LGN · Legence Corp.

Current Thesis

The AI-datacenter MEP/HVAC pick-and-shovel call has converted from sell-side deck slides into booked revenue. Q1 2026 (reported 2026-05-14) put up $1.04B revenue, +105.1% YoY against $506M a year ago, with adjusted EBITDA +132% and record awards of $5.4B. Management raised FY2026 guidance the same day. The old dossier framed the ~May print as the binary that would prove whether hyperscaler capex translates into LGN's order book; that binary resolved, and it resolved loudly. What changed since: the stock ran from the $54 April secondary strike to a 52-week high of $107.24, then rolled into a ~22% correction back to ~$84 (2026-06-05, -5.86% on the day). The theme is ACCELERATING, but the name's own tape is correcting so a fresh entry here is a pullback-in-uptrend, not a clean breakout, and the recurring Blackstone sell-down still caps how far it extends.

Bull Case

• 2026-05-14 Q1 print: revenue $1.04B, +105.1% YoY (vs $506M Q1'25); net income $16M; adjusted EBITDA ~$118M, +132% YoY with margins +~130bps on project execution.
• 2026-05-14 FY2026 guide RAISED: revenue to $4.1–4.3B (from $3.7–3.9B); adjusted EBITDA to $470–490M (from $400–430M). A guide-up printed alongside a beat is narrative acceleration, not maintenance.
• Record backlog $5.4B as of Q1, +104% YoY, book-to-bill ~1.2x; data-center fabrication orders booked into late 2028 multi-year revenue visibility a contractor rarely gets to show.
• 2026-05-14 Q2 guide set: revenue $1.05–1.1B, adjusted EBITDA $115–125M sequential growth guided, no air-pocket signaled into the next print.
• 2026-05-19: Tigress raised PT to $125 (from $85). Consensus sits ~$105 with roughly 19 Buy / 1 Hold (Strong Buy).
• Sell-side coverage cluster post-secondary: GLJ initiated Buy $99 (2026-04-21); Loop Capital initiated Buy $96 (2026-04-29); Goldman Buy $72 and Tigress $85 on 2026-04-16. New institutional coverage stacking inside ~6 weeks confirms the narrative is being adopted upstream.
• Bowers acquisition contributed >$240M to Q1 revenue the M&A roll-up engine compounding on top of organic data-center demand.
• 2026-05-28: $995M term loan amended, rate cut to SOFR + 2.00% lenders re-rating the credit; lower interest expense flows straight to a thin net line.

Bear Case

• Thin margins under the AI multiple: $16M net on $1.04B revenue is ~1.5% net margin; EBITDA margin runs ~11–12%. This is a labor/services contractor, while the ~$9B market cap (~19x forward EV/EBITDA) borrows a multiple from equipment OEMs with structurally fatter margins (VRT ~18% op, ETN ~22%).
• Recurring Blackstone supply: ~74% voting power post the 2025-09 IPO; serial secondaries already (Dec 2025 ~$45, April 2026 $54). The sell-down is a multi-year program, and each ~90-day window is a fresh supply event that caps parabolic extension. Next window watch ~2026-07-07.
• The easy money is spent: $54 secondary → $107.24 52-week high (+98%), now a ~22% drawdown to ~$84, still down 5.86% on 2026-06-05 with no confirmed higher low. The clearing trade and post-earnings rip are behind the tape; what remains is a correction that has not yet based.
• Backlog concentration is opaque: a $5.4B award figure can carry cancellation/timing risk tied to a handful of hyperscaler projects; outside visibility into customer concentration is limited.
• derate risk: any 2026/2027 capex trim from MSFT/META/GOOGL/AMZN hits data-center-levered contractors hardest because LGN has no revenue diversification buffer.

Setup & Price Structure

• IPO 2025-09-11 at $28; 52-week range $26.96–$107.24; market cap ~$9.04B (2026-06-05).
• $107.24 = post-Q1 52-week high; supply/resistance on any retest.
• ~$84 = current zone (2026-06-05, -5.86% on the day) actively correcting, no higher-low confirmation yet.
• ~$76 area = estimated rising 20-EMA / post-Q1 breakout shelf; a weekly close below flips the uptrend to broken.
• $54 = April 2026 secondary strike, deep structural floor / arb cost-basis only in play if the entire move unwinds.
• $105 consensus PT / $125 Tigress = upside targets if the pullback bases and the next print confirms the raised guide.
• Entry character: a MATURING-style pullback inside an accelerating theme, not a fresh ACCELERATING breakout. The disciplined add is a confirmed higher low / 20-EMA reclaim around $80–84 on rising volume, not a falling-knife grab while the name is down ~6% intraday.

Catalyst Calendar (next 30 days)

• ~2026-07-07 (est.): Blackstone supply-window watch ~90 days from the April 8 secondary; a fresh 424B/block could print. Supply risk rather than a guaranteed event, sitting at the edge of the 30-day window. Prior secondaries were absorbed at rising prices; a discounted block the bid can't absorb is the tell that flips supply from "demand" to "distribution."
• Ongoing: hyperscaler capex commentary and any data-center capex revision from MSFT/META/GOOGL/AMZN are the live upstream read-through between now and the next print.
• ~2026-08-13 (est.): Q2 2026 earnings (Q1 reported 2026-05-14 → Q2 ~mid-August) guided revenue $1.05–1.1B / adjusted EBITDA $115–125M; first test of whether the raised FY guide holds. Outside 30 days but the next scheduled binary.
• No company-scheduled catalyst confirmed inside the next 30 days; the July supply window is the only near-term mover.

What Would Change Our Mind

• Weekly close below ~$76 (rising 20-EMA / post-Q1 breakout shelf) → the pullback became a top and the uptrend is broken.
• Q2 2026 revenue below the $1.05B guided floor, or any FY2026 guide cut from $4.1–4.3B → the acceleration stalls and the rich multiple loses its justification.
• A fresh Blackstone 424B/block at a discount the bid fails to absorb (unlike the prior absorbed secondaries) → supply overwhelms demand.
• A hyperscaler capex guide-down on the next mega-cap prints → derate with no diversification buffer to cushion it.
• Bullish flip: a higher low that holds ~$80–84 and reclaims the 20-EMA on rising volume re-rates this from a probe to high-conviction continuation toward the $105–125 target band.

Correlation Notes

• Industrial-power-AI / data-center bucket peers: VRT (Vertiv), ETN (Eaton), PWR (Quanta), nVent, plus the power complex (CEG/VST/TLN). LGN trades as a high-beta contractor proxy on data-center capex same factor, thinner margins, more single-name supply risk.
• Cluster check: when VRT/ETN/PWR break out together the theme is healthy; if equipment/power peers roll while LGN holds, suspect a single-name Blackstone supply effect rather than theme strength.
• Concentration discipline: keep any combined LGN + VRT + ETN exposure inside ~2x a single-name limit all three re-rate and derate on the same hyperscaler-capex headline, so stacking full size triples one factor bet.
• Upstream signal: a hyperscaler capex-cut headline hits LGN harder than the diversified OEMs the MSFT/META/GOOGL/AMZN capex lines are the leading tell.