Lennox International Inc. (NYSE: LII) — Institutional Research Memo

Independent fundamental equity research. As-of date: June 8, 2026. Target: Lennox International Inc. | NYSE: LII | CIK 0001069202 | GICS: Industrials → Building Products & Equipment (HVACR) Date: June 8, 2026 | Price (ref.): $517.78 (2026-06-08; close $508.43 on 2026-06-05) | Market cap: ~$18.0B | EV: ~$19.4B Fiscal year-end: December | HQ: Richardson, Texas

⚡ Claude’s Take

This is the author’s own subjective opinion and general information, not investment advice. It is the single place in this article where a position and a directional valuation zone are taken; the analysis that follows is deliberately position-free and carries no recommendation and no price target.

Verdict: HOLD / quality-compounder-at-the-wrong-part-of-the-cycle — not a short, not a buy here. Accumulate on weakness toward the low-$400s (≈ the 52-week low of $434), where you are paying for a normalized high-teens-margin earnings base rather than a refrigerant-transition peak; fair-to-fully-valued around the current ~$500–570 base-case zone where it trades today; demanding-to-rich above ~$600 until positive residential unit volume proves the 20% margin is structural. Conviction: medium.

Tag: “A fine compounder nursing a refrigerant hangover — wait for the volume, not the price.”

Lennox is a genuinely good business hiding inside flattering optics. The honest profitability signal — ROIC of ~33% (≈41% pre-tax; ~29% capitalizing operating leases) — is excellent and well above its cost of capital; the earnings are clean and cash-backed (the FY25 dip in cash conversion is a one-time A2L inventory build, not an accrual game); the balance sheet is conservatively levered (~1.2x net debt/EBITDA, ~1.5x capitalizing leases) despite a buyback-hollowed equity base that makes the headline ~76% ROE meaningless; and capital allocation is disciplined-to-good, with a 16-year dividend-growth streak and a ~37% share-count reduction since FY10. The industry is structurally attractive — a consolidating top-five oligopoly feeding off a non-discretionary, ~63%+ replacement demand base, with regulation (SEER2, then the AIM Act refrigerant phase-down) functioning as a recurring license to raise prices. What the market is pricing correctly is that the record 20.0% operating margin (up from 13.9% in FY22) was driven more by price than by durable cost-out, during a one-time forced refrigerant transition, while organic unit volume has been collapsing (residential volume −17% in FY25, −21% in Q1-2026). FY25 revenue actually fell 2.7%; the stock is ~25% off its high; and an unusually large ~8.4% short interest is, on a clean filing record, a coherent peak-margin/pre-buy-payback bet rather than a forensic one.

The reason this is a HOLD and not a short is that the bull optionality is real and the downside is buffered: if BCS (commercial, +17% organic volume in Q1-2026) keeps carrying the mix, the Samsung ductless JV scales, and HCS unit volume troughs and turns in H2-2026 with margins intact, the modest ~22.7x trailing / ~19.4x forward multiple is too low for a 30%-ROIC compounder — and the crowded short becomes squeeze fuel. The reason it is not a buy here is that the base case sits modestly above spot, the scenario spread is near-symmetric (~low-$300s bear / ~$540–570 base / ~$775–800 bull), and on today’s evidence the burden of proof (Marathon’s mean-reversion prior, the Q1-2026 −390 bps HCS margin crater, the reaffirm-don’t-raise guidance) is unmet. You are being asked to pay a fair price for an unproven assumption. Wait for the volume. The single piece of evidence that would flip me bullish: two consecutive quarters of positive HCS organic volume with operating margin sustained ≥19%. The single piece that would flip me bearish (toward “avoid / short candidate”): another quarter or two of HCS margin compression below ~18% on still-negative volume, especially if a major (Daikin/Goodman or Carrier) breaks pricing to chase share.

1. Executive Summary

Lennox International is a North-America-focused (≈93% U.S., ≈7% Canada) pure-play HVACR manufacturer and distributor that, since exiting Europe in late 2023, reports two segments: Home Comfort Solutions (HCS, residential; ~64% of sales) and Building Climate Solutions (BCS, light-commercial + Heatcraft refrigeration). The revenue base is replacement-dominated and equipment-led — structurally attractive because ~63%+ of installs are non-discretionary, failure- and weather-driven replacements of a ~15–20-year installed base, not credit-sensitive new construction. Its most distinctive asset is a company-owned, direct-to-dealer residential distribution network (200+ Lennox Stores; 76% of HCS revenue sold direct), which captures a distribution-margin layer that Carrier and Trane cede to independent distributors like Watsco.

The central fact pattern is a paradox: margins at a record while volume collapses and the stock de-rates. Operating margin expanded ~610 bps in three years (13.9% FY22 → 15.9% → 19.5% → 20.0% FY25) and diluted EPS rose to $22.79, yet FY25 revenue fell 2.7% to $5.20B, the stock trades ~25% below its 52-week high of $689, and short interest is elevated at ~8.4% of float. The resolution: the FY24 R-410A refrigerant pre-buy (~$125M of pulled-forward demand) and the January-1-2025 AIM Act transition to low-GWP A2L (R-454B) refrigerant let Lennox push ~+9–10% price/mix that nearly fully offset a −13% consolidated (−17% residential) volume decline. Reported revenue and margin held up on price, not units — the most reversible kind of “growth” — and the cushion is already fraying: in Q1-2026, HCS organic volume fell −21% and HCS segment margin compressed ~390 bps to 13.3% as price could no longer carry the volume loss.

Business quality is real but the moat is narrower than the brand implies. The ~33% ROIC clears the moat test on a current-returns basis, and the local-scale + dealer-captivity advantages of the owned residential channel are genuine. But management’s own 10-K calls its distribution a “low-barrier business,” Lennox sits behind/alongside Trane and Carrier (with Daikin/Goodman a scale-and-price threat at the value/ductless end), and — tellingly — FY25 growth and margin leadership have shifted to the commercial BCS segment, which does not run on the residential owned-store moat. The financials are high-quality and capital-light (capex ~2.3% of sales, ~95–100% FCF/NI over a cycle; the FY25 dip to 79% is a deliberate A2L inventory build, not an accrual flag), and capital allocation is disciplined — though management deployed its largest buyback in years (~$482M at ~$536/share) into a cyclical/valuation peak before cutting repurchases to ~$20M in Q1-2026.

Valuation embeds a single load-bearing bet. At ~$517.78 (~22.7x trailing EPS, ~17x EV/EBITDA, ~3.5% FCF yield), Lennox is the cheapest name in a premium HVAC group precisely because the market doubts the durability of the 20% margin. A no-growth earnings-power value is only ~$230/share, meaning the majority of the price pays for margins-hold-plus-growth-plus-buyback-compounding. The 83rd-percentile price/sales (the market has re-rated to the new margins) sitting beside the 30th-percentile P/E (it discounts the earnings multiple on doubt) is the quantitative signature of the unresolved debate. Consensus is a Hold (~5 Buy / 10 Hold / 2 Sell, average target ~$575), underwriting roughly flat earnings on ~8% revenue growth — i.e., it has stopped paying for further margin expansion. Our read is modestly more cautious than consensus on earnings quality: the ~20% operating margin should be modeled as a peak-of-transition figure to stress-test, not a base to extrapolate, and the air-pocket may run longer than the H2-2026 trough the guide implies. This is a genuinely good business at a fair multiple where the entire risk/reward hinges on one unproven assumption — margin durability — and, on current evidence, the burden of proof is unmet. (No recommendation or price target is expressed in this body.)

