, an ownership transfer explains the entire sample increase. Neither conclusion requires a price forecast or a claim about a particular company's profitability.

Our related production-measurement guide addresses gross, marketed and dry gas volumes. Here the question is who is included in the sample. Once both the measurement and population are consistent, company disclosures can inform a regional analysis with their coverage and limitations visible. The cover is an existing conceptual research illustration, not a chart of observed producer shares.