Sterling Infrastructure: The Margin Story the Market Got Wrong (Upgrade)

I've been baffled by Sterling Infrastructure. The company just delivered a third straight beat-and-raise — revenue, EPS, and EBITDA all exceeded consensus — then raised its full-year guidance by an aggressive margin. And what did the market do? It dumped the stock -13.98% in a single session and hasn't stopped selling since.

The official explanation is "the mix." Investors are apparently worried about margin compression in the E-Infrastructure segment and a 20% decline in transportation revenue. The headline reads like a deterioration story. The numbers say something very different.

The arithmetic error driving the selloff

E-Infrastructure's adjusted operating margin fell from 28.3% in Q2 2025 to 24.1% in Q2 2026. That's the number anchoring the selloff. But here's what that comparison hides: in Q2 2025, the segment consisted almost entirely of site development, which ran margins in the upper 20s. There was virtually no electrical services revenue to dilute the blend.

This quarter, electrical services — acquired via the CEC Facilities Group buy — are a massive and growing part of the mix. Electrical services earned an adjusted operating margin of 11.4%, which is real but structurally lower than site development. Blending the two together mechanically compresses the headline E-Infrastructure margin, even though every single component actually improved.

Site development margins expanded year-over-year and sequentially. CEC's electrical margins also improved year-over-year and sequentially. The sequential E-Infrastructure margin actually increased from 23.5% in Q1 to 24.1% in Q2. The margin didn't deteriorate — it's being averaged with a lower-margin, faster-growing sub-segment that management expects to close the gap over time.

That matters because it means the market is pricing in fundamental decay where there is none. The margin story is a blend effect, not a profit-power problem.

Management guides for CEC to add 300 to 500 basis points of margin improvement over 12 to 18 months. The CEO has expressed an ambition to grow electrical margins toward 20% over three to five years. That's not a return to 28% site development margins — electrical will never be that profitable. But it's a trajectory, not a collapse.

The transportation "decline" was intentional

Transportation revenue fell 20% year-over-year. Taken in isolation, that sounds like a shrinking business. It isn't. Management has been deliberately winding down low-bid highway work and reallocating those resources to higher-margin E-Infrastructure opportunities. The transportation segment's adjusted operating margin expanded by more than 500 basis points to 19.5% as the company exited its least profitable contracts. Revenue contracted by design.

What the guidance actually says

The company raised its full-year 2026 guidance across the board. Revenue guidance now sits at $4.08 billion at the midpoint, a midpoint implying 64% year-over-year growth. Adjusted diluted EPS guidance of $19.70 to $20.30 implies 84% growth. Adjusted EBITDA guidance of $891 million to $916 million implies 79% growth.

For context, Sterling's Q2 revenue of $1.17 billion represented roughly 50% organic growth, with acquisitions contributing the remainder. Signed backlog more than doubled year-over-year to US$4.3b. The total visible addressable pool — signed backlog, unsigned awards, and future-phase opportunities — exceeds $7 billion, up more than $2.5 billion since year-end 2025.

Operating cash flow for the first half of 2026 was $328 million. The company ended the quarter with $464.5 million in cash. This isn't a growth story funded by leverage. It's being funded by the business.

The concentration risk that's real

I'm not going to pretend the setup is pristine. More than 92% of E-Infrastructure backlog was tied to mission-critical projects — data centers, semiconductor fabrication facilities, and advanced manufacturing. That's an extraordinary concentration in AI infrastructure spending.

If hyperscaler capex cycles slow, Sterling is more exposed than the average infrastructure contractor. The broader selloff across AI-adjacent names in late July reflected that exact anxiety. Sterling's decline outpaced the sector rotation because of the margin concern, but the sector downdraft amplified the damage.

That said, 92% mission-critical concentration is also why this company grew revenue 90% in a single quarter and can raise guidance with this much conviction. The backlog is $7 billion and growing. Book-to-burn ratios of 1.4x suggest more than adequate visibility into 2027 work. The concentration is a two-edged sword — and the blade pointing up is longer.

Where the stock sits now

Sterling Infrastructure is trading at $547, roughly 45% below its 52-week high of $1,006. The stock has fallen about 20% over the past 20 trading days. The RSI sits at 39, approaching oversold territory. The MACD is deeply negative. The price has dropped below the 50-day moving average of $744 but remains well above the 200-day moving average of $509.

Forward earnings estimates cluster around $20 per share for 2026, which puts the stock at roughly 27 to 28 times next-year earnings. For a company guiding for 64% revenue growth, 84% earnings growth, and a visible backlog exceeding $7 billion, that multiple is arguably more reflective of pessimism than the business reality warrants.

I must admit that I've been puzzled by the market's refusal to reward this company. The third straight beat-and-raise was the kind of execution that normally triggers accumulation. Instead, investors fixated on an arithmetic margin blend and a deliberate strategic shift in transportation, and sold the stock through its 200-day moving average.

The call

Sterling Infrastructure is a battered growth stock that has been misread, not impaired. The margin story the market is trading is wrong. The guidance raise tells the real story. The concentration in mission-critical work is a risk I acknowledge, but it's also the engine that makes this business compound at this pace.

I'm upgrading Sterling Infrastructure to a Buy. The risk/reward at 45% below highs, with RSI approaching oversold territory and selling exhaustion building, is increasingly asymmetric. I wouldn't chase a gap up, but the setup for patient accumulation is improving with each session.

The reassessment trigger: if E-Infrastructure sequential margins decline for two consecutive quarters, or if the mission-critical backlog share drops materially below 80%, the thesis would need rethinking. Until then, the market's focus on the mix has created a buying opportunity that deserves attention.