2. Business Overview

What Lennox is. Lennox International (NYSE: LII) is a North-America-focused manufacturer and distributor of heating, ventilation, air conditioning and refrigeration (“HVACR”) equipment, founded in 1895 in Marshalltown, Iowa around Dave Lennox’s riveted-steel coal furnace [FACT — FY2025 10-K, Item 1 “The Company”, l.149]. After exiting Europe (4Q23) and reorganizing into two reportable segments (effective FY24), it is now effectively a pure-play North American HVAC business: ~92.8% of FY25 revenue was United States, ~7.2% Canada, and international is essentially nil [FACT — FY2025 10-K, revenue-by-geography footnote: US $4,821.9M, Canada $373.4M of $5,195.3M total, l.3463-3489]. This geographic concentration is a deliberate post-divestiture simplification, not a weakness to gloss over — it removes the loss-making European exposure but also removes a structural growth lever and leaves the company fully levered to the U.S. residential and light-commercial replacement cycle.

The business sells replacement-dominated, equipment-led revenue. Management states “a substantial portion of the sales in each of our business segments is attributable to replacement business, with the balance comprised of new construction” [FACT — FY2025 10-K, Item 1A risk factor, l.749]. Lennox does not quantify the split in the 10-K; industry convention and the company’s prior disclosures place residential replacement at roughly two-thirds-plus of demand [INTERPRETATION — corroborated by the replacement-heavy framing in the risk factors; exact ratio is an OPEN QUESTION]. Replacement demand is non-discretionary and tied to the ~15-20 year life of an installed unit, which is the single most attractive structural feature of the revenue base; new-construction demand (the cyclical minority) is explicitly tied to interest rates, financing and consumer confidence “over which we have no control” [FACT — FY2025 10-K, l.357]. Revenue is recognized at a point in time on equipment shipment, not over time — there is essentially no contracted/subscription backlog [FACT — FY2025 10-K, revenue recognition note, l.3568].

Reportable Segments (FY24 reorganization)

Since FY24 Lennox reports two segments. Home Comfort Solutions (HCS) is the residential business — furnaces, air conditioners, heat pumps, packaged units, indoor-air-quality and comfort-control products, parts and supplies, sold under Lennox, Dave Lennox Signature Collection, Armstrong Air, Ducane, AirEase, Concord, MagicPak, ADP, Allied, and the newly-acquired Supco/LINEBACKER brands [FACT — FY2025 10-K, segment table, l.156-160]. Building Climate Solutions (BCS) is the light-commercial and refrigeration business — rooftop units (2-30 tons), split/air-handler systems, VRF, controls, National Account Services (install/service/maintenance), the AES curb/recycling business, and the legacy Heatcraft Worldwide Refrigeration line (condensing units, unit coolers, refrigeration racks for supermarkets, cold storage and data-center process cooling), under brands including Lennox, Model L, Energence, Raider, Lennox VRF, Allied Commercial, Duro Dyne, AES and Heatcraft/Bohn/Larkin [FACT — FY2025 10-K, segment table, l.161-164]. The pre-2024 “Corporate and Other” segment housed the divested European operations (manufacturing in Germany, France, Spain), which generated $247.6M of revenue in FY23 before the 4Q23 sale and zero thereafter [FACT — FY2025 10-K, segment footnote and l.202-203, 2858].

Segment financials (from the Note 3 segment footnote and segment MD&A):

(All figures $M; FACT — FY2025 10-K Note 3, l.2776-2887, and MD&A segment tables l.964-1025, 1108-1166. “Corporate & Other” carries unallocated corporate costs post-Europe-exit, hence the persistent reported “loss.”)

Two things in this table matter for the thesis. First, BCS is now the higher-margin segment (23.4% vs HCS 21.8% in FY25) and the only one still growing — BCS revenue rose 5% in FY25 (price/mix +8%, Duro Dyne +2%, organic volume −5%) while HCS fell 7% [FACT — FY2025 10-K MD&A, l.991, 1024]. The commercial book is doing the heavy lifting on growth and margin, which complicates the “residential moat” narrative the bull case leans on. Second, the FY24→FY25 HCS line is the clearest fingerprint of the refrigerant pre-buy: HCS net sales fell 7% on a −17% volume / +10% price-mix split [FACT — FY2025 10-K MD&A, l.991], with FY24 having previously risen 11% on +7% volume / +4% price-mix [FACT — FY2025 10-K MD&A, l.1135]. Volume swung from +7% to −17% in a single year. That is not normal HVAC volatility; it is demand pulled forward into 2024 ahead of the R-410A-to-A2L (R-454B, GWP ≤700) refrigerant transition that became effective January 1, 2025, then paid back in 2025 [FACT — refrigerant transition effective 1/1/25, FY2025 10-K Environmental Regulation, l.268; INTERPRETATION — the volume reversal is the pre-buy payback]. Critically, HCS segment margin still rose (21.3%→21.8%) despite the volume collapse, because +10% price/mix more than offset the −17% volume [INTERPRETATION]. The durability of that price/mix is the central question for the rest of the memo (and it is already fraying: in Q1-2026 HCS organic volumes were down ~21% and HCS segment margin compressed ~390bps to 13.3% as price/mix could no longer fully carry the volume loss — [FACT — LII Q1-2026 results, 8-K/press release, sec.gov filing 000106920226000052, reported 4/29/2026]).

Distribution Model — the structural differentiator

Lennox’s most distinctive feature is its hybrid distribution architecture, and management itself frames it as the competitive edge. In the North American residential market it uses three channels [FACT — FY2025 10-K, l.213-216]:

1. Company-owned, direct-to-dealer (“Lennox” brand): Lennox manufactures and sells the flagship Lennox-branded equipment directly to independent installing dealers, bypassing the independent two-step distributor layer. It operates a network of Lennox Stores across the U.S. and Canada — third-party location trackers count ~245 U.S. sites (commonly described as “200+” owned stores) plus regional distribution centers — giving contractors local access to equipment, OEM and universal parts, and supplies [FACT — FY2025 10-K, Item 1, describes “a network of Lennox Stores across the U.S. and Canada” but does not enumerate a count; the ~245-store / ~30-DC figures are third-party (LennoxPros “About Stores”, ScrapeHero), accessed June 2026 — INTERPRETATION]. This makes Lennox “one of the largest wholesale distributors of residential HVAC products in North America” — i.e., it owns the distribution margin a competitor cedes to an independent distributor.
2. Independent two-step distribution (Allied Air / “value” brands): Armstrong Air, AirEase, Concord, Ducane, MagicPak, ADP, Duro Dyne and Supco are sold through independent wholesale distributors who resell to contractors — the conventional Carrier/Trane model, including private-label OEM supply [FACT — FY2025 10-K, l.175-176, 214].
3. Direct sales to end-users / national accounts (concentrated in commercial).

Direct sales (channels 1 and 3 combined) were 76% of HCS revenue in FY25 (74% FY24, 75% FY23); independent distributors the remainder [FACT — FY2025 10-K, l.567]. In BCS, equipment was 80% of revenue (80% FY24, 86% FY23) with the remainder from the growing service book — the service mix is rising as NAS and AES scale [FACT — FY2025 10-K, l.571-573]. The commercial segment relies on conventional commercial contractors and direct national-account selling, not company-owned stores.

