Sterling Infrastructure (STRL) Q2 2026 Earnings Call Transcript

Sterling Infrastructure reported Q2 revenue of $1.17B (+90% YoY, including $250.8M from acquisitions) and adjusted diluted EPS of $5.80 (+116% YoY), with adjusted EBITDA of $256.7M more than doubling and margins up 150bps to 22%. Signed backlog rose to $4.3B (+116% YoY) and combined backlog to $5.6B (+150%), while management raised FY2026 guidance to revenue of $4.0B–$4.15B and adjusted diluted EPS of $19.70–$20.30. The company increased CapEx guidance by $30M to $130M–$140M and expanded its $1.5B revolver (maturity extended to July 2031), alongside $35M of share repurchases in 1H 2026 at an average $511.18/share.

Analysis

STRL is less a one-quarter beat than evidence that mission-critical construction is becoming an operating system business: whoever controls scarce labor, equipment, and customer relationships can reprice mix upward even when headline revenue is noisy. The competitive moat is widening because the limiting factor is no longer demand, it is execution capacity; that favors integrated players with recruiting, prefabrication, and tuck-in M&A capability. Second-order winners are hyperscalers and semiconductor developers such as AMZN, GOOGL, and NVDA’s ecosystem, because a contractor that can compress schedules and absorb scope creep lowers project friction and helps preserve rollout timelines.

The near-term risk is not demand collapse but sequencing: a softer award quarter can trigger a backlog optics miss even while underlying pipeline remains intact. Over the next 1-3 months, the stock may trade on whether management can keep converting unsigned work and whether margin mix stabilizes as electrical grows faster than site development; if the market fixates on mix compression, that is the main setup for volatility. Over 6-18 months, the key falsifier is if labor scarcity or municipal permitting starts extending start dates enough to slow revenue burn, or if CapEx fails to translate into incremental throughput.

Consensus looks too focused on current-quarter margins and too skeptical on the durability of the backlog. The more important variable is that each acquired or newly staffed region creates a local franchise with recurring phase work; that makes the earnings stream more compounding than cyclical. I would treat any post-earnings weakness tied to backlog timing or margin mix as an opportunity, not a thesis break, unless Q3 awards materially miss and management cuts 2026/2027 visibility.

Key Decisions for Investors

• Buy STRL on any 3-5% post-earnings pullback over the next 1-2 weeks; thesis is that the market is underappreciating capacity-constrained revenue durability and phase-work expansion. Risk/reward improves if the tape reacts to backlog timing rather than fundamentals.
• Hold/add to STRL versus a housing-sensitive short such as ITB/XHB for the next 3-6 months; this isolates secular data-center/semi capex from the weak residential end-market, which is a separate drag and likely to remain soft.
• Set a watch item on STRL after the next award cycle: if Q3 awards are softer but Q4/early-2027 awards reaccelerate and backlog stays above roughly $5B combined, stay long; if combined backlog falls materially for two consecutive quarters, reassess the growth thesis.
• Use NVDA and GOOGL as indirect beneficiaries of the same infrastructure cycle, but only on dips and only as a broader AI capex basket overlay; do not chase them solely on this print because the linkage is second-order, not immediate.