Sterling's Perfect Quarter, 62% Drawdown: The Market Has Stopped Paying for Mix
EPS doubled to $7.58, scoped work hit ~$8 billion, guidance went up again — and the stock still sits 62% below its June high. The fight isn't about demand anymore. It's about what an electrical-heavy revenue mix is worth.
STRL · Earnings Call · 2026-10-29
The quarter nobody argued with, and the tape nobody can explain away
Sterling printed one of the cleanest quarters in its history: adjusted diluted EPS of $7.58, up 118%; revenue up 127%; E-Infrastructure revenue up 212%; guidance raised to $4.63–$4.73 billion of revenue and $21.70–$21.96 of adjusted EPS, whose midpoints imply 88% revenue and 101% adjusted EPS growth for the year. Total scoped work — signed backlog plus unsigned awards plus the next phases in design — now sits at nearly $8 billion, up roughly 90% year over year, with E-Infrastructure alone approaching $7 billion, up more than 130%.
And the stock is down 56.9% over the last 90 trading days, 62.4% from its June 4 peak of $993.74. The last filed quarter valued the business at 7.5x revenue, and 56x net income. Those are numbers that require perfection, and perfection is exactly what the company delivered — which is why the disconnect is the story, not the results.
The mix fight: a CEO losing patience with his own shareholder base
The sell-side's first question went straight at margin guidance for Q4 (~19% versus 21.6% in Q3). Joe Cutillo did not do the polite analyst-day answer.
We have an electrical business which is very different than our site civil business. It requires a heck of a lot less capital, and we've been able to grow it significantly faster than we ever imagined. That is a great thing. We've more than doubled the business, yet everybody is looking at it as a negative because it's bringing down the total margins... it's really baffling to me.
The arithmetic behind the anger: CEC did $426 million of revenue in the quarter, "more than quadrupling its pro forma full quarter revenue," and went from 10% to a third of segment revenue. Segment adjusted operating margin of 23% was therefore down year over year even though segment operating income rose 171%. The blended margin is a denominator problem dressed as a quality problem — and management knows it, having already banked "over 150 basis points of improvement in CEC" against a promised 300–500 basis points over 18–24 months. This is not a new argument, either; back in May the framing was identical: “We have more opportunities than we have capacity to get to with CEC, so, like everything else, we will focus more where we get joint awards.” — Joseph A. Cutillo, Chief Executive Officer · 2026-05-05 Three quarters of the same explanation, and the stock has de-rated anyway.
The metric they told you to stop watching
The most revealing passage wasn't about margins. It was about backlog.
“On the backlog, I wish somehow we could get out of backlog altogether because it is really a meaningless number for us.” — Joe Cutillo, Chief Executive Officer · 2026-10-29
That is a genuine rhetorical turn. For several quarters the equity story leaned on combined backlog, book-to-burn and phased releases; the keyword history shows Combined backlog and sign backlog as recurring top-of-page items in 2025 and early 2026. This quarter those phrases slide down the decliner list while scoped work ascends. The reframing is coherent — Transportation's signed backlog rose 31% year over year while its combined backlog fell 22%, and the segment's revenue actually declined 3% as crews were redeployed — but it lands precisely when backlog optics weaken. The substitute is land: “we literally have thousands of acres of land that is permitted, powered, and ready to go... Those represent over five years of work for us.” — Joe Cutillo, Chief Executive Officer · 2026-10-29 Bullish, believable, and unauditable.
Permitting, the election, and the 2030 semiconductor alibi
A new keyword this quarter is permitting, and it arrives with an unusual admission of political limbo: “There is so much noise out there, and I can't wait for next week to be over, on permitting and everything else.” — Joe Cutillo, Chief Executive Officer · 2026-10-29 The TCEQ water-permit mess in Texas — automated approvals issued, then challenged, then a standoff over whether manual review even exists — is the kind of fog that freezes customer decisions for a quarter or two. Cutillo's honest framing: "are those projects going to go? Absolutely... is there confusion in the system right now? Absolutely there is."
Meanwhile the Micron job, pulled in on the back of an ahead-of-schedule New York fab, is explicitly a bonus rather than a base case. The February framing is unchanged: “I don't think we're going to see a major phase of semiconductor come on until 2030 or so just with the length of time of the announced capital spend that it takes to get a project ready to break ground.” — Joseph Cutillo, Chief Executive Officer · 2026-02-26
Now put that beside the global tape. AI data center as a market theme is still up over a year but has turned hard over 90 days — 49 decliners against 7 advancers, with adjacent clusters like High bandwidth memory and data center capacity down 7–8%. Sterling's drawdown is not an indictment of its execution; it is a sector-wide repricing of the physical AI build-out, and Sterling happens to be one of the highest-beta ways to express it. Note the tell in the capacity strategy: the new top-ranked keyword is traveling crews, tied to deals designed to free up a traveling crew and thereby release $500 million-plus of redeployable capacity. That is a company optimizing for a boom it expects to continue while the market prices the boom's duration downward.
What would actually change the narrative
Three things. First, a margin floor: management won't call Q4 the trough, citing seasonality and a conservatively assumed winter after 2025's "easiest, softest winter we've had." Site development margins held flat at first-half levels despite $600 million of Micron work carrying 4 million cubic yards of low-margin rock — a quiet proof point that the core is not leaking. Second, the 2027 framework: 20–30% organic profit growth plus acquisitions supporting roughly 50% earnings growth, which is not aspirational so much as extrapolative — “over the last six years, we've grown at 48% on the bottom.” — Joe Cutillo, Chief Executive Officer · 2026-10-29 Third, cost credibility: “Nobody's heard us talk about diesel. Nobody's heard us talk about inflation.” — Joe Cutillo, Chief Executive Officer · 2026-10-29 That's the phased-contract structure doing real work, and it is the single best argument that the mix is a price rather than a decay.
The balance sheet supports the waiting: $217 million of cash against $95 million of debt, a $1.5 billion revolver extended to 2031, and $267 million left on the buyback — though effective net cash fell 55% year over year to $179 million, and operating margin reached 18.8%. The bear case isn't that Sterling is wrong about demand. It's that a business whose growth is increasingly electrical, less capital-intensive, and structurally lower-margin deserves a contractor's multiple, not a software one — and that 56x earnings was the anomaly, not the drawdown.