Customer types are therefore: independent installing dealers/contractors (the core residential customer, served direct or via distributor), independent wholesale distributors (value brands), and commercial/national-account end users (light-commercial buildings — offices, restaurants, retail, schools — plus refrigeration end-markets: supermarkets, cold storage, food service, and increasingly data-center process cooling) [FACT — FY2025 10-K, l.179-189]. No single customer was ≥10% of consolidated revenue in FY23-25 — customer concentration is low [FACT — FY2025 10-K Note 3 footnote (2), l.2891].

Recent structural moves (FY23-25)

The portfolio has been actively reshaped: the 4Q23 European divestiture (removing ~$248M of low/negative-margin revenue); the 2024 Samsung JV (49.9%, distributing Samsung ductless mini-splits and “Lennox powered by Samsung” products through Lennox Stores — a capital-light entry into ductless, where Lennox historically under-indexed vs Daikin) [FACT — FY2025 10-K, l.198-200]; the Ariston JV (49.9%, water heaters through Lennox Stores — a cross-sell/attachment play) [FACT — l.195-197]; the AES acquisition (Oct-2023, commercial service/recycling, in BCS); and the Oct-2025 Duro Dyne + Supco acquisition (HVAC parts/supplies, split across both segments) [FACT — l.186, 191-192]. The direction is consistent: shed low-margin geography, deepen the North American parts/supplies/service attachment around the installed base, and use JVs to add adjacent product (ductless, water heaters) without owning the manufacturing.

Manufacturing & inputs. Lennox manufactures in North America (with a new commercial factory in Mexico ramping), runs lean/Six-Sigma, and notes that several products are made at single-location facilities — a concentration/operational risk [FACT — FY2025 10-K, l.217-218, 341]. Key purchased components are compressors, motors and controls; key raw materials are steel, aluminum and copper — exposing gross margin to commodity and (post-2025) tariff inflation [FACT — FY2025 10-K, l.220].

Employees. As of December 31, 2025 Lennox employed approximately 12,900 people — ~8,500 in the U.S. and ~3,200 in Mexico; of these ~5,400 were salaried and ~7,500 hourly, with ~3,400 union-represented [FACT — FY2025 10-K Human Capital, l.244-245]. The third-party “5,400 employees” figure is wrong — it captured only the salaried subset; total headcount is 12,900 [OPEN QUESTION resolved — third-party data-feed error].

Verdict

This is a structurally attractive, well-run North American HVAC pure-play with a replacement-heavy, equipment-led revenue base and a genuinely differentiated owned-distribution channel — but the headline financials are presently distorted and the growth picture is more nuanced than the brand narrative implies. The replacement skew and the company-owned direct-to-dealer model are real, durable features that should produce above-peer margins through a cycle. However: (i) FY25 revenue fell 2.7%, and the decline is overwhelmingly a residential-volume phenomenon (HCS volume −17%) that is the payback of a 2024 refrigerant pre-buy, not a secular deterioration — but it is still ongoing (HCS organic volume −21% in Q1-26); (ii) the growth and margin engine has quietly shifted to the commercial/refrigeration segment (BCS), the part of the business that does not run on the owned-store moat; and (iii) the reported margin resilience to date rests on price/mix offsetting volume, a lever that visibly weakened in early 2026. The business overview supports the bull case on quality and structure, but it equally flags that FY24-25 earnings are pre-buy-flattered and that the moat lives in the residential channel even as the growth lives elsewhere. Both of those tensions are carried into the Competitive Position, Financial Quality and Valuation sections.

3. Industry Dynamics

The investment case for Lennox rests less on company-specific brilliance than on the structure of the North American HVAC industry it operates in. That structure is unusually favorable — a consolidated oligopoly, a largely non-discretionary replacement demand base, and a decade of regulation that has serially forced the installed base to be replaced with higher-priced equipment. Understanding how much of Lennox’s 600+ bps of operating-margin expansion (13.9% in FY22 → 20.0% in FY25; verified baseline) is industry rent versus company self-help requires first understanding the industry. The conclusion below is that the industry is genuinely good — but that recent results have also been flattered by a one-time regulatory pull-forward that is now reversing.

3.1 Market structure, size, and the oligopoly

North American HVAC is a large, mature, consolidated market. Third-party estimates put the North America HVAC market at ~$55.1bn in 2024, growing ~4.5% with projections to ~$85.9bn by 2032 (deallab, North America HVAC Market 2024); the broader U.S. HVAC market (including services) is sized at ~$65.2bn in 2025 growing ~4.9% (Mordor Intelligence, United States HVAC Market, 2025). [FACT — third-party market sizing; estimates vary by definition and should be treated as order-of-magnitude.] The U.S. residential unitary equipment slice specifically was ~$4.7bn in 2024 (GMInsights, U.S. Residential HVAC Market, 2025). [FACT]

The supply side is a textbook oligopoly. The same third-party source estimates 2024 North American market shares as follows:

Source: deallab, North America HVAC Market 2024 (accessed 2026-06-08). [FACT — third-party estimate; share figures are directional, not audited, and exclude private players Rheem/Paloma and Bosch, which Lennox itself names as significant competitors.] Lennox’s own FY2025 10-K names the identical competitive set verbatim — Carrier, Trane, Paloma (Rheem/Ruud), Bosch (York/Luxaire/Coleman), Daikin (Goodman/Amana), and Madison Industries in residential; Carrier, Trane, Daikin, CSW Industrials and AAON in light commercial; Hussmann, Emerson/Copeland, Carrier, GEA and others in refrigeration (FY2025 10-K, Item 1, “Competition”). [FACT]

The structural takeaway: this is a handful of well-capitalized incumbents, most either public conglomerates or units of large strategics (Daikin is a ~$30bn-revenue Japanese parent; Bosch private). Lennox is the smallest of the five majors and the only N.A.-pure-play of comparable scale — a double-edged fact explored below. The market is consolidating further, not fragmenting: the top five gained combined share in 2024, and management commentary across the group emphasizes M&A-driven consolidation (deallab, 2024). [INTERPRETATION] In Greenwald’s framework this points to economies of scale plus customer captivity as the operative advantage type for the industry as a whole — scale in manufacturing/distribution, captivity through the dealer/contractor channel (the Lennox-specific moat is developed in the Competitive Position section).

