There are, broadly speaking, two ways to make money selling oilfield services. You can be better than the other guy. Or you can arrange things so that there is no other guy. For about thirty years, the service sector tried option one and mostly failed. Competing on quality in a commoditized market is a treadmill: you invest in differentiation, your competitor copies it, pricing resets to marginal cost, and everybody goes home with a 6% EBITDA margin and a management team that describes the year as “challenging but encouraging.” The operator, meanwhile, runs the RFP, picks the cheapest bid that meets spec, and moves on to the next line item. This was a comfortable arrangement for everyone except the people selling the services.

What I want to walk through today is how three major cost categories on your AFE independently arrived at option two: make supply scarce. They did it through completely different mechanisms. They did not coordinate. And the cumulative effect on your well economics is not the sum of the three individual price increases. It is the product. That distinction matters and I will come back to it.

Game theorists have a term for what happens when independent players converge on the same strategy without coordinating: dominant strategy equilibrium. It means that constraining supply is the best move for each player regardless of what the other players do. You do not need a cartel when the incentives are this clean. You just need every participant to independently read the same page of the same textbook, which, as it turns out, they all did.

The frac companies chose it deliberately.

Liberty, ProPetro, Patterson: the diesel fleets are being actively retired even when they are still mechanically capable. 1.1 million hydraulic horsepower retired in 2024 and 2025. Zero replacement equipment built. Active frac spreads have dropped from roughly 280 at the 2022 peak to approximately 200 today. That sounds like a demand story until you look at completion volumes, which have not declined proportionally. What declined is the number of providers willing to show up and compete at spot rates.

The mechanism is elegant in a way that deserves a moment of appreciation. You take the top of the market, lock it into long-term contracts on e-fleet and simul-frac, and call it “technology differentiation” (which it partly is). This strands the bottom of the market with too few bidders to sustain spot pricing. The spot pricing collapse then justifies retiring the diesel fleets (why keep a fleet running that earns 8% returns?), which constrains total capacity, which supports the premium on the differentiated equipment, which funds more retirements. Game theorists call this Kreps-Scheinkman: firms commit to capacity levels first, then compete on price given those commitments. The frac companies figured out that the optimal capacity commitment is less. Just, less. Make less of the commodity. Charge more for what remains. This is supposed to be hard to sustain because someone always cheats and adds capacity. Nobody is cheating. The returns on cheating (deploying a diesel fleet at spot rates into a market with 200 active spreads) are worse than the returns on not cheating. The equilibrium holds.

(If this sounds familiar, it should. Liberty is now spending $3.50 on power generation capital for every $1 on frac. KGS is spending roughly $1.70 on power for every $1 on compression. Two different service categories, same arithmetic: the core business has become the harvest asset that funds a power generation buildout. The vendors converged on the same capital allocation without coordinating, which tells you everything about where the returns are.)

The OCTG mills had it done for them, which is even better.

Tariffs do not reduce global tubular supply. They reduce the *substitutable* supply available to a U.S. operator, which from your perspective is the same thing. Section 232 duties, anti-dumping and countervailing duties on Korean and Chinese product, and the cascading effect on trade flows from allied mills have compressed the viable import pool to a fraction of what it was in 2018.

There is a classic game theory model for this called Bain-Dixit entry deterrence, where the incumbent has to do something expensive to keep competitors out: build excess capacity, price aggressively, invest in switching cost barriers. The OCTG mills got a much better deal. The government did the deterring for free. The domestic mills with integrated threading and LML capability now have something close to a regional monopoly on production casing, not because they earned it through competitive excellence, but because the trade policy environment handed it to them. (This is not a criticism. If someone offers you a monopoly, you take it. That is also in the textbook.) One market became two: premium grades requiring specifications, certifications, and institutional relationships became calcified domestic territory. Commodity grades remained contestable but heavily penalized on imported substrate. The strategic question (and the one worth watching) is what the mills do with the rents. The answer is visible in every quote you have received in the last twelve months.

The compression vendors had it done for them, and then, remarkably, chose to make it worse.

Caterpillar’s 3600 in-line engine backlog created the constraint. Cat is reallocating production capacity toward power generation turbines, where a single hyperscaler signs a contract worth more than the entire annual output of the compression aftermarket. The compression companies did not ask for a 180-week lead time. But they recognized its strategic value immediately: when nobody can build a new unit for three and a half years, every installed unit is a monopoly franchise that reprices at renewal.

The textbook response to an upstream supply constraint is to diversify your supply chain. Find alternative inputs. Develop substitutes. Invest in the bottleneck. What the compression vendors actually did was the opposite. KGS spent $675 million buying a distributed power company that competes for the same Caterpillar engines that build its compression units. (Read that sentence again slowly.) USAC acquired J-W Power and redirected its manufacturing capacity toward internal fleet growth rather than third-party sales. Archrock pulled back EMD orders because the grid cannot support them, further concentrating demand on gas-drive engines with the longest lead times. The exogenous constraint arrived, and every vendor chose to deepen it.

In game theory terms, this is a Cournot model with an exogenous capacity ceiling, except the ceiling is set by someone (Caterpillar, responding to Microsoft) who does not know or care what a compression unit costs to rent per horsepower per month. The compression vendors are price-takers on capacity and price-setters on rental rates. That is an extraordinarily comfortable position to find yourself in, and they are behaving accordingly.

Three cost categories. Three mechanisms. One dominant strategy. The operator’s instinct to treat each of these as a separate procurement problem, to run three different RFPs and manage three different vendor relationships as though they operate in independent markets, misses that they are three expressions of the same structural shift. The supply side read the textbook. Every chapter. Including the one about what happens to the buyer.

What happened to the buyer is this: for two decades, the operator was the Stackelberg leader. You moved first (set the budget, defined the program, issued the RFP) and the service stack’s best response was already constrained by your choice. That is over. The service providers and their upstream input markets now set the equilibrium. You observe and respond. The math has not changed. The player who moves first still captures the surplus.

The player who moves first is no longer you.

This is the context in which Q1 2026 compression earnings should be read. Not as a quarterly update on three companies. As the most developed case study of a structural shift that has already reached your frac spread and your casing program, and that will define your F&D trajectory for the next three to five years.

Subscribe now

The Lead Time

The single most important number in the Q1 filings is not a margin or a utilization rate.

Caterpillar 3600 in-line engine lead times extended from 110 to 120 weeks last quarter to 160 to 180 weeks this quarter. An additional full year added in 90 days. A new order placed today arrives in Q4 2029. Solar turbine lead times followed: 102 to 106 weeks in Q4, 146 to 150 weeks in Q1.

This is not a temporary disruption. This is Caterpillar structurally reallocating production capacity toward power generation, where a single customer (Microsoft, Google, Meta) signs a contract worth more than the entire annual output of the compression aftermarket. The compression industry became a price-taker in its own supply chain, and the price is set by a hyperscaler whose cost of downtime is denominated in lost AI training runs rather than deferred production.

(Next week, I am publishing what might be the most fun piece I have written for this newsletter: a complete breakdown of Caterpillar, their supplier network, and a bottoms-up supply side model of where the engines are actually going. The lead time number is the symptom. The model is the diagnosis.)

The quarterly scorecard reflects the downstream consequence.

Archrock

$24.27/HP/month (up 8% YoY). Contract ops margin at 72%, highest clean quarterly record. EBITDA $221 million. Utilization 95%. Leverage 2.6x. CFO Doug Aron retiring by year-end.

Kodiak

$23.31/HP/month (up 3.7% YoY), targeting $24 by year-end. Contract services margin at 70.6%, seventh consecutive quarterly record. EBITDA $190 million. Utilization 98%. 10-year contract extension signed with a top customer (unprecedented in the industry). Zero share repurchases.

USA Compression

$22.73/HP/month (up 8% YoY; 4.7% from J-W accretive mix, 3.2% organic). EBITDA $188.6 million, company record. Utilization 91.9%, diluted by 200,000 HP of J-W idle inventory. Leverage 3.74x, hitting the 3.75x target a quarter early. 2026 new HP now 90%+ contracted, up from 50% at Q4.

Blended

$23.44 average revenue per HP per month. 95% average utilization. 69% average adjusted gross margin. All at or near record levels.

In Q4, I told you to quote Brad Childers’ earnings call language in every negotiation. “Your CEO told investors the industry has caught up with inflation.” We called it the comp. We said to use it like a real estate agent uses a below-asking sale.

That weapon lasted 90 days.

Childers’ Q1 tone shifted. No more “more modest.” Instead: “very happy with overall pricing in the market.” Revenue per HP per month moved higher sequentially and year over year, contradicting the Q4 moderation signal. KGS went further. Mickey McKee stated pricing power “will continue to 2027 and beyond.” Supply chain physics overwrote the earnings call narrative in a single quarter. Cat lead times added a full year. The supply constraint that underpins pricing power got dramatically worse between Q4 and Q1.

The lesson is not “find a new quote.” The lesson is: stop anchoring to vendor rhetoric entirely. Earnings call language is a lagging indicator with a 90-day shelf life. Lead times are the leading indicator. And lead times just told you the supply response is not coming before 2029.

The Bifurcation

There is a number in KGS’s Q1 filing that tells you more about the next two years of compression negotiations than anything management said on the earnings call.

Month-to-month horsepower exposure jumped from 9% of fleet in Q4 to 14% in Q1. A 500 basis point increase in a single quarter. That is 615,000 HP now operating on contractually renewable terms, up from roughly 400,000 HP three months ago.

In the same quarter, KGS signed a 10-year compression services contract extension with a top customer. The longest primary term ever disclosed in the contract compression industry. A second 10-year extension is in process.

These two facts are only reconcilable if the customer base is splitting in two.

Tier 1: The Lock-In.

The top 10 KGS customers (S&P 500 constituents, investment-grade credits, 50%+ of contract services revenue) are voluntarily locking in 7 to 10-year terms at peak-cycle rates. These are not naive counterparties. They have modeled the supply constraint. They have concluded that current rates are the floor, not the ceiling, because the supply response is 4+ years away and the engines that would build new units are going to data centers. Certainty of supply at current rates beats the optionality of going short when the alternative is a 3.5-year wait for replacement equipment.

Tier 2: The Walkaway.

Mid-market operators are rolling off primary terms and not re-upping at the terms offered. They are going month-to-month. That 615,000 HP on MTM at 98% utilization is not idle iron. These are operating units generating revenue. The operators attached to those units are choosing short-term flexibility over long-term commitment. Some cannot commit because their planning horizon does not extend to 2033. Some are balking at the rate. Some are simply declining to sign paperwork.

USAC tells the same story from a different angle. Month-to-month revenue now represents 25.4% of contract operations revenue ($78.5 million in Q1), a 74% increase year over year. The highest MTM exposure in the sector.

And there is a third signal, quieter but worth noting. KGS booked $1.2 million in credit loss provisions in Q1, up from zero in Q1 2025. The allowance for credit losses rose from $12.6 million to $14.2 million. At 98% utilization and record margins, who cannot pay?

Here is why this matters for your next negotiation, and why Tier 2 is the better position to be in right now.

Read more

This is not a criticism. It is a business model. David Einhorn figured this out at Greenlight Capital back in 2018, when his investors wanted to leave and he said (I am paraphrasing) no. The investors complained to the Wall Street Journal. Einhorn kept the money (To be clear - if your investors are complaining to the press about how onerous your liquidity terms are, you made the right call on the liquidity terms).

Private credit scaled this insight to $2.1 trillion. The pitch was elegant: we raise long-term locked-up capital, we lend it long-term to borrowers, and because the money is locked up, we cannot be forced to sell assets at the worst possible moment. We are not a bank. We are not runnable. Sleep well.

Then Blue Owl’s investors stopped sleeping well.

In February 2026, Blue Owl permanently gated redemptions on its $1.6 billion OBDC II retail vehicle after withdrawal requests surged 200%. The firm sold $1.4 billion in loans at 99.7 cents on the dollar to CalPERS, OMERS, BC Investment Management, and (notably) its own insurance arm, Kuvare. Blue Owl’s co-president, Craig Packer, went on CNBC and said: “We’re not halting redemptions, we’re just changing the form, and if anything, we’re accelerating redemptions.”

One way to read that sentence is that Blue Owl found a creative solution to a liquidity problem. Another way to read it is that the structural run-proofing that justified the entire asset class is being tested, and the test is producing sentences like that one.

Blue Owl was not alone. BlackRock TCP Capital reported a 19% NAV decline in Q4 2025. Blackstone’s BCRED faced $3.8 billion in redemptions (7.9% of total assets) in Q1 2026. Moody’s revised its entire BDC sector outlook to negative. Payment-in-kind interest hit 12.8% across BDCs. Median debt-to-EBITDA ratios in private credit climbed from 5.2x in 2020 to 6.5x. Portfolios are estimated to be 85% covenant-lite.

And so $543 billion in undeployed private credit capital needs somewhere to go. Blue Owl’s own management described the destination: asset-based finance, which they called “a $7 trillion market only 5% penetrated by private solutions today.” KKR scaled its ABF book to $74 billion (40% year-over-year growth). Blackstone’s infrastructure and asset-based credit platform hit $107 billion (35% growth). Ares, Apollo, Carlyle: all building permanent platforms to deploy capital into hard-asset-backed structures.

The rotation is not subtle. And one of its most interesting destinations is a corner of the oil and gas capital markets that almost no one outside of structured finance is paying attention to.

Subscribe now

The Thing About Cash Flows

If you have seen *The Big Short* (and if you are reading this newsletter, the probability is high), you know the basic architecture. An asset-backed security is what happens when something throws off cash predictably enough that an investment bank can package those cash flows into a bond, get a ratings agency to bless it, and sell it to an insurance company that needs yield.

The “something” can be almost anything. Mortgages, obviously (that one went well). Music royalties: David Bowie securitized 25 of his pre-1990 albums in 1997 for $55 million at a 7.9% coupon with a AAA rating. Then Napster arrived, the cash flows evaporated, Moody’s downgraded the bonds to Baa3 (one notch above junk), and the music ABS market went dormant for a decade. Then streaming happened, consumption patterns stabilized, and Apollo structured a $1.765 billion securitization for Concord Music in 2025 backed by 1.3 million copyrights (including catalogs from The Beatles to Taylor Swift), achieving an A+ rating from KBRA. The asset class came back from the dead because the underlying cash flows became predictable again.

Aircraft leases. The $4 billion Airplanes Group transaction in 1996 was the largest ABS deal ever completed at the time. By 2025, the market produced 16 transactions totaling $9.5 billion. Solar panels: SolarCity issued the first solar ABS in 2013 ($54.4 million, BBB+, 4.8% coupon). The sector now clears over $2 billion annually. Railcar leases. Trade receivables. Auto loans. Equipment finance.

And now: proved developed producing (PDP) oil and gas wells.

The structure is more elegant than it sounds. An operator transfers PDP assets into a bankruptcy-remote special purpose vehicle via a true sale, which legally isolates the cash flows from the parent company’s credit risk. The SPV issues monthly-pay amortizing notes with weighted-average lives of 7 to 10 years (legal finals out to 2037-2044), sculpted to mirror the natural production decline curve of the underlying wells. The issuer hedges 80% to 95% of expected production for five to eight years, which transforms volatile hydrocarbon revenue into something that looks, to a fixed-income investor, remarkably like a mortgage payment.

The waterfall is strict: taxes and LOE first, then hedge settlements, then senior interest and scheduled principal, then reserve top-ups, then junior tranches, then (and only then) residual equity to the management team. Performance triggers tied to debt service coverage ratios, loan-to-value, and production tracking tests accelerate amortization if things deteriorate. The structure does not trust anyone. That is the point.

Share

Where Oil and Gas Sits

Diversified Energy Company executed the first upstream ABS in November 2019: a $200 million, 10-year amortizing note with a 5% coupon. It was a proof of concept. Six years later, DEC has issued over $3 billion in ABS notes, maintains 85% rolling hedge protection on a five-year basis, operates at a 10% base decline rate with a gas breakeven of $1.80 to $2.00/mcf, and carries a 99% financial compliance track record across its securitized portfolio.

The market followed. Jonah Energy has completed seven issuances generating over $3 billion in total proceeds, including the largest single PDP securitization to date ($750 million, October 2022). Cumulative upstream ABS issuance has reached approximately $20 billion across more than 15 issuers. Annual volume scaled from $0.5 billion in 2020 to $4.3 billion in 2025.

The spread compression tells you where the market is on the maturity curve. DEC’s inaugural deals (2019-2020) priced at approximately 425 basis points over benchmark. The multi-tranche structures from 2021 to 2023 came in around 325 basis points. The broadly marketed variable funding notes from 2024 to 2026 are clearing at roughly 250 basis points. If that trajectory looks familiar, it should. Solar ABS followed almost exactly the same path: early proof-of-concept deals pricing above 300 basis points, compressing to 200-250 basis points as institutional comfort deepened. Late-cycle aircraft ABS trades at approximately T+125 basis points. Oil and gas is early-to-mid-cycle, tracking the solar playbook.

The economics are straightforward. ABS A-tranche coupons price at 6.0% to 6.5%, versus 8% to 10% for high-yield corporate debt. Advance rates run 70% to 75% of PV-10 value, compared to 60% to 65% for traditional reserve-based lending. The debt is non-recourse. There is no semi-annual borrowing base redetermination. No refinancing cliff.

The buyer base has institutionalized rapidly. Insurance and financial services firms now represent 56.9% of allocated capital. Asset managers account for 37.4%. Private credit and alternative funds hold 4.1%. Banks: 1.6%. The insurance bid is the structural anchor, driven by the same thing that drives every insurance allocation decision: long-duration, investment-grade, self-amortizing cash flows that match multi-decade liability profiles. DEC’s most recent issuance (ABS XI) attracted over 30 unique institutional investors. In 2019, the first deal went to a single buyer.

Kalibr Partners

The Management Math

Here is the part that matters if you are an operator, a service provider, or anyone who negotiates with either.

ABS entities are not normal counterparties. The financial structure creates a management team that is economically levered to operating cost discipline in a way that traditional E&P equity holders or PE-backed operators simply are not.

The mechanism: every dollar saved in lease operating expense flows directly into securitized net cash flow, which raises the debt service coverage ratio. Higher DSCR lowers the mandatory cash sweep percentage, which releases trapped cash to the management team as residual equity distributions.

Take a base case: 1,200 MMcfe/month production, $3.50/Mcfe realized price (with hedges), $1.40/Mcfe base OPEX, $2.20 million/month debt service. Securitized net cash flow: $2.42 million. DSCR: 1.10x. At that level, the structure mandates a 100% excess cash sweep. Residual to management: zero.

Now reduce OPEX by 10%. SNCF increases by roughly $0.17 million per month. DSCR rises to 1.18x. The sweep drops from 100% to 50%. Residual to management: $2.3 million per year. And if you can nudge DSCR to 1.25x or above, sweep rates fall to 0% to 25%, and the equity take expands dramatically.

This is something close to equity beta on cost control. Fixed debt service plus variable net cash flow equals a high-leverage outcome for the people running the assets.

DEC’s acquisition of Camino Natural Resources (announced May 2026, $1.18 billion) demonstrates the programmatic model. DEC contributed $210 million from its credit facility. Carlyle took a 60% majority equity interest in the SPV. DEC retained 40%, plus 100% of the undeveloped acreage, plus operating and management fees. The ABS debt is deconsolidated from DEC’s balance sheet. DEC retired $92 million in outstanding ABS principal in Q1 2026 alone. The machine acquires, securitizes, optimizes, builds equity, and repeats. In 2025, DEC signed 57 NDAs, submitted 24 bids, and closed one deal. The pipeline is $125 billion.

Subscribe now

What This Means

For operating teams, the implications are structural, not incremental.

1. Cost underwriting becomes the ballgame. I know from experience on the A&D side that fixed and variable cost analysis typically gets a quick pass using the LOE statements in the virtual data room. Most of the analytical energy goes to PUDs and alpha created at the bit. That approach made sense when the capital structure rewarded production growth. In an ABS structure, where management’s economics are a direct function of DSCR, the cost basis is the single most important variable in the deal model. Get it wrong and the sweep traps all the residual.

   Kalibr recommends a triangulation method: LOE statements from the data room, a physical-asset-based bottoms-up reconstruction of the cost structure (I am biased here, since this is what I have invested to build, but I also think it is correct), and public offset proxy analysis to reduce error. Three independent estimates. Converge on a defensible number. The legal and advisory fees on these deals are substantial, so the deal size needs to justify the aggregate cost. You cannot afford to get the underwriting wrong on a structure where every dollar of OPEX error compounds through the waterfall.

2. Systematic cost reduction is not optional. It is the primary equity lever. The only way management makes money in an ABS structure is driving cost out of the system. Every improvement in LOE (compression optimization, chemical renegotiation, automation, predictive maintenance) reduces the sweep and waterfalls more cash to management. DEC demonstrated this when it integrated Maverick Natural Resources: the team identified and eliminated $150,000 per month in wellhead compression rentals by leveraging existing Diversified infrastructure. That is $1.8 million per year, flowing directly through SNCF into management’s residual.

3. G&A is under a microscope. These deals are scrutinizing overhead in ways traditional A&D never did. The question is: how many wells can I run per engineer while holding to the cost discipline above? This is where AI belongs. Not as a buzzword in an investor deck. In the hands of subject matter expert engineers who can use it to monitor more wells, pattern-match best practices from adjacent industries, and reduce the technical headcount required to operate a given portfolio. Fewer engineers running more wells, armed with better tools. That is the structure this capital stack rewards.

For service providers, the calculus is different but equally clear.

1. Understand the growth vector. Companies embracing ABS will keep doing deals. It is purely a cost-of-capital game: once you have proven the structure works, you acquire more assets into the platform. DEC sees $125 billion in actionable deal flow over the next five years. Identifying these companies early and securing first right of refusal on new assets gives you a revenue stream that is diversified by basin (Appalachia, Oklahoma, Permian, with Bakken, DJ, and Eagle Ford on the horizon) and anchored to a counterparty with structural incentives to maintain the relationship.

2. Understand the incentive structure. These operators do not want your rate sheet. They want solutions that lower their LOE, because every dollar you help them save expands their residual equity through the waterfall. A vendor who helps the management team push DSCR above 1.25x is creating measurable economic value. That is a fundamentally different commercial conversation than negotiating monthly rates. Come with the cost reduction plan. The capital stack will do the rest.

Let's Go Deeper

The capital rotating out of private credit direct lending and into asset-based finance is not a temporary dislocation. Blue Owl, KKR, Blackstone, Ares, Apollo, and Carlyle are building permanent platforms. The banks are leaving. The void is being filled by a capital stack that is structurally different from the one operators and service providers are accustomed to negotiating against.

The counterparty across the table increasingly lives inside an ABS waterfall. Their economics, their incentives, their tolerance for cost, and their appetite for growth are all functions of that structure. If you understand it, you can position for it. If you do not, you are negotiating against a set of incentives you cannot see.

Understand the capital stack. It is the counterparty.

Nominally Hedged goes out to a small list. If you know someone thinking about improving sales / lowering CAPEX and OPEX — or who should be — forward it. That's the whole ask.

Share

There is a concept in corporate finance called the conglomerate discount. A company that owns a collection of unrelated businesses trades at a lower multiple than each business would command independently, because investors don’t trust management to allocate capital as well as the market would. The prescription is simple. Sell the non-core assets. Return the capital. Focus.

For the past decade, the Street has applied conglomerate discount logic to the Denver-Julesburg Basin.

Colorado was the non-core asset. Senate Bill 19-181 rewrote the state’s mandate from fostering energy development to aggressively regulating it. 2,000-foot setback rules. “Severe” ozone nonattainment classification. Permitting timelines of seven to eight months while Texas secured equivalent approvals in three days. Investors applied a “Colorado discount” to every DJ-weighted producer, demanding higher risk premiums and lower multiples. The prescription was the same one every activist gives every conglomerate: sell the DJ. Focus on the Permian. Simplify.

The echo for portfolio rationalization hasn’t stopped. Kimmeridge released a letter to Devon’s board just this week demanding “an accelerated program of non-core asset divestitures.” The playbook is evergreen. The logic is clean.

The conclusion might also be wrong.

The thing about conglomerate discount logic that nobody mentions in the activist presentation: the prescription assumes fungibility. Sell the “bad” asset, redeploy into something “better,” watch the multiple expand. This works precisely as long as “something better” exists. The Permian is running out of Tier 1 rock. Not catastrophically, but in the way that matters to anyone building a five-year drilling program: the next well is slightly worse than the last one. Bernstein called it in early 2026. Eagle Ford growth is unlikely through 2030. The Bakken is price-sensitive and declining. Delaware Basin productivity declined 6% year-over-year broadly, 15% for specific operators.

When the “better” alternative is deteriorating, the discount thesis collapses. The DJ isn’t dragging down your multiple because it’s a bad asset. It’s dragging down your multiple because the market hasn’t recalculated what “better” means in a world where everyone is running out of good rock at roughly the same time.

I’ll be honest: I thought Chevron and Civitas would eventually exit the DJ. Two deals changed my mind.

SM Energy acquired Civitas for $4.6 billion not because SM loved Colorado (SM’s CFO cited “comfort with Colorado” at the Bank of America conference, which is the kind of thing you say when you’re trying to convince yourself as much as the audience). SM bought Civitas because SM’s Midland inventory was severely depleted. Devon acquired Coterra for the same structural reason. Not synergy. Inventory duration. When your Delaware Basin reserve life is 8.1 years against an 11.5-year peer average, you’re not buying efficiency. You’re buying time.

These aren’t synergy deals. They’re admissions that the replacement hypothesis failed. The Street demanded asset rationalization. The geology demanded inventory preservation. The geology won.

And the DJ Basin (the asset everyone was told to sell) produces 40 barrels of water per well compared to 400 in the Permian. Recent vintage wells show slight productivity increases versus the seven-year average, bucking degradation trends across every other major play. 75% of Chevron’s DJ locations break even below $50/barrel. Laterals have lengthened 29% since 2018. The economics aren’t aspirational. They’re proven, improving on the margin, while the Permian’s marginal well gets slightly worse every quarter.

We score this probabilistically. Across five evidence streams, our assessment is that DJ Basin sentiment is structurally strengthening at 70% probability (confidence interval 60-80%). Three of five dimensions trending “improving.” The catalyst is not Colorado getting better. It is competing basins getting worse.

The legislative truce helps. Senate Bills 24-229 and 24-230 paused anti-drilling ballot initiatives through 2028. Permitting times declined 50% in two years. Colorado permit approvals grew at 23% CAGR since 2021. The “Colorado discount” is still priced in. The regulatory risk that justified it is largely gone (for a couple of years, atleast).

This is Kalibr’s first basin-level compression intelligence report. We mapped every active compressor in the DJ: 1,150 units, 476 facilities, ~1.275 million HP. We classified the fleet by vendor. We built a bottom-up demand model showing 177,000 HP of new demand in 2026. What we found is a market where compression demand has structurally decoupled from rig count, and where the four largest vendors are positioned very differently for what comes next.

See The Platform

The Part Where We Count the Compressors

1,150 active units. 476 facilities. ~1.275 million HP. Average engine age: 14.7 years.

Context: KGS’s entire national fleet is 4.5 million HP. AROC runs 3.7 million. USAC, 3.9 million. The DJ represents 5-7% of the US contract compression market, concentrated in a single state with the most stringent emissions regime in the Lower 48. The Permian has more iron. The Haynesville has bigger units. No basin has Colorado’s air quality rules.

The fleet has a split personality. Unit counts are eerily even across HP bands (~25% each), but the 1,500+ HP band carries 52% of all installed horsepower from 24% of units. If you think of the DJ as a “small wellhead compression” market, you’re missing that half the basin’s capacity sits in centralized stations with 5+ compressors per facility (10% of sites, 48% of HP). Caterpillar powers 73.8% of the fleet. Waukesha runs the big iron at 16.1% (averaging 1,581 HP per unit). Cummins handles gas lift at 5.9% (averaging 359 HP).

The fleet is old, and that’s not an accident. When operators view a basin as temporary (the “we’ll exit Colorado eventually” thesis), they redeploy older units and run them until the contracts end. That logic persisted for a decade. Now the contracts aren’t ending. The fleet is still old. 294 units from the 2010-2014 horizontal boom are approaching major overhaul economics. 121 units have already crossed or are approaching the 20-year replacement threshold.

The annual additions data shows a market in transition. 2024 saw 244 units set (68% above average) but average HP collapsed to 858 as gas lift installations surged to 43 units. 2025 corrected to 120 units with average HP recovering to 1,118. The question: was 2024 an anomaly or the front edge of a structural gas lift build cycle as the 2018-2019 vintage wells (1,200-1,400 spuds/year) enter the decline phase where artificial lift becomes economic?

The basin splits 50/50 on HP between midstream-owned and E&P-contracted equipment, but the character is completely different: 313 midstream units averaging 2,105 HP versus 837 E&P units averaging 768 HP. The 287 midstream-owned units (~593,000 HP) sit permanently outside the contract compression market. Which means the market AROC, KGS, and USAC actually compete in is roughly 780 units and 596,000 HP. Smaller than it looks.

Who Wins That Market

The most valuable piece of intelligence in compression is the one nobody has: which vendor operates which unit, at which facility, for which operator. Every basin-level analysis you’ve seen treats the fleet as one undifferentiated pool because that’s as far as the data gets them.

Kalibr’s proprietary classification model resolves this. We’ve built a machine learning system that attributes vendor identity across compression fleets using facility-level spatial patterns, operator relationship signatures, and equipment configuration data. In the DJ, the model classifies 84% of units by vendor. This is the foundation that makes everything else possible: market share, runtime associations, contract term exposure. Without vendor attribution, you have a fleet count. With it, you have a competitive intelligence map.

Subscribe now

Read more

Devon Energy can call any compression vendor in the Permian and get a yes. KGS has 4.5 million horsepower. Archrock has 4.8 million. USAC, post-J-W, has 4.4 million.

