$LMB: Limbach missed the data center boom. Is that the opportunity? | 1 Main Capital

Summary

• Yaron Naymark of 1 Main Capital is double-dipping on Limbach—first pitched in June 2023, up about 6.5x over the next 2½ years, now down roughly 75% from its peak but still up 80% from the original pitch. He has added substantially, arguing the selloff created a “triple whammy” of multiple expansion on the core business, value creation from M&A in a fragmented market where targets trade at 5-6x EBITDA, and a “cherry on top” if Limbach wins data-center work.
• The stock fell roughly 50% after Q2 while EBITDA guidance only fell from $90M to $80M; Naymark says the gap reflects guidance credibility more than fundamentals. H1 EBITDA was down about 30%, implying H2 growth of roughly 20% year over year, which “seems unrealistic” to the market. He thinks guidance is achievable, and even a disastrous $65M EBITDA print could produce roughly $45M of FCF, or about $4 per share, leaving valuation support on a clean balance sheet.
• Limbach deliberately missed the data-center boom—its exposure is effectively zero while some MEP peers have 30-60% of revenue there—and Naymark concedes it missed a major trend in hindsight. Andrew notes FIX is up roughly 800% and EME 250% over three years versus Limbach’s 16%. Naymark says data-center demand has inflated Limbach’s labor and materials costs while its core customers push back on price, squeezing margins from both sides.
• The Simcore acquisition ($30M for a data-center program-management business expected to produce about $4M of EBITDA) could provide an entry point into data centers. If its historical healthcare pull-through of roughly 20x is replicated, the dream case is a few hundred million dollars of data-center revenue. Winning $100M-$200M of data-center work next year could support $100M-plus, or even $120M, of organic EBITDA; Naymark sees scenarios where the stock doubles or triples over six to nine months.
• Against the short case that management knowingly bid low-margin backlog, Naymark says the gross-margin decline was primarily attributed to lower-margin Pioneer Power being consolidated and fewer project write-ups, while fixed-cost deleveraging further hurt EBITDA margins. Organic revenue fell roughly 5-6% while EBITDA fell 30%, which he finds explainable. H2 gross-margin expansion will test whether new bookings were knowingly lower margin. Even if Limbach becomes “general contracting in another name,” GC peers trade at 10-25x EBITDA amid the data-center tailwind.
• Management was genuinely surprised—on the Q1 call it said it was “comfortable with Q2 consensus,” which Naymark says got him and other longs in trouble—but CEO Mike McCann never sold a share even when worth roughly $40M on paper. The problem was backlog burning more slowly than expected because customers paused projects amid tariffs, macro concerns and war, or could not source electricians. Comparable public and private companies appear to have seen similar non-data-center softness, which may be starting to normalize.
• Naymark does not think Limbach is using the $50M buyback currently, preferring M&A: buying MEP targets at 5-6x EBITDA can add scale, diversification, operating leverage and a lower cost of capital, even if the stock itself trades around 6x EBITDA. His three-year math is roughly $10 per share of FCF by 2030 at 20x, or $200 versus about $40 today. Potential strategic buyers provide another backstop: EMCOR could be a consolidator, while Naymark doubts Comfort Systems would fit because it is largely nonunion.
• Zooming out, Naymark’s portfolio filter is to buy AI-neutral or AI-winner businesses not yet valued that way: Limbach has data-center optionality, IWG could benefit as uncertain headcount makes long leases less attractive, and KKR—reinitiated this year—could benefit as mega-alts with long track records receive “a pass for a bad vintage or two” while smaller mid-market firms consolidate away.

