Sterling's Breakneck Data-Center Growth Collides With a Mix Problem — and a 48% Drawdown
Record margins and backlog hide a margin-suppressing revenue mix that has the market asking whether the runaway train can keep its valuation.
STRL · Earnings Call · 2026-08-04
A Quarter That Should Have Soared
Sterling Infrastructure delivered what would be a career-defining quarter for most CEOs: revenue up 90%, adjusted EPS up 116%, and total revenue reaching $826M, driven by a 192% surge in E-Infrastructure. Yet the stock, after a spectacular run that peaked at $993.74 on June 4, is down 48% from that high. The disconnect isn't demand — it's mix. As CEO Joe Cutillo put it bluntly:That's the crux. Cutillo insists every individual business — site development, CEC electrical, even transportation — improved margins year-over-year. But the revenue pie is shifting heavily toward electrical (CEC) and away from the ultra-high-margin site development, depressing overall reported margins. The market is pricing in the optics rather than the underlying trajectory.The only downside, which is the only reason I could see that our stock is not up exponentially today and looks like it's down is the mix makes our margin look like it's lower.
The CEC Machine and the Electrician Bottleneck
The CEC acquisition has been a revelation. It added a massive electrical capability that management believed would take a year to fill; it took 90 days. The business grew 140% in the quarter, and its combined backlog jumped $1.7 billion since year-end. "We filled their capacity up in about 90 days," Cutillo said, and the only constraint is people: "If we had 1,000 or 2,000 more electricians, we'd be growing it even faster." That capacity crunch is now the key strategic bottleneck. Across the mission-critical universe — data centers, semiconductors, EV plants — Sterling is being asked to do more, but the shortage of skilled electricians and the broader labor market limits how fast it can scale. Management is responding with investments in training (Sterling Academy, CEC University), prefabrication facilities, and a more aggressive M&A program, including a recent $1.5 billion revolver expansion to fund acquisitions. This isn't a new theme — prior calls were full of the same "if we had more electricians" refrain — but the tone has shifted from opportunism to urgency. "We are going to have to make more acquisitions," Cutillo said in Q2, "not only just from a strategic geographic standpoint, but from a pure capacity add standpoint to keep up with that demand."Margin Dilution, Not Margin Compression
The market's biggest worry is that margins are peaking or eroding. Management's response is that the mix math is obscuring real gains. "We're not losing margin in our businesses," Cutillo insisted. "This is purely a mix of revenue." The data supports the claim if you decompose it: site development margins sit in the high-20s, CEC around 12%, and the former is growing slower than the latter. As long as that growth differential persists, overall E-Infrastructure margins will look flat or down even while every unit improves. That's precisely what happened in Q2. E-Infrastructure adjusted operating margins were 24% — strong, but below the prior quarter's level. Analysts pressed on the trajectory, and management guided to continued margin expansion over the medium term: "We see 300 to 500 basis points on top of some things we're doing internally of margin improvement ... over 12 to 18 months." The path is three-fold: exiting low-margin legacy CEC work, scaling data-center projects, and gaining operational leverage in new geographies like Texas. The concern is whether the mix drag is transitory or structural. That the market is punishing the optics of a mix shift rather than the underlying returns suggests a valuation that leaves little room for error. At $4.3x price-to-sales (down from 4.7x at peak), STRL still trades at a premium that assumes flawless execution and continued 30%+ organic growth.What's Coming Next: Visibility and Caveats
The company sees a transformational runway ahead. Future phase opportunities exceed $1.4 billion, and there's "visibility into a total addressable pool of work of more than $7 billion." The semiconductor campus in New York is running ahead of schedule, and they've landed the initial scope on an EV plant in Atlanta. Management even hinted that some projects could last 8-12 years if customers keep expanding. But they also injected a note of caution: "We could see softer third quarter awards with higher levels in the fourth quarter and early 2027." That statement was likely a trigger for selling, as investors feared demand was stalling. Cutillo came out forcefully to counter that: "This would reflect award timing, not a change in demand." The market may be too quick to read the worst into any hiccup, especially after a run that put the stock at nosebleed levels. The street's skepticism is understandable — the guidance raise implies H2 revenue of roughly $2.1B, up 64% year-over-year, but margins are guided to mid-20% in E-Infrastructure, down from the 24% reported. If mix continues to mask margin gains, the story becomes harder to sell.That's the longer-term thesis: AI isn't just boosting demand for data centers; it's also boosting Sterling's own productivity through better project management, which compounds the margin story once the mix headwind fades. Prior to this quarter, the company was already talking about "geographic expansion" and "Pacific Northwest" as key drivers (see Q1 2026 call, component 2368274023080917239). The new twist is the willingness to add capacity at any cost — including through M&A at a time when valuations are high. "We are seeing more high-quality acquisition targets in the market today than a year ago," Cutillo said, echoing the M&A interest from the May call. The financial foundation is solid: $223M of cash net of debt, revolver expanded to $1.5B, and free cash flow conversion expected to remain over 80%. The company is being conservative in guidance, and the stock's 48% drawdown from its June peak presents a valuation that, while still rich, better reflects execution risk. Geographic expansion is the next chapter, and the market will be watching to see if Stone Ridge (Pacific Northwest) and additional Texas capacity can replicate the CEC success story. If margins break out once the mix stabilizes, this dip could be the opportunity of the cycle — but the market needs to see evidence, not just guidance.I think people would be shocked what we've done and what we're doing in and around AI and how we're tying that to not only the estimating and bid process, but the project execution.