3.2 The profit pool: replacement is the prize, and it is non-discretionary

The most attractive feature of this industry is where the demand comes from. Roughly 63% of U.S. HVAC equipment installations in 2024 were retrofit/replacement, with new construction ~37% (GMInsights, 2024). [FACT] Other industry estimates skew the replacement share even higher (~75–85%) once the installed-base math is run: there are ~130mn installed U.S. residential HVAC systems with a typical ~6% annual replacement rate (~8mn units/yr) versus only ~1–1.5mn new-construction units/yr (industry estimate via market research summaries, 2025). [FACT — estimate; the precise split varies by source and by residential vs. commercial mix.] Lennox’s 10-K confirms it serves “both the residential replacement and new construction markets” and that replacement parts, supplies and the Lennox Stores network are a core part of the model (FY2025 10-K, Item 1). [FACT]

Why this matters: replacement demand is non-discretionary and weather-driven, not credit- or sentiment-driven. When a 15–20-year-old furnace or compressor fails in a Texas July, the homeowner replaces it within days — the only choice is brand and efficiency tier, not whether to buy. Lennox itself states the industry “ships roughly twice as many units during June as it does in December,” and that cooler summers/warmer winters directly depress sales (FY2025 10-K, Item 1, “Seasonality”). [FACT] This makes the replacement portion of revenue far more resilient than the new-construction portion, which Lennox explicitly ties to interest rates, financing availability and consumer confidence — “cyclical factors over which we have no control” (FY2025 10-K, Item 1A). [FACT] The profit pool is therefore split between a sticky, recurring, semi-annuity replacement stream and a smaller, genuinely cyclical new-build stream. The replacement stream is the structural prize and the reason HVAC majors earn high-teens-to-low-20s operating margins through cycles. [INTERPRETATION]

3.3 Regulation as a pricing engine — SEER2 and the AIM Act refrigerant transition

The single most important industry-specific dynamic is regulation, which in HVAC functions less as a cost burden than as a recurring, mandated price-up event. Two regulatory waves dominate the recent period:

SEER2 (DOE efficiency standards, 2023). The Department of Energy raised minimum efficiency standards effective January 2023, forcing redesign of the entire residential lineup. Lennox states it “successfully updated our lines of HVAC equipment to comply with the latest energy conservation standards” in 2023 (FY2025 10-K, Item 1, “Energy Efficiency”). [FACT] Higher-efficiency equipment carries higher price tags — a structural tailwind to average selling prices.

The AIM Act / EPA refrigerant phase-down (the big one). The 2020 bipartisan American Innovation and Manufacturing (AIM) Act gave the EPA authority to phase down high-GWP hydrofluorocarbons. Effective January 1, 2025, most new commercial and residential HVAC products must use refrigerant with a global-warming potential ≤700, forcing the transition from R-410A to low-GWP A2L blends (principally R-454B) (FY2025 10-K, Item 1, “Refrigerants”). [FACT] This is a hard, dated cut-over of the entire product catalog — and it confers pricing power. Industry pricing data confirm the magnitude: Carrier’s CEO stated R-454B base pricing runs ~10% above R-410A, escalating to 15–20% over two years; Lennox’s own A2L systems are priced “over 10% more” than the R-410A equivalents they replace (ACHR News, Less R-410A, Higher Equipment Prices, 2025). [FACT] Lennox CEO Alok Maskara estimated ~65% of 2025 end-market demand would shift to low-GWP product (vs. Trane’s CEO at ~75%) (ACHR News, 2025). [FACT — management estimate, treat as hypothesis.]

The 2024 pre-buy and the 2025 payback. Anticipating the higher-priced A2L cut-over, the channel (distributors, dealers, contractors) over-ordered R-410A equipment in 2024 to bank cheaper inventory — a classic pre-buy. The financial fingerprint is unambiguous in Lennox’s own segment bridges:

Source: FY2025 10-K, MD&A, “Results by Segment.” [FACT] The +7% resi volume in 2024 (pre-buy build) reversing to −17% volume in 2025 (pre-buy payback plus underlying softness) is exactly the pull-forward/payback signature. Consolidated FY2025 net sales fell 3%, decomposed by management as −13% volume, +9% price/mix, +1% from the Q4 Duro Dyne/Supco acquisition (FY2025 10-K, MD&A, “Consolidated Results”). [FACT] Critically, price/mix more than offset the volume collapse at the margin line: HCS segment margin actually rose 50 bps (21.3%→21.8%) despite losing 17 points of volume, because $256mn of favorable price/mix swamped $224mn of lost volume profit (FY2025 10-K, MD&A). [FACT] This is the clearest evidence that the recent margin story is price-led, and that the pricing power is regulation-enabled. AHRI shipment data corroborate the industry-wide volume normalization: U.S. AC shipments were −13% YoY in May 2025 with a 3-month rolling average of ~−8%, while heat pumps (+5.9%) and furnaces (+15.7%) rose — a mix shift consistent with A2L cut-over churn (HARDI, AHRI Shipment Report May 2025, Aug 6 2025). [FACT]

A regulatory tail risk now sits over the model: in 2025–2026 the EPA faced industry pressure (HARDI flagged an ~$8bn stranded-inventory concern) to extend the R-410A installation/sell-through deadline, which could either prolong the transition or strand inventory (ACHR News / HARDI commentary, 2025–2026). [OPEN QUESTION — the precise final EPA sell-through timeline and any reversal materially affect the 2026 demand bridge and should be tracked.]

3.4 Input costs, tariffs, and the cost side

The cost side is commodity-exposed but manageable. Lennox states steel, aluminum and copper account for the bulk of raw-material purchases, while compressors, motors and controls are the largest component buys (it sources compressors partly through a JV) (FY2025 10-K, Item 1, “Strategic Sourcing”). [FACT] Tariffs are a live overhang: the company flags U.S. tariffs on Canada/Mexico goods not covered by USMCA, and notes the USMCA is up for review/renewal in 2026 — directly relevant given Lennox manufactures in Mexico (new commercial HVAC factory) (FY2025 10-K, Item 1A). [FACT] In FY2025 the cost side was a partial headwind: gross margin rose only 10 bps (33.3%→33.4%) as +290 bps of price/mix was nearly offset by −160 bps higher product costs and −120 bps higher freight/distribution (FY2025 10-K, MD&A). [FACT] The structural point: HVAC majors have historically passed commodity inflation through with a lag via annual price increases — pricing power that the regulation-driven cut-overs reinforce. [INTERPRETATION]

3.5 Distribution — the channel is the moat-maker

HVAC distribution is two-tiered and the channel structure is where industry economics get captured. Most manufacturers sell through two-step independent distribution (manufacturer → distributor → installing contractor), where the dominant independent is Watsco (WSO), a pure-play distributor with no comparable in-house manufacturing — a model that lets it aggregate multiple OEM lines. Lennox is structurally different: it runs three channels simultaneously — (1) company-owned direct-to-dealer for the flagship “Lennox” brand via ~250+ owned Lennox Stores across the U.S./Canada; (2) traditional two-step independent distribution for Armstrong Air, Ducane, AirEase, Concord, MagicPak, ADP, Duro Dyne and Supco; and (3) direct sales to end-users/national accounts in commercial (FY2025 10-K, Item 1, “Distribution”). [FACT] Direct sales were 76% of revenue in 2025 (74% 2024, 75% 2023), with independent distributors the remainder (FY2025 10-K, Item 1). [FACT] Lennox describes itself as “one of the largest wholesale distributors of residential HVAC products in North America” (FY2025 10-K, Item 1). [FACT — management characterization.]