Deep dive

1. From Enron castoff to ODR story: the setup, round two

• Naymark’s history: Limbach is over 100 years old and an MEP contractor specializing in HVAC for mission-critical assets—hospitals, advanced manufacturing and similar facilities. It passed through Enron’s bankruptcy, private equity and a 2016 SPAC merger intended to roll up local contractors trading at “four to five times EBITDA. Now it’s maybe crept up to five to six.” After major new-construction project write-downs, it became distressed into COVID, then recovered through earnings growth, free-cash-flow generation and deleveraging.
• The transformation under Mike McCann, COO and then CEO from early 2023: shifting from general-contractor work with “blow-up risk” to owner-direct (ODR)—upgrades, retrofits, repairs and “adding a wing to a hospital.” The mix went from roughly 80% GC/20% ODR at IPO to about 75% ODR/25% GC last year, while EBITDA margins expanded from low single digits to low double digits.
• Andrew’s framing of the trade: pitched in June 2023, the stock rose about 6.5x over the next 2½ years, then fell roughly 75% from its peak while remaining up about 80% from the original pitch. Naymark is back in the stock and has added substantially.

2. Why down 50% when EBITDA guidance fell 12%? The credibility discount

• The air pocket: trade-war and tariff issues, “the Build Back Better bill” and its Medicaid cuts affecting healthcare, the Israel-Iran war, higher oil prices and broader macro concerns caused discretionary projects to pause. H1 organic revenue fell roughly 5-6%, while EBITDA dropped about 30% because the company retained a fixed-cost base it viewed as necessary for a temporary slowdown.
• Naymark’s decomposition of the selloff: with effectively no leverage, EBITDA down 30% could justify a stock decline of about 30%. The extra decline came from the H2 guide, which implied a roughly 30% H1 decline flipping to about 20% year-over-year growth in H2. That “seems unrealistic,” leaving little valuation support if investors expect further misses or guide-downs.
• His downside math, hedged but specific: “If they come out and print a 70 or a 65, the stock’s going lower for sure.” But $65M of EBITDA less $5M of stock compensation and $5M of capex leaves roughly $55M of pretax earnings and about $45M of FCF, or approximately $4 per share. At roughly 10x that figure, he believes the business is worth more than 10x earnings even after a major EBITDA haircut.

3. The bear case—panic-bid low-margin backlog—and the rebuttal

• Andrew presents a VIC short report and the broader bear thesis: management allegedly panicked during the summer 2025 order air pocket and accepted very low-margin bookings. Revenue guidance rose—from Andrew’s rough shorthand of $750M to about $780M—while EBITDA guidance fell, raising concern that management either knowingly bid low-margin work or did not recognize its economics.
• Naymark says he would have been more concerned if headline revenue had risen while margins fell substantially. Instead, organic revenue declined, the fixed-cost base delevered, and Pioneer Power was added—a lower-margin business that diluted consolidated margins. He says the company attributed the gross-margin reduction primarily to Pioneer and fewer project write-ups, while fixed-cost deleveraging further hurt EBITDA margins. He is “not certain” that new bookings were knowingly lower margin, and H2 gross-margin expansion will test that claim.
• His concession on what longs missed: as the business shifted toward ODR, investors became “probably a little overly dismissive of weak bookings” because short-duration, intra-quarter work did not always appear in backlog. “The bears turned out to be right over the short term.”
• The reframe on GC risk: “General contracting in another name sounds bad if you’re talking about Limbach,” but other GC stocks are trading at 10-25x EBITDA because of the data-center tailwind.

4. Missing the data-center boom cut both ways

• Andrew’s puzzle: FIX is up roughly 800%, EME roughly 250% and Limbach 16% over three years. Shouldn’t the rising tide of data-center demand still help a contractor focused on healthcare? Naymark’s answer: only “boats who are playing that tide.” Data-center demand raises labor and material costs, while core customers without the same demand uplift push back on price.
• Andrew extends it: ODR may have actively hurt Limbach as technicians could earn “$20 more per hour” on data-center work, leaving Limbach with wage inflation and owner relationships where it could not readily pass costs through. Naymark agrees that both wages and materials are rising.
• The strategic error, stated plainly: management was “vocal about not benefiting significantly from data-center business,” focusing on ODR while peers built 30-60% data-center exposure—“we effectively have zero.” Winning a fair share could support $100M-plus, or $120M, of organic EBITDA next year with acquisitions on top.