Devon chose NGS. NGS has 563,000.

The CEO was specific about what drove the expansion: “as they understood some of the capabilities of our units and some of the data that they would be able to get off of that, that was the primary driver.” That is the management narrative. There is an alternative explanation that fits the data at least as well: when two customers account for 59% of your revenue, you deliver exceptional service to those two customers because the alternative is going out of business. The technology story and the concentration story produce identical observable outcomes (great uptime for Devon). They produce very different predictions about what happens when NGS tries to serve customer number 61 with the same 196 field employees.

That distinction tells you something about this market that the earnings call does not.

We have spent the last three weeks taking apart the Big Three. The conclusion was not subtle: the strongest vendor BATNAs in compression history. If you stopped there, you would conclude that the only game is learning to negotiate within those three relationships. For 87% of the public compression market, that conclusion is correct.

The other 13% is where the leverage lives. And those vendors are, to put it precisely, weird. Not “might go bankrupt” weird. “Each one has a structural vulnerability the Big Three have spent two years eliminating” weird. The interesting kind.

This week: three niche players that do not fit the Big Three template and do not fit each other. A compression manufacturer with a rental fleet it is trying to grow into something the market takes seriously, whose $30/HP/month headline metric collapses the moment you decompose it by fleet composition (Slide 16 of the investor presentation is doing a remarkable amount of work for a single chart). The fastest-growing public compression fleet in the sector, growing EBITDA at 34% CAGR on the strength of a technology narrative the market is valuing at a 24% discount to peers, which is the equity market’s way of saying “we are not sure that is technology.” And a production optimization platform built around HPGL and VRUs that IPO’d 15 months ago, carries four material weaknesses in internal controls (which have already produced financial statement revisions, a detail the investor presentation omits), and just bought an ESP company to get upstream in the artificial lift lifecycle, which is actually the smartest strategic move any vendor in this analysis has made.

None of them are the Big Three. All of them have pressure points the Big Three spent two years eliminating. For operating teams building a BATNA, these are the three phone calls that change every other negotiation you have this year.

What This Means

If you are an operating team, three things:

Read more

Subscribe now

There is a specific problem that only successful companies have, and it does not look like a problem when it is happening. It looks like success, which is why it is so easy to misread.

Suppose you run a business that does one thing extremely well. Every quarter, the business generates a pile of cash. The first few years, the pile disappears back into growth: more units, more customers, more of the thing you are good at. This is healthy. This is what you want. But the core business has a size, and you do not. Eventually the rate at which you generate cash exceeds the rate at which the core business can absorb it, and you are left with a pile of money and nowhere inside the business to put it.

You have three options. Give it back to shareholders, which is honest but concedes that you have become a terminal-value company. Hoard it on the balance sheet, which your shareholders will punish you for. Or find a new market adjacent to the core, where the company’s existing capabilities give it an edge and the next dollar has a better home than a brokerage account. If the CEO has any ambition at all, door number three is the only door. The incentive structure guarantees it.

In 1997, Boeing walked through door number three. Commercial aerospace had a size, Boeing had outgrown it, and McDonnell Douglas offered a defense adjacency that used the same airframes, the same engineering, and the same customer logic. The pitch book was immaculate. The deal was not stupid. It was the textbook answer to the textbook problem of excess capital in a bounded market.

What Boeing did not price into the deal was what happens when a single-answer capital committee becomes a two-answer capital committee. Before the deal, every quarter the committee had one question: what does commercial aerospace need? After the deal, it had two. And the second question was always reasonable, always had a pipeline behind it, always had executives in the room making a perfectly defensible case for why this quarter’s marginal dollar should fund their program. No single decision was wrong. A rebuild on the Renton line gets deferred six months because defense has a tooling deadline. A component interval gets stretched from eighteen thousand hours to twenty-two thousand because the data says you can probably get away with it. Each trade is defensible in the meeting. It is only in the aggregate, over twenty years, that you see what happened, which is that the discipline of the single-answer committee (the one that had produced the fleet that had produced the franchise) had been quietly liquidated one capital allocation decision at a time. The 737 MAX was not a single failure. It was the terminal installment on a bill that had been accruing since 1997.

I want to be precise about what I am saying here, and what I am not. KGS just walked through door number three. The DPS acquisition is a genuinely smart deal in a genuinely attractive market. I admire almost everything Kodiak has built, and I think their AI-driven maintenance program is the most sophisticated thing happening in compression right now. What I am saying is narrower than “this will go badly.” I am saying: the arithmetic of a two-question capital committee creates a specific corridor of risk, and the most likely place that risk materializes is a line item most people are not watching.

Let me show you the math.

Kalibr Partners

The Winning

Here is what winning looks like in contract compression: $1.308 billion in revenue, $715 million in adjusted EBITDA, and a 69.2% Contract Services adjusted gross margin that represents a 247 basis point expansion year-over-year. Fleet utilization at 97.7%, with the large horsepower units (which account for 80% of the fleet by horsepower, because KGS figured out years ago that big iron on long-term contracts is where the economics compound) running north of 99%. Revenue per horsepower per month at $23.10, with management guiding toward $24 by year-end 2026. Free cash flow of $230 million, up 87.7% from 2024. Discretionary cash flow of $462 million, up 23.7%.

The EQT AB private equity sponsor fully exited its 76% stake during 2025. That is the kind of detail people skim over, but it matters: the PE overhang that once gave E&P negotiators a talking point about “sponsor pressure” is gone. The leverage ratio hit the long-stated 3.5x target. The company issued compression’s first 10-year bond at 6.75% and left $1.5 billion of ABL capacity undrawn. The balance sheet has never been cleaner.

And the AI-driven maintenance program, the one I am going to spend considerable time discussing in the context of risk, deserves credit. KGS has shifted from time-based 90-day service intervals to condition-based intervals of 120 to 150 days using a Fleet Reliability Center that monitors units in real-time and custom LLMs that provide field support. This is not marketing language. They are running fewer touches per unit per year, maintaining higher utilization, and doing it at a lower maintenance cost per horsepower than any peer in the sector. We will get to what that means in a moment. First, the report card.

I pulled the data and calculated the same metrics the same way. No adjusted-adjusted numbers, no investor deck cherry-picking. The accountants’ math.

The number that jumps out is not the one you expect. KGS’s 6.4% ROIC trails Archrock’s 10.4% by a wide margin, but that is largely a function of leverage (debt is a denominator, and KGS has more of it). The Contract Services margin is best-in-class, and fleet utilization leads the sector. Those are genuinely impressive operating results.

The number that matters for this article is the maintenance CAPEX row. KGS spends $17.05 per horsepower annually. Archrock spends $23.12. That is a 36% gap. On similar-sized fleets serving similar customer bases in similar basins. Archrock’s approach is the traditional one: spend more, overhaul more frequently, maintain buffer. KGS’s approach is the technological one: spend less, use AI to extend intervals, trust the data. One of these approaches is right and the other is wrong, or (more likely) both are defensible within their respective capital structures, and the question is what happens when the capital structure that supports the lower-spend approach gets stretched by a second mandate.

Share

The Door

On February 5, 2026, KGS announced the acquisition of Distributed Power Solutions for $675 million: $575 million in cash drawn from the ABL facility and $100 million in newly issued stock. DPS brings a 384-megawatt fleet of reciprocating engines and turbines, exclusively powered by Caterpillar, serving data centers, midstream operators, and industrial microgrid customers.

I want to say this clearly: this deal makes sense. The Caterpillar 3516J engine that powers KGS’s compression fleet shares mechanical DNA with the 3516H generator set engine. The 700-plus Cat-certified technicians already know the platform. The spare parts inventory overlaps. The customer adjacency (Permian operators who need behind-the-meter power because the grid will not show up for seven years) is real. RBC raised its price target from $45 to $64. The market cap expanded by $420 million. Northland Securities noted the market is capitalizing the acquired EBITDA at a premium 12x multiple, which is the market’s way of saying: we like this.

The deal was executed at 7.4x 2026 estimated adjusted EBITDA, implying roughly $91 million in near-term earnings. Of the acquired 384 megawatts, 250 are dedicated to data centers, including a 100 MW Virginia contract extending 15 years. Management’s demand pipeline is weighted 9-to-1 in favor of data centers over traditional oil and gas. In a world where grid interconnection queues stretch five-plus years and hyperscalers will pay almost anything for reliable behind-the-meter power, this is a market where the demand curve is doing the selling for you.

But (and there is always a but, this is a newsletter about compression, not a pitch book) there are specifics worth noting. The DPS fleet is concentrated around 16.5 MW Caterpillar SMT130 units. It lacks the larger 35 and 38 MW turbines that entrenched competitors use for utility-scale data center work. Much of the industrial and microgrid portfolio sits on one-to-two-year contracts requiring immediate repricing and extension upon closing. Piper Sandler flagged this directly: outside of the primary 100 MW data center agreement, the portfolio needs urgent commercial work on day one. And pro forma leverage, per Seaport Research Partners, spiked to roughly 4.0x.

This is a fleet that needs capital on arrival. The question is where the capital comes from, which requires meeting the neighbors.

Leave a comment

The Neighbors

Here is something I want people to understand about the distributed power generation market, because the analyst coverage treats it as a homogenous growth opportunity and it is not. The companies already operating at scale in this space are not middling operators collecting a tailwind. They are extremely good at what they do, with structural advantages that took years to build and cannot be replicated by writing a check.

Liberty Energy is the one I keep coming back to. Liberty’s power business (Liberty Power Innovations, through the LAET subsidiary) does something that no other oilfield power entrant does: it manufactures its own equipment. The Forte modular scalable generation systems and Tempo power quality systems are built in-house. When Caterpillar lead times for reciprocating gas gensets stretch past 100 weeks and OEMs are enforcing strict customer selection on allocation slots, Liberty is not standing in the queue. It is running its own factory.

That manufacturing integration matters more than most people appreciate. Liberty has 1.3 gigawatts of power agreements in place, a 400 MW reservation with Vantage Data Centers on 10-to-15-year energy service agreements, and a 3 GW capacity target by 2029 supported by an $813 million 2026 CAPEX budget. Piper Sandler notes that up to 70% of that budget can be funded through project-specific financing, which is a polite way of saying Liberty can build power gen capacity without putting the core frac balance sheet at risk. And the balance sheet: Total Debt to Total Stockholders Equity between 0.14x and 0.30x. I want to make sure that registers. Kodiak is running at 3.5x net debt to EBITDA. Liberty’s leverage ratio has a decimal point in front of it.

Liberty’s 14-year average Cash Return on Capital Invested runs 23%. In 2025 it compressed to 13% as completions activity softened (commodities do that), but 13% on a 0.30x levered balance sheet is a company with the margin of safety to absorb a bad quarter, a pricing war, or a new competitor who wants to buy their way into the neighborhood. The ROIC-to-WACC spread (the gap between what the business earns and what the capital costs) is wide enough to fund mistakes. KGS, at 6.4% ROIC on a 3.5x levered capital base, does not have that same width.

The rest of the power gen field is not empty either. Solaris is targeting 2.2 GW by early 2028 with $300,000 in annualized EBITDA per megawatt. ProPetro’s PROPWR division has a 60 MW hyperscaler contract and hybrid gas-battery systems. Atlas signed a 1.6 GW Global Framework Agreement with Caterpillar at high-teens IRR targets. These are not pilot programs. They are funded businesses with contracted cash flows.

None of these companies are better than KGS at running compression. I am not saying that. What I am saying is that KGS’s compression excellence was built over a decade of single-answer capital allocation, where every dollar was a compression dollar. These power gen competitors have been building their own version of that single-answer discipline in their own market. KGS enters as the new entrant against incumbents who have the fleet scale, the contract relationships, and (in Liberty’s case) the manufacturing capability that KGS will need to spend capital to match.

The risk is not that KGS will be bad at power generation. It is that being competitive might cost more than the capital structure can comfortably absorb. And when you are trying to close a gap against operators who have a 500 basis point CROCI advantage and a leverage ratio with a decimal point in front of it, each incremental dollar of investment feels more urgent than the last.

Subscribe now

The Rigid Triangle

I have been circling the arithmetic. Let me land the plane.

KGS operates under three financial commitments that management has publicly treated as non-negotiable. I believe them on all three, which is precisely the problem.

1. The leverage ceiling. 3.5x consolidated net leverage or below. The DPS acquisition pushes pro forma leverage to approximately 4.0x, with Goldman projecting a return to 3.5x by year-end 2026. This is not guidance you can quietly revise. The 3.5x target underpins the midstream-like valuation multiple the market assigns KGS, the pricing on $2.6 billion in outstanding debt (including compression’s first 10-year bond), and $1.5 billion of undrawn ABL capacity. Breach it and you are not just missing a target. You are repricing the equity.

2. The dividend. $0.49 per share quarterly, $1.96 annualized, recently increased 20%. Approximately $170 million per year. Dividend coverage at 2.6x (versus Archrock’s 4.9x, which is the financial equivalent of never having to worry about your rent). Cutting the dividend in this sector is like a restaurant failing a health inspection: technically survivable, but the customers remember.

3. The growth mandate. 150,000 new large horsepower per year in compression, 750,000 total through 2030. Growth CAPEX guided at $235 to $265 million for 2026 compression alone, before DPS power gen spending. The 2026 new-build order book is 100% pre-contracted. Engine deliveries are being secured for 2027 and 2028 at 100-plus-week lead times. You do not call Caterpillar and cancel because your Q2 EBITDA came in light. This capital is committed.

Three immovable objects. One budget. If growth costs more than planned (because power gen scaling requires incremental capital, or because Cat raises prices again, or because the macro softens and EBITDA undershoots guidance), the money has to come from somewhere.

You cannot borrow it (leverage ceiling). You cannot redirect shareholder returns (dividend). You cannot slow the new-build pipeline (committed).

Maintenance CAPEX, guided at $75 to $85 million for 2026, is the line item with give. It is, if you want to be clinical about it, the only remaining degree of freedom in the capital allocation framework.

Maybe it does not need to give. Maybe KGS’s technology-optimized maintenance approach is genuinely sustainable at $17.05 per horsepower while Archrock’s $23.12 represents structural over-spending. The AI is real. The efficiency gains are documented. KGS management has earned the right to that benefit of the doubt.

But the margin between “everything works” and “something has to give” is narrower than I would like. So I built a model.

Kalibr Partners

The Model

Three scenarios. All three hold the dividend at $170 million and the leverage ceiling at 3.5x. Fleet: 4.62 million horsepower.

1. Base Case. Guidance met. EBITDA: $765 million. Growth and other CAPEX: $300 million. Implied maintenance CAPEX: $80 million ($17.31 per horsepower). This is within baseline. Overhauls on schedule. AI extends intervals without degradation. Everything works. I think this is the most likely outcome.

2. Stress Case. KGS leans into the data center opportunity (management described the behind-the-meter market as “vast” and expressed intent to deploy “as much capital as is reasonable”). Growth CAPEX increases $100 million for power gen. Leverage ceiling prevents ABL borrowing. Dividend is sacred.

   Maintenance CAPEX absorbs the shortfall: $35 million, $7.57 per horsepower. A 53.6% deviation from baseline. At this intensity, major engine overhauls get deferred. Spare parts pools deplete. The Fleet Reliability Center keeps monitoring, but the budget to act on its findings is not there. The technology still works. The budget does not.

3. Bear Case. Macro softens. Archrock’s CEO has already noted that repricing increases going forward will be “more modest,” which is the compression sector’s version of saying the pricing cycle has peaked. EBITDA drops to $680 million. Growth CAPEX, locked by 100-week-lead-time commitments, cannot flex. Maintenance CAPEX is forced to $18 million, $3.89 per horsepower.

   At $3.89 per horsepower, you are not optimizing maintenance schedules with artificial intelligence. You are hoping nothing breaks.

Now here is the layer that connects the competitive landscape to the budget. If KGS’s power gen ROIC comes in at 10% (versus Liberty’s 13% incumbent benchmark, which is itself a down year), the EBITDA gap on the initial $675 million investment is approximately $23.5 million. Closing that gap at degraded unit economics requires $235.8 million in incremental capital.

That is 51% of KGS’s annual discretionary cash flow. On a balance sheet that is already at 3.5x leverage. With a dividend that cannot be cut.

I do not think this is the most likely outcome. But I note that the distance between “base case” and “we need more capital” is uncomfortably short, and the consequences of landing in that corridor are disproportionate to the probability.

Share Kalibr Compression

The Part Where the Contract Becomes the Story

I spend a lot of time reading compression contracts (someone has to, and based on the redlines I see, it is not always the people signing them). The availability guarantee is the clause that matters here, and it is worth understanding precisely because the mechanism is a sequence, not an event.

Every KGS contract contains a mechanical availability guarantee: 95.0% to 98.0%, calculated against all downtime including mechanical shutdowns, routine repairs, and scheduled overhauls. Three consecutive months below threshold and the customer can terminate. That is not ambiguous language buried in a rider. It is the core commercial term.

Deferred maintenance does not produce a dramatic failure. It produces a slow erosion that nobody notices until everyone notices.

First, spare parts pools thin out. The localized inventory that enables a three-to-five-hour field cylinder head swap starts getting allocated instead of stocked. Second, the Fleet Reliability Center keeps flagging potential issues (the AI works regardless of the budget), but when the money for parts and labor is not there to act on the flags, response shifts from proactive to reactive. Third, minor issues that were same-day fixes stretch into multi-day outages because the part is not on location and OEM components have their own lead-time queue. Fourth, unplanned downtime hours compound. The 98% number softens.

Industrial software specialists note that even the best predictive dashboards miss roughly 5% of sporadic failures. When budgets are healthy, that 5% gets absorbed by spare inventory and rapid field response. When budgets tighten, the 5% becomes the leading edge of a reliability decline. And KGS has already pushed intervals from 90 to 120-150 days. The buffer is thinner than it used to be.

The revenue exposure: approximately 9.0% of KGS’s fleet sits on month-to-month contracts (30-to-90-day termination). Management indicates 25% to 30% comes up for renewal annually. Against the $1.26 billion 2026 Contract Services revenue midpoint, roughly $460 million is exposed over the next 12 months.

$460 million in revenue governed by availability guarantees that are directly connected to a maintenance budget that is the designated relief valve for a two-front capital strategy. Maybe the technology holds. But if I had $50 million in annual KGS spend and a contract rolling in 18 months, I would want to understand this math.

Subscribe now

How the Peers Solve This

The question is not whether dual-growth strategies work. Baker Hughes pivoted into power and data center infrastructure and the market rewarded it with a premium multiple. SLB is building a billion-dollar data center solutions business. Companies that can fund adjacencies without straining the core tend to be rewarded handsomely.

The question is whether KGS’s specific capital structure has the flexibility to fund both mandates without one cannibalizing the other.

• Archrock stayed home. No power generation entry (explicitly declined; assets did not meet return and quality criteria). The result: 2.69x leverage with 0.8 turns of headroom below target, $110.7 million in maintenance CAPEX ($23.12 per horsepower, highest in sector), 4.9x dividend coverage, 6.0% senior notes (lowest coupon in compression history), no maturities until 2032. Archrock’s capital allocation answer is: grow the core, protect the maintenance budget, let the balance sheet absorb shocks. It caps the upside. It also structurally bounds the downside.

• Liberty walked through door number three with a fundamentally different balance sheet. 0.14x to 0.30x debt-to-equity. When Liberty needs power gen capital, it decelerates frac growth CAPEX, slows buybacks, and uses project-specific financing. RBC notes explicitly: Liberty has outlined a clear funding path that preserves the estimated $200 million annual frac maintenance budget. Liberty never touches maintenance. Its release valves are everything else.

The difference is not about management quality. Mickey McKee at KGS is as good as anyone running a compression business. The difference is structural. Archrock and Liberty have release valves that are not their core fleet. KGS, at its maximum stated leverage with a recently increased dividend, has fewer options. The capital structure funnels any variance from the base case toward the one budget with flex.

And that budget is the one that produces the uptime that produces the revenue that produces the cash flow that funds the dividend that cannot be cut.

Share

What to Watch

Five metrics. All leading indicators of whether this thesis is playing out.

1. Maintenance CAPEX per horsepower. Baseline: $15.00 to $17.50. Below $15.00 signals departure from predictive schedules. Context: Archrock runs $23.12 per horsepower and generates a 10.4% ROIC. KGS runs $17.05 and generates 6.4%. The AI narrative has to keep working.

2. Capital classification shifts. KGS split growth CAPEX into “growth” and “other” in Q1 2025. If maintenance activities migrate into “other” while the headline maintenance line stays flat, the accounting is doing the work that the spending is not.

3. Operating expense per horsepower. Management credits AI for lowering engine repair costs and lube oil consumption. If per-horsepower OPEX rises sequentially, deferred maintenance is surfacing as reactive field cost. Technology delays maintenance. It does not eliminate it.

4. Large HP utilization language. KGS cites 99%+ for large units. If the descriptor shifts from “operating” to “contracted” or “revenue-generating” utilization, mechanical downtime is entering the fleet. Words are leading indicators.

5. The AI efficiency narrative. If management’s emphasis on AI-driven maintenance intensifies precisely as per-horsepower budgets flatten or decline, ask yourself: is the technology supporting the budget, or is the narrative substituting for it?

What This Means for Operating Teams

I want to end where I started. KGS has built something exceptional. The utilization numbers are real. The Cat relationship is the deepest in compression. The AI maintenance program is genuine competitive advantage, and it may be the thing that makes this entire article a theoretical exercise. I would not bet against this management team.

But I would not bet my contract terms on a capital structure, either. The rigid triangle (leverage, dividend, growth) creates a corridor where the maintenance budget absorbs variance. That corridor may never be entered. If it is, the consequences compound through a specific mechanism: deferred overhauls, thinning spare parts, lengthening response times, softening availability, and eventually, customer leverage during $460 million in near-term renewals.

Operating teams with KGS contracts in the renewal window should be doing three things. First, structuring escalating performance credits tied to trailing three-month mechanical availability (not just the headline guarantee, the running average that triggers termination rights). Second, building month-to-month conversion clauses that activate if trailing availability drops, preserving optionality without forcing premature vendor changes. Third, documenting insourcing economics (replacement at current lead times: approximately $1,250 per horsepower, 7-year payback on new, shorter on opportunistic used 3516s) and surfacing the analysis in renewal discussions. Not as theater. As a BATNA.

This is the analysis that moves from interesting to actionable when you map it against your specific contracts, basins, and renewal timelines. The Leverage Matrix was built for exactly that purpose.

Get a Plan

The 737 MAX was not a design flaw. It was what happens when a two-question capital committee defers one small decision per quarter for twenty years.

KGS is not Boeing. The deal is smart, the team is excellent, and the AI-driven maintenance program may genuinely solve the problem I have spent 4,000 words describing. I hope it does. Permian compression is better because KGS is in it.

But the arithmetic of a two-question capital committee is the same arithmetic regardless of who is running it. And when the margin between “everything works” and “something gives” narrows to the width of a single line item on a quarterly filing, someone should be watching that line item.

Watch the line item. Not the headline.

Nominally Hedged goes out to a small list. If you know someone running compression negotiations this cycle — or who should be — forward it. That's the whole ask.

Share

In real estate, when one house sells below asking in a hot market, every listing on the block gets repriced overnight. The sale itself might be explainable (divorce, relocation, bad roof). Doesn’t matter. The comparable is in the system. Every buyer’s agent in the zip code pulls it up on their iPad at the next open house and says, “Well, the comp says...”

This is obviously compression I’m talking about.

On February 25, Archrock CEO Brad Childers told investors that pricing increases going forward would be “more modest” because the industry has “caught up with inflation.” He meant this as a positive. We are operating at peak profitability. Things are great. The machine is running.

What he actually did was create a comp. The largest compression fleet in the United States (4.79 million horsepower, 70-year operating history, every major basin) just told the public record that the repricing cycle is decelerating. That’s not Archrock’s problem. That’s everyone’s problem. When your KGS rep tells you rates are going up 5% next year, you pull up the Archrock transcript on your iPad and say, “Well, the comp says...”

Last week we took Kodiak apart. Peak BATNA, funded walk-away, systematically eliminated vulnerabilities. KGS is the vendor that does not need your contract.

This week: Archrock and USA Compression. Next week: The rest of the publics (Enerflex, FlowCo, NGS). The picture is more interesting than Kodiak, because these two have cracks Kodiak does not.

Archrock is the national incumbent. It is also the vendor whose CEO just handed every operator in America a negotiating weapon they did not have three months ago. USA Compression just doubled its fleet through acquisition and now faces the largest recontracting wave in its history while simultaneously integrating 594 new employees, a new ERP system, and 300+ new customer relationships it has never managed before.

Two vendors. Two different structures. Two very different pressure points.

What This Means (The Short Version)

If you are an operating team negotiating compression contracts, three things:

Read more

Large horsepower lead times doubled in six months. Roughly 55 weeks at mid-year 2025. 110 to 120 weeks today. The proximate cause is not a temporary supply chain disruption. It is a structural reallocation of industrial manufacturing capacity toward power generation, where demand visibility extends a decade, contract terms run 10 to 15 years, and payment economics are simply more attractive than anything the oilfield can offer.

Kodiak spent $675 million to buy a distributed power company. Permian gas processing plants, unable to access the electrical grid for seven to eight years, are converting what would have been 75,000 horsepower of electric motor-driven compression into gas-engine-driven units, consuming the same engine slots the oilfield needs. The compression industry quietly became a price-taker in its own supply chain.

Meanwhile, the Big Three reported Q4 results that belong in a private equity pitch deck.

Subscribe now

Blended fleet utilization at 95 to 98%. Revenue per horsepower at all-time highs: $23.10 at KGS, $21.69 at USAC, an estimated $23.50 at Archrock. And $2 billion in acquisition capital deployed in six months: KGS absorbing Distributed Power Solutions, USAC folding in J-W Power, Archrock integrating NGCS, and Service Compression acquiring AXIP Energy’s Chapter 11 estate. The competitive landscape has meaningfully thinned in a market where new equipment cannot physically arrive for two years.

What this means for your next renewal is straightforward and uncomfortable: your vendor’s best alternative to your deal has rarely been stronger. At 95%+ utilization with no credible supply response before 2028, the threat to walk away is not a negotiating posture. It is free. Every public vendor deleveraged this year. Archrock dropped from 3.3x to 2.7x and issued the first 10-year bond in compression history at 6%, then used it to push debt maturities past 2032. Enerflex hit 1.0x net leverage on record free cash flow. KGS delivered on its IPO commitment of 3.5x and bought back over $100 million in stock. When your counterparty is simultaneously improving margins, retiring debt, raising dividends, and repurchasing shares, the credibility of their walk-away is not hypothetical. It is funded.

The Signal the Earnings Calls Are Broadcasting

Here is the part the prepared remarks do not want you to hear: vendors are pushing aggressively for longer contract terms, and that behavior is deeply informative.

Kodiak is in active discussions with customers about seven- and ten-year renewals. Terms that would have been unthinkable three years ago. Archrock’s fleet now averages 73 months on location, up 61% since 2021, with large horsepower units averaging eight years. Eighty-five percent of Archrock’s 2026 new-build capacity is pre-contracted. KGS’s order book is full through 2026 and they are booking into 2027.

If vendors believed today’s conditions would persist indefinitely, they would be indifferent to contract length. The same terms would be available at renewal. By expending negotiating capital to lock in current positioning, they are pricing in deterioration they can see but you cannot.

This is the quarterback problem. When the party with peak current value demands a long-term commitment, the rational counterparty goes short.

The structural demand runway is real. Raymond James estimates 26 Bcf/d of incremental U.S. gas growth through 2030, implying 15 million additional compression horsepower, a 25% increase in the installed base. Demand will be enormous. But demand is not the same as vendor leverage. The supply response will come, the utilization plateau will break, and the operators who preserved renegotiation optionality through this constrained window will capture disproportionate value on the other side.

Key Performance Indicators | Q4 2025

Read more

I have officially launched the paid tier of this Substack, with the first deep-dive dropping this Friday. The content pipeline will stay fluid. I would rather give you the best and latest from our proprietary dataset than stick to a rigid calendar. The rough weekly cadence will look like this:

1. Quarterly breakdowns of the market and every service provider in the space

2. Basin-level reports: HP setting trends, potential primary term expiry pools, performance benchmarking by vendor

3. Relative positioning of CAT and other OEM suppliers feeding the ecosystem

4. General trend analysis on compression-adjacent activity

These will refresh every quarter. My goal is to arm you with the data to close more deals and lower LOE. I have always believed that great commercial negotiation is not about one side beating the other on book cost. It is about creating better vendor-operator pairings based on their relative positioning and strategy that, as a byproduct, produce better cost outcomes for both sides. If there is something you want me to dig into, just e-mail me. I am here to serve my readers.

Subscribe now

Easter marks the launch into spring sports in my household, which means I am currently operating as an unpaid Uber driver shuttling between batting practice and field hockey, curating walk-out music for a 7U baseball team whose tastes span Hakuna Matata to Sicko Mode, and trying to maintain the thin membrane of sanity that separates a functioning adult from someone who has just explained to a group of 7-year-olds why we cannot all play shortstop.

It also means explaining to a 7-year-old boy why we cannot do baseball practice, ride our bike, *and* watch a movie all in one night.

This is, whether my son appreciates it or not, game theory. Not the John Nash, Beautiful Mind, write-equations-on-windows version. The real kind. The kind that governs every resource allocation decision from a 7-year-old’s Tuesday evening to a Fortune 500 capital plan.

The principle is simple enough for a child to understand and difficult enough that most companies get it wrong: every time you say yes to something, you are saying no to everything else. Not just the thing you considered and rejected. *Everything* else. The entire universe of alternatives you did not evaluate, did not model, did not even see. All of them are now off the table. And if you say yes to everything? Well, you end up like this. And here is the part that makes it interesting: those alternatives do not vanish. They become available to everyone around you. Your no becomes someone else’s yes. Their yes reshapes a market you may not be watching.