This owned-distribution model is the operative competitive-advantage candidate (developed in the Competitive Position section): by capturing the distributor margin and owning the dealer relationship, Lennox internalizes a profit layer that Watsco earns independently and that Carrier/Trane share with their distributors. It is also a barrier to entry — replicating a national owned-store/direct-dealer network is capital- and time-intensive. [INTERPRETATION]

3.6 Barriers to entry and the capital cycle (Marathon lens)

Barriers to entry are moderate-to-high in equipment, low in parts/service. Lennox concedes that in “parts and supplies, distribution, and service of commercial heating and cooling equipment, barriers to entry are lower,” producing highly competitive, fragmented markets; it also warns that “manufacturing operations in low-cost countries could provide cost advantages to existing and emerging competitors” (FY2025 10-K, Item 1A). [FACT] In core equipment, however, barriers are real: established brands, DOE-certified product portfolios that must be re-engineered with each regulatory wave, dealer/contractor relationships, and distribution density. The serial regulatory cut-overs (SEER2, AIM Act) raise barriers — each transition requires R&D and recertification that favors incumbents with scale. [INTERPRETATION]

In Marathon/Capital Returns terms, the supply side appears disciplined: no new entrant has cracked the U.S. residential oligopoly in decades; the five majors are gaining, not losing, share; and capital is flowing into M&A consolidation and efficiency (Mexico capacity, JVs) rather than greenfield capacity wars. High returns have not yet attracted destabilizing new capacity — the classic precondition for durable returns. The watch-item is the demand side, not supply: the asset-growth/mean-reversion risk here is that peak margins (20% operating) reflect a peak-cycle, pre-buy-flattered demand environment rather than a permanent step-change, and that as the A2L price-up annualizes and volume normalizes, margins could mean-revert toward the high-teens. [INTERPRETATION — this is the central peak-margin question the valuation and financial sections must resolve; the FY25 volume −17%/price +10% split shows the rent is real but the volume base is correcting.]

Verdict

Structurally a good industry — among the better in industrials — but the goodness accrues to incumbents collectively, and recent profitability has been amplified by a one-time regulatory pull-forward that is now unwinding. The favorable structure is genuine: a consolidating top-5 oligopoly (~84% N.A. share), a ~63%+ replacement-driven, non-discretionary, weather-pulled demand base, and a regulatory cadence (SEER2, AIM Act/A2L) that functions as a recurring license to raise prices — visible in FY25’s +9–10% price/mix offsetting a −13–17% volume air-pocket. Barriers to entry in core equipment are real and rising with each regulatory wave; supply discipline is intact in Marathon terms, with no destabilizing new capacity. The industry is good for the incumbents — Carrier, Trane, Daikin, JCI, Lennox — and for the dominant distributor (Watsco); it is hostile to subscale entrants and to the lower-barrier parts/service fringe.

The caveats are two, and they are material to Lennox specifically. First, the 2024 R-410A pre-buy borrowed demand from 2025–2026; the −17% resi volume decline is partly payback, and the price-led margin print should not be naively annualized as a clean run-rate. Second, Lennox is the smallest of the five majors and a North-America pure-play — it enjoys the oligopoly’s pricing umbrella but carries less scale than Daikin/Carrier/Trane to absorb tariff and commodity shocks, and zero geographic diversification (Europe divested Q4 2023). The industry’s structural quality is real; the question the rest of this memo must answer is how much of Lennox’s peak 20% margin is durable industry rent versus a regulatory sugar-high that is already fading.

4. Competitive Position

The question this section answers is narrow and unforgiving: does Lennox have a durable competitive advantage — a barrier that lets it earn returns above its cost of capital that competitors cannot compete away — and if so, what is the mechanism? Applying Greenwald’s “Competition Demystified” taxonomy, there are only three genuine advantage types: (1) supply-side cost advantage, (2) demand-side customer captivity, and (3) economies of scale combined with captivity. Brand, “quality” and “reputation” are not on the list unless they translate into one of those. I conclude Lennox has a moderate, real-but-narrow moat concentrated in its residential channel — a blend of (a) local economies of scale + dealer captivity in the company-owned direct-to-dealer system and (b) the replacement-driven installed base — but it is not a wide, Trane-class moat, the moat does not extend cleanly to the commercial/refrigeration segment that is now driving growth, and management itself describes its own distribution as a “low-barrier” business. The high ROIC is real; the durability of it is the open question.

The ROIC test — the moat shows up in the numbers

A moat must surface as returns. It does. FY25 NOPAT ≈ $842M (operating income $1,041.5M × (1 − 19.2% tax)) on invested capital of ~$2,517M (total debt $1,388.4M + equity $1,163.1M − cash $34.2M), implying ROIC ≈ 33% (~29% capitalizing operating leases) and a pre-tax ROIC of ~41% [INTERPRETATION built on FACTs: operating income $1,041.5M and 19.2% tax rate, FY2025 10-K MD&A; total debt $1,388.4M, equity $1,163.1M and cash $34.2M from the FY25 10-K balance sheet/debt footnote and EDGAR XBRL]. That is comfortably 3-4x any plausible WACC (~8-9%) and would be far higher excluding goodwill/intangibles from acquisitions. The reported ROE of ~76% (net income $805.8M on average equity ~$1.06B) is real arithmetic but largely an artifact of the buyback-shrunken equity base (see the Capital Allocation section) and should be ignored as a moat signal — ROIC is the cleaner test, and it passes decisively [INTERPRETATION; ROE inputs FACT — NI $805.8M, equity $962.1M→$1,163.1M, EDGAR XBRL]. So the financial outcome a moat should produce is present. The analytical work is establishing whether it is durable (Greenwald’s real test) or a peak-cycle, pre-buy-flattered artifact (Marathon’s mean-reversion warning).

Mechanism 1 — Company-owned direct-to-dealer distribution: local scale + dealer captivity (the real moat)

The strongest moat element is the Lennox-branded, company-owned direct-to-dealer channel and its ~200+ Lennox Stores [FACT — FY2025 10-K l.172-174; ~245 U.S. locations per third-party trackers, June 2026]. Two Greenwald advantages plausibly operate here, and they are local, not national:

• Local economies of scale. HVAC distribution is a density game: stocking equipment and parts close to contractors so a same-day replacement install is possible. Within a given metro, the distributor with the most contractor relationships and the densest local inventory has lower per-unit logistics cost and higher service levels than a sub-scale entrant. Lennox’s owned-store density in its strong markets creates a local scale advantage that a national average understates [INTERPRETATION via Greenwald local-scale lens]. This is the same structural logic that makes Watsco (WSO) — a pure HVAC distributor — a durable franchise.
• Dealer/contractor captivity (switching costs). Independent Lennox dealers are trained on Lennox equipment, carry Lennox-specific parts/controls (iComfort, Comfort Sync), build their consumer-facing reputation partly on the Lennox brand, and rely on the local Lennox Store for same-day parts. Switching to Carrier/Trane means retraining, re-stocking, and re-marketing. Management explicitly says the model “creates a large installed base” and “opportunities for longer-term service, monitoring, solutions and retrofit revenue” — i.e., pull-through captivity [FACT — FY2025 10-K l.339]. By owning the distribution margin a competitor cedes to an independent two-step distributor, Lennox captures economics across the whole chain on its flagship brand.

Pressure-test / the critical caveat. Management’s own risk language is the most important disconfirming evidence in the entire filing: it warns of “entry of new competitors into the low-barrier distribution business” [FACT — FY2025 10-K l.339, emphasis added]. Lennox is telling investors the distribution layer it relies on is low-barrier. That is fatal to a “wide moat” claim. Direct-to-dealer is better characterized as vertical integration that captures margin and creates moderate local captivity, not an impregnable structural barrier. A well-capitalized rival (Daikin’s Goodman, or distributor consolidators) can and does build competing density. The moat is therefore local and contestable, not national and durable in the Trane sense.

