On August 3, SterlingSTRL-- Infrastructure — the Texas builder that rebuilt itself around data centers — reported revenue of $1.17 billion, up 90% from a year earlier, with adjusted earnings per share of $5.80, up 116%. It booked a record backlog and raised full-year guidance for the third straight quarter. The stock still trades roughly half off its 52-week high. That gap, between record visibility and a halved share price, is the whole story — and it comes down to what a backlog actually buys you.
A record backlog is contracted revenue, not a pipeline
A "backlog" sounds like a wish list, but in construction it is the opposite: signed contracts for work the customer has already committed to pay for. That is why the record matters. Sterling's combined backlog of $5.62 billion is roughly 1.4 times the $4.08 billion midpoint of its own full-year revenue guidance — more than a year of sales already under contract. Management pointed to a visible pool above $7 billion, with more than 90% of the e-infrastructure backlog tied to mission-critical work: data centers, semiconductor campuses, and manufacturing. Bookings in the quarter ran ahead of the work the company burned through, so the pile keeps building.
There is fine print, though. A meaningful share of that growth was bought, not earned. Combined backlog rose 150% year over year, but roughly 36% of that on an organic basis — the gap is the CEC electrical business and Stone Ridge acquisitions. The "record" is real. It is also partly purchased, which matters when you are deciding how much of the future is already spoken for.
Quick Backtesting Tool
The market is not arguing about backlog — it is arguing about the mix
If the near-term work is contracted, why did the stock get cut in half? Because two things the market actually cares about sit outside the backlog. The first is margins. Sterling's fastest-growing segment, e-infrastructure, boosted revenue 192% yet saw its operating margin slip to roughly 24% from about 28% a year earlier. That is not operational decay — it is accounting mix. The newly acquired electrical business grows faster but earns thinner margins than the legacy site-development work it is layered onto, so the blended number falls even while every piece improves. Management is counting on that electrical margin to climb toward 20% as it exits low-margin legacy contracts scaled up. That is a multi-year ambition, not a certainty.
The second is durability. Sterling's engine is the hyperscaler data-center build-out, and investors are asking the honest question for any builder in a supercycle: does the 2025-26 construction boom turn into AI revenue by 2027-28, or does it get digested first? Management itself flagged a "lull" in the third quarter as award timing gets lumpy. The backlog smooths the next two to three quarters; it cannot vouch for the years after.
What the reset leaves in the price
Here is where the numbers have meaning, not just the story. On trailing earnings the stock still carries a P/E near 34 and an enterprise value about 21 times EBITDA — the priciest of the major construction names on those trailing multiples, including MasTec, EMCOR, and Primoris. But Sterling also earns the highest return on invested capital in that group, near 32%, and is growing revenue vastly faster. On the company's own freshly raised guidance, the forward multiple lands near 24 times adjusted earnings — expensive if the growth ends this year, reasonable if the electrical margin ramp and the cycle cooperate.
The momentum read is the uncomfortable part. The stock sits below both its 50- and 200-day averages, with an RSI near 41 — a broken chart sitting underneath top-quartile fundamentals. Positive earnings revisions and record returns on one side, a downtrend on the other. That is the signature of a derating, not a collapse: the company got stronger and cheaper at the same time. AInvest's aggregate signal currently labels the stock a Buy, though that is an opaque cross-check, not the load-bearing evidence.
The two bets still in the price
The record backlog settles the near-term question: Sterling's "more growth" is contracted before the quarter ever starts. What you are deciding now is the premium for what happens after the contracts run — whether electrical margins climb as management plans, and whether hyperscaler spending endures long enough to refill the pipeline at these growth rates. Those two variables, not a price target, are the live ones.
In a factor framework, this is the aggressive-growth sleeve, and a concentrated one at that: it earns its keep only in a regime where AI infrastructure capex keeps compounding, and it hedges nothing on the downside. After the derating the shares trade for something closer to what the growth justifies than what the 2025 euphoria printed, which makes the near-term risk-reward more balanced than it was at the peak. But a derated price is not the same as a cheap one — the full value still rides on the margin ramp and the cycle both working. Those are the two things to watch, not the chart's rebound.