A professor I had in business school used to say: “Tell me how often you say no, and I will tell you how strategic you are being.” I have thought about that line roughly once a week for a decade. It sounds like something you’d embroider on a throw pillow, but it is actually a precise statement about competitive positioning. Strategy is not the things you choose to do. Anyone can make a list of initiatives. Strategy is the things you choose *not* to do, and whether you understood, at the moment you declined them, the full map of what you were giving away.

The hard part is not the choosing. The hard part is tracing the consequences. Because trade-offs do not stay local. They cascade. Your product exit becomes an operator’s procurement constraint. Their constraint becomes a competitor’s market opportunity. That opportunity reshapes demand in an adjacent service category that nobody in the original decision room was even thinking about.

Which is why certain rumors percolating around ESPs could have implications for compression demand.

Let me explain, starting with the aircraft manufacturer that has generated more MBA case studies than a compression vendor has reasons they cannot lower your rate.

Share

Boeing’s Hundred-Billion-Dollar Gift to Airbus

Boeing stopped producing the 757 in late 2004. This was, on paper, a straightforward product rationalization. Sales had slowed. The 737 and 787 programs were where the growth capital was going. The 757 was a tweener: too big for short-haul economics, too small for long-haul prestige. Management made the trade-off.

Nobody panicked. The planes still flew. Airlines had years of useful life left in their existing fleets. The consequences would arrive later.

The 757 fleet aged. Meanwhile, the industry shifted toward flexible point-to-point network strategies, exactly the routes where a 757-class aircraft earned its keep. Boeing had read the market for the 757 as declining. What was actually declining was the hub-and-spoke model the 757 was built for. The aircraft category itself was about to matter more than ever.

Airlines that needed to replace their aging 757s did what airlines do: they bought the best available product for the route economics. They did not care whose logo was on the fuselage. Airbus, which had stayed in the category, was happy to oblige. The A321neo, the A321LR, the A321XLR. Airbus built an entire product family to fill the vacuum that Boeing had voluntarily created.

The mechanism is clean enough for a physics textbook. Manufacturer exits product. Operators run aging fleet. Fleet ages out. Operators shop for replacement. The only credible replacement is a competitor’s product. Market share shifts permanently. What looked like a production decision was actually a gift to Airbus that compounded for two decades. The financial magnitude, measured in lost orders, ceded market share, and strategic positioning surrendered, is comfortably north of a hundred billion dollars.

I thought about the 757 when I heard whispers this week about SLB.

Subscribe now

The RC(1000) and Me

Specifically, I heard rumblings that SLB is discontinuing production of certain RC series pumps as well as other legacy gas handling equipment.

That one caught my attention. Let me tell you why.

Back in my days as an engineer in industry (let me have my millennial “back in the day” oilfield story), we picked up a large acreage package in the North DJ Basin that came with an existing DUC. This DUC was critical to underwriting the economics of the acreage, which heightened the criticality of getting the frac design and production profile right. The challenge: no gas infrastructure to support gas lift, rods would not cut it on rate, and we had to embrace ESPs. Trade-offs.

That is a tall order in 5.5-inch monobore within a basin whose best ESP run time to date was something around six months. So I called some really smart people (shout out Jeff Dwiggins) and asked the question you ask when the well economics depend on getting artificial lift right: what is the most reliable system you would put downhole if your career depended on it? The answer, without hesitation, was the RC series. Not because it was the newest or the most technically exotic. Because it was the workhorse. The RC family was the thing you ran when you could not afford to be wrong, the go-to pump that every completions engineer in my network reached for when the wellbore was difficult and the margin for error was thin.

We ran an RC1000. The result: 18 months of continuous run time in a basin where six months was bragging rights. I fully intend to tell my kids about it someday, if they ever ask, which they will not, because I am just the Uber driver.

So when I hear that the RC series, the workhorse of the ESP world, the pump that a generation of engineers defaulted to in their hardest wells, might be getting rationalized, my instinct is to pay attention.

But we do not make commercial decisions on instinct. The whole premise of Kalibr is that anecdotal pattern recognition (”I ran this pump once and it worked great”) is not a strategy. It is a starting point. The question is what signals we can pull from public filings, analyst coverage, and operational data that either corroborate or contradict the rumor. What does the data say that the hallway conversation cannot?

Now I want to be clear. These are rumblings, not press releases. No formal announcement from SLB confirms the explicit discontinuation of specific legacy ESP models.

But oilfield rumors are a lot like the scale in the bathroom. They often tell the truth, even when you would rather they didn’t.

So let’s see what we find.

Learn More About Kalibr

$407 Million Worth of No

Before I show you the financials, I want to say something that may surprise you given the direction of this article: I think SLB’s strategic repositioning is smart. Genuinely smart. Not “smart for the press release” smart. Smart in the way that a well-constructed BATNA is smart, where the move looks obvious only after you understand the positioning it creates.

Here is the thesis. We are in a world of geological headwinds, maintenance CapEx drilling cycles, and PDP-structured entities that need to squeeze every barrel out of every producing well. There are multiple ways to win the next decade in oilfield services. You can build bigger frac fleets and compete on power generation technology, and some companies will win that way. But the capital intensity is brutal and the commoditization cycle is already well advanced. There is another way to win, and it requires a fraction of the capital: embed yourself in the production lifecycle. Every barrel optimized. Every failure predicted before it happens. Every well monitored, chemically treated, and artificially lifted by an integrated system that the operator cannot easily unbundle.

SLB looked at the board and picked that square. From a corporate finance perspective, the math works: production optimization is less capital intensive than drilling and completions hardware, stickier than equipment rental, and, if you can stay ahead of the commoditization cycle, structurally higher margin. The CapEx tells the story. Total capital investments declined to $2.4 billion in 2025 from $2.6 billion in both 2024 and 2023, with management guiding to 5-7% of revenue going forward. You spend less on physical manufacturing. You spend more on software, chemicals, and digital surveillance. The capital intensity drops. The recurring revenue grows. Finance departments love this.

The $4.9 billion ChampionX acquisition (finalized July 2025, all-stock, which is the corporate finance version of paying for dinner by offering to merge your bank accounts) was the enabling move. SLB already held 21% of the global artificial lift market. ChampionX added 5%, plus a production chemicals business, plus the XSPOC optimization platform. Combined: 26% market share and the ability to bundle lift hardware, chemical treatment, and digital monitoring into a single integrated offering that is extremely difficult to unbundle.

Two product catalogs. Two sales organizations. Two overlapping sets of legacy pump lines that were, until recently, competing against each other. You do not maintain both indefinitely. I mean, you *could*. But no one who has ever sat through an integration planning meeting would recommend it, and the bankers who sold you the deal certainly modeled the synergies assuming you would not.

The financial evidence of that rationalization is not subtle. It is, in fact, rather loud if you know where to listen

$407 million in workforce restructuring. $166 million in inventory write-downs. A Production Systems segment where the 12% revenue growth vanishes once you strip out ChampionX and the margin story (”lower profitability in legacy artificial lift,” in management’s carefully hedged language) reads like a quarterly memo from the finance department explaining why the thing they are sunsetting needs to be sunsetted faster. There is a version of the SLB story where the old product lines are a beloved heritage. This is not that version. The incentive to accelerate is not theoretical. It is quarterly.

The replacement products, and the pricing power logic.

This is the part most people will miss, and it is the part that matters most.

SLB is not just swapping old pumps for new pumps. It is doing something more deliberate: introducing products that specifically close the escape routes operators currently use to avoid ESP pricing. This is classic commoditization avoidance, the same playbook you have seen in [frac]() and [compression](), where the vendor introduces differentiated technology that no one else has, commands a pricing premium, and reduces the buyer’s credible alternatives.

Consider what each new product actually competes against.

The Reda Agile compact ESP pushes the pump node deeper into the wellbore and increases drawdown, which makes the ESP option competitive in well profiles where operators were previously concluding that gas lift was the better path. It does not just replace the RC series. It attacks the reason operators were leaving ESPs in the first place.

The Reda PowerEdge ESPCP closes the other exit: the “single run, then convert to rod or gas lift” lifecycle that (full disclosure) I used routinely as an engineer. By extending ESP economics deep into the mature well phase, it eliminates the conversion event that ended the ESP revenue annuity.

And sitting on top of both: the Lift IQ and Lumi platforms, wrapping every pump in AI-driven digital surveillance. SLB’s Digital segment now surpasses $1 billion in annual recurring revenue at 35% accretive margins. The hardware is becoming the delivery mechanism for the software. If you squint, this is less of an oilfield services business and more of a SaaS company that happens to involve things that go downhole.

The sell-side agrees. Unanimously. RBC, Jefferies, Capital One, Bernstein, Piper Sandler, Evercore ISI, TD Cowen, HSBC. Every major coverage name reaches the same conclusion, which is notable given that these are people who are professionally incentivized to disagree with each other. A 26% market share entity with overlapping product portfolios will consolidate. The only question is pace.

Good strategy. Good corporate finance. Genuinely. Some E&Ps will welcome this. If you have been fighting drawdown limitations or burning through ESPs on single-run cycles before converting to rod, the Agile and PowerEdge solve active problems you are already spending money on. Those operators will adopt quickly.

But not every operator thinks linearly inside the ESP category. The ones running tried-and-true legacy systems that work for their wells are not going to see new, unproven technology as an upgrade. They are going to see it as risk. And because they are engineers, not ESP salespeople, they will not ask “which ESP should I run instead?” They will ask a broader question: “Is this the right lift mode at all?” Once you open that aperture, the evaluation is no longer Reda Agile versus RC1000. It is ESP versus gas lift versus rod pump, holistically, with all the upstream implications for casing design, surface equipment, and operating philosophy that come with switching lift modes entirely.

That is where the cascade starts.

Fewer Doors, Same Hallway

If this were just SLB rationalizing a product line, you could shrug. Companies prune catalogs constantly. Operators migrate. Life goes on.

But SLB’s catalog is narrowing inside a supplier market that is simultaneously consolidating at the top, fracturing at the bottom, and walling itself off digitally at every tier. The operator who decides to re-evaluate lift mode will discover that the hallway has fewer doors than it did two years ago.

The top of the market is a fortress. SLB and Baker Hughes now operate with balance sheets clean enough (0.9x and 0.4x net debt-to-EBITDA, respectively) and revenue diversification deep enough (data center power, geothermal, critical minerals) that they can walk away from low-margin ESP tenders without flinching. Post-ChampionX, SLB manages over 20,000 connected assets through Lift IQ and XSPOC. Once your algorithms manage someone’s drawdown in real-time, the switching cost is no longer the pump. It is the data. These are not desperate suppliers. They are choosing their customers.

The middle is gone. ChampionX was the last standalone, publicly traded artificial lift pure-play. SLB absorbed it. The mid-tier negotiation option that operating teams historically played against the Big 3 no longer exists.

The independents face a tariff wall. Most US-based ESP assemblers source pump stages, motors, and cables from Tianjin and Jiangsu. The “Made in USA” label gets applied at final assembly. The core metallurgy is Chinese. Under current tariff structures, these assemblers have no margin buffer. Halliburton itself has acknowledged that artificial lift is its “largest component of tariffs” and that rewiring the supply chain will take quarters.

And in the Permian, where 80-90% of ESPs are deployed under rental agreements and the supplier captures a replacement revenue annuity every 6 to 9 months, the operator’s primary negotiation lever against all of this is the credible threat of converting to gas lift. Jefferies identifies it explicitly: the gas lift BATNA is what forces ESP pricing concessions. It is the only credible exit most operators have left.

Gas lift requires compression.

The buyer’s negotiation leverage in the ESP market and the compression demand signal are the same mechanism viewed from two angles. As the supplier landscape consolidates and the exits narrow, that BATNA gets exercised more frequently. Not because operators want to switch. Because it is the only credible threat they have left.

Subscribe now

So What Do You Do With This

The cascade does not require every operator to convert. It does not require SLB to formally discontinue a single SKU. It requires only that the aggregate probability of gas lift conversion ticks upward at the margin, across hundreds of wells, across multiple basins, across the next 18 to 24 months, as legacy ESP options narrow and the economics of staying versus switching shift.

The mechanism is the 757. Manufacturer rationalizes legacy product. Operators run existing fleet. Fleet ages out. Some migrate within the new catalog. Others discover that their well conditions push them toward an alternative that lives in an entirely different service category.

High-pressure gas lift requires compression.

For compression service providers:

Start having longer-term conversations with operators who lean heavily on SLB ESPs, particularly in basins with limited gas lift infrastructure. An operator whose ESPs are failing more frequently, whose replacement parts are harder to source, whose OEM is steering them toward a product that does not fit their wellbore geometry is going to start asking about HPGL. When they do, they will need horsepower. The obvious demand drivers are the ones everyone tracks. The subtle ones are where the edge lives.

In the Kalibr Compression platform, this is relatively straightforward. We have every compression site mapped. We can flag ESP-specific operations. Overlay that with workover frequency from our signals and you have a targeting map for conversations that no one else is having yet.

For E&P Operators:

If you are running SLB ESPs today, evaluate the new product lines now, before a failed pump and no replacement parts forces the decision for you. If HPGL is even a potential path, begin the compression procurement conversation early. And if you are using the gas lift BATNA as negotiation leverage against your ESP provider, understand that credibility cuts both ways: the more you signal willingness to convert, the more you should actually be prepared to source the compression.

Go Deeper

The Larger Point

The SLB example is interesting on its own. But it is not really about ESPs.

It is about the difference between having data and interpreting it. The world is not short on data. Every public company files quarterly. Every earnings call is transcribed. Every analyst publishes a note. The data is everywhere. What is scarce is interpretation through a lens that actually matters to the person sitting across the negotiating table.

Here is how we think about market intelligence at Kalibr, and it is the framework that made this article possible.

First, aggregate. And if you know the lens you are interpreting through, you can build datasets that do not exist anywhere else. We did not find a public database of every compression site in the Lower 48 mapped to artificial lift method, workover frequency, and vendor attribution. We built one. Because we knew the commercial questions we needed to answer, and the data infrastructure followed the interpretation, not the other way around. This is the part most people think is hard. It is not. It is plumbing with a thesis.

Second, interpret through the lens of negotiation. Not “what does this data mean for the market” in the abstract. What does it mean for the commercial levers available to a specific operating team sitting across from a specific seller? The $407 million in restructuring charges is interesting to an equity analyst. To an operating team negotiating an ESP contract, it is a signal about product availability timelines, spare parts inventory, and the vendor’s willingness to negotiate on legacy equipment they are actively trying to sunset. Different lens. Different output.

Third, apply it to your counterparty. Then apply it to your peers, because your counterparty’s alternatives are dictated by what your peers are doing. If three operators in your basin are already evaluating gas lift conversion, your ESP vendor knows that. Your negotiation position is not independent of theirs.

Fourth, cascade it outward. Upstream. Downstream. Adjacent service categories. The SLB story started as a product rationalization inside an artificial lift division and ended as a compression demand signal. That connection is invisible if you are only watching your own market. It is obvious if you are watching the full topology.

This is the power of interpreting data through the commercial negotiation lens. Not every signal matters equally. Not every rumor deserves a response. But when you can quantify and qualify the signals that do matter, sort them by impact, and map them to specific actions for specific counterparties, you stop being the operating team that reacts to the cascade after it arrives. You become the one that saw it building.

My son cannot do baseball practice, ride his bike, and watch a movie in one night. He has to choose. And his choice, picking baseball, means his sister gets the TV, which means he is catching the tail end of KPOP Demon Hunters for the hundredth time when he finally walks through the door.

Trade-offs cascade. The question is whether you see it coming.

Subscribe now

Nominally Hedged goes out to a small list. If you know someone running compression negotiations this cycle — or who should be — forward it. That's the whole ask.

It worked. D&C costs fell roughly 25 points on a 2015-indexed basis. Genuinely impressive, especially considering the service sector was extracting every possible margin on the way up.

Then everyone looked up from their completions dashboards and noticed that LOE had moved about five points over the same period. Five. The gap between those two numbers is not explained by the relative difficulty of the problem. It is explained by where the industry decided to point its analytical infrastructure.

The oilfield decided that finding and drilling hydrocarbons was the interesting problem, and then proceeded to solve it with genuine intensity. The fact that nobody built the same infrastructure for running the wells afterward is not a failure of imagination. It is a choice, legible in where the industry pointed its people, its capital, and its analytical resources for fifteen years. Drilling and completions teams got more of all three. Production teams got the same headcount and a bigger footprint.

The Kalibr staffing index makes that choice visible in a single data point: the same production team is now managing twice the operational footprint, with tools still pointed almost exclusively at the D&C side of the ledger. The industry did not just underinvest in LOE intelligence. It underinvested while the problem was compounding.

This is the gap. Compression is where it starts.

The Part Where My Last Piece Matters

Last week’s piece established the structural constraint on your compression vendor’s pricing. The short version: their $/HP/month is not a negotiating position. It is the number their entire asset base gets marked against on a consolidated basis for the equity story. When you ask for a rate reduction, you have not asked for a discount. You have asked them to call their auditors.

That analysis revealed where the flexibility actually lives: off-book accommodations that compensate without touching portfolio valuation. Free months, mobilization credits, escalator caps. The accommodations that never appear in the blended $/HP/month figure they report, but absolutely appear in your total cost of ownership if you know to ask for them.

That was half the problem. Here is the other half, and it has three pieces.

The first issue is performance. Compression is a critical-path service. A unit running at 78% mechanical availability is costing you multiples of whatever you saved on rate. The cheapest compressor is the one that runs. Everyone knows this. The problem is that there is no way to benchmark it. Vendors report mechanical availability in their investor presentations, but those are fleet-wide blended numbers that tell you nothing about the specific equipment class or basin where your wells actually sit. You have no third-party performance data. You have the vendor’s word and whatever your field team has observed firsthand.

The second issue is stickiness. Most operators mentally discount their ability to switch vendors because the demob economics look prohibitive. Downtime windows, new contract minimums, transition logistics, lost production. But consider the transaction from the other direction. Once that unit leaves your pad, your incumbent faces a maintenance capex event, uncertainty about whether the unit redeploys in-basin or gets trucked to another region, and carrying cost if it sits in the yard. Their switching cost is not necessarily smaller than yours. Depending on the unit configuration and current supply/demand balance for that HP class in your basin, it may be materially larger. A new vendor can flex on free months or a rebate structure that offsets your transition cost. Your incumbent knows this. And if you can estimate their redeployment risk by compressor configuration — which Kalibr tracks — you know exactly how much leverage you have before you walk into the room.

The third issue is the unit of analysis. Your invoice says dollars per horsepower per month. But horsepower is not what you are buying. You are buying compression duty: the ability to move gas from your wellhead to your sales meter at the pressure differential and flow rate your reservoir demands. Horsepower is how you get there. It is not the destination.

Nobody evaluates a frac spread on raw horsepower. You evaluate whether it delivers the treatment rate and BHTP your completion design calls for with enough redundancy to pump continuously. The HHP figure matters only insofar as it gets you there. Compression works identically. An oversized unit running at 60% load is not a good deal at any rate. An undersized unit as GOR rises is a production constraint dressed up as a line item. Neither shows up as a problem in the $/HP/month contract. Both show up in your net revenue.

The correct unit of analysis is $/MCF moved — what it actually costs to transport a molecule from your wellhead to the meter. This number integrates equipment sizing, utilization, runtime performance, and contract rate into a single figure you can compare across vendors, basins, and well types. It is also the number that nobody in your organization currently has. Getting there requires matching equipment specs to gas volumes and pressures, correlating runtime by vendor and basin, and having contract timing intelligence that tells you when your negotiating window opens.

That is what Kalibr built.

What Kalibr Compression Intelligence Actually Does

The platform is built on three layers of data that do not currently exist anywhere in a single integrated product. Each layer addresses one of the problems above.

Layer 1 is financial — the pricing benchmark. Every public compression provider discloses consolidated blended metrics: fleet HP, blended $/HP/month, reported utilization. What they do not disclose is how those numbers disaggregate by HP band, basin, fuel type, or contract vintage.

This is not an accident. The consolidated number supports the equity story. The disaggregated number would support your negotiation. So you get earnings calls that reference record pricing and 97%+ utilization without any detail on which equipment classes or basins are driving it.

Kalibr reconstructs that disaggregation. The methodology anchors to the one provider that breaks out HP band revenue in their 10-K and infers the band-level rate structure across the industry. The result is a benchmark matrix: $/HP/month by HP band, basin, vendor, and fuel type, with confidence intervals. This is the foundation — knowing what you should be paying before you walk into the room.

Layer 2 is operational — performance and timing. Kalibr’s proprietary asset dataset matches compression equipment in the field — by HP rating, set date, vendor attribution, and the gas volumes and pressures associated with it — to the producers operating them.

This match enables three things public financials cannot. First, regional runtime correlations by provider: which vendors are delivering mechanical availability above or below their marketed guarantees, by basin, by unit size, in aggregate across Kalibr’s observation set. This is the performance benchmark that does not exist anywhere else.

Second, contract timing intelligence: primary term exposure at the vendor level. This tells you when your negotiating window opens and when your counterparty is feeling renewal volume pressure — the information you need to structure flexibility into the deal.

Third, throughput economics. With equipment specs matched to production data, you can calculate actual compression duty and identify where you are paying for capacity that is not moving molecules. This is the $/MCF analysis.

The asset data is what turns financial benchmarking into operational intelligence.

Layer 3 connects vendor financial health to your operational risk. This is the part that functions as a leading indicator — a signal that precedes the outcome by one to three quarters in the historical data.

The logic is straightforward. A vendor deferring maintenance to protect near-term EBITDA margins looks, in the income statement, like they are getting more efficient. Maintenance intensity is falling. Cost per HP is improving. This is the story the equity analyst is writing. It is also the story your account manager is telling you at renewal. What is actually happening is an accumulation of reliability risk that will surface in your operational results before it surfaces in their reported metrics.

The same dynamic applies to labor. A vendor stretching field technicians across more horsepower — HP per technician climbing, disclosed in the filings if you look — looks like productivity improvement. It is actually a response time and preventive maintenance backlog building in the field. The unit on your pad is not getting less attention because it needs less attention.

Kalibr tracks these signals through a composite Reliability Risk Score (RRS) that combines maintenance intensity, labor leverage, and fleet renewal velocity into a single 0–100 metric. A vendor whose RRS is deteriorating is one you want performance guarantees from (or optionality to exit) before the downtime window opens. The financial signal precedes the operational outcome by one to three quarters. That lag is the intelligence.

Layered on top: operational metrics from the asset dataset (HP set intensity, run hours between service, runtime correlations by configuration) that amplify the financial signals with field-level observation. Next week, we will start using this data to benchmark real-time performance in specific basins.

Subscribe now

What This Means for How You Negotiate

Put it together. You have a rate benchmark: $/HP/month by HP band, basin, vendor, and fuel type, so you know whether you are above or below market before you open your mouth. You have contract timing intelligence, which tells you whether your vendor is facing renewal volume pressure in your basin right now. You have a $/MCF analysis, which tells you not whether your rate is competitive but whether your contract is sized correctly for the work it is actually doing. And you have the RRS, which tells you whether the vendor’s forward reliability posture warrants performance guarantees or an exit option.

None of these are rate conversations. Rate is the thing your vendor explained is structurally immovable, and for structural reasons they explained honestly. All of these are contract structure conversations: mobilization credits, performance guarantee provisions, escalator caps, early termination rights. The flexibility exists. It has always existed. The question is whether you showed up knowing where to find it.

The practical application works on three levels.

Benchmark the rate. Know whether your $/HP/month is above or below market for your HP band and basin before you walk into the renewal. This sets the frame.

Evaluate the vendor. Check the RRS. A vendor with deteriorating reliability metrics is a vendor you want performance guarantees from — or a vendor you want optionality to exit.

Right-size the contract. Once you know your actual compression duty, you can identify where you are paying for capacity that is not moving gas. That is not a discount conversation. It is a right-sizing conversation, one your vendor can have without touching their portfolio valuation.

The combination of understanding their model (where the off-book flexibility exists) and understanding your own position (where the optimization opportunities are) is what makes this a better partnership,

Share Kalibr Compression

Kalibr Compression

Starting this quarter, Kalibr is launching the paid tier of Nominally Hedged, anchored by Kalibr Compression, a monthly compression intelligence brief. Each issue covers the following.

Vendor benchmark update. $/HP/month rate movements by HP band and basin across public providers and the independent market. Flagged against prior quarter and trailing four-quarter range.

Reliability Risk Score update. RRS movements for each public vendor, with interpretation of what directional changes mean for operators currently in or approaching renewal.

Contract expiration exposure map. Kalibr’s view of aggregate primary term exposure across the L48 fleet. Regional concentration flags. Where renewal windows are clustering, and which basin supply/demand dynamics are shifting.

Compression intelligence feature. One deep analysis per month. The first issue covers the $/MCF calculation: how to run it against your existing fleet, what the typical range looks like across well types and GOR profiles, and which units to target first.

The free version of Nominally Hedged continues as it has. The paid tier is for operators who want to do something with the analysis.

Leave a comment

The Cost Structure Thesis

D&C intelligence was built for the era when drilling was where the value creation happened. That era delivered 68 on the D&C index against a 2015 baseline of 100. The gains compounded. The margins on the service side compressed. The engineers moved on to the next problem.

LOE is a decade older and largely unchanged, The same systematic intelligence applied to compression, chemicals, artificial lift, and production system design will take it lower; not because production engineers are doing their jobs poorly, but because they are doing those jobs without the tools that D&C built over fifteen years. The operators who close that gap first will have a cost structure advantage that does not depend on commodity prices, inventory quality, or how many rigs their neighbor is running. This is not a cyclical edge. It compounds.

Your vendor has been doing this math since before they picked up the phone. The only question is whether you walk into the next renewal as the operator who brought data, or the operator who confirmed what they already suspected.

Nominally Hedged goes out to a small list. If you know someone running compression negotiations this cycle — or who should be — forward it. That's the whole ask.

Here is a useful theory of the oil and gas industry: it is subsurface real estate development with better PR.

You assemble acreage. You fund development. You sell production at a significant multiple to capital invested, assuming the reservoir cooperates and your completions engineer read a couple of SPE papers at HFTC rather than tapping out the happy hours. This happens underground, which is why most people miss the parallel.

Commercial real estate runs the same play above ground. You assemble land. You fund development. You sell or lease at a multiple to capital invested, assuming the market cooperates and your architect was right about the floor plate. The developer does not own the excavators. They do not own the tower cranes or the concrete batch plants. Just as the developer would never own the equipment moving dirt between the foundation and the first floor, the E&P does not own the compression moving gas between the wellhead and the sales meter. They contract for that capacity, use it, and, in theory, send it back.

The useful part is this: in both cases, the comparative advantage is finding acreage, reading the market, and funding projects. The investors backing that developer, or that E&P, are paying for geological and financial judgment. They are not paying for compressor maintenance margins or crane depreciation economics. So each industry evolves, naturally and correctly, toward a structure where every function is owned by whoever can finance it cheapest.

Now extend the analogy one step further, because this is where it gets interesting.

The compression vendor is not the developer. The compression vendor is the REIT. They own the building. They lease it to you. Their entire business model, the capital structure, the investor base, the return thresholds, the covenant calculations, is built around the assumption that the rent holds. A REIT cannot selectively cut dollars per square foot for one tenant without that conversation immediately becoming a conversation about what every other unit in the portfolio is worth. The auditors ask. The lenders ask. The covenant calculations update. What looked like a one-off accommodation is now a portfolio-wide markdown.

Compression works identically. The dollars per HP per month your vendor quotes is not a negotiating position. It is a valuation input, the number their entire asset base gets marked against. Ask them to cut it, and you have not asked for a discount. You have asked them to reprice their balance sheet.

Which means the first step in any serious compression negotiation is not asking for a lower rate. It is understanding the model well enough to know what they actually can give you, and what the accounting will and will not allow. That requires reconstructing their return thresholds, their leverage math, and the specific utilization floor below which the arithmetic stops working.

What the oilfield supply chain actually is, and almost nobody talks about it this way, is a tranched capital structure wrapped around a set of operational risks. Think of it like a CLO, except instead of slicing credit risk into senior and junior tranches, you are slicing oilfield operational risk into layers and selling each layer to the investors best positioned to hold it. The E&P equity sits at the top of the stack: reservoir risk, commodity price risk, execution risk, in exchange for the IRRs that justify the geological and financial judgment the investors are actually paying for. The compression vendor sits several tranches lower. Contracted cash flows. Hard asset collateral. Multi-year lease terms with investment-grade counterparties. Predictable maintenance cycles. This is, functionally, an infrastructure yield instrument, and it attracts infrastructure yield investors. Pension funds. Insurance companies. The same capital that buys Prologis warehouses and toll road concessions. These investors accept 12 to 15% unlevered returns because the risk profile warrants it and because their liability structures demand duration, not upside.

The tranching is elegant. The tranching is correct. Modigliani and Miller had a theorem about why vertical integration destroys value here: combining high-beta reservoir exposure with low-beta contracted equipment yield produces a cost of capital that is wrong for both businesses simultaneously. The practical version: the E&P that owns its compression fleet has confused its equity investors, annoyed its lenders, and made its CFO’s job harder, all at once.

The ecosystem is not an accident. It is the industry’s solution to a genuine capital allocation problem. And for compression vendors in particular, it has worked almost suspiciously well.

Share

The Part Where the Vendors Start Making Embarrassing Amounts of Money

It has been a good time to own a compressor.

To understand why, you have to start with what happened to the customer. In a previous piece, we traced how E&P consolidation quietly transformed compression from a commodity rental into a critical-path service. The short version: as operators consolidated, pads got bigger, laterals got longer, and simul-frac operations demanded more horsepower delivered to a single location. The architectural decision to run gas lift, driven by capex efficiency and slimmer casing designs, locked compression into the production system the way a conveyor belt gets locked into a factory floor. Rising gas-oil ratios across maturing Permian fields deepened the dependency further. You were no longer renting a compressor. You were building your production system around one.

