Home

STRL — Deck

Sterling Infrastructure · STRL · NASDAQ

Sterling Infrastructure is a U.S. site-development contractor that pours the dirt and pads under data centers, highways, and Texas home foundations; data-center work now drives 59% of revenue and roughly 75% of operating profit.

$476
Share price
$14.6B
Market cap
$2.49B
Revenue (TTM)
14.8%
Operating margin
Listed since 1991; bottomed near $4 in mid-2016 during a six-year operational wreck, then compounded to $476 — a 100× run from the trough and +1,500% over five years.
2 · The math behind the multiple

The FY26 guide quietly demands 400 bps of E-Infra margin lift with no source in the data.

  • The guide implies ~18% op margin. The $13.45–$14.05 FY26 EPS range on $3.05–$3.20B revenue maps to a consolidated operating margin near 18%, up from FY25's 14.8% — a 370 bps step-up in twelve months.
  • CEC blends the other way. The deal documents disclose CEC at ~13% EBITDA margin versus E-Infrastructure at ~26%. CEC is EPS-accretive because it was bought with cheap capital — it is structurally margin-dilutive at the mix level.
  • Q4 already showed it. Consolidated op margin compressed from 16.6% (Q3) to 13.4% (Q4) in CEC's first full quarter. The Street filed it as seasonal mix; consensus EPS has since been cut 4% in 30 days while ratings stayed at 5/5 Buy.
Either core E-Infrastructure lifts from 23.6% to ~28% in one fiscal year, or the guide doesn't land. The 30× multiple is anchored to the first; the disclosed segment math points at the second.
3 · Six days to a binary

One week resolves more than the prior six months of debate; nothing material follows until August.

  • Apr 29 – May 1. Microsoft, Meta, and Amazon print Q1 capex guides. Any single hyperscaler easing 2026 capex language resets STRL estimates and the multiple — 70%+ of incremental dollars come from this cohort.
  • Mon May 4, after close. Sterling Q1 FY26. Five lines do the work: consolidated op margin (>14% bull / <13% bear), organic book-to-burn ex-CEC (>1.2× / <1.0×), CEC standalone segment margin, margin-in-backlog vs the 17.8% Q4 mark, and whether the FY26 guide is reaffirmed.
  • Thu May 7. Annual meeting — first chance for direct commentary on CFO permanence (three CFOs in twelve months) and on the CEO's 10b5-1 sale cadence after the April 23 tranche.
The window is the window. After mid-May the next hard date is the Q2 print in early August — three months of price action with no information.
4 · Best-in-class business, top-of-range price

Every quality dimension checks out; every valuation dimension says priced for the data-center cycle to extend indefinitely.

14.8%
Op margin PWR 5.7%, MTZ 4.6%
32.1%
Return on equity ROIC 17.3%; net cash
$3.0B
YE25 backlog +78% YoY (CEC ~$489M)
30×
EV/EBITDA 5-yr mean 11×; 20-yr mean ~7×

Sterling earns roughly twice the operating margin of PWR or MTZ at one-tenth the revenue, and finished FY25 in net cash even after a $443M cash deal for CEC. The 5-year FCF/NI ratio of 1.83× — most of the cash-quality premium the multiple is built on — is largely a $432M working-capital release from FY23–FY24 that already reversed to a $54M drag in FY25 as project size grew. EV/EBITDA reversion to the 5-year mean on flat EBITDA implies ~$170/share, a 64% drawdown with no recession required.

5 · The CEO is harvesting into the thesis

$94M of CEO selling in four months — more than the prior five calendar years combined.

  • Position cut roughly in half in 16 months. CEO stake fell from ~566K shares (YE24) to 290,593 after the April 23, 2026 sale at $497. He still owns $146M of stock — but the velocity is the signal, not the absolute level.
  • Plan adopted at the December peak. The 10b5-1 was put in place December 8, 2025, with a schedule that takes ~50% of the position out over 16 months. A CEO who structures a plan with that depth at that price is implicitly stating he is a willing seller here.
  • CFO carousel during record results. Three CFOs in twelve months — Ballschmiede retired, Villaverde terminated without cause with $591K "all other," Grindstaff interim since July 2025 — through a year of accelerating M&A and a goodwill-plus-intangibles ratio that jumped from 29% to 43% of assets.
Director and non-CEO executive ownership is 1.6% as a group. The only insider with a real position is the one selling fastest, into the same multiple expansion the bull case depends on.
6 · Bull and Bear

Watchlist — at 30× into a binary print six days out, the calendar does not require forced conviction.

  • For. Highest operating margin and ROE in the U.S. listed infrastructure peer set; net cash after the largest acquisition in company history; backlog up 78% with margin-in-backlog rising 110 bps to 17.8%.
  • For. Customer concentration narrowing — top-4 E-Infra share fell from 40% (FY23) to 27% (FY25); data-center revenue grew 125% YoY in Q3 FY25; PSU plan paid out at the 200% cap on real EPS performance, not handed out.
  • Against. 30× EV/EBITDA is two standard deviations above the 20-year mean; reversion to the 5-year mean on flat EBITDA implies $170 — a 64% drawdown with no recession.
  • Against. FY26 implied 18% margin path has no disclosed source given CEC at 13% EBITDA margin; CEO sold $94M in four months at the December peak; consensus EPS has been cut 4% in 30 days while ratings stayed unanimously Buy.
Lean Long if Q1 prints op margin >14% and organic book-to-burn ex-CEC >1.2× with CEC standalone margin lifting; Avoid if op margin <13% or book-to-burn <1.0×. Either resolves the deck on May 4.

Watchlist to re-rate: Q1 FY26 consolidated op margin vs Q4's 13.4% trough; organic book-to-burn ex-CEC; CEC standalone segment margin disclosure; CEO Form 4 cadence after April 23.