Mechanism 2 — The installed base / replacement annuity

Because HVAC equipment lasts ~15-20 years and ~two-thirds-plus of demand is replacement [FACT — FY2025 10-K l.749], every unit Lennox installs seeds a future replacement and a multi-year parts/supplies/service annuity routed back through Lennox Stores. This is a genuine, compounding demand-side advantage — but it is shared by every incumbent (Carrier, Trane, Daikin/Goodman, Rheem all have large installed bases) and is therefore an industry attribute, not a Lennox-specific moat. It explains why the HVAC oligopoly as a whole earns good returns; it does not by itself explain Lennox out-earning peers. Where it does differentiate is in combination with Mechanism 1: Lennox captures more of the replacement/parts annuity per unit because it owns the distribution, whereas a Carrier replacement flows partly to an independent distributor. The Ariston (water-heater) and Samsung (ductless) JVs are explicit attempts to widen the attachment per installed-base customer [FACT — FY2025 10-K l.195-200] — a sensible moat-deepening move, though unproven.

Mechanism 3 — Brand: real but it does NOT command share leadership

The “Lennox” brand is well-regarded in the premium residential tier (Dave Lennox Signature Collection). But Greenwald’s discipline asks: does the brand command price or share that a generic competitor cannot? On price — partially yes; Lennox indexes to the higher-efficiency/premium end and its +10% FY25 HCS price/mix shows pricing power exists in the channel [FACT — FY2025 10-K l.991]. On share — no. Third-party North American residential share estimates (which vary by source and treat brand families inconsistently) place Lennox roughly second-to-third among the “Big Three,” behind or alongside Trane and Carrier, with the Daikin family (Daikin + Goodman + Amana) collectively large:

[INTERPRETATION — third-party estimates (BuildOps, marketsandmarkets, trade compilations), accessed June 2026; figures are illustrative ranges, not audited, and double-count risk exists where families are split. The “Big Three” (Trane/Lennox/Carrier) collectively ~50-55% of the U.S. market is the most defensible takeaway.] Lennox’s own competitor list in the 10-K names Carrier, Trane, Rheem (Paloma), Bosch (York/Luxaire/Coleman), Daikin (Goodman/Amana) and Madison Industries in residential, and Carrier, Trane, Rheem, CSW, Daikin and AAON in commercial [FACT — FY2025 10-K l.234, 238]. The brand keeps Lennox in the leadership tier; it does not confer dominance, and it competes against Daikin’s superior global scale and Goodman’s price aggression at the value end.

The market-share-stability test (Greenwald’s gold-standard durability proxy)

Greenwald argues stable market shares over time are the best evidence of a moat (volatile shares mean no captivity/barriers). The U.S. residential HVAC oligopoly has exhibited durable, slow-moving shares for decades — the Big Three have held their positions, new national entrants are rare, and the last major structural change was Daikin’s acquisition of Goodman (2012) [INTERPRETATION — industry history, corroborated by stable competitor set in successive 10-Ks]. This is a positive for the industry moat. But it cuts both ways: Lennox is not visibly gaining share — FY24-25 volume swings were refrigerant-driven, not share gains, and Q1-2026 HCS organic volume was −21% (an industry-wide payback, not a Lennox-specific loss) [FACT — Q1-2026 8-K/press release, 4/29/2026]. Stable share supports “durable oligopoly economics”; it does not support a thesis of Lennox-specific competitive gains.

Head-to-head vs the named competitors

• vs Trane (TT): Trane is the premium leader with arguably the strongest brand and a larger, more diversified (incl. transport refrigeration) franchise; it trades at a richer multiple (~35x P/E, ~25x EV/EBITDA vs LII ~23x / ~17x) [FACT — yfinance comps, June 2026]. Trane’s moat is at least as wide; Lennox’s edge is the owned-distribution margin capture and a leaner, more focused North American footprint.
• vs Carrier (CARR): Post-Viessmann/portfolio reshaping, Carrier is larger and broader but less North-America-residential-pure; Lennox’s tighter focus and direct-to-dealer model give it structurally higher segment margins (LII HCS ~22% vs diversified Carrier).
• vs Daikin/Goodman: The genuine threat. Daikin has global scale, leading ductless/heat-pump technology, and uses Goodman/Amana to attack the value tier on price. As electrification/heat-pump adoption grows, Daikin’s technology depth is a real competitive risk to Lennox’s premium-resi position — Lennox’s Samsung ductless JV is a defensive response [FACT — FY2025 10-K l.198-200].
• vs Rheem / Bosch / AAON: Rheem competes on value; AAON is a focused, high-quality commercial/rooftop specialist that out-grows and out-multiples (AAON ~94x P/E) the majors in its niche and is a more dangerous competitor in BCS than Lennox’s residential rivals are in HCS.

Verdict

Lennox has a real but narrow and moderately durable moat — not a wide one. The defensible advantages are (1) local economies of scale plus dealer/contractor captivity in the company-owned direct-to-dealer residential channel, and (2) the replacement-driven installed-base annuity, amplified because Lennox owns more of the distribution chain on its flagship brand. These translate into a genuine ~33% ROIC that clears the moat test on a current-returns basis. But four facts cap the verdict at “moderate”: (i) management’s own filing calls the distribution layer a “low-barrier” business — the single most candid disconfirming datapoint; (ii) the brand keeps Lennox in the leadership tier but does not confer share dominance (it sits behind/alongside Trane and Carrier, with Daikin’s scale and Goodman’s price a structural threat at the value/ductless end); (iii) the moat is residential-channel-specific, yet FY25 growth and margin leadership has shifted to the commercial/refrigeration BCS segment, which runs on conventional contractor/national-account distribution where the owned-store advantage does not apply and where AAON is a sharper competitor; and (iv) through Marathon’s capital-cycle lens, today’s ~33% ROIC (~41% pre-tax) and 20% operating margins are at or near a cyclical/refrigerant-transition peak — high returns attract capital and mean-revert, and the Q1-2026 HCS margin compression (−390bps) is the first evidence that the price/mix lever cannot indefinitely offset volume. The correct characterization: a quality oligopoly participant with a genuine local-distribution edge and durable replacement economics, earning real excess returns — but a contestable, residential-concentrated moat whose current returns are flattered by a non-repeating refrigerant cycle, not a fortress that justifies extrapolating peak ROIC indefinitely.

5. Growth History and Forward Opportunities

The central growth question for Lennox is whether the top-line and margin expansion of 2022–2024 was demand growth or price growth — and the filings answer it bluntly. Over FY21–FY24 revenue compounded at ~8.4% (from $4,194.1M to $5,341.3M), then fell 2.7% in FY25 to $5,195.3M. Decomposing the consolidated bridges shows the growth was overwhelmingly price/mix-led, was flattered in FY24 by a one-time regulatory pre-buy, and gave much of the volume back in FY25–H1’26. The forward story is therefore less about secular demand and more about whether refrigerant-driven price can keep carrying the P&L while volumes normalize and commercial (Building Climate Solutions) and the new Samsung ductless JV pick up the slack.

5.1 Revenue by segment, FY21–FY25

Lennox reports two operating segments — Home Comfort Solutions (HCS), residential, and Building Climate Solutions (BCS), light-commercial unitary plus Heatcraft refrigeration — with European businesses sitting in Corporate & Other until their Q4-2023 divestiture. (FACT — FY2025 10-K, Business and segment notes; FY2023 10-K MD&A.)

(FACT — FY2025 10-K MD&A “Results by Segment” for FY25/FY24; FY2023 10-K MD&A for FY23/FY22. Segment figures as reported each year.)