5. Management was surprised—and why Naymark still trusts them

• Andrew’s sharpest question: Q1 sounded like “a blip, everything’s under control,” but three months later guidance was slashed and 2026 was described as a reset year. Were they blindsided? Naymark: “Yes, I do think they were surprised.” Management had said it was “comfortable with Q2 consensus estimates,” which he says got him and other longs in trouble. He owned the stock through the Q2 blowup and added substantially afterward.
• The mechanism: bookings entered backlog but burned slowly. Customers paused voluntarily because of macro conditions, tariffs or potential war, or involuntarily because they could not source electricians and therefore could not begin the mechanical work. Naymark believes management has scrubbed the numbers and expects burn rates to improve in the back half, though that remains uncertain.
• Public and private competitors appear to have seen similar non-data-center softness over the past six to nine months, and Naymark says conditions seem to be starting to normalize.
• The character evidence: McCann worked his way up the company and, when the stock was $150, was worth roughly $40M on paper but did not sell a share. Andrew separately highlights chairman Josh Horowitz, a fellow small-value investor who owns a meaningful stake, and notes—while qualifying his recollection—that boards on which Horowitz served included BDMS, which sold to private equity at what Andrew recalls as a massive premium, another board that sold, and BKTI, which performed strongly.

6. M&A over buybacks, a $200 target and the takeout backstop

• Why the $50M buyback sits unused: Naymark does not think the company is currently executing it, because buying MEP targets at 5-6x EBITDA with little capex can add scale, diversification, operating leverage and a lower cost of capital. At $750M-$800M of revenue, Limbach remains small versus Comfort Systems at $11B-$12B, EMCOR at tens of billions and private companies at $5B-$8B, leaving room to grow through M&A.
• The three-year math behind his letter’s $200 target: “$10 a share of free cash flow by 2030… and if that trades for 20 times, there are 200.” He regards 20x as reasonable for a clean balance sheet in a durable end market with data-center tailwinds. Peer references include Comfort at 20-plus times EBITDA, EMCOR around 15x, newly public Legence around 13x and midsize players at 12-15x.
• The sale scenario: Naymark thinks EMCOR could be a consolidator and says it probably “kicked the tires” on Limbach more than a decade ago. Comfort is less likely because it is a merit-shop business with very few union employees, while Limbach and EMCOR are union shops. Hostile deals are difficult where talent can leave, but a formal sale process could attract buyers at a premium. An acquirer would also underwrite roughly $90M-$95M of EBITDA after removing public-company costs, rather than just $80M.
• On Andrew’s running joke about the “gravitational pull toward $10” for de-SPACs: Naymark cites QSR and APi Group as occasional SPAC winners and argues that Limbach “bucks the trend.” Andrew notes that the people associated with the 2016 de-SPAC are effectively gone, but Naymark still sees an enduring end market, a low multiple, cost-cutting levers, a clean balance sheet and the possibility of being either consolidator or consolidatee.

7. The broader filter: unpriced AI winners in physical businesses

• Naymark’s portfolio rule in the AI era is to avoid obsolescence risk and buy businesses that are AI-neutral or AI winners without being valued as such. Limbach is one example: data-center-exposed peers have faster growth and higher multiples, while Limbach retains data-center optionality and an M&A opportunity.
• IWG is viewed by some as an AI loser because “all office jobs are going away,” but Naymark sees a potential long-term benefit: if companies cannot forecast headcount ten years out, they may prefer short-term office rentals. Today, only a low-single-digit percentage of office space is rented short term, and he expects that percentage to rise.
• KKR, which Naymark reinitiated this year, has exposure through its portfolio and lending activities. He thinks mega-alts with long track records may receive “a pass for a bad vintage or two,” while smaller mid-market firms with fewer successful vintages consolidate or disappear.
• Andrew adds a proprietary-data angle: mega-alts possess decades of deal, diligence and ownership data that a new private-equity entrant would not have. He also questions whether AI makes passive investing harder by increasing the importance of avoiding AI losers.
• Andrew highlights additional KKR growth avenues: high-net-worth retail, underpenetrated alternatives allocations in Asia and Europe, and expansion in U.S. credit, infrastructure and real estate relative to Blackstone and Brookfield. Both conclude that these are the kinds of businesses that could remain durable over five, ten or twenty years.