That shift did two things to vendor economics that do not show up immediately in the headline numbers. First, it upgraded the counterparty. The operators committing to compression contracts on large centralized facilities are not small privates running a three-well pad on a prayer. They are Devon and Diamondback and Coterra, locking in compression for drilling programs that have already been AFE’d, approved, and scheduled. Investment-grade credit risk on mission-critical equipment. The contract book got longer, the credits got better, and the embedded switching costs got higher, all at the same time.

Second, and this is the part that rarely gets written about: the supply chain collapse did not just constrain new capacity. It restructured how vendors carry risk. Caterpillar engine lead times stretched to 90 to 120 weeks at the peak, call it two years from order to delivery on a large horsepower unit. Vendors responded by requiring operators to commit to contracts before equipment was ordered. The mechanics matter here. The vendor does not pay the balloon payment to Caterpillar or the engine manufacturer until the unit sets on location, which is also when the contract starts and revenue begins. No float. No inventory carrying cost. No speculative deployment. The vendor is running a pre-sold order book against a manufacturer who holds the inventory risk.

The numbers are what you get when all three of those things compound simultaneously. USAC reported revenue per horsepower per month of $21.69 in Q4 2025, a record. Archrock’s operating margins have exceeded 70% for five consecutive quarters. Kodiak’s fleet utilization sits at 97.7%, which in a capital-intensive equipment business is essentially the speed of light. You do not get meaningfully closer to full without something breaking. USAC has made 50 consecutive quarterly distributions without a cut since its IPO, returning approximately $1.9 billion to unitholders.

The bad debt figure is the one that stops the room. Over nineteen years, USAC’s write-offs have totaled 0.06% of billings. Not 6%. Zero point zero six. When your customers are investment-grade operators running production systems they cannot shut down, and your contracts were pre-sold before the equipment existed, bad debt becomes a rounding error on a rounding error.

To put that in context your CFO will appreciate: USAC’s three-year total shareholder return of 54% nearly matched the S&P 500 Value index during one of the most aggressive tech bull markets in history. Compression does not have a ChatGPT moment or a Blackwell chip launch. It has a Caterpillar 3608, a multi-year contract, and a customer who cannot turn off the gas lift without killing the well. It turns out that is a durable competitive position.

Which makes the Axip Energy Services bankruptcy filing, Southern District of Texas, February 22, 2026, genuinely interesting in a way that has nothing to do with sympathy for Energy Spectrum Capital’s limited partners.

In a market this favorable, going broke requires some effort. But here is the thing about a compression bankruptcy in a compression bull market: it is the only controlled experiment you are going to get. Every healthy vendor is running the same fundamental model, same return thresholds, same accounting logic, same structural constraints. But they are not required to show their work. Axip is. The capital structure, the return thresholds, the utilization floor below which the arithmetic stops working, the exact sequence of events that made refinancing impossible, all of it goes into the public record, sworn and filed under penalty of perjury. Most operators will read the headline and move on. That is the wrong read. The Axip filing is a tutorial on the model your vendor is running, and therefore a precise map of where the leverage actually sits when you are sitting across the table from Archrock or USAC or Kodiak at contract renewal.

Share Kalibr - Strategy Research for Oil & Gas

The Part Where We Pull Out a Calculator

On February 22, 2026, Axip Energy Services filed for Chapter 11 in the Southern District of Texas. Energy Spectrum Capital had acquired the company in September 2022, loaded it with $240.5 million in debt, and watched the thesis fall apart in roughly three years. The stalking horse bid from Service Compression LLC came in at $161 million, 33 cents on the debt dollar.

The capital markets press wrote it up as a private equity cautionary tale. That reading is not wrong. It is just not useful.

What the filing actually is, the First Day Declaration, the capital structure schedules, the forbearance amendments, all of it sworn under penalty of perjury, is a detailed post-mortem on what the occupancy model looks like when it breaks. The REIT parallel from earlier is not decorative here. Axip did not fail because they were bad operators. They failed because vacancy crossed a threshold their debt structure could not survive. Understanding exactly how that happened tells you more about your vendor’s negotiating constraints than anything they will voluntarily disclose across a contract renewal table.

The building and the mortgage

Axip operated 940 units totaling 326,070 HP across seven facilities in Texas, New Mexico, North Dakota, and the Gulf of Mexico. Against that asset base sat $240.5 million in funded debt: an ABL facility with JPMorgan Chase at $207.8 million, a superpriority tranche at $13.2 million, and a second-lien held by Permico Inc. at $19.5 million. At roughly $92 million in annual revenue, that is a 2.6x debt-to-revenue ratio. There is no margin for error in that structure. A building owner running 2.6x debt-to-rent does not survive a vacancy spike. Neither does a compression vendor.

Note the second-lien holder. Permico is PE-affiliated. Seller financing at the junior layer is a signal that the leverage could not be fully placed with third-party lenders at closing. It is not conclusive, but it is a tell worth filing away when you are evaluating vendor financial health.

The tenant that did not pay and could not leave

In Q1 2024, a major Gulf of Mexico customer filed Chapter 11 and converted to Chapter 7 liquidation. Twenty-four offshore units, over 15% of Axip’s total horsepower, were stranded overnight. The customer did not pay demobilization costs. The units could not be economically redeployed onshore. Offshore compression does not truck to the Permian. It sits.

The EBITDA impact was immediate and unrecoverable. But the more important lesson is structural: one counterparty event erased 15% of revenue on a balance sheet with no cushion to absorb it. This is the right-tail risk that the occupancy model carries silently until it does not. For operators evaluating vendors, counterparty concentration is not an abstract risk factor buried in a 10-K. It is the variable that determines how much negotiating room your vendor actually has, because a vendor carrying fragile customer exposure needs your stable, investment-grade contract more than their renewal posture suggests.

The building that needed a generator to turn the lights on

More than 25% of Axip’s fleet was electric-motor driven. On paper, a forward-looking strategic position. In practice, a mismatch between strategy and ecosystem.

Electrification is not a bad bet. Archrock and Kodiak are making similar investments, and the long-term direction of the market is not seriously in dispute. But executing an electrification strategy requires your customer base to be able to support it: reliable grid access, infrastructure buildout, power availability that matches compression demand. Axip’s customer mix could not deliver that. When grid availability failed to keep pace with deployment, customers ran generator sets to power units they were already renting, effectively paying twice for the same compression. That math accelerates the decision to return equipment considerably.

The contrast with Archrock is instructive. Archrock can pursue electrification credibly because their customer base includes operators large enough to command grid infrastructure buildout. ExxonMobil does not run generator sets. Axip’s customers did. Strategy requires ecosystem alignment, and a vendor pursuing a strategy their customer base cannot support is a vendor burning capital and absorbing returns. Which is both a risk signal and, if you are in their customer base, a leverage signal.

What the vacancy actually cost

Axip’s pre-stranding revenue ran approximately $92 million annually against a fleet of 326,070 HP, an implied rate of $23.52 per HP per month across the total fleet. But 15% of that fleet was stranded and generating zero revenue. Calculate on deployed horsepower only and the effective rate was $27.67 per HP per month, higher than USAC’s record $21.69. Axip was not losing the rate war. The business model was working on the equipment that was working. What it could not survive was the vacancy.

Post-stranding, revenue dropped to approximately $78 million. Gross profit at industry-standard margins fell to roughly $51 million. Debt service alone consumed most of that. The remaining operating cost structure consumed the rest. There was no path to solvency that did not require either a significant equity injection or a buyer willing to reset the capital structure, and after six months and 85 parties contacted in an Evercore-run process, neither materialized.

What your vendor’s balance sheet is actually telling you

The Axip autopsy is not a story about bad luck or bad management. It is a demonstration of how the occupancy model fails, and therefore a precise illustration of what makes it fragile even when it appears to be working.

Three things to take away.

First, utilization is not a performance metric. It is a solvency metric. For a vendor running 4x leverage against contracted equipment cash flows, the difference between 95% and 80% utilization is the difference between distributable cash and covenant breach. The problem is that in the current market, all three majors are running 94 to 98%. The headline number does not tell you much. What matters is primary term exposure: which contracts are rolling, when, and what that does to forward utilization when the aggregate looks healthy. That is where the fragility hides. It is also where the negotiating window opens, if you know where to look. (Kalibr tracks primary term exposure at the vendor level. More on how to use that in a future piece.)

Second, counterparty concentration is a commodity exposure signal, not just a default risk flag. Consolidation has largely removed outright bankruptcy risk from the E&P customer base. Devon and Diamondback are not going Chapter 7. But concentration still matters, because it tells you what your vendor’s growth outlook is actually tied to. A vendor with a gas-heavy contract book in a weak gas price environment is a vendor whose forward EBITDA is under pressure. If you are an oil-weighted operator in the Midland Basin, you are not just a renewal. You are a hedge against the exposure that is compressing their growth outlook. That asymmetry is leverage, and it is leverage most operators never think to calculate because they are focused on their own commodity exposure, not their vendor’s. Read their 10-K customer concentration disclosures and map it against the current price environment before you walk into a renewal.

Third, know your vendor’s strategic white space and know where you sit inside it. Every vendor has edges where they are strong and edges where they are exposed. A vendor overextended into a technology or geography their customer base cannot support is a vendor absorbing execution risk that will eventually show up in your service quality or their willingness to deal. Understanding the shape of their strategy tells you where the pressure points are before they become your problem.

The practical starting point for all three is the same: model the occupancy floor. What utilization does your vendor need to service their debt at current rates, and how close are they to it? That calculation sets the realistic bands of negotiation, not what you want, but what the arithmetic will actually allow. The inputs are their cost basis, their WACC, their stated IRR targets, and their leverage structure. Archrock, USAC, and Kodiak publish most of this. What they do not publish is reconstructible from public filings. (This is what Kalibr’s market intelligence package is built around: the vendor-level economics that turn a renewal conversation from a negotiation about numbers into a negotiation about constraints.)

Once you have the floor, the last piece is understanding what form a concession can actually take. Rate cuts are a balance sheet conversation your vendor cannot have without repricing their entire portfolio. The auditors ask, the lenders ask, the covenant calculations update. But off-book accommodations are a different category entirely. Free months, mobilization credits, escalator caps, priority dispatch commitments: accounting-friendly concessions that do not touch the portfolio valuation. Understanding which category each ask falls into is the difference between a negotiation that goes somewhere and one that does not. That is where we are going next.

Subscribe now

The Part Where We Tell You What to Do With All of This

Here is what you now know that most operators sitting across a compression renewal table do not.

The rate your vendor quoted is not a negotiating position. It is the number their entire asset base gets marked against. Ask them to cut it and you have not asked for a discount. You have asked them to call their auditors and explain why their portfolio is worth less than it was yesterday. That conversation does not happen. Not because your vendor is unreasonable. Because their accounting will not allow it. The REIT does not discount one floor of the building. The compression vendor does not discount one contract. The math is the same.

What the Axip filing gave you, and what three years of record utilization, record margins, and record distributions from the healthy vendors obscures, is a precise picture of how fragile the occupancy model actually is. One counterparty event. Fifteen percent vacancy. A debt structure with no cushion. The building empties faster than the mortgage adjusts. That fragility does not disappear because USAC is at 94.5% utilization and Archrock’s dividend coverage is 4.9 times. It goes underground. It waits. And it surfaces, quietly, in the primary term exposure of a contract book that nobody outside the vendor’s treasury department is tracking.

Until now, anyway.

Build the model. It does not have to be perfect. It has to be better than walking in blind, which is a low bar to clear. Use their public filings. Use their stated IRR targets and leverage ratios and cost of capital disclosures. Back into the utilization floor below which the arithmetic stops working. Map their counterparty concentration against the current commodity price environment and ask yourself whose problem you are solving by signing a renewal. That is your band. That is the realistic range of what this negotiation can produce, before anyone has said a word about rates.

Then, and this is the part that turns a framework into a result, look for the intelligence that tells you where you sit inside that band. Primary term exposure. Counterparty concentration. Strategic overextension. The signals that tell you whether you are negotiating from the middle of their comfort zone or the edge of it. (Kalibr tracks this at the vendor level, if you would rather not spend a weekend with a Bloomberg terminal and three years of 10-K filings. Either way, the inputs exist.)

What you will find, once you have done this work, is that the negotiation looks different than it did before. Not because the vendor suddenly became generous. Because you stopped asking for things the accounting will not allow and started asking for things it will. There is a whole category of concessions that compress vendor economics without touching the portfolio valuation, concessions that are, in some cases, worth more to an E&P operation than a rate cut would have been anyway.

That is next week.

But here is the thing to sit with until then. Every sophisticated commercial real estate tenant knows that you do not walk into a lease renewal and ask the landlord to reprice the building. You ask for three months free rent, a tenant improvement allowance, and a right of first refusal on the suite next door. The landlord says yes, the portfolio valuation stays intact, and both parties leave the table having gotten something real. The dollars per square foot never moved. Nobody’s auditors had an uncomfortable morning.

Your compression vendor is the landlord. You have been asking them to reprice the building. You need to be asking about the tenant improvement allowance. More to come.

Nominally Hedged goes out to a small list. If you know someone running compression negotiations this cycle — or who should be — forward it. That's the whole ask.

Let’s say a foreign company makes widgets (Yes, I love widget examples). Congress says that’s a national emergency. Three agencies get to work.

Tool one: the antidumping duty. The Commerce Department figures out what the widget should cost in the foreign producer’s home market, then turns that theoretical cost advantage into a tariff rate. Some came in above 100%. Others at 44.93%. Others at 78.30%. These numbers imply very careful math. Whether they correspond to anything observable in the real world is a question the lawyers prefer not to examine, because the lawyers get paid either way. The widget does not have an opinion.

Tool two: the countervailing duty. Same idea, different offense; not “you’re selling too cheap” but “your government is helping you sell too cheap.” Stacked directly on top of the first rate. Two agencies. Two rate-setting processes. Two sets of lawyers in no hurry.

Tool three: Section 232. Different game entirely. The first two tools are about fairness. This one is about national security. A blanket 25% tariff, no dumping finding required, no subsidy calculation necessary. Security matter, done. The widget is now a geopolitical actor.

All three stack. The widget isn’t just expensive. It’s expensive in three directions at once, for three official reasons, administered by three agencies who have never been in the same room and have no intention of changing that.

Then the exceptions. Bilateral carve-outs, hard quotas, tariff-rate agreements; each country negotiating its own arrangement, each deal its own legal instrument, each instrument its own lawyers. One producer negotiated an import allowance calibrated to an exact ton (not a round number, an exact figure) in a document justified by national security. The widget, at this point, probably has its own trade attorney and a LinkedIn post highlighting “The 10 things I learned about B2B sales through tariffs.”

By the mid-2020s, the machine had issued so many bespoke exceptions that it had stopped sorting anything. Then the reset. Every exemption revoked. Blanket tariffs reinstated and doubled. Years of learned behavior, erased overnight.

The widget doesn’t care what you call it. It does care that it is OCTG, and it goes in the ground.

Subscribe now

One Market, Then Two

A stacked tariff doesn’t hurt all pipe equally. It hits premium and commodity grades asymmetrically, and that asymmetry, without a single phone call or industry meeting, sorted an entire market.

At the premium end, the purchase decision was never just about price. There’s a specification, a qualification process, a technical relationship, institutional knowledge built over years. A foreign producer competing here isn’t fighting the tariff alone. They’re fighting the tariff plus the qualification barrier plus the certification cycle plus the accumulated weight of every prior purchase decision. Domestic producers looked at this and made a rational choice: this is where we live. The tariff didn’t just protect premium, it calcified it. Producers with strong balance sheets, domestic manufacturing, and proprietary connections were going to win this segment regardless. The tariff just meant winning it required less effort, at higher margins, against a field that had been administratively thinned.

The commodity end is a spreadsheet. No proprietary spec. No qualification cycle. Just: what does it cost delivered. The tariff bites here too (50% is still 50%) but a foreign producer with structurally lower input costs can absorb enough to stay competitive. Domestic producers ran the same calculation in reverse: why fight a cost-per-unit battle against someone with lower labor and energy costs when the premium end is sitting there uncontested?

Nobody coordinated this. Everyone read the same signal, the tariff gradient, permanent and public, and independently arrived at the same answer. Domestic producers migrated up. Foreign producers concentrated down. One market became two. The government spent fifteen years building a machine to protect domestic producers and accidentally built the market structure that a McKinsey team would charge seven figures to recommend.

McKinsey would call it strategic portfolio optimization. The game theorists would call it a coordination equilibrium sustained by a common focal point. The lawyers would call it Tuesday.

The operations group that needs to catch up is on the buy side, because the old mental model, domestic and import mills competing across the full product spectrum, price tension everywhere, a credible alternative always one call away, is simply no longer accurate. The segmentation is real, structural, and built on a tariff schedule that is not going away. Most operations playbooks were written for the market that existed before it was reorganized.

You Brought a Weathervane

When operations groups use data to make OCTG decisions, it tends to come from two places. Neither was designed for the market that currently exists.

The rig count model. The industry default demand framework, and it was reasonable when it was built. Rig count as a demand proxy fails to capture drilling efficiency gains — fewer rigs are drilling more footage than three years ago, which means rig count systematically understates demand intensity as that trend continues. It cannot account for lateral length inflation, which changes not just the volume of OCTG demanded but the specific product mix. And it has no mechanism for the bifurcated operator environment you’re actually in right now, where a well-capitalized major on a board-approved program and a leverage-constrained PE-backed private respond to the same WTI print with completely different drilling behavior.

A model that cannot distinguish those two cases is not a demand forecast. It is an activity count. It tells you how many rigs are running. It does not tell you what they are buying, from whom, or at what point in the product stack the pressure is actually building.

Pipe Logix. This one deserves careful treatment, because it is widely cited and widely misunderstood — a combination that tends to be expensive for whoever is on the wrong side of it.

Pipe Logix is a distributor sentiment survey. A diffusion index constructed from voluntary responses by a panel of distributors on questions about order books, inventory, and price expectations. The benchmark pricing data is a legitimate spot market reference. The problem is not the index itself. The problem is what happens when you build a term contract around it.

The whole logic of committing volume to a mill is straightforward: you trade flexibility for price certainty, the mill trades margin for utilization predictability, both sides capture something real. That trade only works if the price you lock in actually reflects the market at the time you commit. Pipe Logix indexing quietly voids that logic. Changes take one to two quarters to flow through into contract pricing — Tenaris management confirmed this publicly regarding their own North American contract reset cadence. The index declined four consecutive months through December 2025. An operations group on an indexing agreement in that environment continues paying near-peak rates for six months after the physical market has already corrected.

What that means structurally: you have surrendered the volume commitment, which is real and binding, in exchange for price certainty, which the indexing lag has made illusory. You have locked in the obligation without locking in the benefit. At that point a spot buyer, no commitment, full flexibility, is actually better positioned to capture the market as it moves. The index doesn’t make term contracts bad. It makes indexed term contracts a mechanism for giving up optionality without getting paid for it.

There are two additional construction issues worth noting, though neither is the primary concern. Survey respondents report prices with an inherent incentive to cluster near prior month levels — sharp declines devalue their own inventory — so with distributor sentiment at lows and roughly three months of supply in the market, the index likely overstates true clearing prices. And the seamless segment has actively decoupled from Pipe Logix when the market falls while using it as a floor when the market rises, a dynamic JP Morgan has specifically noted in Tenaris pricing commentary. Both are real. But the more fundamental issue is the one above: the indexing lag structurally inverts the trade you thought you were making.

Pipe Logix is a weathervane. It tells you which way the wind is blowing after it has already arrived. A weathervane is the right tool until the weather system changes. When it does, you need a forecast.

Subscribe now

The Number Under The Number

The right starting point is not what an index says. It is what it actually costs to make steel tube and put a connection on the end of it. This is harder than it sounds, and the difficulty is not accidental, cost opacity is a competitive advantage for the people selling you pipe.

The domestic OCTG market runs on two fundamentally different steelmaking processes with completely different cost structures.

The EAF mill melts recycled scrap using electricity. Primary input costs are two variables: scrap and power. Both are volatile. Both trade on liquid futures that encode the market’s real-time view of where input costs are heading — which means a disciplined analyst can build an EAF cost floor from public data alone, updated monthly, without a single conversation with the mill. The yield dynamics of the seamless piercing process introduce a cost multiplier that most surface-level analyses get wrong, and getting it wrong compounds into meaningful floor estimation error.

The ERW mill starts with hot-rolled coil, forms it into a cylinder, and welds the seam. Better yield. Lower energy intensity. Lower capital requirement. Structurally lower cost floor. ERW and seamless are not the same product competing on the same cost basis. A procurement strategy that treats them as interchangeable is leaving money on the table before the conversation starts.

Then there is the third archetype that most cost analyses miss entirely: the billet importer.

Some domestic mills do not produce their own substrate. They import semi-finished steel, billets or green tubes, and run them through downstream operations domestically. The finished product is technically domestic pipe. The substrate carries the full Section 232 tariff on the way in.

At a 50% Section 232 rate, imported billet adds a meaningful per-ton penalty relative to a domestic EAF competitor sourcing scrap at market. The producer who built their U.S. manufacturing footprint around imported substrate is paying the tariff that was ostensibly designed to protect them. The wall keeps out the competition and simultaneously taxes the raw material. This is the kind of outcome that makes trade lawyers wealthy and procurement analysts confused.

The practical consequence for a buyer: the billet-import producer has extraordinary competitive strength in the premium segment — proprietary connections, rig-direct distribution, deep operator relationships — and a structurally challenged cost position on commodity grades. They are not trying to win commodity casing. Their cost structure is one of several reasons why. Which changes the bundling conversation considerably.

A few layers that matter on top of raw substrate:

• P110 heat treatment. Quench and tempering adds meaningfully to the cost structure.

• Threading and connections. This is where the cost stack becomes a hierarchy. Fully integrated seamless with proprietary in-house connection technology and rig-direct distribution captures the highest revenue per ton and carries the deepest moat. The connection is not just a product — it is a switching cost embedded in every well program you run. Understanding that is the first step toward negotiating it correctly.

• Freight. Systematically underweighted. Mills were sited deliberately for their basins. A Haynesville ERW mill sitting inside the basin has already won on freight to Haynesville before your RFQ goes out. Procurement strategy that ignores geography is fighting a structural advantage instead of using it.

The one variable that is not directly observable — and the one that matters most when the market is turning — is utilization. Fixed cost absorption moves materially with utilization, and it is the variable mills have the most incentive to obscure in public commentary. The difference between a mill at 70% and 90% utilization represents meaningfully different cost floors, and that difference determines whether current market prices represent a floor or a ceiling with room to fall.

There are observable signals that triangulate utilization without requiring the mill to disclose it. Mapping those signals changes the procurement conversation from what does the index say to how close is the market price to the point where this specific producer starts curtailing. Those are different questions. The second one is the right one.

What The Mill Thinks Of You

The supply-side cost floor tells you the bottom. It does not tell you when the market gets there, how long it stays, or what your specific buying power looks like relative to everyone else competing for the same pipe.

The conventional demand model asks: how much pipe does the market need? That is a reasonable question and also an incomplete one. The market is not a single buyer. It is approximately 650 operators across fourteen basins with different capital structures, different board mandates, different hedging books, and fundamentally different propensities to drill when WTI prints the same number.

A model that starts with aggregate rig count cannot distinguish a well-capitalized major on a board-approved five-year program from a leverage-constrained PE-backed private on a borrowing base renewal. Those two operators respond to the same commodity price signal with completely different behavior. Conflating them is how aggregate forecasts systematically overstate demand at the tails and miss inflections at the turns.

The insight that most operations groups have never formalized: your position on the demand curve is not just a function of how much pipe you buy. It is a function of how reliably you buy it, how predictably you can commit to future volumes, and how that reliability compares to every other buyer the mill is evaluating simultaneously.

A large operator with variable drilling behavior is less valuable to a mill than a smaller operator with a credible multi-year commitment — because the mill is not just selling pipe, it is managing utilization, and utilization predictability has real economic value that shows up in pricing. The operator who understands this can negotiate it. The one who doesn’t pays the price of being treated like a spot buyer even when they’re not.

Where you sit on the volume-conviction surface determines not just what discount you should be able to negotiate, but which mills have the most incentive to compete for your business, at what point in their fiscal calendar, and under what contract structure. Most operators have never mapped where they actually sit. The ones who have are having different conversations with their vendors.

The Midland and Delaware Basins together account for the dominant share of 5.5” P110 production casing demand. The top two specs represent a disproportionate share of total footage demand — both concentrated by basin, with bear-to-bull sensitivity in the 10–13% range. That figure is the model’s answer to a question most operations groups haven’t asked: how much does my buying environment change between a $57 WTI world and a $72 WTI world? The answer, at the spec level, is more than most contracting postures have priced in.

Distributor inventory is elevated. Working capital pressure is visible in publicly disclosed financial metrics for anyone paying attention. The difference between an operations group that understands the distributor’s financial position and one that doesn’t runs $100–200 per ton on identical pipe. That gap is not permanent — it closes as the supply overhang works off. The cost model tells you where the floor is. The demand model tells you when the window closes.

Together they answer the one question operations groups actually need answered: lock in now, or wait?

Timing Isn’t Neutral

The rig count model and the Pipe Logix index were fit for purpose when the market they were built to describe actually existed. The market reorganized. Nobody updated the instruments.

The operations group that has mapped its own position on the demand curve, built its own cost floor, and walked into a renewal knowing what the mill’s utilization-adjusted breakeven looks like — that group is having a negotiation. Everyone else is having a pricing conversation where only one side brought data.

The lock-in window is open. How long it stays open is a function of inputs the rig count cannot see. When it closes, it will close faster than the index confirms it.

Ignorance is not neutral. It is a pricing signal the market reads every time you sign a renewal.

This is the first in a series. The framework above is the foundation; supply-side cost architecture, demand conviction, and the analytical connective tissue that actually answers a contracting question. The interesting part starts now.

A few of what we’ll work through in the coming weeks:

What actually happens to demand when two major operators consolidate? The press release says “synergies.” The demand model says something more specific: which specs compress, in which basins, on which timeline, and what that means for the mills who had both operators as customers and now have one.

If Pipe Logix is a structurally compromised indexing mechanism (and we’ve made the case that it is) what does a better one look like? There are observable market inputs that encode real clearing price information without the survey bias, the inventory incentive distortion, or the one-to-two quarter lag. We’ll build one.

What happens to 4.5” P110 demand concentration when a major Bakken operator pauses its drilling program? The spec is already highly concentrated by basin. A single operator decision can move the supply-demand balance materially — and the pricing response may arrive before most buyers realize the shift has happened.

These aren’t hypotheticals. They’re the questions that fall out naturally when the model is granular enough to ask them. The rig count can’t answer any of them.

Same time next week. Bring your AFE.

Then it gets harder. Your tool is not that difficult to replicate, and your client base consolidates faster than your competitors exit, which means more buying power on one side of the table and a price war developing on the other. You respond the way every serious operator responds: tighter geography, vertical integration, proprietary systems, performance contracts. These work. Mostly. But you are still selling a commodity, and commodity sellers eventually compete on price, and competing on price is a terrible way to spend a career.

So you build a better tool. Not incrementally better — fundamentally different. A tool that nobody else has yet, that demonstrably cuts your client’s cost per job, and that requires serious capital commitment on your part to deploy. Here is where it gets interesting: your client is playing a completely different game. They are a cash flow maximization program. They have a board expecting returns, a buyback target to hit, and a CFO who measures success in dollars returned to shareholders, not equipment purchased. They love the story of cost reduction. They will put it in their investor deck. What they will not do is write the check to make it happen. So when you show up with the better tool and ask for term in exchange for making the capital commitment yourself, they are not reluctantly agreeing. They are relieved. The term contract is not a concession — it is the product. You get EBITDA predictability and the investment thesis to deploy capital. They get the technology, the narrative, and none of the depreciation. Everyone shakes hands and actually means it for once.

The problem is the old tool. You still have a fleet of them. In a soft market, clients ask for the old tool and make you compete on price, which pulls your blended margin back toward a number your CFO mentions at every ops review with the same face. You cannot unilaterally retire them without ceding share to competitors who still offer them. You cannot call your competitors and agree to retire the old tools together, because this is fundamentally not how capitalism is supposed to work. Lina Khan may no longer be running the FTC, but the Sherman Act did not leave with her.

What you can do is make a series of carefully worded public statements — in earnings calls, in investor decks, in the particular register of language that sounds like operational housekeeping but lands as strategy — indicating that you are not merely taking the old tools out of service. You are dismantling them. Parting them out. Making the retirement structurally irreversible, not just temporarily inconvenient. Your competitors are listening, because they have the same problem and the same CFO giving them the same look. They may reach the same conclusion independently. The old tools disappear from the market without anyone having to say anything that a regulator would find interesting.

This is, roughly, what happened to conventional diesel frac fleets over the course of 2025. The Q4 earnings cycle just told us whether it worked. The short answer is yes. The more interesting answer is that “it worked” created a follow-on problem that now belongs entirely to E&P operators: where the old market had one commodity tier and some premium options, the new market has three distinct next-generation architectures with different economics, different operator fit, and very different responses to a buyer who shows up without a credible alternative. Someone has to understand what they are. It might as well be you.

Subscribe now

1.1 Million Horsepower Walked Out the Door and Nobody Is Replacing It

Let’s do the ledger first, because the numbers are specific enough to matter.

The D&C index sits at 75 — down from 100 in 2015 — while the staffing index has doubled to 18. Operators are completing more wells per crew, returning more capital per barrel, and running programs explicitly designed to outperform at $61/bbl. That backdrop matters because it explains the behavior on the service side. When your client base is structurally committed to doing more with less, the rational response is to concentrate capital where the returns are highest and liquidate everything else. That is exactly what happened.