Two structural facts jump out. First, HCS — ~64% of sales — is the swing factor and is cyclical-residential, not recurring: it surged +11% in FY24 then dropped −7% in FY25. Second, BCS has been the steadier grower (+16.7% then +5.0%) and is now the engine of the forward plan. The FY25 consolidated decline is not one story but three, which the next sub-section separates.

5.2 Price/mix vs. volume — the decomposition that matters

Management discloses the consolidated growth bridge each year, and it is the single most important growth datum in the file:

¹ FY23 bridge stated as “favorable mix 5% + favorable price 5%, offset by volume −4%.” (FACT — FY2025 10-K MD&A; FY2023 10-K MD&A; Q1-2026 10-Q MD&A.)

This is the crux of the thesis. In FY25, the entire revenue story was a +9% price/mix benefit nearly offsetting a −13% collapse in organic volume, with a +1% bolt-on from the Duro Dyne/Supco deal. (FACT — FY2025 10-K MD&A.) The same shape repeated in Q1’26: +9% price/mix, −8% organic volume, +5% from acquisitions. (FACT — Q1-2026 10-Q MD&A.) (INTERPRETATION) A business whose revenue line holds up only because price is being raised ~9% while unit shipments fall double digits is not enjoying demand growth — it is enjoying a regulatory- and inflation-driven price reset. That is far lower-quality than the headline “flattish revenue, record margins” suggests, and it is the reason the stock can de-rate even as reported numbers look stable.

5.3 The FY24 R-410A pre-buy and the FY25–FY26 payback

The volume swing is largely a manufactured one. The EPA’s AIM Act required a transition to refrigerants with GWP ≤700 effective January 1, 2025, pushing residential systems from R-410A to A2L/R-454B. (FACT — FY2025 10-K, “Refrigerants.”) Ahead of that deadline, the channel pre-bought R-410A equipment in 2024 — inflating FY24 HCS volume (+7% segment volume, +11% segment sales) and segment profit (+24% to $760.5M, margin 19.0%→21.3%). (FACT — FY2025 10-K MD&A, HCS FY24 vs FY23.) That pull-forward then reversed hard:

• FY25 HCS: net sales −7% on a −17% volume decline, only partly offset by +10% price/mix; segment profit −4% to $728.5M. (FACT — FY2025 10-K MD&A.)
• Q1’25 HCS organic volume −32%; Q1’26 organic volume still −21% (a “meaningful improvement” per management, but a second straight year of sharp declines). (FACT — Q1 2026 earnings call, Quenzer, 2026-04-29; these organic-volume figures are management call disclosures, not separately broken out in the 10-Q, which reports the HCS net-sales change.)

(INTERPRETATION) The FY24→FY25 swing — roughly a 7% segment volume gain becoming a 17% loss — is consistent with a meaningful slug of 2024 HCS demand having been borrowed from 2025. Management now says “channel destocking has largely concluded” with “no indication of any pre-buy ahead of price increases.” (FACT — Q1 2026 call, Maskara.) (OPEN QUESTION) Whether residential underlying replacement demand has genuinely stabilized or whether the −8% to −13% volume run-rate reflects soft end demand (new-construction down >30%, ~25% of HCS) is not yet provable from one improving quarter; treat the “destocking is over” claim as a hypothesis.

5.4 Is A2L price a growth lever or just inflation?

The 2025–2026 price increases — annual list increases plus A2L-conversion pricing — are real revenue, but management’s own framing exposes the quality issue. On the Q1’26 call the CFO explained that incremental margins are improving because “there is now more price than mix” in the blend. (FACT — Q1 2026 call, Quenzer.) (INTERPRETATION) Pure price (vs. richer-product mix) can drop through at high incremental margin, but it is also the most reversible form of “growth”: it depends on industry pricing discipline holding through the A2L transition and tariff pass-through, which management itself flags as fragile if a competitor (Carrier, Trane, Daikin-Goodman, Rheem) chooses share over margin. Counting a one-time regulatory step-up in refrigerant cost — passed through as a price increase — as “growth” overstates the durability of the top line. Through a Marathon capital-cycle lens, peak pricing power amid full order books is exactly the condition that mean-reverts.

5.5 Forward opportunities — and how much to credit them

Management raised FY26 guidance on the Q1 call to ~8% revenue growth (from 6–7%), with HCS ~4% (from ~2%) and BCS ~16% (from ~15%), adjusted EPS $23.50–$25.00 and FCF $750–850M. (FACT — Q1 2026 call, 2026-04-29; FY25 8-K guidance, 2026-01-28.) The genuine forward levers, ranked by how much weight the evidence supports:

• BCS / commercial (highest-quality lever). BCS grew sales +38% in Q1’26 (+17% organic volume, +12% acquisition volume, +9% price/mix) with segment margin 16.7%→19.7%. (FACT — Q1-2026 10-Q MD&A.) Management attributes this to emergency-replacement share gains and national accounts and calls the emergency-replacement build-out “the second inning of a nine-inning game.” (FACT — Q1 2026 call, Maskara.) This is volume-led and structurally more attractive than HCS price; it is the most credible part of the bull case.
• Samsung ductless JV (structural adjacency, unproven scale). A 49.9% JV distributing Samsung ductless AC/heat pumps and “Lennox powered by Samsung” products through Lennox Stores and the direct-to-dealer network — addressing Lennox’s historical under-indexing to ductless mini-splits/heat pumps. (FACT — FY2025 10-K, “Joint Ventures.”) Management expects “meaningful impact this year.” (INTERPRETATION/OPEN QUESTION) As a 49.9% equity-method JV, it shows up in equity income, not consolidated revenue, and FY25 income from equity-method investments was only ~$6–8M — so the financial contribution is currently de minimis and unverifiable in scale. Promising adjacency, not yet a numbers story.
• Parts & supplies / bolt-on M&A. Duro Dyne + Supco ($545.0M cash, Oct 2025) add ~1% to consolidated and ~2%/~12% to HCS/BCS volume respectively; this is acquired, not organic, growth and the FY26 plan leans on a high-single-digit M&A contribution in BCS. (FACT — FY2025 10-K business-combination note; Q1 2026 call.)
• Data-center cooling. Frequently cited by sell-side as an option (Lennox’s refrigeration/Heatcraft and cold-climate heat-pump tech), but management gave no data-center revenue commentary on the Q1’26 call and the 10-K does not size it. (FACT — Q1 2026 call contained no data-center disclosure; FY2025 10-K silent on quantification.) (OPEN QUESTION) Lennox’s data-center exposure appears materially smaller than AAON’s; do not credit it in a base case.

Verdict — Low-to-medium quality growth, presently price-not-volume

The growth here is lower-quality than the reported top line implies. Through FY25 and into H1’26 the company is running negative organic volume (−13% FY25, −8%/−21% organic in Q1’26 consolidated/HCS), with revenue held up almost entirely by a +9% price/mix benefit that is itself a regulatory refrigerant step-up plus tariff pass-through — the most reversible kind of “growth,” and one Marathon’s framework would flag as peak-cycle pricing power vulnerable to mean reversion. FY24’s apparent acceleration was substantially a borrowed-from-2025 pre-buy, not durable demand. The credible, volume-led growth sits in BCS (emergency-replacement and national accounts, +17% organic in Q1’26) and, prospectively, the Samsung ductless JV — but the JV is still de minimis financially and data-center exposure is unquantified hype. Net: a structurally fine but cyclical residential franchise whose recent “growth” is mostly price; the burden of proof is on a genuine return to positive HCS unit volume in H2’26 before the top line can be called healthy.

