LGN · Legence Corp. · Stock research

Current Thesis

The AI-datacenter MEP/HVAC pick-and-shovel story is still improving on paper — record $5.38B backlog booked into 2028, revenue doubling YoY, FY2026 guide raised to $4.1–4.3B (all confirmed on the 2026-05-14 Q1 print) — while the underlying hyperscaler buildout theme keeps accelerating. What has changed is the tape: from the $102.27 closing high (2026-05-06) LGN has fallen ~33% to $69.13 (2026-07-16 close, ~$68 intraday 2026-07-17), slicing through the $76 post-Q1 breakout shelf and rolling its 20-EMA over. The stock now trades below every mainstream price target, including Guggenheim's fresh $115 (2026-07-10) and Barclays' cautious $80 (2026-06-22). That is sell-side lagging price, which historically marks a de-rate the analysts have not yet written down. This is a broken structure, not a base — the fundamentals accelerate while the chart makes lower highs and lower lows. A momentum book stands aside here until the name proves a higher low; catching the falling knife into the July Blackstone supply window and an August print is the trap, not the setup.

Bullish and bearish views on Legence Corp.

The model's bull view on Legence Corp. (LGN), in brief: Q1 2026 (2026-05-14): revenue $1.04B, +105.2% YoY; adjusted EBITDA ~$118M, +132% YoY; net income $16M — the deck-slide thesis converted to booked revenue. The bear view: Price has decoupled from the story: ~33% off the 2026-05-06 closing high ($102.27) and ~36% off the $107.24 intraday high, with the $76 breakout shelf now lost and the 20-EMA rolled over — a defined downtrend since May, not a corrective pause. Both cases follow in full.

Bull Case

• Q1 2026 (2026-05-14): revenue $1.04B, +105.2% YoY; adjusted EBITDA ~$118M, +132% YoY; net income $16M — the deck-slide thesis converted to booked revenue.
• FY2026 guide raised the same day to revenue $4.1–4.3B / adjusted EBITDA $470–490M (from $3.7–3.9B / $400–430M) — a guide-up alongside the beat.
• Record backlog $5.38B as of Q1, +104% YoY, with data-center fabrication orders booked into late 2028; book-to-bill ~1.2x.
• Guggenheim raised its target to $115 (Buy) on 2026-07-10 from $106 — ~68% above the ~$68 spot; consensus targets cluster $93–103 with a majority Buy tape.
• 2026-05-28: $995M term loan amended, rate cut to SOFR + 2.00% — lenders re-rating the credit and trimming interest expense off a thin net line.
• Bowers acquisition added >$240M to Q1 revenue — the roll-up compounding on top of organic hyperscaler demand.

Bear Case

• Price has decoupled from the story: ~33% off the 2026-05-06 closing high ($102.27) and ~36% off the $107.24 intraday high, with the $76 breakout shelf now lost and the 20-EMA rolled over — a defined downtrend since May, not a corrective pause.
• Sell-side above spot on every target (Guggenheim $115 on 2026-07-10, Barclays $80 on 2026-06-22) while the stock keeps falling — analysts trailing the tape rather than leading it.
• Thin margins under a premium multiple: ~1.5% net, ~11–12% EBITDA versus equipment OEMs (VRT ~18% operating, ETN ~22%). At the May peak the ~19x EV/EBITDA was borrowed from fatter-margin businesses; a de-rate is that multiple normalizing.
• Recurring Blackstone supply: ~74% voting power post the 2025-09 IPO; serial secondaries already (Dec'25 ~$45, Apr'26 $54 across 15.4M shares for $831M). The ~90-day window from the 2026-04-08 secondary opened around 2026-07-07 — no fresh 424B surfaced as of 2026-07-17, but the bid bled ~19% through it.
• Backlog concentration is opaque; a 2026/2027 capex trim from MSFT/META/GOOGL/AMZN would hit a data-center-levered contractor directly.

Setup & Price Structure

Spot ~$68.24 (2026-07-17 intraday), $69.13 close (2026-07-16), inside a 52-week range of $26.96–$107.24. The all-time closing high is $102.27 (2026-05-06). The sequence since that high is lower highs and lower lows: ~$84 on 2026-06-23, ~$74 on ~2026-07-09 (implied by Guggenheim's 2026-07-10 "~55% upside" note), ~$68 now. The $76 post-Q1 breakout shelf — the prior line that defined the uptrend — has been lost by roughly a full $8, and the 20-EMA that ran through the low-$80s in late June has rolled over with price beneath it. The next visible structural support is the $54 April secondary strike, the last shelf where Blackstone supply was absorbed at a known price. There is no confirmed higher low and no positive divergence yet; oversold does not equal based. With price below Barclays' $80, Guggenheim's $115 and the $93–103 consensus, the divergence is a warning that the analyst community has not repriced — the opposite of the early-narrative edge this playbook hunts.

Catalyst Calendar (next 30 days)

• 2026-08-09 (confirmed): Q2 2026 earnings. Company guide revenue $1.05–1.1B / adjusted EBITDA $115–125M; consensus ~$1.07–1.08B revenue, EPS ~$0.32–0.36. Binary print — avoid fresh exposure in the 3 trading days ahead of it.

Elapsed catalysts

• ~2026-07-07 through late July (window open): Blackstone ~90-day distribution window from the 2026-04-08 secondary; watch for a fresh 424B / S-1 takedown. None confirmed as of 2026-07-17. (passed 23d ago)
• Rolling: sell-side revisions. Guggenheim $106→$115 (2026-07-10) is the most recent; a downgrade cluster becomes likely if price keeps diverging from a $93–115 target band. (passed 30d ago)

What Would Change Our Mind

• Bullish re-arm: a weekly close back above $76 with a confirmed higher low after basing three-plus weeks above $54 — ideally on a 2026-08-09 Q2 beat and FY guide raise — reclaims the broken shelf and turns the drawdown into a fresh entry setup rather than a knife.
• Bearish confirmation: a weekly close below $54 loses the April secondary shelf and turns the post-IPO uptrend into a value trap; a fresh Blackstone block the bid fails to absorb; Q2 revenue under the $1.05B guided floor or any cut to the $4.1–4.3B FY guide; a hyperscaler capex trim for 2026/2027.

Correlation Notes

• Same hyperscaler-capex factor as VRT, ETN, PWR and nVent — LGN moves with the data-center-infrastructure basket, so a theme-wide de-rate (NVDA-led AI/semis rotation) hits the whole MEP/electrical complex at once. Cap combined LGN+VRT+ETN exposure at ~2x a single-name limit.
• Idiosyncratic overhang: the Blackstone selling stockholder is a supply risk peers do not carry — LGN can underperform the basket purely on a secondary print.
• Thin margins make LGN more sensitive to project timing and labor cost than equipment OEMs, so it can lag the basket on execution even when the theme is intact.
• Theme-tag correction retained: the 2026-05-10 auto-tag "biotech-precision-therapeutics" was a discovery error — LGN is an MEP/HVAC data-center contractor.