Over 2024 and 2025, the top five pressure pumping providers retired, impaired, or scrapped more than 1.1 million hydraulic horsepower of conventional diesel equipment. Patterson-UTI removed 600,000 HHP over two years. ACDC retired 400,000 HHP in a single quarter. ProPetro took a $189 million write-down and described the remaining diesel fleet as being “harvested for cash” with zero maintenance CAPEX. Halliburton described their conventional idle capacity as “consciously stacked and price-dated,” which is a slightly more collegial way of saying the same thing. Liberty Energy cannibalized their legacy equipment for spare parts. In no case was any of this equipment stacked while awaiting a better market (with the exception of HAL). It was stripped for transmissions and power ends, written off, and in several cases physically scrapped.

Zero dollars replaced it in kind. Not one provider discussed building new Tier 4 DGB equipment on any earnings call across eight consecutive quarters. The existing Tier 4 fleet is the fleet. Minus annual attrition, which nobody is replacing either.

This does not mean Tier 2 and Tier 4 DGB equipment vanished. It means they became something different: the new commodity floor. The old commodity tier was Tier 2 diesel. The new commodity tier is Tier 4 DGB — the equipment that every provider still has in some quantity, that sets the baseline market price, and that service companies are increasingly trying to deploy in simul-frac configurations rather than single-well operations. The reason is straightforward: running two wells simultaneously with Tier 4 DGB requires more horsepower per location, reduces crew count relative to stages completed, and improves service company margin per job without requiring newbuild capital. Your vendor’s enthusiasm for getting you onto simul-frac with their conventional fleet is not purely about your cost structure. It is also very much about theirs.

Every dollar of newbuild capital went above the commodity tier. And the thing about next-generation equipment in frac is that it is not one thing — it is three distinct things, each with a different power source, a different capital structure, and a different answer to the question of which operators it actually fits.

The Part Where We Explain That “Electric Fleet” Is Not One Fleet

The conventional framing in E&P operations is: there are old diesel fleets, and there are new electric fleets, and you should try to get the new electric fleets because they are better. This framing is approximately true the way “a plane and a helicopter are both aircraft” is approximately true. Useful at 30,000 feet. Unhelpful when you are trying to figure out which one lands in your backyard.

Every dollar of newbuild capital is flowing into one of three distinct architectures. They share a fuel source and a general direction of travel. They do not share a capital cost, an operating envelope, or an operator fit.

HAL Zeus runs aircraft-derivative gas turbines generating ~35MW of continuous electrical output distributed across purpose-built electric pumps. At approximately $60 million per fleet it is the most capital-intensive deployment in the peer group — and build-to-contract only, meaning no spot market exists. The operational payoff is real: Zeus executes simul-frac at ~65,000 HHP versus 100,000 to 120,000 HHP for legacy configurations running the same job design. The switching cost is structural. Zeus deployments embed wellsite power infrastructure, data integration, and choreography that is incompatible with any competitor fleet. Operators who built programs around Zeus efficiency metrics have created a durable relationship with Halliburton whether or not they intended to.

LBRT digiFrac and digiPrime are two distinct platforms that Liberty’s investor materials tend to flatten into a single “natural gas fleet” narrative. digiFrac is a clean-sheet purpose-built electric pump with the best utilization record in the peer group: 7,143 pumping hours per year on a single fleet, representing 96% of the calendar (~19.5 effective hours per day) at operating pressures exceeding 10,000 psi. digiPrime is mechanically different — a variable-speed natural gas reciprocating engine in direct mechanical coupling, eliminating the electrical conversion step. Lower capital cost, simpler infrastructure, and a program fit that does not require Permian-scale pad operations to justify the economics.

PTEN Emerald takes the mechanical drive premise further still. No turbine, no electrical system, no conversion losses — natural gas combustion to pump motion in a single step. Capital runs 25 to 30 percent below a full e-fleet at 100% diesel displacement. The differentiation from Zeus and digiFrac is program fit: Emerald does not require the water logistics of trimul-frac (~330,000 bbl/day versus ~110,000 for a conventional zipper) and works in basins and completion designs where pad scale cannot support electric fleet economics. You change the iron. You do not change the program.

The Part Where We Pull Out a Calculator

The conventional way to track frac market tightness is total horsepower. Count the active fleets, multiply by nameplate HHP, compare to demand. This number is now systematically misleading, and the operators still using it are making procurement decisions with a broken instrument.

Total horsepower deployed across the peer group remained roughly consistent through 2025 even as fleet counts declined. The reason: simul-frac and high-efficiency operations require 30 to 100 percent more horsepower per job than conventional single-well pumping. A market that looks flat in raw HHP terms is actually tightening in job capacity. The retired Tier 2 diesel was completing approximately 60 percent of the stages per month of a next-generation fleet. When you retire 1.1 million HHP running at 60 percent efficiency and replace part of it with equipment running at 95 percent, the aggregate horsepower number goes down but the job capacity does not. What also does not survive the transition is the pricing that attached to the old equipment.

The E&P operations engineer benchmarking their 2026 frac budget against 2023 actuals is comparing against a market that no longer physically exists. The equipment that set those prices has been parted out. What replaced it has a different cost structure, different margin expectations, and a very different answer when a buyer shows up without a credible alternative.

Q4 2025 Earnings: What the Numbers Actually Said

The Q4 2025 earnings cycle is the most useful dataset this peer group has produced in several years, for a straightforward reason: soft markets make financial disclosures honest. The narrative in prepared remarks — “steady,” “resilient,” “disciplined” — runs about two minutes before the analyst Q&A begins disassembling it. The job is to read the transcripts, the 10-Ks, and the broker notes together, and build a vendor picture based on what the numbers require rather than what management preferred to say.

What follows organizes five companies across the contracting dimensions that actually matter before you sign a 2026 completion program. Margins and leverage matter for obvious reasons. Pricing trajectory, customer concentration, capital posture, and management tone matter because they determine how this vendor behaves in a renewal negotiation and what happens to your program if their internal situation changes mid-contract.

Patterson-UTI (PTEN)

Q4 EBITDA margin: 19.2% — up 60bps sequentially on flat revenue of $1.15 billion. Margins expanding while revenue holds flat is a cost structure story, not a pricing power story. The company’s Q4 free cash flow came in at $259 million against a $167 million analyst consensus — a 55% beat. Approximately $15 million of that came from customer prepayments for 2026 work, which is itself a negotiating data point: counterparties who prepay are telling you something about their view of available alternatives. Net debt down to $822 million from $1.06 billion at Q3, at 0.9x net debt/EBITDA. No senior note maturities until 2028. Total liquidity $739 million. Fitch affirmed BBB- with stable outlook in August 2025. A 25% dividend increase to $0.10 per share — a signal of confidence in sustained cash generation that is harder to fake than a management quote.

The vendor that raises its dividend 25% in a flat market has decided it does not need your volume badly enough to discount for it. That is the first thing to understand about negotiating with PTEN in 2026.

What the pricing language actually says: Management described Q4 rates as “steady” — the same word used in Q3, which they also called “steady.” Barclays analyst Sungeun Kim identified the gap: Q1 2026 Completion Services gross profit guidance of $95 million represents a roughly 14% sequential decline against activity characterized as only “slightly” lower. Her question: “That would seem to imply a not insignificant pricing decline. Is that a fair assessment?” CEO Andy Hendricks responded “Not at all” and attributed the gap to weather and mix. Companies that are genuinely holding price explain the arithmetic. Companies managing the narrative change the subject. The gap between the word “steady” and the implied math in the Q1 guide is real and belongs in your model.

Completion services contract structure — what the frac business actually looks like commercially: PTEN’s completion programs reset annually. Dedicated fleet pricing is negotiated at year-end and “locked in place” for the coming year — management’s language, not a summary. Spot work fills calendar white space between dedicated programs, at whatever the market clears. Approximately 10% of completion work has shifted to performance-based agreements, where compensation is tied to efficiency outcomes rather than a fixed stage or hourly rate. That last number matters for a renewal negotiation: the buyer who enters a performance-based discussion expecting to benchmark against a simple dayrate is negotiating the wrong contract type against the wrong variable. The customer prepayment signal — roughly $15 million in advance payments for 2026 work embedded in the Q4 FCF beat — tells you something about how the customer base is sorting. Counterparties who prepay have decided that their alternative is worse than locking in early. Counterparties who did not are either testing the spot market or waiting to see if Q1 weather and pricing headwinds create an opening. Knowing which of those you are is the operational intelligence question.

Where the pricing power is real: PTEN was explicit that natural gas-capable equipment — roughly 85% of active fleet horsepower by year-end 2026 — is operating at near-full utilization with “minimal spare capacity.” Management predicted the gas fleet would be “essentially sold out” by Q2 2026. If that proves accurate, and Haynesville and Appalachian activity responds to LNG-driven demand as modeled, the pricing leverage dynamic for gas-capable units shifts sharply toward the vendor in H2 2026. The Permian is a different market. West Texas pricing is, per PTEN’s own words, “certainly still competitive” relative to gas basins. Buyers sourcing Permian oil-weighted programs have more room in 2026 than buyers competing for gas basin capacity.

What management conceded under pressure: When calendar white space appears, PTEN fills it at spot, and spot means a lower price. Exact language: “if there’s any white space in the calendar... we have to fill some dedicated work with some short-term spot work, maybe we take a little bit lower price to do that.” That admission is a procurement playbook. A buyer who delays award decisions or deliberately creates scheduling gaps is creating exactly the condition that triggers PTEN’s spot market behavior. Whether that tactic makes sense for your program depends on how much program continuity risk you are willing to carry in exchange for a rate concession.

Customer concentration: Top customer at 12% of consolidated revenue for full year 2025. Top five: 39%. Top ten: 57%. Roughly 60% of revenue from the 15 most active U.S. E&Ps — large public and large private, structurally resilient but also the customers PTEN cannot afford to lose, which provides leverage to that cohort in renewal discussions. Performance-based agreements now cover approximately 10% of completion work — a contract structure that protects PTEN’s revenue from dayrate deflation by tying compensation to efficiency outcomes. The buyer who pushes purely on rate against a performance contract is negotiating the wrong variable.

Halliburton (HAL)

Q4 adjusted operating margin: 19.8% — the highest in the peer group, up 130bps sequentially. Before using this as a benchmark for anything, read the Q&A. CFO Eric Carre confirmed that over half of the guided 300bps margin compression for Q1 2026 is explained by the non-repeat of year-end completion tool sales — a transactional event that inflates Q4 and disappears from Q1. Anyone using HAL’s Q4 C&P margin as a run-rate for North American frac profitability is using the wrong input. The underlying run-rate is closer to the Q1 guide than the Q4 print.

Balance sheet: Net debt $4.95 billion at year-end, down from $5.4 billion in Q3. Net debt/EBITDA 1.2x at investment-grade (BBB+/A3). Total liquidity approximately $5.7 billion. Q4 FCF $875 million; full year $1.9 billion. The company returned 85% of annual FCF to shareholders and repurchased $1 billion in stock during 2025. The 2026 CAPEX budget is down 30% from 2025 — the largest percentage reduction in the peer group.

The contracting implication of the North America guidance: HAL explicitly guided to a high-single-digit decline in North American revenue for 2026. The highest-margin vendor in this peer group is deliberately shrinking its North American completions business. CEO Jeff Miller: “Pricing reaches a point where it’s not... companies aren’t investing in it. We are moving equipment away from it.” That statement functions as a credible commitment to supply withdrawal only because HAL actually has an alternative — 59% of total revenue is international, growing 7% sequentially in Q4. A vendor who genuinely does not need North American volume to cover overhead can execute a withdrawal threat. Most vendors cannot. HAL can. This changes the negotiating dynamic in a specific way: HAL is not a vendor who will chase your program at a concession to maintain utilization. The question for E&P procurement is whether the premium over comparable conventional services is justified by the performance differential — because HAL is not offering a discount path to find out.

Customer and counterparty risk — one number that belongs in any complete HAL analysis: The company entered into $750 million in Credit Default Swaps to hedge receivables exposure to Pemex, which represents 11% of total receivables. This reflects a working capital risk in the consolidated entity that does not appear in the North American completions P&L but affects overall financial positioning. The significance for E&P buyers is indirect but real: a company managing concentrated international credit risk may price NAM work differently than its standalone NAM margins would suggest.

What the stacking strategy means operationally: HAL stacked fleets in Q3 and continued stacking in Q4 — voluntarily, at rates that management characterized as “not economic.” The crews associated with stacked fleets do not stay on payroll indefinitely. The personnel who run HAL’s most complex completion systems are specialized and mobile. A sustained period of voluntary capacity withdrawal has implications for the speed and cost of reactivation if demand recovers sharply. The vendor who confidently withdraws supply in a soft market is also the vendor who faces the steepest reactivation cost if the market inflects faster than they modeled. That asymmetry is worth tracking.

Liberty Energy (LBRT)

Q4 EBITDA margin: 15.2%, up from Q3’s 13.5%. Revenue grew 10% sequentially — the strongest sequential revenue print in the peer group. Balance sheet: net debt $219 million at year-end at approximately 0.4x net debt/EBITDA — the lowest leverage ratio in the group. Total liquidity $281 million, up from $146 million at Q3.

The pricing disclosure that should be in your file: Liberty was more specific than any other company in this peer group, and the specificity is analytically useful precisely because the content is unfavorable to the vendor. CEO Ron Gusek quantified the Q4 RFP season impact as “low to mid-single digits” below H2 2025 rates — confirming that dedicated customers who demonstrably value Liberty’s operational record still extracted price concessions during the annual contract reset. His description of the trend: “a slow and steady trend downwards” from the mid-2022 peak. Goldman Sachs analyst Ati Modak pushed directly on the contradiction between “flight to quality” and concurrent pricing concessions. Gusek’s answer: competitors “find white space” and defend share with price, “we are not immune to these market forces.” That sentence is worth its own line item in any Liberty renewal discussion. The vendor who publicly acknowledges they cannot hold price against commodity competitors is giving you the benchmark for your next negotiation.

Why this matters for contracting beyond the rate: The RFP process that drove these concessions ran during Q4 2025, when activity was soft and buyer leverage was highest. The rates set in that process are locked for 2026. If the gas demand inflection that LBRT (and PTEN) are predicting for H2 2026 materializes, those rates will look increasingly favorable to Liberty — and they will have the utilization data and market conditions to push for higher rates at the next annual reset. Operators who locked term in Q4 2025 captured the trough pricing. Operators renewing in Q4 2026 may be negotiating in a different market. The timing of your commitment is itself a pricing decision.

Capital allocation — the most consequential signal in this quarter’s disclosure: 2026 completions CAPEX approximately $250 million, of which roughly $175 million is maintenance. Power generation CAPEX: $725 million to $900 million, targeting 3 gigawatts of deployed capacity by 2029. That ratio — $3.50 to $3.60 of power capital for every dollar of frac capital — is a clear statement of where growth investment is going. CFO Michael Stock confirmed capex is “markedly shifting” toward power with completions in maintenance mode. The specific phrase: “No digi builds in 2026.”

What this means for a multi-year completion contract: Liberty’s frac business has formally entered harvest mode to fund a power infrastructure buildout. The business runs well today — the Q4 operational numbers support that. The question for an E&P operations team evaluating a 2026-to-2028 term contract is whether frac continues to receive the management attention and equipment investment that the current operational record reflects. The $800 million per year flowing to LPI is not flowing to frac fleet quality. Maintenance CAPEX budgets on existing equipment set the reliability trajectory. The gap between “maintenance mode” and “adequate maintenance” is what your completion cost per foot eventually measures.

Customer concentration — monitor the ExxonMobil exposure: Occidental Petroleum and ExxonMobil each exceeded 10% of 2025 revenue. The ExxonMobil position warrants specific attention given ProPetro’s parallel and disclosed non-renewal situation for two FORCE electric fleets contracted to the same operator. Whether the ExxonMobil programs in Liberty’s portfolio carry equivalent renewal risk is not publicly disclosed — but a procurement team that knows ExxonMobil is actively restructuring its vendor commitments across multiple service providers should be asking the question directly rather than waiting for it to appear in a 10-K.

ProPetro (PUMP)

Q4 EBITDA margin: Estimated at approximately $32 to $34 million total — a run rate suggesting roughly 11-12% margin on the completion services business. The organizing fact for ProPetro in 2026 is not the margin. It is the customer concentration.

The ExxonMobil situation in plain arithmetic: ExxonMobil accounts for approximately 19.7% of total ProPetro revenue. Per ProPetro’s own prospectus filing dated January 2026, the agreement governing two FORCE electric fleets for ExxonMobil expires in late 2026 and is “not expected to be renewed.” Permian Resources: 14.9% of revenue. EOG Resources: 10.6%. Top five customers: 58.8% of revenue. Top ten: 75.3%. ProPetro derives 98.5% of fleet revenue from the Permian Basin. When approximately one-fifth of your revenue from a single customer is disclosed as non-renewing in your own SEC filing, the follow-on question for every other customer in the portfolio is whether their program pricing reflects a vendor under competitive pressure to replace that volume. It usually does.

The equity raise tells the story that the earnings call could not: ProPetro launched a 12.5 million share equity offering in January 2026 specifically to fund ProPWR growth capital. The timing — equity raise at cyclical lows, simultaneously with a disclosed anchor-customer non-renewal — reflects a vendor whose organic cash flow from the core completion business is insufficient to fund its strategic pivot. Management has stated the completion business is in “maintenance mode.” The ProPWR book includes 220 MW of contracted power, including a 10-year Permian operator deal and a data center contract. The strategic logic is sound. The contracting implication for E&P buyers is more immediate: a vendor simultaneously managing a 20% revenue hole from a non-renewing anchor customer and redirecting capital to an adjacent business is a vendor whose internal resource allocation decisions deserve explicit conversation before any program is committed. The best equipment and most experienced crews go to the anchor programs. Understanding where your program ranks in that allocation is not a theoretical question in ProPetro’s current situation — it is the operational risk question.

The Permian density is real and belongs in the analysis: ProPetro’s operational density in the Permian Basin — 98.5% of revenue, infrastructure concentrated around a single geography — creates genuine logistical advantages that distributed competitors cannot easily replicate. Mobilization costs, crew familiarity, and local supply chain depth all improve at this concentration level. For a buyer running a Permian-only program with no geographic flexibility in their vendor set, ProPetro’s basin expertise offsets some of the customer concentration risk. The relevant question is what rate premium that density justifies relative to the negotiating leverage available from the current financial situation.

What “disciplined” actually means in this context: Management characterized Q4 pricing as “disciplined” — the word service companies use when they are trying to hold a line they find harder to hold than the word implies. Q3 pricing language was “weak.” “Disciplined” is the upgrade. The arithmetic behind the upgrade is that ProPetro reduced active fleet count from 13-14 in Q2 to 10-11 in Q3-Q4 — deliberately idling fleets to avoid working at sub-economic rates. A vendor who idles 20-25% of their fleet to protect pricing has made a specific judgment about where the pricing floor is. It also means 20-25% of that vendor’s revenue-generating capacity is sitting idle, which changes the math on how aggressively they need to fill what remains at what rate.

ProFrac (ACDC)

Start with the credit event, because everything else is organized around it: S&P Global Ratings lowered ProFrac’s issuer credit rating to CCC in January 2026 before withdrawing it at the issuer’s request. The withdrawal does not reverse the downgrade or the underlying conditions that produced it. Net debt: approximately $1.04 billion against $58 million in cash at Q3, at roughly 3.4x net debt/EBITDA — the highest leverage ratio in this peer group by a factor of approximately three. Q3 free cash flow: negative $29 million. The company executed a fourth amendment to its credit agreement in December 2025, reducing amortization payments and deferring leverage ratio testing to March 2028. The ABL facility becomes current in early 2026.

What 3.4x leverage at CCC-equivalent capital costs means operationally: The interest expense on $1.04 billion in net debt at current borrowing costs is consuming operating cash flow at a rate that limits the capital available for equipment maintenance, crew quality, and program investment. The $100 million annualized cost-saving initiative — covering SG&A (reduced 17% sequentially in Q3) and a headcount reduction executed October 2025 — is the organizational response to that constraint. Cost-cutting preserves liquidity. It does not improve the equipment maintenance posture. Q3 Stimulation Services segment EBITDA margin: 6%. West Texas pricing described as “competitive” — the word in this industry that means the vendor is not setting the clearing price, the market is.

The cost structure advantage is real and belongs in this analysis: ProFrac’s in-house manufacturing produces frac equipment at approximately $540 per HHP versus an industry average of approximately $861 per HHP — a 37% cost advantage on newbuild equipment. The company also owns sand mines and chemical manufacturing, capturing margin at multiple points in the completion cost chain that non-integrated competitors pay through. This is not trivial. The provider who can build at 63 cents on the dollar has a structural cost floor that competitors without equivalent integration cannot match. If ProFrac’s financial structure were different, this cost basis would translate into durable pricing power on the low end of the market.

The problem the cost advantage cannot solve: The interest expense on $1.04 billion in net debt at CCC-equivalent capital costs is consuming that operating advantage faster than the business regenerates it. The fourth credit agreement amendment is a deferral of the leverage ratio test, not a resolution of the underlying leverage. The buyer who negotiates the deepest rate concession from ProFrac — and the leverage for that negotiation is genuinely high, because the vendor needs cash flow to service debt and will price work accordingly — is the buyer with the highest exposure if a restructuring process interrupts mid-contract. Those two facts describe the same entity observed from opposite ends of the same problem.

The commercial pivot and its execution risk: Management’s 2026 strategy is to move from spot-heavy to dedicated programs, targeting 90%-plus of active fleet capacity on committed work. The rationale is sound — spot market pricing has converged with dedicated pricing, eliminating the premium that previously made spot market exposure attractive, and the “head fakes” in Q3 spot activity caused material operational inefficiency. A dedicated program base provides cash flow stability that the debt structure requires. The execution risk: restructuring commercial terms from a position of financial distress limits how aggressively ProFrac can negotiate the terms of those dedicated agreements. A counterparty who knows the vendor needs the volume to service its debt is a counterparty with a structural negotiating advantage in every discussion about price, equipment quality, and crew assignment.

The specific data point most buyers overlook: Spot pricing and dedicated pricing have converged. Per management: “spot and term pricing [are] pretty in line relative to each other.” The historical premium for accepting spot market volatility has disappeared. This means the buyer who previously used spot market pricing as a lever in dedicated contract negotiations has lost that benchmark. The market has done the compression for them. What remains is leverage derived from the vendor’s financial position — and that leverage is substantial and real, but it comes with the program continuity risk described above.

The Vendor Scorecard

Five observations the matrix reveals that the individual dossiers do not.

1. HAL and PTEN score identically in total, but for structurally different reasons. HAL’s score is anchored by balance sheet strength and willingness to credibly withdraw supply — genuine leverage backed by an international revenue base that does not require North American volume. PTEN’s score reflects operational consistency and margin quality. Where they diverge is on frac management attention: HAL’s active capital investment is migrating toward Zeus, which is a different negotiation than conventional tier. PTEN still has conventional capacity, even if it is being managed at the margin.

2. LBRT’s composite score (23) understates the operational quality of the frac business and overstates the risk — the score reflects the capital allocation signal, not a current performance problem. The business runs well today. The question is the 2026 and 2027 forward-looking answer when $800 million per year in growth capital is flowing to LPI rather than frac. That uncertainty loads the “management attention” and “program continuity” dimensions.

3. PUMP’s score of 18 reflects a specific, identifiable event rather than a general posture of weakness: the ExxonMobil non-renewal. A vendor managing a 19.7% revenue hole while raising equity to fund an adjacent pivot is a vendor whose negotiating position has narrowed for a specific window. The basin depth and operational density in the Permian are genuine, and they matter for a Permian program.

4. ACDC scores higher on buyer pricing leverage (5) than any other vendor and lower on program continuity (1) than any other vendor. Those two numbers describe the same entity observed from opposite ends of the same problem. The operator who negotiates the deepest rate concession from ProFrac is the operator with the highest exposure if a restructuring process interrupts mid-contract. That trade-off does not resolve without a specific view on whether the credit trajectory stabilizes.

5. The peer group’s aggregate management attention score on frac is the number that deserves the most scrutiny from an E&P operations team doing multi-year planning. Three of the five vendors in this peer group are actively redirecting growth capital away from frac. The equipment in the commodity tier is the equipment that exists today, managed on maintenance budgets that are at or below the floor required for standard PM intervals. That is the market you are contracting into in 2026.

TLDR: The Tool Manufacturer Won

Remember the tool manufacturer from the beginning of this piece. They had an old tool, a new tool, a Sherman Act problem, and a CFO with a face. The Q4 2025 earnings cycle is the closing chapter of their story: the old tool is gone. Not stacked. Not temporarily idled. Gone. HAL pulled the last of the conventional excess capacity off the market and is holding it in reserve to be reactivated at a price of their choosing, which is a different thing than capacity that is available at any price. When the seller controls the release valve, the buyer does not control the pressure.

The cascade from here is straightforward if you follow it in sequence. Conventional diesel is off the table as a competitive tier. Tier 4 DGB is now what diesel used to be — the commodity floor, the equipment receiving maintenance-only capital, the fleet that every service company still runs but none of them are growing. Next-generation electric and mechanical drive is where newbuild capital is going, and most of that capital is chasing power infrastructure and international markets, not more frac fleets. The service companies have spent four years methodically eliminating their own worst alternative — cheap conventional supply they could not profitably stack — while simultaneously building yours. That is the complete picture.

Here is what it means in practice.

If you are running DGB equipment in 2026, stop optimizing for rate and start optimizing for reliability. The capital that used to maintain your vendor’s fleet is leaving. Simul-frac customers get the maintenance CAPEX that remains, because simul-frac is where the margin is. A further concession on stage rate from a vendor whose fleet is being systematically underfunded does not reduce your completion cost. It increases it — it just moves the cost from the invoice to the NPT report. Build performance standards and equipment quality provisions into the contract before you negotiate price. The vendor running a properly funded maintenance program signs those terms without objection.

Start with a should-cost model. You do not need Kalibr Pulse to build one. Public financial disclosures give you fuel consumption benchmarks, maintenance cost ranges, crew cost data, and the margin math at current activity levels. What you are looking for is whether the bid in front of you is fundable at the performance standard your program requires, or whether the vendor is borrowing from future reliability to win today’s award. The gap between a sustainable margin and the clearing price in your RFP is the leading indicator of future downtime. Build the model. Read the gap.

Long-term, the question is optionality. The technology choices you are making today in development planning have a longer shadow than they appear to. A Zeus-optimized program is a program designed around pad configurations, water logistics, and well sequencing that is incompatible with conventional fleet economics. That is a feature when Zeus is available and performing. It is a constraint when it is not. Before you build pad architecture around any single provider’s fleet specifications, understand what you are giving up and for how long.

If you have a consistent completion program outside the Permian and you are not currently in conversations with ProPetro, have the conversation. The Permian service market is facing structural headwinds — too many fleets concentrated in one basin with anchor customer commitments unwinding. ProPetro has genuine operational density in the basin that does not travel easily, but the financial math of their current situation creates negotiating conditions that will not persist indefinitely. A non-Permian E&P with a reliable program is exactly the customer they need right now, which means you have leverage you will not have in eighteen months.

Finally: if this situation looks familiar, it should. What happened in the frac market over 2023 to 2025 — systematic retirement of commodity capacity, capital concentration in premium equipment, service company leverage rebuilding as the commodity alternative disappeared — is structurally identical to what happened in compression in the early 2020s. The compression playbook that worked then works here. Think about what you did differently when you understood that the old commodity compressor fleet was not coming back, and apply the same logic to what you are signing in frac today.

The tool manufacturer built three new tools, retired the old one, and handed you a market where the commodity option no longer exists. The question they are waiting to see answered is whether you noticed before you signed the renewal.

Subscribe now

The visible mistake gets all the attention. A drilling optimization tool that blows up a motor. A completions design that underdelivers on EUR. An AI model that tells you to pump faster and the well screens out. These are loud. Expensive. Career-defining in the wrong direction. The post-mortem gets a slide deck. Someone presents it at the next ops review. Lessons learned. Process updated. Controls tightened. I’ve been part of many!

The invisible mistake is worse, and nobody writes a post-mortem for it, because nobody knows it happened. You paid $0.50 per gallon too much on surfactant for eighteen months because nobody modeled the feedstock spread. You renewed a compression contract at rates 15% above where the market had moved because the Ops team didn’t have visibility into vendor utilization. You let an escalation clause compound for three years because the original negotiation happened before you had data on basin-level pricing dynamics. These aren’t dramatic. They don’t show up in a single quarter. They just bleed. Slowly. Across every line item on the LOE statement and AFE, every month, compounding into millions that never make it to the distribution line.

Most of the AI conversation in oil and gas is focused on avoiding the visible mistake. How do I make sure this model doesn’t break something? How do I validate the output before it touches an $8 million well decision? Those are real concerns. But they are concerns about acting on something that turns out to be wrong.

The question nobody is asking is: what am I missing by not acting? What margin am I leaving on the table because I don’t have the information to act on it? What is the cost of not knowing what my contracts should look like relative to market fundamentals?

That asymmetry is the whole game. And it’s where the real moat lives.

A week ago I argued that enterprise AI in oil and gas is mostly a Larry David carpool lane scheme. You pay a vendor to sit in the passenger seat, keep the same headcount, add a license fee to the G&A line, and call it transformation. The car costs more and moves no faster. The real path is convergence: collapse the SME and the builder into a single node, build the automation internally, reduce headcount with a clear ROI, and keep the G&A trend moving in the direction the last decade established. No vendor. No license fee. Fewer people in the car.

That gets you a first-mover advantage. But first-mover advantages in oil and gas have a shelf life. Once someone proves a workflow automation works, everyone copies it. The 530-to-750 wells-per-engineer ratio becomes table stakes the way the simul-frac became table stakes. The real competitive moat comes from something different: a differentiated view that nobody else has, embedded in a tool that repeats it at scale.

I promised three buckets. Here they are.

