Sterling Stock Plunges 35% in 3 Months: Is the Sell-Off Overdone?

Sterling Infrastructure shares have fallen 34.8% over three months despite Q2 revenue rising 90% year over year, adjusted EPS increasing 116% to a record $5.80, and adjusted EBITDA margin reaching 22%. E-Infrastructure revenue surged 192%, signed backlog rose 116% to $4.3 billion, and total signed, awarded and prospective future-phase work exceeds $7 billion. Management expects E-Infrastructure revenue to grow more than 100% in 2026, while 2026 consensus EPS has risen to $20.06, implying 84.4% growth; however, weak technicals, housing-related pressure, execution risks and a 20.9x forward P/E versus a 17.62x five-year median support a Hold stance.

Analysis

The relevant debate is not whether STRL has demand, but whether it can convert unusually concentrated, lumpy project exposure into cash earnings without giving back margin. Data-center civil/electrical work is typically front-loaded in labor, equipment and working capital; a rapidly growing backlog can therefore depress near-term free-cash conversion even as reported EBITDA accelerates. Any evidence of receivables build, unfavorable change orders, subcontractor inflation, or a falling book-to-bill ratio would likely matter more to the equity than another headline backlog increase.

STRL’s valuation discount to PWR is defensible only if its E-Infrastructure mix proves durable and repeatable across customers and geographies. PWR has broader utility and transmission exposure, while STRL is increasingly a higher-beta proxy for hyperscaler/semiconductor capex; that creates upside if AI-campus construction remains capacity constrained, but exposes it to abrupt project phasing or customer concentration. The acquisition-led capacity build also raises integration and bid-discipline risk precisely when competitors are likely chasing the same skilled electrical labor and equipment.

Near term, the stock needs a catalyst beyond favorable estimates: stable backlog after revenue conversion, sustained mid-20s E-Infrastructure margins, and cash conversion that supports growth without meaningful leverage. Over 6-18 months, successful prefabrication and fleet utilization could justify a structurally higher multiple because they shift value from commoditized site work toward scarce, schedule-critical delivery. Contrarily, the sharp drawdown may reflect a legitimate multiple reset from an unsustainably high growth rate rather than a technical dislocation; consensus targets are not an independent validation of project economics.

Key Decisions for Investors

• Initiate a starter long STRL only after the next earnings release confirms positive book-to-bill and no material deterioration in operating-cash-flow conversion; add if E-Infrastructure margin holds above 20% and management reiterates 2026 growth. Target a 6-12 month rerating toward PWR/MTZ-relative valuation; exit on a material guidance cut or a two-quarter decline in mission-critical backlog.
• Express the relative thesis as long STRL / short MTZ in equal dollar amounts for 3-6 months, rather than outright long STRL. STRL offers greater direct data-center operating leverage, while MTZ has similar valuation but more telecom and energy-infrastructure cyclicality; cover if STRL’s backlog conversion weakens or MTZ’s communications segment shows a clear recovery.
• Do not chase PWR as a sympathy beneficiary: its premium embeds superior execution and diversification. Use a long STRL / short PWR pair only following confirmation that STRL’s cash conversion and margins are stable; otherwise PWR remains the better-quality defensive exposure to electrification capex.
• Set an event-driven alert around customer concentration, working-capital days, acquisition consideration, and net-debt trajectory at the next filing. A move from net cash toward leverage without proportional EBITDA accretion would invalidate the thesis and can compress STRL below its historical valuation range.