Archrock's Best Bookings In Three Years Met Its Worst Tape In One
Record-adjacent compression demand, an eighth straight 70% margin quarter — and a stock down 22% while the buyback sat idle.
AROC · Earnings Call · 2026-11-03
The Disconnect Nobody On The Call Could Name
Archrock's third quarter was, on the operating line, close to flawless. Net income of $84.5 million ($0.48/share), adjusted EBITDA of $228.4 million, contract operations utilization at 94.3%, adjusted gross margin of 72% — the eighth consecutive quarter at or above 70% — bookings the highest since the second quarter of 2023 with year-to-date volumes up 13%, leverage down to 2.5x from 3.1x a year ago, and a dividend raised roughly 10% with 3.8x coverage. The leverage of 2.5x and the dividend coverage are both, by management's own framing, comfortably inside the long-term target rather than at it. And yet the stock closed the day 22% below where it started the 90-day window, 25% off its June 26 peak of $41.95. Jim Rollyson of Raymond James asked the only question that mattered, and asked it plainly — whether, given "all the noise kind of going on around data centers going into elections today and obviously, the recent retreat in your share price," anything had changed through 2030. Brad Childers' two-word answer: “In short, no.” — D. Childers, President and Chief Executive Officer · 2026-11-03 He then added the line that should anchor the whole thesis: “If I were to take the over-under, I still believe the market is more likely to be short of compression equipment than long on compression equipment.” — D. Childers, President and Chief Executive Officer · 2026-11-03Two Things Management Quietly Stopped Saying
The buyback went silent. This is the sharpest tell on the call. Archrock ended the quarter with $113.2 million of remaining repurchase authorization and, by its own CFO's admission, didn't touch it: “we are also in an open window. It just did not happen in Q3.” — Mohit Singh, Chief Financial Officer · 2026-11-03 That is a notable departure from a company that has been loudly price-sensitive about its own equity. A year ago the CEO described the buyback posture as opportunistic to the point of aggression — “We are definitely price sensitive on that. And currently, we might be ambitious.” — D. Bradley Childers, President and Chief Executive Officer · 2025-08-05 The cash trail is unambiguous: repurchases ran $29M in 2025Q2, $25M in Q3, $16M in Q4, $4M in 2026Q1 and $0 in Q2. Notably, the phrase attractive buyback opportunity still appears in the company's Q3 keyword set even as the dollars vanished. Language persisted; capital did not. Either the open window was never really open, or management's own view of fair value is less aggressive than its commentary implies. The lead-time tautology finally broke. For four straight quarters Caterpillar engine lead times were the company's favorite proof of a structurally tight market: 60 weeks in October 2025, 110–120 weeks by February 2026, 160 weeks in May, then 195 weeks in August with the CEO declaring “their lead times are now out where we're ordering for 2029.” — D. Childers, President and Chief Executive Officer · 2026-08-05 On this call, for the first time, the number moved the other way — "over 180 weeks" in the prepared remarks, and in Q&A: “Lead times have come down.” — D. Childers, President and Chief Executive Officer · 2026-11-03 Management credits Caterpillar throughput investment, and that is plausible. But the Caterpillar lead time was the single cleanest proxy this company had for physical scarcity, and it just stopped tightening.The P&L Noise, And What It Hides
Contract operations revenue rose about 1.5% year-over-year to $331 million despite the sale of ~112,000 non-strategic horsepower, with revenue per horsepower per month up 4% to $24.38. The offsetting wound was aftermarket services: revenue of $40 million versus $56 million, a 29% decline, as customers again deferred maintenance activity. This is now a multi-quarter refrain rather than a surprise. Six months ago the CEO was already calling it a pay-me-now-or-pay-me-later business, and this quarter he was candid that the clock is unknowable: “I don't really have a clock or a calendar that is going to give us that insight.” — D. Childers, President and Chief Executive Officer · 2026-11-03 The segment held 23% margins, so the deferral is a revenue problem, not a profitability one. Costs were the other grind. Lube oil and make-ready cost to return idle equipment to service both pressed on the quarter, and the company's own keyword set tilted defensive — cost headwind, cost pressure and disciplined cost management all rank near the top of the Q3 theme list. That is what a margin-defense quarter sounds like. Meanwhile SG&A flattered the print via a mark-to-market credit on long-term incentive comp; the CFO was explicit that “Q3 SG&A should not be taken as the ongoing run rate.” — Mohit Singh, Chief Financial Officer · 2026-11-03 The underlying trajectory is still intact — revenue of $371M, down 3% year-over-year through the July quarter with a gross margin of 46.6%, up 1.2 points year-over-year — but it is a business now priced for the future, not the present. Price to FCF less SBC sits at 26.7x, down two-thirds year-over-year, against effective net cash of -$2.3 billion.The Real Innovation Is The Delay, Not The Demand
The demand narrative is well-worn and the market has stopped paying for it. Bcf per day ranked first in the company's Q3 vocabulary — 2026 gas volumes up ~2.3 Bcf/d, LNG-related demand from ~18 Bcf/d today to 34 by 2030 — yet compression demand and LNG export both fell out of the company's keyword ranks this quarter. The market's own AI data center basket has been sold hard across windows, and Archrock is being marked as collateral to that unwind despite gas compression for power being second-order exposure at best. What genuinely changed in Q3 is the fuse length. With lead times out to 2029, bookings no longer convert on the old cadence:So the very headline bulls are quoting — highest bookings since 2023 — is a 2028–2030 revenue signal, not a 2027 one. That is precisely what timing from booking to revenue captures, and it's why the $1.4–1.6 billion of growth capital through 2030 reads as a commitment to a thesis the equity market is no longer willing to finance at 30x. One genuine positive nuance: growth is broadening. The Permian supplied only 30–40% of Q3 growth, with the Rockies the largest single contributor. That is a healthier mix than the 60% Permian concentration management described a year ago — but it also means compression equipment demand is increasingly tied to dry-gas and power-linked basins where the customer capex is less certain. The verdict. Nothing broke at Archrock this quarter. What broke is the alignment between the story and the tape. A company with 3.8x dividend coverage, 2.5x leverage and 8 straight quarters of 70%+ margins is trading as if the gas thesis had cracked — and management, sitting on $113 million of buyback authorization, declined to bet its own capital on the gap. Until that changes, the bookings are a promise and the silence is a statement.Normally speaking, if we had a 52-week lead time, the timing from booking to revenue would be something like instead of 1 year, it would be a year and a quarter... But with lead times out through 2029, the time frame for bookings -- that booking cycle to first revenue is way extended.