The Three Buckets

The industry is focused on maximizing cash flow. If you accept that premise, and at this point there is no alternative premise, then every AI application has to connect to one of three levers:

1. Reduce Capital Intensity. Think of this as maintenance capital. How do I inject the minimum amount of capital required to meet my relatively flat production profile? Two sides: drilling and completion efficiency, and contracting those services.

2. Increase Topline Production. Better recovery per well, improved production uptime, artificial lift optimization, completions design that actually moves EUR.

3. Boost Bottom Line by Cutting LOE. Reduce the money it costs to produce the well. Chemicals, compression, water handling, workover frequency. Every dollar off the LOE line goes straight to free cash flow.

Simple enough framework. Three buckets. But the buckets are not created equal, and the reason they’re not has everything to do with the structure of the errors inside them.

Buckets 1 and 2: The Service Provider Problem

Reducing cycle time and increasing production are the categories that get the most airtime at conferences. Drill faster. Frac smarter. Optimize your ESP settings with machine learning. Design completions with AI. These are exciting. They are also, for an E&P building in-house, structurally disadvantaged.

Oil and gas relies on outside service providers for most of the physical work that drives Buckets 1 and 2. The drilling contractor runs the rig. The pressure pumper executes the frac. The ESP manufacturer designs and installs the lift system. These companies don’t just have more data than you. They have better data. A model trained across 30 operators and 4,000 wells sees more geological variation, more equipment configurations, more failure modes, and more edge cases than your internal dataset ever will. Your 40-well model doesn’t just lose on volume. It loses on statistical robustness. Diversity of input reduces overfitting, and that is a structural advantage you cannot replicate internally. And because service providers develop the tools, they disseminate quickly to your competition. Your neighbor gets the same algorithm next quarter.

This is the relativity problem I’ve written about before. You don’t move relatively. And relativity is the only thing that matters when you’re trying to build a competitive moat. If everyone gets the same drilling optimization tool from the same service provider, nobody has an edge. You have spent money and effort to arrive at the same place as your peers. The carpool lane, but everyone has a passenger now.

But the second problem is more important than the service provider data advantage. It’s the structure of the risk.

The Two-Sided Bet

Using AI to improve drilling times, completion cycle times, completions design, or ESP run settings is an $8 million decision per well where the errors cut in both directions.

If the model tells you to drill more aggressively and you blow out a motor, that’s the visible mistake. You acted on something that turned out to be wrong. AFE over budget. Post-mortem. Slide deck at the ops review. If the model says a completions design will boost EUR by 15% and it delivers 5%, you just allocated capital to a design that underperformed. Expensive. Measurable. Everyone knows about it.

But if the model could have told you something useful and you didn’t trust it, you left barrels in the ground. That’s the invisible mistake. The well that could have been 15% better but wasn’t, because you drilled the conservative design. Nobody writes that post-mortem because nobody knows the counterfactual existed.

Both errors are expensive. Both are hard to detect in real time. Both impact capital decisions where the feedback loop is measured in months, not minutes. And critically, both have to be managed simultaneously. You can’t just optimize for one without exposing yourself to the other.

This is a two-sided bet. And two-sided bets are hard to build AI around, because the tolerance for error on the downside constrains how aggressive you can be on the upside. The risk management overhead is enormous. The validation requirements are extreme. And the ROI case is muddy because you’re always weighing the cost of the mistake you might make against the value of the improvement you might capture.

That’s Buckets 1 and 2. Important. Real. But structurally difficult for in-house AI to win on, and structurally likely to get commoditized by service providers who have the data advantage.

Bucket 3: The One-Tailed Bet

Now contrast that with AI that helps optimize your cost structure through automated market monitoring, bid optimization, and contracting intelligence.

Here, the error structure is fundamentally different. And that difference is the entire strategic argument for where AI creates a defensible moat rather than a replicable efficiency.

If my model overestimates my buying power, if it tells me conditions are more favorable than they actually are and I bid too aggressively, the decision arc is self-correcting. I submit an aggressive bid. The vendor says no. I adjust the model or explore another option. The cost of being wrong in this direction is a failed bid and a phone call. Feedback is immediate, binary, and cheap. This is also, by the way, one of the reasons I keep beating the drum about disintermediating the middle man in procurement. Every intermediary between you and a “no” from the vendor is a delay in the feedback loop that makes your model smarter.

The mistake that should keep you up at night is the other one. The invisible one. Where you think your position is weaker than it actually is and overpay relative to market fundamentals and your unique buying power. There is no vendor calling you to say, “Hey, you could have gotten this 12% cheaper.” There is no natural correction mechanism. The overpayment just compounds. Quarter after quarter. Across chemicals, compression, water hauling, workover rigs, every line item on the LOE statement.

And here’s the critical distinction that makes this a strategically different AI problem from Buckets 1 and 2: you have to buy this stuff from someone. Production chemicals, compression services, water hauling. These are not optional. The question is never whether to spend the money. It’s whether you’re spending the right amount. The only error that actually costs you money is the one where you leave margin on the table. It’s a one-tailed bet.

When you only need to worry about error in one direction, you can build systems that are far more aggressive, far more precise, and far more deployable. You don’t need the model to be perfect. You need it to be directionally correct on the side that costs you money to be wrong. The other side corrects itself on contact with the market.

One-Tailed Does Not Mean Simple

I want to be careful here because the argument so far might sound like Bucket 3 is the easy button. It is not. The error structure is favorable. The moat dynamics are favorable. But the actual work of building the intelligence layer is enormous, and for reasons that are specific to how oil and gas operates.

The first reason is the service-based nature of the business itself. Roughly 80% of the CAPEX and OPEX budget flows through service-based categories. That means what you are actually trying to optimize is Total Cost of Ownership, not just price. And TCO is a fundamentally harder problem. It depends on uptime performance, staffing quality, equipment condition, contract primary term, infrastructure proximity, employee competency. The vendor quoting you $5,000 per month less on compression might cost you more in aggregate if their uptime is 95% instead of 97%. Apples-to-apples comparison across vendors requires performance benchmarking data, operational reliability data, and contract structure data that most operators do not systematically collect. You have to go find unique data sources that allow you to quantify this in a way that makes the comparison objective rather than anecdotal. That is a lot of work before the model even starts learning.

The second reason is what I think of as counting cards. You have to completely assess the deck. A BATNA framework helps here, but deploying it requires cataloging and tracking every potential counterparty across every line item on your AFE and OPEX budget. That alone is a substantial data exercise. But it gets harder. Because the counterparty’s BATNA is other oil and gas companies, you also need a data-driven approach to assessing every operator they could sell that service to instead of you. Who else in the basin needs compression? Who else is buying methanol at volume? Where does your vendor’s next best customer sit relative to you on reliability, volume, and payment terms? And a large portion of both the vendor universe and the operator universe is private, which means you need a novel approach to assessing companies that don’t file 10-Ks or host earnings calls. This is not trivial. This is building an intelligence map of an entire market structure, most of which does not want to be mapped.

The third reason is the fundamental cost decomposition. To know whether a price is fair, you have to understand what goes into it. What are the raw components and feedstocks? What is a reasonable proxy for that market? What other industries besides oil and gas consume those inputs, and how are those end markets performing? Where are the logistical hubs and what are freight dynamics doing? For production chemicals, this means tracking natural gas to methanol spreads, MTO run rates in China, Gulf Coast utilization, hazmat freight premiums into Appalachia. For compression, it means understanding manufacturing lead times, steel and engine costs, fleet age demographics, and the vendor’s cost of capital. This is a full decomposition of the supply chain for every major cost category. And by the way, this analysis fits naturally with the direct sourcing and insourcing evaluation. If you understand the cost structure deeply enough to negotiate intelligently, you also understand it deeply enough to evaluate whether you should be sourcing directly or bringing the capability in-house. The synergies between the intelligence layer and the strategic sourcing decision are real.

So yes, this is a lot. Three interlocking data problems, each one substantial on its own.

But here is where the one-tailed nature of the bet rescues you again. I do not need to solve any of these problems in the absolute. I need to solve them in the relative. I do not need to know exactly what my buying power is. I need to know how it fits in the relative pecking order. I do not need to know exactly what my leverage is with Vendor A. I just need to know that it is greater than my leverage with Vendor B and Vendor C, so I can sequence my negotiations correctly and construct my bids with the right level of aggression for each counterparty.

Relative accuracy is a dramatically lower bar than absolute accuracy. And in a world where most operators have zero systematic intelligence on any of these dimensions, even a directionally correct model puts you in a fundamentally different negotiating position than your peers.

To be clear, I am biased here. This is what Kalibr does. This is why we have spent a significant amount of time studying these markets, building Kalibr Pulse, and developing the algorithms that help on these cost centers. But I think the logic is founded independently of who builds the tools. And I am largely putting my money where my mouth is.

Why This Is the Moat

In Buckets 1 and 2, your drilling data is roughly as good as your neighbor’s. Your type curves look the same. Your service providers are iterating on the same models for everyone. The service provider has a structural data advantage and a structural incentive to spread the gains across their customer base. Anything you build gets commoditized.

Bucket 3 is the opposite. Your cost structure is proprietary. Your vendor relationships are specific to your acreage position, your basin, your volume commitments, your contract history. Your buying power is a function of your scale, your operational rhythm, your pad complexity, your credit quality. None of that is replicable by your neighbor because none of it is the same as your neighbor’s. And unlike drilling or completions, where wider data dispersion across operators makes a model better, in contracting it makes it worse. What compression should cost you given your volume, your vendor history, and your basin dynamics is not improved by knowing what someone in the Haynesville paid. The “bias” in your procurement data isn’t bias. It’s the signal.

The data that feeds an AI contracting model is inherently proprietary: your invoices, your contract terms, your bid history, your vendor performance data, overlaid on market intelligence about feedstock costs, utilization rates, freight differentials, and vendor financial health. This is the kind of data that I wrote about in the Kalibr Sweep last October when we broke down methanol economics. Know what it costs the producer to make the molecule, what it costs you not to buy it, and how both numbers move with feedstock, utilization, and freight. The side with better information captures the margin.

The companies that build this capability will have structurally lower LOE that their competitors cannot explain and cannot replicate. Because you can copy someone’s frac design. You cannot copy their intelligence on what compression should cost them in the Delaware Basin in Q3.

What’s Next

So we have the strategic logic. G&A convergence gets you lean. Bucket 3 contracting intelligence gets you a moat. The error structure is asymmetric and favorable. The data is proprietary. The complexity is real but tractable because you only need relative accuracy, not absolute.

What we have not done yet is the financial case. Over the last two weeks, we have covered two distinct AI applications: the G&A convergence play from Part I and the contracting intelligence play from today. Next week, I want to put both of those into a financial framework. How do you differentiate an organization from a fundraising perspective when it employs AI across both of these dimensions? What does the cash flow impact actually look like when you model G&A reduction alongside structurally lower LOE? And how does that story land with the capital providers, whether PE sponsors, family offices, or public market investors, who are deciding which teams get backed and which ones don’t?

Then we will decompose an actual cost center and walk through building one of these tools. Not the theory. The data inputs, the architecture, and the product. Because the invisible mistakes are not going to find themselves.

This is not a company managing volatility. This is a company that has decided volatility is someone else’s problem.

For anyone sitting across the table from a vendor who counts COP as a customer, that posture matters more than any line item in the 10-Q.

The Score

COP ticked up two-tenths of a point, which in Fortress territory is like gaining altitude at 35,000 feet. You’re already above the weather. The improvement came from two places: balance sheet strengthening (cash swelled by $1.1 billion despite buybacks, net debt dropped) and the efficiency story getting harder to argue with (15% D&C improvement, 30% fewer rigs and crews than the two pre-merger companies combined). The only thing holding the score below 9.0 is the unhedged commodity posture, which is a feature, not a bug, from management’s perspective. It does, however, mean that a sustained sub-$50 WTI environment would test the “we don’t whipsaw” commitment faster than Lance’s tone suggests.

What They Said (And What They Meant)

The Anti-Reaction Function

Every analyst on this call was fishing for the same thing: at what price does COP blink? Lloyd Byrne at Jefferies asked it directly. Lance’s answer was the most revealing sentence of the quarter: “We don’t like to whipsaw these programs up or down, and we’ll use the balance sheet in the downside case if we need to.”

Read that again. He’s not saying they won’t cut. He’s saying the balance sheet takes the hit before the rig count does. That is a fundamentally different operating philosophy than most of COP’s peers, and it has direct implications for anyone selling into their supply chain. A vendor servicing COP’s 24-rig program can model that revenue with higher confidence than a vendor servicing a peer running the same rig count but hedging 40% of production and using words like “flexible” on every earnings call.

Last quarter, the language was about “optionality” and a “highly flexible portfolio.” This quarter it shifted to “resilience” and avoiding “whipsaw.” That’s not a subtle change. Optionality implies you might do something different. Resilience implies you won’t. For anyone modeling COP’s forward activity, Q4 just gave you permission to remove the downside scenario from the base case.

The Efficiency Machine Is Real

Betty Jiang at Barclays did the smart thing and pushed on the quality of the efficiency gains. Was the $1.4 billion pro forma capex reduction real savings or just an activity cut dressed up in synergy language? The answer, honestly, was both. Nick Olds broke it down: some of it is structural (Marathon’s rigs got folded into COP’s steady-state program), some is synergy capture ($500 million), some is operational improvement (more wells per rig), and about $200 million is plain deflation.

Here’s why the composition matters for BATNA analysis. The deflation component ($200 million, driven by 30% tubular cost reductions and lower pumping rates) tells you that COP is capturing service cost reductions in real time. They’re not passing that through as margin to vendors. They’re banking it. If you’re a pressure pumping company or a tubular distributor, COP’s procurement team is quantifying the deflation you’re experiencing and adjusting their cost models accordingly. That $200 million is coming out of someone’s revenue line.

The M&A Door Is Closed (And Bolted)

Leo Mariani at Roth asked if COP was still in the market. Lance’s response was as close to “absolutely not” as a public company CEO gets: “We’ve done our heavy lifting on the M&A side over the last 4 to 5 years. We’ve been there, done that.”

This matters for two reasons. First, the $5 billion disposition target (with $3 billion already closed) means COP is a net seller of assets, not a buyer. That capital is flowing into buybacks and debt reduction, not new acreage. Second, a company in organic execution mode has a different relationship with its vendors than a company in integration mode. Integration creates disruption, which creates procurement opportunity. Execution creates cadence, which creates stickiness. COP’s vendor relationships are hardening, not loosening. If you’re already in, that’s good news for your contract duration. If you’re trying to break in, the window is narrower than it was 12 months ago.

The Delta: What Changed Quarter Over Quarter

Free cash flow compressed by 48%, from $2.5 billion to $1.3 billion, entirely driven by realized price deterioration ($42.46/BOE vs. $46.44/BOE prior quarter) and continued Waha gas basis punishment. That’s the number that makes headlines. It’s also the number that doesn’t matter for activity resilience, because COP’s cash balance still grew by $1.1 billion to $7.4 billion thanks to $1.6 billion in asset sale proceeds. Total liquidity expanded to $12.9 billion. Net debt fell $763 million. Management’s tone on capital commitment was firmer, not softer, despite the cash flow compression. The message: the balance sheet absorbed the hit so the operating machine didn’t have to.

The 2026 opex guidance dropped $400 million to $10.2 billion. Willow’s cost estimate crept up to $8.5 to $9.0 billion (from $7.0 to $7.5 billion), mostly North Slope inflation. One number went right, the other went wrong, and neither changed the rig count.

The Kalibr Take

The consensus narrative post-call is that COP is a “safe haven” in a volatile commodity environment. That’s true and it’s also not the interesting part. The interesting part is what COP’s posture does to the rest of the game board. When the largest independent E&P tells the market it will not cut activity in a downturn, it creates an asymmetric information problem for every vendor in the Permian and Eagle Ford. COP’s 24-rig program is “priced in” to every service company’s revenue forecast. That makes COP’s volume reliable but also makes it non-negotiable. The vendors know COP won’t leave. Which means the real leverage for COP isn’t the threat of cutting. It’s the efficiency: they’ll keep drilling, but they’ll keep squeezing unit costs every quarter, and they’ve got the data to prove the deflation is real. COP doesn’t negotiate with a gun. They negotiate with a spreadsheet. That’s harder to defend against.

Kalibr BATNA Translation

COP scores an 8.7 ARI, Fortress classification. For vendors, this is the most durable demand signal in the Lower 48. That sounds like good news for the service company, and it is, right up until the procurement meeting. COP’s activity resilience means they’re not going anywhere. It also means they know they’re not going anywhere, and they know you know it too. The leverage here isn’t about volume risk. It’s about price. COP captured $200 million in service cost deflation this quarter and guided to another $400 million in opex reductions for 2026. Someone is funding that.

If COP is your vendor’s largest customer, their order book is secure but their margins are under pressure. If COP is your peer and you share vendors, understand that those vendors are benchmarking your efficiency against COP’s 15% D&C improvement. The standard just moved.

Act accordingly.

Known Unknowns is Kalibr’s earnings intelligence series. Each post takes a single earnings call, layers it against our Kalibr Pulse activity data, and scores the company through our Activity Resilience Index (ARI): a 7-dimension, 1-to-10 framework that quantifies how likely a company is to sustain its spending on oilfield goods and services over the next 24 months. Most earnings analysis is written for equity investors. This is written for the people negotiating contracts. If you’re an E&P, the ARI tells you how your vendor’s other customers are trending and what that means for your leverage. If you’re a vendor, it tells you which revenue in your book is durable and which is one bad quarter from disappearing. We track the delta because direction matters as much as level.

Season 4, Episode 6. Larry David needs to get to a Dodgers game. Traffic is terrible. The HOV lane is wide open, but you need two people in the car to use it. The rational solutions are obvious: leave earlier, take a different route, accept the drive. Larry does none of these things. He picks up a stranger to sit in his passenger seat so he technically qualifies for the carpool lane.

This is a very Larry David solution to a problem. He has not reduced his travel time through any structural change. He has not altered his route, his departure time, or his driving habits. He has paid someone to sit next to him so that he can access a lane reserved for people who have actually solved the transportation problem. The appearance of efficiency, purchased at a premium, layered on top of a system that remains completely unchanged.

And, predictably, the shortcut creates more problems than it solves. The passenger becomes a subplot. The subplot becomes a catastrophe. Larry arrives at the game having spent more time, more money, and more emotional energy than if he had just sat in traffic like a normal person.

This is enterprise AI spending in oil and gas.

Most companies right now are Larry David in the carpool lane. They have an efficiency problem. Too many people doing manual, repetitive work across procurement, finance, operations, engineering, regulatory. The real solution is to build internal AI capability that automates those workflows and right-sizes the team. But that is uncomfortable. It means making hard org design decisions, restructuring staff, and doing the unglamorous work of rearchitecting how the business actually runs.

So instead, they pay a vendor to sit in the passenger seat. They license an AI platform, bolt it on top of existing workflows, keep the same headcount, and call it digital transformation. The vendor shows up in the investor deck. The conference panel gets a talking point. The HOV lane sticker goes on the bumper.

Meanwhile, the car has the same number of people in it, costs more to operate, and is not moving any faster.

The reason this matters is specific to where oil and gas sits right now, and to understand that, you have to understand the decade that just happened.

From Body Shop to Factory Floor

I talk a lot about the commoditization of the technical fields in Oil and Gas, which, like the commoditization of anything, is generally bad for the thing being commoditized (in this case, my fellow engineer).

This long journey started with the rise of unconventional development around the 2010s but amplified in the period of 2015 to 2019 and was precipitated by the commodity price collapse of 2014-2016. Kimmeridge wrote a lot of great stuff on this (hell, I wrote my MBA application on their model). E&P operators pivoted from “growth at all costs” to strict capital discipline and efficiency. This fundamentally altered G&A cost structures and workforce compositions, moving the industry toward a manufacturing model of standardized processes and reduced reliance on bespoke technical oversight.

The simplest way to understand what happened is this: oil and gas used to be a custom body shop. Every well was a bespoke project. Different geology, different casing programs, different completion designs, different problems that required different expertise to solve. You needed people who could think on their feet because the work genuinely required it. Casing seat hunts, gravel pack designs, true exploratory drilling where you were not sure what you were going to find at total depth.

Then unconventional development turned it into a factory. The wells started looking the same. The completion designs converged. The drilling programs became statistical exercises in repeatability. And when the work looks the same, the inputs become narrower, simpler, and easier to switch out. You do not need an artisan when you are running a production line. You need an operator. And operators, by definition, are more interchangeable than artisans.

The G&A reduction that followed plays out in four stages, and each one made the next one inevitable.

Stage 1: Standardization (2015-2019). The entire industry converged on the same factory model. Cube development. Standardized well designs. Pad drilling. The language varied by operator but the economics were identical: make the wells the same, make the people interchangeable, make the G&A per barrel go down every quarter. Pioneer’s G&A journey from 2015 to 2017 tells the story in three numbers: $4.39 per BOE, $3.80, $3.28. Management’s stated strategy? “Not replacing personnel who have left the Company.” Production volumes climbed 15-16%. Headcount didn’t. The math was elegant and merciless. Continental hit $1.57 per BOE by 2019. Extraction collapsed from $4.85 to $3.05 in a single year.

From a negotiation and BATNA perspective, this was bad for technical engineers and generally good for companies. If I need an engineer to drill an extended reach well off the coast of Sakhalin Island, there may be a handful of people I can call (and I got a buddy if you need it). But if you need someone to run a rig in North Dakota, there are generally more people to switch out for that.

Stage 2: Capital Discipline (2020-2023). This really took off post-COVID, as companies fought relentlessly against the flight of stockholders to more ESG-friendly stocks and the general appeal of technology companies. Remember, when you buy a stock, you are buying the discounted cash flows into the future, and if folks are uncertain of that industry’s future, you don’t want to own the stock. The industry combated this through dividends and buybacks: “Don’t worry about the cash 20 years from now, we are going to give it to you now.” Cash for dividends had to come from somewhere. Executive compensation structures got scrutinized. “Right-size G&A” became a boardroom mantra. More jobs out.

Stage 3: Consolidation (2023-present). This chapter starts when my old mothership, Exxon, bought Pioneer. That was the shot that started the race. And what followed was the most concentrated wave of upstream M&A in modern history. Chevron-Hess. Diamondback-Endeavor. ConocoPhillips-Marathon. Oxy-CrownRock. The synergy math on every single one of these deals started with G&A, because G&A is always the first and easiest thing to cut. It is certain and the ramifications are felt later. Clear books on salary. Cut this, save this. The number shows up in your next quarter. The industry has eliminated $2.0 to $2.5 billion in pure annual corporate overhead through this cycle, within a broader total synergy pool exceeding $8 to $10 billion annually. Exxon doubled its Pioneer synergy target from $2 billion to $4 billion. ConocoPhillips doubled Marathon from $500 million to over $1 billion. Chevron hit its Hess target six months early and raised it. Every single acquirer guided conservatively and then beat and raised. This is not coincidence. This is a playbook.

Stage 4: The Build. All of this adds to a more than 100% increase in wells per engineer between 2015 and 2025. The low-hanging fruit for G&A reduction has been harvested. Standardization, consolidation, headcount attrition. The next horizon in G&A reduction is a bit more complicated. It is building new capability on top of what is now one of the leanest operating foundations of any capital-intensive industry in the country.

And this is where AI actually matters. But for it to matter, it has to respect the direction of the trend. It has to continue driving G&A down. Not sideways. Not up. Down.

That turns out to be a very specific constraint, and it is the one that almost every AI vendor currently fails.

The Passenger Seat Has a License Fee

Let me come back to Larry.

The reason the carpool lane scheme fails is that Larry has added a cost to the system without removing one. He still has the car. He still has the commute. He still has himself. Now he also has a passenger he is paying for, a passenger who creates complications, and a net increase in the total cost of getting to Dodger Stadium. He is in the fast lane. He is spending more to be there.

AI software sold to an E&P company works the same way. It sits on the G&A line. The license fee is G&A. The implementation cost is G&A. The internal resources dedicated to managing the vendor relationship are G&A. And the person whose work the software was supposed to automate? Still there.

Here is why that person is still there. The work that AI currently automates well in oil and gas tends to be partial. It handles a slice of a role, not the whole thing.

Take gas and production reconciliation (we built something around this last year). An agent can pull statements from the buyer, cross-reference production against the E&P’s internal data, map contractual agreements with midstream counterparties, and generate a reconciliation report with automated state submission. That is real, functional automation. It works.

In most cases, that task is handled by a production tech or production accountant making roughly $150K all-in. But reconciliation is maybe 25% of their job. So you are automating $37,500 of addressable labor. The other 75% of that person’s role is untouched.

You are not going to reduce staff and show an ROI on the investment over $37,500 of automated work. The person does other things. You are going to want a human reviewing the output before it goes to the state, because the downside of a bad filing is regulatory, not cosmetic. So the person stays. The software gets added. G&A goes up.

This is the structural problem with selling AI. The roles that are easiest to partially automate are the roles where the person does multiple things, the salary is moderate, the addressable slice is small, and the output still requires oversight before it is done. At every step, the math pushes you toward G&A addition, not G&A reduction. The vendor’s business model requires a license fee that exceeds the value of the automated slice. The operator’s risk tolerance requires a human check on the output. The result is a passenger in the seat, a sticker on the bumper, and a larger G&A number than you started with.

The companies writing seven-figure checks to AI vendors without reducing staff are not transforming anything. They are paying someone to sit next to their existing problem so they can say they are in the fast lane. And yes, there are AI applications improving cycle time, production optimization, and reservoir characterization that live outside the G&A line entirely. I will touch on those next week. But for the current wave of enterprise AI sold against the overhead budget, the arithmetic is what it is.

And just like Larry, they are going to end up spending more, causing more chaos, and arriving no faster than if they had done the hard thing from the start.

Fewer People in the Car, No Vendor in the Passenger Seat

The company that actually captures AI value is the one that skips the carpool lane gimmick entirely. No vendor riding shotgun. No inflated software line item. Just fewer people in the car and lower cost per mile.

The way you get there is by shrinking the distance between the subject matter expert and the coder to a single node. Not replacing the SME. Not bolting a chatbot onto their workflow. Collapsing the two roles into one person who builds and runs the automation themselves, with no vendor in the middle. This is relatively straightforward for most O&G technical personnel who graduated in the 2010s with some background in Python and data.

The vendors are selling separation: “Your engineer does engineering, our product does the automation, you pay us a license fee.” That is three nodes. That is more G&A, not less. I am arguing for convergence: “Your engineer does engineering and the automation, you pay nobody a license fee, and you reduce headcount with a clear ROI.” That is one node. G&A goes down. The trend continues.

The separation model is a product. The convergence model is a capability. Products are G&A. Capabilities are not.

Let’s set a target. Average wells per engineer is somewhere around 16 to 18 today. Diversified Energy is the industry proof of concept for what technology-driven asset management looks like when you build it internally. DEC runs approximately 75,000 wells with 1,589 total employees, yielding a wells-per-engineer ratio somewhere between 530:1 and 750:1. DEC is a fundamentally different animal. No-drill model, 10% decline rates, mature conventional assets, 99.9% held-by-production acreage. They built a proprietary platform called SAM that decouples well count from headcount, pushes decision-making to field operators via handheld diagnostics, and monitors 75,000 wells through Integrated Operations Centers using management-by-exception. Their lease operators manage roughly 150 wells each versus the industry standard of 50.

DEC is not the comp for an active driller. But it is the proof of concept that the gap between 16 wells per engineer and 500+ wells per engineer is not a technology problem. It is an architecture problem. DEC did not buy SAM from a startup. They built it because they understood their own operations better than any third party could.

My hypothesis is that with the right AI stack and a subject matter expert who knows what to build, you can replicate something approaching DEC’s ratio in an active development context. Not by buying software. By building agentic workflows that couple technician, regulatory, and entry-level engineering tasks with AI agents: automated state reporting (Quad O-A, SPCC, stormwater), production surveillance and exception-based alerting, lease revenue reconciliation, decline curve screening, and regulatory compliance tracking. The goal is: how many assets can a single engineer and limited staff manage effectively, with no vendor overhead and continued downward pressure on G&A per BOE? This, of course, is an optimization problem. Building these things in AI is not void of cost, so one must factor that in. But an example of a base REPO that we have worked on looks something like this:

Small Companies, Big Leverage

Engineers at small companies and startups will have a leg up for three reasons.

First, breadth of exposure. If you have been part of a startup, you are keenly aware of the early days filling state reports yourself, handling production accounting, managing regulatory submissions. You know the full surface area of operations because you had to do all of it. At big companies, that work gets siloed. The engineer who has never filled Air Emissions report does not know it can be automated.

Second, it is much easier to start lean and never hire than to restructure an existing workforce. ConocoPhillips is cutting up to 3,250 people to capture its Marathon synergies. Chevron is reducing headcount by 15-20% globally. That is brutal, expensive, and operationally disruptive. Starting with a small team and agents on day one is not.

Third, large companies are largely constrained to CoPilot AI through Microsoft enterprise agreements, and these are meaningfully behind the alternatives in the market.

Table Stakes and the Three Buckets

Everything I have outlined above gets you a first mover advantage. It is real and it matters. But it is also replicable. Like most D&C techniques, once someone proves it works, others copy it. The 530-to-750 wells-per-engineer ratio becomes table stakes the way the extended lateral became table stakes.

The true competitive moat comes from something different: having a differentiated subject matter expert view that no one else has, and building a tool that repeats that view at scale. Not generic automation of generic tasks. Proprietary insight embedded in proprietary workflow. The person who understands both the domain and the tooling does not need a vendor. They are the vendor.

I think that moat has to fall into three specific buckets. Next week I will go into detail on what those are, the incentive case for building them (both for fundraising and running an asset), and an actual product we have built and deployed.

The next version of the oil and gas company might be a $500MM PDP machine run by a small group of people. They will not be interchangeable. They will be interdisciplinary. Fluent in reservoir economics, capital planning, procurement, and AI workflows. When talent is commodity, the edge becomes synthesis.

This industry did not spend a decade getting lean because it lacked ambition. It spent a decade getting lean because it is pathologically good at adapting to constraints. The companies that survived are the most operationally disciplined firms in any capital-intensive sector in this country. That is not a customer base that needs a vendor in the passenger seat. That is a customer base that needs fewer people in the car, moving faster, at lower cost.

Larry paid for a shortcut and ended up with a longer, more expensive trip and a story he would rather not tell. The companies that skip the gimmick, build the capability internally, and actually reduce the headcount and the overhead?

They’ll be….Pretty, pretty, pretty good.

Someone on LinkedIn disagreed.

The pushback went something like this: compression providers have economies of scale, specialized labor, and OEM purchasing power that E&Ps cannot replicate. Insourcing is a distraction. Stick to your knitting.

This is a reasonable-sounding objection. It is also the exact objection your compression vendor hopes you find persuasive. They have, after all, spent considerable effort making sure it sounds reasonable. Their investor decks are filled with reassuring charts about “customer stickiness” and “barriers to entry.” I would be worried about barriers to entry too, if my business model depended on customers not doing arithmetic.

Since I lean contrarian on almost everything, let’s do the arithmetic.

The Part Where We Establish That Something Funny Is Going On

Here is what we know. USAC reported average rental rates of $21.46 per horsepower per month in Q3 2025, up 4% year-over-year. Natural Gas Services Group, which focuses on high-spec fleets, reported $27.08. Jefferies uses $25.00 as their base modeling assumption. Kalibr typically pegs large unit rentals at $40,000 to $45,000 a month for 1,500-2,000 HP packages.

Meanwhile, Kodiak Gas Services achieved an adjusted gross margin of 68.3% in its Contract Services segment. USAC reported 69.3%.

I want to dwell on this for a moment. Sixty-eight percent gross margins. In a business that involves owning steel and sending technicians to remote locations. For context, Apple’s gross margin is 46%. Hermes, the luxury goods company whose customers pay $10,000 for a handbag because it has a horse on it, runs 72% gross margins. Your compression vendor is approaching Birkin bag economics.

Large horsepower utilization is running at 97% to 99%. Lead times for new Caterpillar 3600 series engines have stretched to nearly 60 weeks. Kodiak management has explicitly referenced a 15% to 20% premium for new unit deployments, reflecting what they call the cost of “excess optionality.” That is a nice way of saying “you need this, you need it now, and we both know you have no alternative.”

Put differently: you are paying record rates, your vendor is earning record margins, and there is no credible supply response coming anytime soon.

The question is not whether compression providers have built a good business. They obviously have. The question is whether that good business is built, in part, on the assumption that you will never pull out a calculator.

The Part Where We Pull Out a Calculator

Let’s build a baseline. A standard 1,000 HP compression package at current newbuild costs runs approximately $1,200 per horsepower, or $1.2 million per unit. Electric motor drive packages are cheaper still: $2.0 to $2.2 million for large units versus $2.5 million for gas-fired equivalents, a $300,000 to $500,000 savings because electric motors are simpler than Caterpillar engines.

Operating costs for owned compression vary by drive type. A safe, Kalibr assumption pegs direct operating cost for large gas-fired units at approximately $30,000 per month. Electric units run $15,000 to $20,000 per month because you eliminate engine oil changes and reduce mechanical complexity. Here is the fun part: in the first few months of operation, electric unit OpEx can run as low as $3,000/month. The $30,000 figure is a lifecycle average that includes major overhauls in Years 5-7. Year 1 is considerably more profitable than Year 7.

Now let’s talk about what you are actually paying to lease. Kalibr’s market standard is around $25.00 per HP per month. And here is the detail that most models miss: KGS management explicitly noted that CPI pass-throughs have historically lagged market rate increases. Real escalation has been running 4-5% annually, not the 3% your contract nominally references. Your Year 7 rate is not what your spreadsheet says it is. It is worse.

The comparison:

Lease Scenario (10 years, 1,000 HP)

• Monthly rental rate: $25.00/HP/month

• Annual escalator: 4.0%

• Total undiscounted cash outflow: $3.6 million

Buy Scenario (10 years, 1,000 HP)

• Upfront CapEx: $1.2 million

• Operating cost: $6.50/HP/month (electric/blended fleet assumption)

• Residual value at Year 10: 40% of initial CapEx ($480,000)

• Total undiscounted net cash outflow: $0.78 million

The Net Present Value of “Buy” versus “Lease” at a 10% discount rate:

• 7-year horizon: $285,000 — Buy wins comfortably

• 10-year horizon: $580,000 — Buy wins decisively

• 12-year horizon: $820,000 — At this point you’re just lighting money on fire by leasing

The Internal Rate of Return on the insourcing investment:

• 7 years: 16.2%

• 10 years: 19.4%

• 12 years: 21.1%

Simple payback: 4.8 years. For operators running 10,000 HP or more, that is $2.8 million in NPV capture at 10 years. Scale that to a 50,000 HP program and you are looking at $14 million of value creation. For a 100,000 HP program, $29 million.

This aligns with Jefferies’ analysis of Kinetik Holdings, which estimated a roughly 6x build multiple on owned compression, saving approximately $30 million per year on $180 million of CapEx. That is a 16.7% return on compression capital. In perpetuity. Achieved by a company with concentrated Delaware Basin acreage and a long development runway.

Now, I can already hear the objections forming. “But Ian, surely it can’t be that simple. Surely the compression providers have thought of this.” They have. They think about it constantly. The word “insourcing” has appeared over 200 times in compression industry earnings transcripts since 2023. It is the monster under their bed. They mention it because their investors keep asking about it, and their investors keep asking about it because the math is becoming obvious to everyone except, apparently, the people writing the checks.

What the People Who Actually Know Things Say

The provider economics are interesting. But what do the people who run these businesses think about insourcing?

Start with the industry rule of thumb. Per a VP in the Compression Industry'; “It’s typically looked at on a seven-year model. If it’s over seven years, the owner would want to buy the compression because that’s going to be the most economical for them.”

Seven years. Not ten. Not twelve. Seven. That is the compression industry’s own internal benchmark for when ownership wins. It is a fascinating number because it is not in any investor deck. It is just something everyone inside the industry knows.

On operating costs, the numbers are more favorable to ownership than headline figures suggest. The $30,000/month figure is a lifecycle average incorporating major overhauls. For new equipment, costs are a lot lower. Year 1 operating costs for a well-maintained unit run closer to $18,000-22,000/month before the first major PM cycle.

Electric motor drive units change the math entirely. The research pegged electric operating costs at $15,000 to $20,000 per month over the lifecycle, versus $30,000 for gas-fired. If you have grid access and can deploy electric motor drive compression, your NPV advantage versus leasing is not $580,000 per 1,000 HP. It is closer to $820,000. The spread widens because you are saving on both CapEx and OpEx simultaneously.

On switching costs, the data cuts both ways. Mobilization runs $40,000 to $100,000 depending on the unit. All-in swap costs approach $120,000 when you include crane, crew, and downtime. This stickiness is the core of the provider business model: once you are in, you are in. But the same stickiness protects owned assets from competitive displacement. Every month you operate past the mobilization sunk cost is pure margin capture. The switching cost that traps you in a lease is the same switching cost that protects your owned equipment from being swapped out.

Where the Critics Are Right, Where They Are Wrong, and Where They Are Arguing in Bad Faith

The LinkedIn skeptic raised three objections. Two are reasonable. One is something your vendor’s IR team crafted specifically to sound reasonable.

Objection 1: “Providers have superior purchasing power and supply chain access.”

This is true. Kodiak noted that lead times for new engines have stretched to 60+ weeks, creating a barrier to entry for individual buyers. But this objection assumes E&Ps cannot plan ahead. Kinetik successfully secured deposits for incremental compression delivery in 2026-2027 by integrating procurement into their long-term infrastructure planning. If your development model is a “manufacturing model” with multi-year visibility, you can plan ahead. You are not buying a compressor for next Tuesday. You are buying it for 2028.

More importantly, the supply chain constraint cuts both ways. If you cannot get equipment quickly, neither can your vendor. The scarcity premium embedded in rental rates is the same scarcity premium that would protect the value of your owned assets. Scarcity is not a reason to lease. Scarcity is a reason to own.

Objection 2: “Owned assets distract from core drilling returns (ROIC dilution).”

This is the standard corporate finance objection, and it holds if the E&P is capital constrained. If you have $100 million and can only either drill wells or buy compressors, drill the wells. Wells are your business.

But for investment-grade operators, the lease-versus-buy spread has widened to the point where the implied payback on ownership is approaching 5 years. That is a viable infrastructure return profile for operators with long-life inventory. Devon Energy allocates $100 to $115 million annually solely for compression and gathering infrastructure in the Delaware Basin. At some point, you stop being capital constrained and start being margin constrained. At 68% vendor margins, you are margin constrained.

And here is the thing that gets surprisingly little airtime: buy versus lease is not binary. There is a third path.

Some operators work with equipment financing groups or bank leasing structures that essentially outlay capital and rent back at cost. The financing entity takes the depreciation, the operator gets a fixed monthly rate without the 68% margin baked in, and the residual risk is shared or transferred. Think of it as synthetic ownership without the balance sheet impact. It is the DEC playbook applied to oilfield equipment: turn a capital purchase into a structured finance product, fund at investment-grade rates, and capture the spread between your cost of capital and the provider’s required return.

Another strategic pathway in the gray is co-investment. Kodiak Gas Services disclosed a recent agreement with an investment-grade E&P to co-invest in a new compressor station under a 50/50 ownership structure. The E&P preserves capital while Kodiak operates the entire facility. The E&P gets ownership economics without full operational burden. Kodiak gets contracted utilization and a committed customer. Everyone wins except the pure-play outsourcing model.

The point is that capital structure creativity exists. If your CFO is framing this as “own it or rent it,” the analysis is incomplete. The correct framing is “own it, rent it, finance it, co-invest it, or some combination thereof.” The worst option is not knowing which one is right.

Objection 3: “Technical complexity requires specialized 24/7 support.”

Providers cite the shortage of skilled technicians as a persistent operational risk they are better equipped to manage through centralized training academies.

This is the validity check. And it is fair. Kinetik asserts they have achieved a scale where they possess the mechanics, capability, and spares to maintain high uptime internally. But that narrative only works for E&Ps with concentrated geographic footprints where technician travel time is minimized.

Here is what the critics miss: the counterargument to insourcing assumes you are replicating the provider’s entire business model. That is the wrong frame. The question is not whether you can become Kodiak. The question is whether you can own the baseload horsepower required for your core production wedge while leasing the variable peak horsepower required for flush production and new pad bring-online.

Own the minimum. Lease the surge. Capture the margin on what you know will run for a decade, and preserve flexibility on what might not. This is not “insourcing versus outsourcing.” It is more portfolio optimization.

The Utilization Sensitivity Is More Forgiving Than You Think

The base case assumes 95% utilization, consistent with current tight market conditions. At that level, the Buy scenario saves approximately $220/HP/year versus leasing.

But here is the counterintuitive finding: the break-even utilization is far lower than most operators assume.

• At 90% utilization: Buy still wins with positive NPV at 7 years

• At 85% utilization: Buy breaks even at 8 years, wins thereafter

• At 80% utilization: Buy breaks even at 10 years

The reason the threshold is so forgiving is the margin embedded in rental rates. When your vendor is earning 68% gross margins, you have a lot of room to operate less efficiently and still come out ahead. You do not need to run compression like Kodiak runs compression. You need to run it 80% as well, and you win.

This is where the manufacturing model becomes operationally determinative. The shift to multi-well pads allows a single, owned central compression station to service 10-20 wells. Even if individual well variance exists, the aggregate pad volume remains stable enough to justify ownership.

EQT’s “Combo-Development” strategy exemplifies this: contiguous acreage drilled in large-scale pads generating 600+ MMcfe/d of volume impact. That operational density allows shared infrastructure planning and long line-of-sight on logistics. When you know what your gas volumes will look like in 2032, you can own the compression that will move them.

The risk framework is straightforward: concentration de-risks utilization. If your development model has shifted from wildcat delineation to industrial-scale harvesting, you have already built the production wedge that makes ownership rational.

The Hybrid Playbook for Operators Who Actually Think About This

The most sophisticated operators are not choosing between lease and buy. They are choosing both.

Kodiak’s 50/50 co-investment structure is one example. But the hybrid playbook extends further.

Base-Load/Peak-Shave Model: Own the minimum continuous horsepower required for your core production wedge. Lease the variable “peak” horsepower required for flush production and new pad bring-online phases. This aligns capital commitment with production certainty and insulates your base volumes from rental rate inflation. It is also the structure that most resembles how you already think about hedging commodity exposure.

Bank Leasing Structures: Work with equipment financing groups that outlay capital and rent back at cost-plus. The financing entity takes depreciation, you get a fixed monthly rate without the 68% provider margin, and residual risk is transferred. This is synthetic ownership for operators who cannot or will not carry compression on the balance sheet. It is also the structure most likely to make your compression vendor nervous, because it unbundles the financing from the operating margin.

Managed Services with Owned Assets: Buy the equipment but outsource maintenance to a provider under a fixed-fee contract. Comprehensive maintenance contracts are typically priced at 4% to 8% of the unit’s ex-works price annually. That is $48,000 to $96,000 per year for a $1.2 million unit, compared to $90,000+ in embedded margin you pay in a full-service lease. You capture the ownership economics while outsourcing the operational complexity.

The purest expression of the insight is not “insource everything” or “outsource everything.” It is to own the infrastructure whose utilization you control and can predict, and to lease the infrastructure whose utilization depends on factors outside your control.

For a Permian operator with concentrated acreage, rising GORs, and a decade of development runway, baseload compression looks like infrastructure. Own it.

For a Haynesville operator with steep decline curves and commodity-sensitive drilling activity, variable compression looks like a service. Lease it.

For most operators, the answer is some of each. The wrong answer is not knowing where the line is.

The Kalibr Takeaway

The math has shifted. At current rental rates, escalation trends, and provider margins, insourcing compression is no longer a niche strategy for integrated midstream players. It is a legitimate capital allocation decision for any operator with concentrated acreage, predictable production profiles, and a 7+ year development horizon.

The numbers are stark: 19% IRR at 10 years, $580,000 NPV per 1,000 HP, and a break-even utilization threshold below 85%. For a 50,000 HP program, that is $14 million in value creation. For a 100,000 HP program, $29 million. At some point, this stops being a negotiating tactic and starts being fiduciary negligence to ignore.

But the more interesting point is strategic. Running a serious insourcing analysis is not the same as committing to insource. It is about knowing where the line is. It is about understanding which horsepower classes matter, where reliability actually drives value, and how switching costs compound across your system.

At a minimum, the analysis creates leverage. Your vendor should not assume you are captive. They should believe you have a Plan B. And if the Plan B pencils out at 19% IRR, maybe it should be Plan A.

Because here is what the compression providers know that you might not: at 68% margins and 98% utilization, every incremental point of market share lost to insourcing costs them real money. They mention “insourcing” on every earnings call. It is keeping them awake at night.

Maybe it should keep you awake too. Not because insourcing is the answer, but because not knowing the math means you are negotiating with one hand tied behind your back.

And in a market where the conveyor belt costs more every quarter, ignorance is not neutral.

It is a pricing signal your vendor reads every time you sign a renewal.

Say you start a business manufacturing widgets. You survey the landscape to evaluate which widgets are in the market: red, blue, green, yellow. Rather than make all of them, you decide to focus on one. Red widgets. This is good. The market tends to reward specificity, so they like you focusing on a color. Index funds want “pure-play red widget exposure.” Analysts write initiations with “Red Widget” in the title. Your stock gets a ticker symbol. Things are going well.

You focus extremely hard on driving down the unit cost of that red widget compared to other red widget makers. You make the widget better. Cheaper. Faster. Wall Street loves this. Your multiple expands. Everything is great.

Except.

There’s a large caveat to your manufacturing process versus most others. There is not an infinite amount of raw material to make your red widget. The stuff you need? It’s finite. And you’ve been using the best stuff first, because of course you have. Year one, you make widgets from premium-grade raw material and each widget comes out perfect. Year five, you’re still technically making red widgets, but you’re using slightly worse raw material, and the widgets are measurably, quantifiably 6% less good than the widgets you made in year one. By year ten, you’re down 15% from peak widget quality.

So a natural first step might be to go to other red widget makers and combine. Now you have more red widget raw material to show investors. “Look, we can make widgets for longer now!” you announce on the earnings call. Additionally, you have more scale to drive down your cost to make the red widget. Merge the supply chains, fire duplicate widget inspectors, buy your widget-making machines in bulk.

This doesn’t solve your red widget scarcity problem. But it does buy time. You’ve extended your runway.

Eventually, you’re at a fork in the road. Do you continue the strategic focus you’ve been rewarded for, making red widgets, or do you decide to get into the blue widget business? After all, people need widgets. And isn’t it better to have red and blue widgets than no widgets at all?

Wall Street hates this. “You’re a red widget company,” the activists yell. “Divest blue! Return to core competency! Simplify the story!” But you know something they don’t. The red widget raw material is running out faster than anyone wants to admit. And when it runs out completely, you’d rather be the company making multiple colors than the company making nothing.

This is obviously oil and gas I’m talking about. Devon Energy and Coterra Energy. Barrels instead of widgets. Tier 1 Delaware Basin rock instead of premium raw material.

And look, the past few years have been exactly that first-step playbook. Red widget makers calling other red widget makers. Exxon-Pioneer. Chevron-Hess. Diamondback-Endeavor. Consolidating contiguous positions, eliminating duplicate costs, drilling longer laterals across combined acreage. Classic operational harvest of the best remaining rock.

But something shifted. SM Energy and Civitas Resources was the inflection point. That deal wasn’t about contiguous acreage or operational synergies. SM was Permian and Eagle Ford. Civitas was Midland and DJ Basin. Geographic complexity that activists hate. Most analyst called it achieving scale at a steep cost with “imited operational synergies. But SM did it anyway, because standing alone with inventory questions was worse than adding basins to the map.

Now we’ve got Devon-Coterra merger talks. Different companies, slightly different basins, but the same underlying thesis. It doesn’t sing exactly like SM-Civitas. But it rhymes. And the rhyme scheme is: we’re running out of the good stuff faster than efficiency gains can offset, so let’s combine piles of progressively-less-good stuff and call it duration.

The market is starting to notice the raw material is depleting. And the strategic playbook is changing in response.

Kicking The Can Down A Longer Road

Jefferies published research in January 2026 quantifying what everyone in the Delaware Basin already understood intuitively. Basin-wide productivity declined 6% year-over-year. Operator-specific? Both Devon and Coterra clocked roughly 15% productivity declines in their Delaware assets between 2024 and 2025.

If you’re in the industry, this isn’t news. You’ve seen it in your own well results. You’ve watched EUR guides get quietly revised downward. You’ve sat in engineering meetings where the pressure depletion maps look progressively worse and the parent-child interference is no longer a variable you can model around. It’s just the reality of developing a mature basin.

But here’s what matters. The gap between what the Street thinks is happening and what’s actually happening in the subsurface keeps widening. Sell-side models still assume something close to flat productivity with modest efficiency gains offsetting geological degradation. That worked in 2019. It worked in 2021. It stopped working somewhere in 2023, and by 2025 the math just doesn’t close anymore.

Same rigs. Same or better completion designs. Longer laterals. Bigger fracs. Simul-frac, trimul-frac, every optimization technique that completion engineers spent the last decade perfecting.

The wells are just worse.

This isn’t an execution problem. It’s the rocks. The Tier 1 sweet spots got drilled first because they were the highest-return intervals. High permeability, thick pay, minimal water, no parent-child interference because you were drilling parent wells. What’s left is everything else. Tighter rock, thinner zones, more water, wells that interfere with each other, locations that clear your hurdle rate on a spreadsheet but deliver 15% less oil in practice.

Coterra CEO Tom Jorden acknowledged this in August 2025 when he talked about the industry entering the “final chapter of Tier 1 inventory.” He positioned Coterra as hitting that wall later than peers. But the acknowledgment itself matters. CEOs don’t typically announce their primary product is depleting unless the internal forecasts leave them no choice.

The numbers tell the story. Devon’s reserve life sits at 8.1 years. Peer average is 11.5 years. Coterra claims 15 years in investor decks, but most analysts calculate it at 9.2 years based on actual 2024 production levels.

Both below average. Both acquiring externally to fill gaps. Devon spent $5 billion on Grayson Mill in the Williston, explicitly to extend oil inventory. Coterra spent $3.9 billion on Franklin Mountain and Avant in the northern Delaware, explicitly to boost oil weighting.

When you’re buying externally what you used to develop internally, you’ve crossed a threshold. You’re not growing organically anymore. You’re managing depletion. And the playbook changes.

Balance Sheet Engineering for the Geologically Constrained

So what do you do when organic replenishment can’t keep pace with depletion?

You call your neighbor who has the exact same problem and propose combining countdown clocks.

Devon has 8.1 years of reserve life. Coterra has 9.2 years (Although management claims 15-years). Neither can extend that runway organically at the pace Wall Street expects. Wall Street wants flat-to-growing production and 10% free cash flow yields. The math doesn’t work if you’re drilling progressively worse wells each year.

So they merge. Not to drill monster laterals across newly contiguous acreage. Not to eliminate duplicate finance teams or consolidate gathering contracts. Those savings exist but they’re rounding errors. The actual strategic value is portfolio duration extension through asset aggregation.

This is balance sheet engineering masquerading as industrial logic. You have an 8-year bond. Your neighbor has a 9-year bond. Stapling them together doesn’t create a 17-year bond. But it does create a blended maturity profile that looks more durable than either standalone, especially to investors who don’t stress-test your reserve replacement assumptions every quarter.

This deal would be less about operational synergies and more about securing remaining Tier 1 resource life.

To be clear, I am not knocking the deal. That’s reality. When the constraint shifts from capital to geology, the strategic response shifts from optimization to aggregation. You’re not buying synergies. You’re buying time. And in a depleting resource business, time measured in years of drilling inventory might be the only asset worth acquiring.

The Strategic Value of Complicated (Ask SM Energy)

Devon-Coterra isn’t unique. SM Energy and Civitas Resources announced their merger in November 2025 with the same fingerprints all over it.

Many analysts called it “achieving scale at a steep cost” and noted “limited operational synergies” because there’s virtually no contiguous acreage. SM’s positions were Permian and Eagle Ford. Civitas was southern Midland and DJ Basin. The deal added geographic complexity, the thing activist investors spend entire presentations screaming about.

SM did it anyway. Because standing alone with compressed inventory depth and fielding quarterly questions about how many Tier 1 locations you really have left is worse than adding the DJ Basin to your map and buying 18 months of narrative breathing room.

Both deals have identical architecture. Geographic complexity over simplification. Inventory aggregation over operational synergy. Duration extension over growth creation. Scale as defense over efficiency as edge.

It is clear current M&A prioritizes absolute scale over substantial operational synergies.

That’s the shift. The 2022-2023 consolidation wave was about operational harvest. Exxon-Pioneer, Chevron-Hess, Diamondback-Endeavor. Those were deals where companies bought their neighbors, eliminated costs, drilled longer laterals across combined acreage, and printed cash from premium rock.

This wave is different. This is inventory accumulation when the inventory itself is degrading. You’re not optimizing red widget manufacturing anymore. You’re buying other red widget companies because everyone’s running out of premium red widget material at roughly the same rate, and having two piles of mediocre material sounds better than one pile.

Rational adaptation to a binding geological constraint.

The Venezuela Cost Curve (Under New Management)

Now let’s talk about the external threat that every E&P CFO is modeling privately but nobody discusses on earnings calls.

January 3, 2026. U.S. forces capture Nicolás Maduro. Trump’s response comes in two parts. First, impose a total blockade on Venezuelan oil tankers. Short-term supply tightness, WTI spikes briefly, everyone feels good. Second, signal intent to restore Venezuelan production under U.S. oversight. Long-term supply threat, prices compress, Permian margins evaporate.

The economics matter because they’re structural, not cyclical.

Venezuelan lifting costs run $20 to $30 per barrel once infrastructure gets operational. Permian Tier 1 costs, back when Tier 1 actually existed in meaningful quantities, were $45 to $55 per barrel. Permian Tier 2 costs today run $65 to $75 per barrel. Marginal Permian locations push $75 per barrel and higher.

Trump’s stated price target to combat inflation? $50 per barrel.

Restoring Venezuelan production to meaningful levels isn’t immediate. It requires an estimated $100 billion in infrastructure investment. The country’s oil sector has been decaying for two decades. You don’t flip a switch and produce 2 million barrels per day. Early projections are modest. Maybe 0.5 million barrels per day growth over three to five years. Not a flood. But not nothing either.

Here’s the thing though. Firms don’t plan for base cases. They plan for tail risks. And the tail risk here is that Venezuelan barrels currently flowing to China via shadow tankers redirect to U.S. Gulf Coast refiners, creating a new anchor point on the global cost curve at $25 per barrel just as North American operators are nursing marginal costs toward $80 per barrel.

You can’t hedge that operationally.

You can’t complete-optimize your way out of a structural cost curve disadvantage.

You can’t drill your way to profitability if the global marginal barrel costs $30 and yours costs $75.

The only hedge available is balance sheet scale and durability. Build an entity large enough to survive multi-year commodity price compression if international supply dynamics shift unfavorably.

A combined Devon-Coterra at $53 billion enterprise value provides that buffer. Two separate $25 billion entities don’t.

Many estimate that if WTI hits $45 per barrel, cash flow revisions hit 21% across the E&P sector. Companies with $20 billion market caps and 8-year inventory runways get picked apart by activists or acquired at distressed multiples. Companies with $40 billion market caps and blended 10-year runways survive, retrench, and live to drill another cycle.

You can argue this is pessimistic. Maybe Venezuelan restoration takes longer. Maybe Trump’s $50 target is rhetoric. Maybe Chinese demand growth offsets any new supply. All plausible. But if you’re running an E&P company in 2026 and you’re not modeling the downside case where your marginal cost of production sits $40 per barrel above a credible new supply source, you’re not doing your job.

Scale isn’t just about procurement leverage or index inclusion. It’s about surviving scenarios where the rules change and the cost curve shifts beneath you.

The Service Provider Put

There’s another angle here that matters directly to the bottom line, and it’s the mirror image of what compression providers have been doing to E&Ps for the last three years.

A combined Devon-Coterra controls $6.4 billion in annual capital spend. That’s 9 rigs from Coterra and 14 to 16 from Devon in the Permian Delaware alone. Call it 23 to 25 active rigs. Add 6 to 8 dedicated frac crews across the basin. Plus consolidated Anadarko activity, currently fragmented across two separate 1-rig programs.

That is what we at Kaligr like to call monopsony leverage. The ability to influence market rates through concentrated buying.

Coterra management noted in August 2025 they were witnessing “reductions in market rates driven by increased rig and frac availability leading to competitive pricing in our bids.” Strip away the earnings call language and here’s what that means: as E&Ps consolidate into fewer, larger buyers, service providers face customers with real negotiating leverage.

Devon’s strategy with it’s cost reduction initiatives has been to decouple bundled services to utilize a diversified vendor universe. Operational translation: bypass service integrators. Maybe buy tubulars directly from mills at volume pricing. Or perhaps source sand directly from mines. Use that $6.4 billion buying power to extract concessions from providers who can’t afford to lose a customer that size.

For compression providers, pressure pumpers, drilling contractors, this is the nightmare that started in 2022 and keeps getting worse. Your largest customers are merging into entities with enough scale to dictate terms while your cost structures remain fixed. You’re selling a commodity service to a concentrated buyer base.

For Devon and Coterra, it’s margin defense when you can’t control the revenue line. If you can’t control WTI because some guy in Caracas might decide to start producing again, at least control your well costs. If revenue is threatened from above, squeeze costs from below.

You can’t expand the pie. Fight harder for every basis point of what’s left.

Market Capitalization as Competitive Moat

Devon’s current market cap sits around $22.4 billion. Coterra’s around $19.2 billion. Combined equity value hits roughly $42 billion. Enterprise value pushes $53 billion.

That $50 billion threshold matters. A lot. It’s the difference between “regional mid-cap E&P that requires conviction from active managers” and “investable large-cap with S&P 500 Energy index relevance that attracts passive flows automatically.”

Passive capital flows into energy ETFs and index funds get allocated based on market cap weighting. A $42 billion equity entity pulls materially higher passive inflows than two separate $20 billion entities. The capital doesn’t evaluate rock quality or reserve replacement ratios or inventory depth. It just buys the index weight and rebalances quarterly.

Scale the size of CTRA/DVN would highlight the valuation difference versus large caps, potentially closing Devon’s multiple gap. The company currently trades around 4.7 times EV to discretionary cash flow. Large-cap peers trade closer to 5.5 to 6.0 times.

Is that multiple expansion justified by better operations? No. It’s justified by size as a quality factor in a financialized market. Larger companies aren’t necessarily better operators. But they’re easier to own at institutional scale, more liquid for large position sizes, less volatile for passive strategies, more likely to stay in indices through rebalancing.

This is oil and gas financialization in its purest form (and was at play for SM-CIVI too). When organic growth is constrained, you optimize for capital markets’ structural preferences. And capital markets structurally prefer scale. They prefer names they can put $500 million into without moving the stock. They prefer constituents that won’t get kicked out of indices. They prefer companies large enough that analyst coverage is mandatory, not discretionary.

None of that has anything to do with how good your rocks are or how efficiently you drill wells. But it has everything to do with your cost of capital. And in a capital-intensive, commodity-price-exposed business, cost of capital is the game.

The Optionality Bet: Fighting the Pure-Play Religion

Here’s where Devon-Coterra diverges from what every activist investor with a Wharton MBA thinks they should do.

Kimmeridge Energy has been publicly pressuring Coterra to divest gas assets and become a Permian oil pure-play. Their pitch is clean. Focus creates value. Conglomerates get discounted. Simplify the story. Make it easy for generalist investors to understand. “You’re an oil company or a gas company. Pick one.”

Devon-Coterra is betting the exact opposite.

The combined entity gets oil exposure in Delaware and Williston. Gas exposure in Marcellus and Anadarko. Commodity flexibility to toggle capital allocation between oil and gas based on relative economics and macro conditions.

Both companies have successfully used Anadarko as a flexible dial. Increase gas allocation when NGL and gas spreads justify it. Treat the basin as a high-margin cash generator instead of a growth engine. When oil crashes and gas holds, you have options. When specific basins face regulatory pressure or operational issues or infrastructure constraints, you have alternatives.

The multi-basin model isn’t elegant. It’s messy. Harder to market to investors who want a clean Permian pure-play thesis. Probably commands some conglomerate discount in the multiple.

But it creates commodity optionality. And when external threats are numerous: Venezuelan supply restoration, Trump price targets, regulatory uncertainty, infrastructure bottlenecks, optionality has value. It might not maximize your multiple in bull markets. But it improves survival probability in sustained downturns.

You’re not trying to win the quarter. You’re trying not to lose the decade.

The Kalibr Read

The consensus narrative will frame this as strategic scale creation and enhanced shareholder returns through operational synergies and capital efficiency.

The contrarian reality is simpler. This merger is an acknowledgment that the organic growth phase is over, and the companies that survive the next decade will be the ones that recognized geological constraints earliest and adapted fastest.

For companies facing rock quality degrading 6% to 15% annually in core acreage, reserve life below peer averages and compressing, potential supply competition from international sources at structurally lower cost points, and investor expectations for 10% free cash flow yields with stable-to-growing production, aggregation might be the most rational path available.

Devon and Coterra would be buying time. Measured in years of inventory duration, reserve life extension, runway to generate cash before the next strategic inflection point forces another adaptation.

In a depleting resource business where the best raw material is running out, time might be the most valuable asset left to acquire.

You can’t make more Tier 1 Delaware rock. But you can combine what’s left and manage the depletion curve more efficiently than two separate entities could. You can build enough scale to survive commodity price shocks. You can create enough procurement leverage to squeeze service costs when oil prices don’t cooperate. You can cross market cap thresholds that attract passive flows and lower your cost of capital.

None of that changes the underlying geology. The wells will keep getting worse. The rock will keep depleting. The industry will keep transitioning from growth to harvest.

But if you’re going to harvest, you might as well be the largest, most efficient harvester with the longest runway and the best balance sheet.

That’s not pessimism. That’s adaptation. Every resource basin eventually transitions from discovery to development to maturity to harvest. The Permian is just transitioning faster than anyone expected in 2019. The companies that survive won’t be the ones pretending it’s still 2019. They’ll be the ones that saw the transition coming and built the right structure to manage it.

The less obvious answer is that you are not the first industry to confront this. Airlines and steel manufacturers have been navigating nearly identical dynamics for decades. Their playbooks are worth studying, not because compression is exactly like jet engines, but because the structural economics rhyme in ways that should make operators uncomfortable.

Let us start with aviation.

Power by the Hour: The Original Outsourced Uptime Model

Airlines face a version of your problem every time a jet engine needs maintenance. An engine on the ground is an aircraft not flying, which is revenue not earned. The economics are brutal: a shop visit can take 120 to 210 days. Downtime is existential.

So carriers outsourced.

Rolls-Royce pioneered a model called TotalCare, essentially “power-by-the-hour.” Airlines pay a fixed fee per engine flight hour. Rolls-Royce assumes the maintenance cost and uptime risk. Advanced telemetry monitors engine health in real time. Availability becomes someone else’s problem.

Sound familiar? Fixed monthly fees. Uptime guarantees. Predictive maintenance through remote monitoring. Contract compression is TotalCare for wellbores.

The numbers are instructive. Rolls-Royce has been systematically extending time-on-wing through durability upgrades, including an upgraded Trent 1000 high-pressure turbine blade that more than doubles service intervals (EMD Compression anyone?). The economic logic is self-reinforcing: longer intervals mean fewer shop visits, lower lifecycle cost, and happier customers willing to sign longer contracts. Compression providers have absorbed this lesson. Average time on location now exceeds six years, with 60 to 65 percent of contracts structured for periodic repricing.

But here is where the airline parallel gets interesting. Carriers are not passively accepting this arrangement. They are selectively pulling work back in-house.

When Airlines Decided to Build Leverage

easyJet recently announced it is acquiring SR Technics’ Malta operation and the Adria Tehnika facility in Slovenia. The carrier already handles roughly a quarter of its heavy maintenance internally. It has also purchased additional spare engines to improve resilience and reduce disruption risk.

Why? Because in a market where MRO capacity is constrained and shop slots are scarce, control matters. Deutsche Bank characterizes the current environment as one where demand for flight services and maintenance exceeds industrial capacity. Jefferies notes pricing tailwinds on spare parts. Goodbody describes an “engine-MRO super cycle” with GE, Rolls-Royce, and Lufthansa Technik investing heavily to expand overhaul capacity.

The translation for compression is direct. Utilization across scaled fleets sits in the mid-90s. New equipment faces approximately 60-week engine lead times. Capacity reservations, not just capital, determine start dates.

easyJet’s response is not full insourcing. It is a mixed model: maintain power-by-the-hour coverage for certain fleets while building internal capability and spare engine pools that provide flexibility and negotiating leverage. On its fiscal 2025 call, management acknowledged GE PBH coverage today but flagged the need to plan for future LEAP shop visits independently.

Emirates took a different path. It secured rights to perform MRO on its Trent 900 engines in Dubai while simultaneously extending TotalCare contracts into the 2040s. The hybrid approach preserves risk transfer while building internal capability and slot access within Rolls-Royce’s network.

Both carriers arrived at the same insight: you do not need to own everything to shift the negotiating dynamic. You need to own enough to set an internal benchmark and demonstrate credible alternatives.

The Exchange Engine Trick You Should Steal

Here is a cost-reduction lever that deserves more attention in oil and gas: modular exchange programs.

FTAI Aviation operates a model where carriers can swap an entire engine or module in as little as 48 hours versus lengthy shop-visit induction queues. Their turnaround times run 45 to 80 days compared to 120 to 210 days for traditional MRO. Customer downtime during an exchange is zero to three days.

The magic is inventory positioning. FTAI maintains rotable pools of engines and modules, combined with proprietary PMA parts and in-house repair capabilities. Jefferies describes the integrated inventory and repair ecosystem as aimed at becoming the “lowest-cost producer of aftermarket power.”

Compression operators should be thinking about this. Standardization around CAT 3400, 3500, and 3600 engines with Ariel frames preserves resale and redeployment value. A small exchange pool of standardized skids, combined with module and component swap capability, could materially accelerate restart times and reduce downtime revenue loss. Typical outsourced availability guarantees run 95 to 98 percent. A 4 percent availability delta can swing operator revenue by approximately $225,000 per unit on gas-to-sales and roughly $450,000 on gas-lift applications.

The aviation playbook is clear: if downtime is expensive and capacity is scarce, hedge your slot access. Pre-book maintenance windows. Stagger interventions. Build a rotation capability that keeps production humming while equipment cycles through service.

Industrial Gases: The Vertical Integration Precedent

Aviation offers one template. Industrial gases offer another, arguably closer to the structural reality compression providers inhabit.

Large industrial gas users face a classic make-versus-buy decision. Steel plants, refineries, and chemical facilities need massive quantities of oxygen, nitrogen, and hydrogen. The question is whether to outsource under long-term “over-the-fence” contracts or own captive production.

Air Liquide operates approximately 9,700 kilometers of oxygen and hydrogen pipelines across Northern Europe and the U.S. Gulf Coast. Contracts typically run 15 to 20 years with take-or-pay protection. Energy is the single largest cost, so majors embed escalation and pass-through clauses while driving productivity to stabilize margins.

The contract structures are remarkably similar to compression. Fixed monthly commitments. CPI or market-rate adjustments. Uptime responsibility transferred to the supplier. Revenue visibility for the provider, cost predictability for the customer.

But here is where the industrial gas story diverges. Large steel makers are increasingly choosing to own captive air separation units rather than outsource.

Linde India notes in its recent annual report that the BOO (build-own-operate) outsourcing model is losing appeal as large users favor control over supply, energy sourcing, and margins. When you are running a blast furnace at scale, the economics of owning your oxygen plant start to pencil differently than when you were smaller and less certain about utilization.

That sounds a lot like unconventional oil and gas today. Early shale was uncertain. You did not know what the factory would produce, so you rented flexibility. That uncertainty is mostly gone. Wells are predictable. Development is geographically concentrated. The seven-year ROI on purchasing compression equipment outright that midstream players have long enjoyed starts looking achievable for E&Ps with sufficient scale and density.

The Anchor-Plus-Merchant Model

Industrial gas majors figured out something compression operators should study: you do not have to choose between full insourcing and full outsourcing. You can blend.

Air Products’ fiscal 2025 revenue mix shows 52 percent on-site production, 44 percent merchant sales, and 4 percent equipment sales. The anchor-plus-merchant model commits core production to dedicated customers while enabling redeployment of incremental capacity to monetize merchant volumes and spread fixed costs.

Linde India’s new air separation units near Tata Steel will supply both the anchor customer and local merchant demand. Same infrastructure, multiple revenue streams, higher utilization.

The translation for compression is straightforward. Commit core horsepower to critical wells under long-tenor, minimum-take constructs. Enable redeployment of surplus capacity to nearby users during off-peak periods. Run the compression fleet more like a pipeline network and less like a rental fleet.

The compressed gas majors also figured out that network scale matters. Headered, shared infrastructure pools loads across multiple demand points, raising utilization and lowering unit cost. Brownfield expansions piggybacking on existing pipelines offer efficient capital deployment. Air Liquide recently targeted $50 million in brownfield additions on the Gulf Coast to extend hydrogen capacity efficiently by leveraging existing infrastructure.

Operators with geographic concentration should be asking whether headered compression across multiple pads could achieve similar economics. The physics are different from gas pipelines, but the principle of pooling load to raise utilization is identical.

Comparative Economics: What the Numbers Actually Say

Let us be concrete about the numbers driving these decisions.

Aviation maintenance costs: Traditional MRO shop visits: 120 to 210 day turnaround Exchange/modular repair: 45 to 80 day turnaround Customer downtime during exchange: 0 to 3 days Cost reduction from used serviceable material and exchange programs: roughly 25 to 35 percent over time

Industrial gases contract structures: On-site/pipeline tenor: 10 to 20 years Minimum take-or-pay provisions: Standard Energy cost pass-through: Embedded in contract pricing formulas Air Products revenue mix: 52 percent on-site, 44 percent merchant

Compression market dynamics: Fleet utilization: Mid-90s to high-90s across scaled providers Monthly tariffs: $21 to $24 per horsepower Engine lead times: Approximately 60 weeks Average time on location: Greater than six years Provider gross margins: Roughly 68 to 73 percent

Transaction multiples for context: USAC acquiring J-W: Approximately 5.8x 2026E EBITDA, roughly $1,015 to $1,050 per horsepower Archrock acquiring TOPS: Approximately 7.3x at roughly $1,695 per horsepower Archrock acquiring NGCS: Below 7.0x at roughly $1,130 per horsepower

The market is telling you that compression assets trade at infrastructure-like multiples because they behave like infrastructure. The question is whether you want to be a buyer at those multiples or find another way to capture the value.

The Hybrid Strategy Playbook

Experts across compression, aviation, and industrial gases converge on a consistent recommendation: mixed models outperform pure strategies.

From compression industry experts: Consolidation and longer five- to seven-year terms have strengthened pricing power. Electric units often move to month-to-month after initial terms, providing flexibility but exposing buyers to prevailing market rates. Fleets are highly utilized, with larger mobile units favored and quick redeployment into premium basins.

From airline executives: Large carriers insource heavy work but consistently outsource a tranche based on availability and price under long-term arrangements. The approach reflects risk aversion and capacity hedging.

From aviation MRO specialists: Exchange engines and module swaps can trim turnaround times by about 20 percent when combined with used serviceable material availability. Operators often prioritize turnaround time over price to keep aircraft productive.

From industrial gas veterans: Large users adopt on-site or pipeline supply with 15 to 20 year contracts and minimum take-or-pay to recover capital, with margins earned on volumes above the floor. Once plants are paid down, customers push back on pricing, but the efficiency risk sits with the supplier.

The actionable guidance for E&Ps synthesizes these perspectives into three pillars:

Contract design: Anchor a take-or-pay baseline for critical horsepower with CPI and energy pass-throughs. Add efficiency “reopener” clauses. Negotiate month-to-month tails post-initial term for flexibility.

Capacity hedging: Pre-book maintenance windows and stagger interventions. Build a small exchange pool of standardized skids. Allow module and component swaps to accelerate restarts and reduce downtime.

Mixed model: Insource a targeted share to set internal benchmarks and surge-response capability. Outsource the balance under long-term arrangements to transfer uptime risk and preserve capacity options.

The critical insight is that you do not need to own everything. You need to own enough to establish a credible internal reference cost. That reference cost becomes your leverage in every subsequent negotiation.

What This Means for the Insourcing Decision

Last week we posed the question: what is your plan? This week we have the framework for answering it.

The insourcing decision scorecard should evaluate several factors:

Horsepower intensity and mobilization economics: Rule-of-thumb mobilization and demobilization for a large horsepower unit exceeds $100,000. Multi-unit pads can exceed roughly $500,000 per site. Ownership wins when redeployments are frequent or locations are long-lived.

Uptime criticality: Typical outsourced guarantees run 95 to 98 percent mechanical availability. Justify insourcing only if internal operations and maintenance can credibly beat those guarantees.

Lead-time exposure: Outsourced providers report approximately 60-week lead times on larger orders. Owning does not eliminate supply constraints unless you pre-buy or secure used inventory.

Capital and O&M capability: Outsourced fleets are standardized around Caterpillar and Ariel with telematics across units. Match this in-house to avoid higher lifecycle cost and lower uptime.

Energy strategy: Electric motor drive offers superior uptime and lower maintenance where power is available. Grid constraints can cap realizable benefits. Plan for dual-drive flexibility.

A simple comparison: Annual outsource cost equals tariff times horsepower times twelve, minus SLA credits. Insource total cost of ownership equals depreciation plus financing plus planned maintenance plus fuel or power plus field labor plus expected downtime revenue loss. Include mobilization costs and lead-time slippage penalties.

Run the math. Airlines and industrial gas users did not guess their way into hybrid models. They quantified the inflection point where partial ownership changed the economics.

The Contrarian Take

The consensus view is that compression consolidation leaves operators with fewer options and higher costs. That is true but incomplete.

The more interesting observation is that the same consolidation creating pricing pressure is also creating the conditions for insourcing to pencil. When three companies control 75 percent of the outsourced market and earn 70 percent margins, the economics of vertical integration improve mechanically. The spread between outsourced cost and insourced total cost of ownership widens. The internal rate of return on ownership assets goes up.

Aviation and industrial gases teach us that this dynamic has a natural endpoint. At some threshold of supplier pricing power and margin capture, large users begin bringing production back in-house. Not all of it. Enough to establish benchmarks, create options, and restore leverage.

The compression market is approaching that threshold. The USAC acquisition of J-W is not the end of consolidation. It is the beginning of the response.

Over the coming weeks, we will run the math on different insourcing entry points, model the hybrid strategies that airlines and gas users have deployed, and outline specific contract structures operators can use to respond.

Because when your conveyor belt provider starts earning margins that look like your margins, rational manufacturers do what they have always done.

They bring the factory home. One module at a time.

Thanks for reading Kalibr - Strategy Research for Oil & Gas! Subscribe for free to receive new posts and support my work.

Now imagine the plant is an oil well.

Every well is a purpose-built factory. You choose the architecture up front, and that choice locks in how the system behaves for years. In unconventional oil, the critical fork in the road is artificial lift. ESPs can deliver high rates, but they pull you toward larger casing strings and more steel in the ground. The industry, in pursuit of capex efficiency and free cash flow, has been moving the opposite direction. Slimmer designs push operators toward gas lift, and gas lift pushes you, inevitably, toward compression. These are not modular decisions. They are factory design choices.

Which is why outsourced compression is such a strange equilibrium. Compression is not an accessory. It governs backpressure, stabilizes lift performance, and determines whether production hums or stumbles. Switching it is not like swapping a pump. It is like changing the conveyor belt after the plant is already running. Historically, outsourcing made sense because early unconventionals were uncertain. You did not know what the factory would produce, so you rented flexibility.

That uncertainty is mostly gone. What replaced it is something more interesting. Since 2020, compression providers have culled fleets, consolidated, and rediscovered a lesson E&Ps learned earlier: scarcity plus discipline equals distributions. The result is a business earning roughly 70 percent margins on equipment that is operationally critical and structurally sticky.

That dynamic has accelerated consolidation, particularly around large, electric, high-horsepower fleets. Archrock and Kodiak have been explicit about this playbook. Fewer units. Higher utilization. Cash returned to shareholders. Compression, it turns out, wants to be an E&P when it grows up.

Which brings us to USAC and J-W. The deal has the same family resemblance, but the value creation tightens a different part of the system. That distinction matters more than it looks at first glance.

Let’s briefly put on the investment banking hat.

The transaction values J-W between $860 million and $890 million, split evenly between cash from the revolving credit facility and new common unit issuance. The implied valuation is roughly 5.8x 2026E EBITDA and approximately $150 million of adjusted EBITDA in 2026. That multiple sits meaningfully below recent compression comps.

USAC effectively arbitraged its own 9.0x trading multiple to acquire assets at 5.8x, creating immediate balance sheet accretion and supporting a path to sub-4.0x leverage. Part of that discount reflects J-W’s mixed fleet composition, only 46 percent large horsepower versus roughly 75 percent for USAC, and its geographic diversity across the Rockies, Mid-Continent, Northeast, and Gulf Coast.

That difference is critical. Most recent third-party compression consolidation has been tightly focused on large horsepower, electric motor drive, and the Permian Basin. This deal is not.

So what does this mean for E&Ps?

Kalibr takeaway #1: The “Permian pricing power” story is about to get less Permian and more everywhere else

The market has grown comfortable with a tidy narrative. Consolidate large horsepower in the Permian, keep utilization tight, push price. Archrock and Kodiak have executed this well, particularly around large horsepower and electric-drive adjacencies, and the returns reflect it.

USAC’s acquisition of J-W is a reminder that pricing power cannot stay bottled up in one basin and one product class. The deal clears at roughly 5.8x 2026E EBITDA and about $1,015 to $1,050 per horsepower, a clear discount to Archrock’s recent transactions. TOPS transacted at roughly 7.3x and about $1,695 per horsepower. NGCS cleared below 7.0x and around $1,130 per horsepower. That discount is not just asset quality. It is also geography and mix.

J-W is only 46 percent large horsepower and carries meaningful medium and small horsepower exposure. It expands USAC’s footprint well beyond the Permian. If compression pricing is increasingly a function of scarcity and discipline, then USAC just bought the mechanism to export that dynamic into basins and horsepower classes that historically cleared on shorter contracts, higher churn, and more vendor substitution.

The Permian will not lose pricing power. It may lose exclusivity. Operators outside the Permian, and operators with smaller horsepower needs, should assume the Permian experience is coming to a basin near them.

Kalibr takeaway #2: This is not a horsepower acquisition. It is a contract repricing acquisition disguised as M&A

The headlines will focus on triopoly dynamics. The real economics sit inside J-W’s contract book, which reads like a list of correctable terms.

J-W’s average contract tenor is roughly 15 months, versus USAC’s preferred 30 months. Gross margins are about 60 percent at J-W versus roughly 69 percent at USAC. That nine-point gap does not come from a secret operational trick. It comes from pricing, contract structure, and inflation protection that actually works. USAC management has already noted that CPI pass-throughs have historically lagged market rate increases. Translation: the next contracts will look different.

The strategy is straightforward. Acquire a fleet where contracts roll quickly, then migrate those contracts to longer tenor and higher margin under USAC’s standard form. That is value creation you can model without heroics. Layer in the operational option value. The deal includes approximately 50,000 horsepower of idle capacity that can be redeployed with minimal capital, roughly $100 to $300 per horsepower. In a tight market, cheaply deployable horsepower is not an accounting detail. It is a pricing lever.

For E&Ps, this is how consolidation shows up in practice. Not as a dramatic price reset on day one, but as a steady conversion of short-duration, more negotiable contracts into longer, stickier agreements with tighter escalators and higher base rates. Vendor choice does not disappear overnight. Leverage erodes quietly, at renewal.

Kalibr takeaway #3: USAC is positioning itself as the dry gas compression utility, and that pressures peers where they least want pressure

Most observers will frame this as USAC simply getting bigger. The more interesting point is what USAC is becoming.

Pre-deal, roughly 74 percent of USAC’s exposure was Permian. J-W is only about 21 percent Permian. Pro forma, the combined company is meaningfully less Permian-concentrated and far more present in the Northeast, Rockies, Mid-Continent, and Gulf Coast. The deal positions USAC as a go-to provider for dry gas compression, aligned with investor focus on gas supply growth outside the Permian.

This matters because electrification is real, but the grid is not always cooperative. Electric compression offers lower maintenance and emissions, but power availability and reliability remain binding constraints. USAC’s dual-drive approach monetizes the reality that electrification is uneven and infrastructure arrives on its own timeline. That flexibility is commercially valuable in non-Permian basins where power buildout can be a gating factor.

The broader implication is subtle but important. Pricing power does not disappear. It diffuses. It spreads into other basins, into smaller horsepower, and into a more utility-like model where the largest players do not need to win every Permian bid to win the pricing cycle. Meanwhile, the mid-tier and fragmented end of the market becomes structurally more vulnerable. The removal of J-W as an 850,000 horsepower standalone bridge widens the gap, leaving the next cohort of roughly 500,000 horsepower players and an estimated 2.7 million horsepower fragmented private fleet operating in a world where scale, service infrastructure, and cost of capital matter more than ever.

For operators, the trade-off is familiar. Potentially better reliability and availability from scaled providers, paired with fewer credible alternatives when pushing back on price or terms. Compression was already sticky. This deal makes it stickier and extends that stickiness well beyond the Permian.

Which brings us to the practical question this transaction forces back onto the table. What, exactly, is your plan?

Kalibr’s New Year’s resolution for every oil and gas operator is simple and uncomfortable. Run a serious insourcing analysis. Not because you should immediately own compression, and not because outsourcing is wrong, but because without a quantified inflection point you are negotiating blind. You do not know where ownership begins to win on total cost, reliability, or operational control. In a market where credible third-party alternatives are shrinking, ignorance is not neutral. It is a pricing signal.

At a minimum, an insourcing framework creates leverage. It clarifies which horsepower classes matter, where reliability actually drives value, and how switching costs compound across drilling, completions, and production. It forces answers to basic questions. Do you start with smaller horsepower at the edge of the system, or large horsepower on the critical path? What utilization threshold makes ownership rational? What does incremental uptime actually buy you in barrels, not anecdotes?

Over the coming weeks, we will explore how other capital-intensive industries pulled critical infrastructure back in-house, run the math on different insourcing entry points, and outline strategies operators can use to respond to the growing pricing power of third-party compression. This is not theory. It is arithmetic.

Because in a world of tightening geological runway and relentless scrutiny on LOE, a 70 percent margin on a mission-critical conveyor belt stops looking like a service and starts looking like an opportunity. At some point, rational operators do what manufacturers have always done when a supplier becomes too essential and too profitable. They bring the conveyor belts home.

It’s peak Larry David: wasting more energy avoiding waste than the waste itself was ever worth.

Which, incidentally, is how most operators treat chemical spend.

Ask any production or operations engineer. Chemicals are the most resented line on the LOE. They’re invisible when they work, loud when they don’t, and somehow always the first target when management needs to trim costs. Every month there’s that one cell in the spreadsheet flashing cut me. The better your chemicals perform, the less evidence you have they were ever necessary. Economists call it a counterfactual. Engineers call it Quarterly Board Meeting prep.

At Kalibr, this is one of the most common categories we get asked to untangle. The issue isn’t that chemicals are unnecessary. It’s that no one can prove the current cost structure is rational. So, let’s walk through a simple framework to fix that.

Five Steps to Rational Chemical Spend

1. Feedstock Decomposition and Cost Modeling – Break every chemical into its component inputs, map each to transparent price indices, and identify where cross-industry demand drives cost volatility.

2. Supplier and Inventory Dynamics – Quantify pricing power by looking at balance sheets, inventories, and short-cycle demand.

3. Should-Cost Benchmarking – Combine the inputs into a baseline model and measure immediate savings potential.

4. Forward-Looking Monitoring – Build a living system that tracks feedstocks, freight, and demand signals with automated triggers.

5. Direct Sourcing – Eventually, stop paying other people to buy your chemicals for you.

Let’s ground this in the Permian Basin.

1. Feedstock Decomposition and Cost Modeling

Engineers like process, so we’ll start there. The goal is to break each chemical into its core molecular precursors, link those to public indices, and understand which ones are domestic, which are imported, and who else in the economy needs them.

Take foaming agents. These are typically alcohol ethoxylate surfactants made from ethylene oxide (EO) and C12–C14 fatty alcohols.

• EO is traded locally, with only about one percent imported, benchmarked through CMA or OPIS.

• Fatty alcohols are mostly palm-based imports from Asia, roughly 70 to 90 percent of total supply.

• EO demand comes mainly from polyester and MEG production, not oil and gas, which matters when China’s textile plants sneeze and your chemical supplier catches a cold.

Once the components are mapped, we regress the end-product price against its feedstock indices using two to three years of historical data. Add freight, packaging, and a normalized margin, and you’ve built a defensible should-cost curve in dollars per gallon.

The result is a feedstock-to-finished cost curve. It shows exactly how index movements flow through to delivered cost. Alcohol ethoxylate prices, for example, move with EO outages on the Gulf Coast and palm-based alcohol volatility in Asia.

The point isn’t precision for its own sake. It’s to turn “the chemical budget looks high” into “ethylene oxide moved up six cents a pound and your supplier passed through seven.”

2. Supplier and Inventory Dynamics

The Kalibr approach borrows from game theory. We model supplier leverage across multiple dimensions: customer exposure, basin demand, inventory strain, credit posture, and cross-industry pull. But for simplicity, let’s focus on one that’s easy to visualize — inventory.

Inventory isn’t free. It carries a financing cost that erodes supplier margins and reveals who’s most motivated to move product. Knowing that cost helps you quantify where the real leverage sits.

Illustrative inventory carrying costs:

• Baker Hughes: $5.036bn × ~4.18% ≈ $210m annual (~$52m/qtr). Lower inventories reduce carry and support negotiation flexibility.

• Halliburton: $3.095bn; with investment‑grade profile and fixed debt, apply illustrative 4–5% → $124–155m annual (assumption)

• ChampionX (Pre-SLB) — Debt cost benchmark SOFR+2.75%; post‑close funded within SLB structure.

• Innospec: No revolver usage; financing carry de‑minimis; focus on DSI mix effects

Once quantified, the BATNA range becomes visible. Anything above the supplier’s BATNA and they have better options. Anything below yours and you do. That turns “we think the price is high” into a concrete, data-backed leverage map.

3. Should-Cost Benchmarking and Savings

Now we combine the models. Feedstocks define the floor. BATNA defines the ceiling. Somewhere in between lies the right price.

Example:

• A methanol-based H₂S scavenger is roughly 60 percent methanol by cost. A 10 percent decline in methanol prices should cut delivered cost by about six percent. Adjust freight from $0.18 to $0.10 a gallon and you’ve got another few cents of margin.

• EO/PO surfactant blends can shed 4 to 6 percent through freight normalization and modest markup compression.

Across a typical portfolio, that’s a 3 to 8 percent savings window at time zero, with freight-heavy categories reaching double digits. On a $25 million chemical budget, that’s three quarters of a million to three million dollars a year of cash margin.

4. Forward-Looking Cost and Demand Monitoring

Optimizing once is clever. Doing it continuously is strategy.

We track market data—feedstock indices, manufacturing PMIs, freight rates, and supplier commentary—to flag when conditions justify a renegotiation.

Sample triggers:

• Methanol index drops 15 percent in 30 days: open new RFPs for scavengers.

• PGP down 10 percent quarter-on-quarter: reset escalation clauses.

Then we layer on a dashboard for visibility:

• Current invoice versus should-cost

• 30-, 60-, and 90-day feedstock deltas

• Freight indices from Houston to Midland and Odessa

• Supplier exposure and alert status

Finally, we forecast the next quarter using a simple model with external variables like ethylene, methanol, freight, PMIs, and energy curves. The point isn’t to predict the future; it’s to know when the present is out of line with it.

5. Direct Sourcing

Eventually, you stop optimizing the middleman and start eliminating him.

That’s a long-term move, like insourcing compression or owning your own slickline trucks. But it’s also where structural margin lives. Operators with visibility into Gulf Coast producers—Dow, BASF, LyondellBasell, INEOS, Sasol, Shell—and regional blenders like Seatex or Monument can realistically capture mid-teens margins that currently accrue to service companies.

The savings from the first four steps usually fund the capability to make this happen. At that point, you’re not just cutting cost. You’re changing where in the value chain it accrues.

Conclusion

No one has a tougher job in Oil and Gas than Production Engineers. Spread thin, tasked with ensuring the basic cash flow of the business and constantly answering questions about their cost structure while being required to show proof no one can concretely show. IMO, this job gets even more critical as we move forward as an industry, with much of the focused around cost reduction being squarely on the D&C side or the easy G&A numbers of mergers. But, as we continue to face significant Tier 1 acreage headwind, it becomes more critical.

At a minimum, the workflow above ensures a justifiable cost structure and the ability to quickly capture market wins in an already strained environment. And that alone, as stated by the infamous Larry David, would be “pretty, pretty, pretty good.”