6. Financial Quality

Lennox is, on the numbers, a high-return industrial: ~33–34% gross margin, ~20% operating margin, ROIC in the 30s, modest leverage, and free-cash-flow conversion that has historically run at or above net income. The harder questions — and the ones this section resolves — are (a) whether the three-year operating-margin jump from 13.9% to 20.0% is structural or cycle/regulatory-flattered, (b) why FY25 cash flow fell below net income, and © why the headline ~76% ROE is economically meaningless. The short answers: the margin gains are mostly real but partly borrowed from a refrigerant pre-buy and from price increases that are now decelerating; the FY25 cash dip is a deliberate A2L inventory build, not an accrual game; and ROE is a treasury-stock artifact, not a measure of franchise quality. ROIC is the honest profitability signal, and it is genuinely excellent.

6.1 Multi-year P&L, cash flow and the margin trajectory

(Sources: FY25 10-K, Consolidated Statements of Operations and Cash Flows, filed 2026-02-17 [lii-20251231.htm]; FY23 10-K filed 2024-02-13; share counts from EDGAR XBRL WeightedAverageNumberOfDilutedSharesOutstanding.)

FACT — gross margin expanded ~620 bps and operating margin ~610 bps in three years (27.2%→33.4% GM; 13.9%→20.0% OM, FY22→FY25). This is the central bull point and the central skeptic’s target.

INTERPRETATION — the expansion is roughly 60–70% structural and 30–40% price/regulatory tailwind, and the price piece is fading. Management’s own segment bridges are the best evidence. In the FY24-vs-FY23 Home Comfort Solutions bridge, of a $149M segment-profit gain, $122M came from “higher price and favorable mix,” $90M from volume, and only $9M from “factory productivity and favorable product costs” (FY25 10-K, MD&A, Results by Segment). In other words, price/mix did the heavy lifting, productivity/cost-out was small. The FY25-vs-FY24 bridge is even more revealing: Home Comfort Solutions net sales fell 7% as a 17% volume decline was offset by a +10% price/mix, and segment profit fell only $32M because $256M of favorable price/mix nearly absorbed a $224M volume hit plus $64M of “product costs and factory inefficiencies” and $47M higher freight (FY25 10-K, MD&A). Building Climate Solutions told the same story: +5% sales on +8% price/mix and a +2% acquisition contribution against a −5% organic volume decline. Conclusion: the margin structure has been carried by pricing power during the A2L transition, while organic volumes have been falling. That is a fragile mix — pricing decelerates and “factory inefficiencies” appear once the pre-buy demand reverses.

OPEN QUESTION — is 20% the new floor or a cycle peak? Q1-2026 is an early warning. Net sales rose 5.8% to $1,135.1M, but operating income fell to $163.5M from $168.0M, and gross profit rose only $10.4M on $62.5M more revenue — a ~17% incremental gross margin versus the 30%+ implied by the FY22–FY24 ramp (Q1-2026 10-Q, filed 2026-04-29, lii-20260331.htm). One quarter is not a trend, but margins are no longer expanding; they are flat-to-down even as revenue recovers. The own-history valuation percentiles (P/E 30th percentile, composite 39th) and the ~25% drawdown are consistent with a market that has stopped paying for further margin upside. (Treat the percentile as a third-party signal, not evidence.)

6.2 ROIC vs. ROE — read the right number

FACT — ROE is ~76%, ROIC is ~33%, and only the latter is interpretable. The ROE gap is a capital-structure artifact, not franchise quality. The balance sheet shows treasury stock at cost of −$4,923.4M against retained earnings of $4,891.1M (FY25 10-K, Consolidated Balance Sheet). Lennox has repurchased 52.4M of its 87.2M issued shares cumulatively (~60%), and the treasury contra-equity has almost exactly offset two decades of retained profit, leaving total equity at just $1,163.1M. Equity was actually negative −$203.1M at FY22 (EDGAR XBRL StockholdersEquity) before the FY23–FY25 earnings surge rebuilt it. Dividing $806M of net income by a near-zero, buyback-suppressed equity base mechanically produces a 70–80% ROE that says nothing about incremental returns. The honest read is ROIC ~30–34% on a properly defined invested-capital base — still excellent, comfortably above any reasonable WACC, and the real evidence that the business earns economic profit. Per Greenwald, sustained ROIC this far above cost of capital is the financial fingerprint of a barrier to entry; the durability of that spread (not its current level) is the live question (see the Competitive Position section).

6.3 FCF conversion and the FY25 quality-of-earnings flag

FACT — FY25 CFO ($757.6M) was below net income ($805.8M); CFO/NI fell to 94% from 117% in FY24. The cause is fully visible in the cash-flow bridge and is not an accrual red flag: a deliberate inventory build of ~$300M (inventory rose from $853.0M to $1,152.6M year-over-year; EDGAR InventoryNet) was the dominant driver. Netted against the offsets — a +$108.9M receivables collection and +$58.6M income-tax timing, less accounts-payable and accrued-expense outflows — the aggregate working-capital cash drag was ~$228M (FY25 10-K, Statement of Cash Flows).

INTERPRETATION — the inventory build is the A2L/R-454B transition, not channel stuffing. The 2025 mandated shift from R-410A to low-GWP A2L refrigerant (R-454B) required Lennox to carry dual product lines and pre-position new-refrigerant inventory; the FY24 R-410A pre-buy had pulled resi volume forward, so FY25 simultaneously saw weak shipped volume and a higher inventory peg for the new platform. This is a one-time transition cost to cash, and it is the correct interpretation of puzzle #5: earnings are cash-backed; the FY25 dip is timing, not quality. It should reverse as A2L inventory normalizes — worth confirming in FY26 working capital. Over four years, FCF/NI averaged ~74% (dragged by FY22’s 40%, itself a $249M post-COVID inventory build), and CFO/NI averaged ~99% — clean. D&A of only $112.5M against capex of $118.8M and FCF of $639M confirms a capital-light assembly model (capex ~2.3% of sales), where margin gains convert efficiently to cash.

6.4 Balance sheet, leverage, SBC, tax, one-time items

• Leverage is modest. FACT: total debt $1,388.4M ($226.0M commercial paper + $800M senior unsecured notes + $300M term loan + $50.6M finance leases + $18.3M current maturities − $6.5M issuance costs), cash $34.2M → net debt ~$1,354M; with EBITDA of ~$1,154M (EBIT + D&A), net debt/EBITDA ≈ 1.2x (≈1.5x capitalizing $382M of operating leases) (FY25 10-K, debt footnote and Balance Sheet). This is investment-grade headroom; the deliberately thin equity (not the debt load) is what makes the company look levered on a debt/equity screen.
• Pension is a non-issue. FACT: net funded status −$14.8M (FY25) on a fully matched plan — immaterial to the thesis (FY25 10-K, retirement-plans note).
• SBC is immaterial and not a hidden dilution lever. FACT: stock-based comp expense was $29.1M (FY25), $28.5M (FY24), $30.1M (FY23) — ~0.56% of revenue, far below software/growth-industrial norms, and diluted share count actually fell (37.5M FY21 → 35.4M FY25). Dilution from grants is more than offset by buybacks; this is a genuine ~8% five-year share-count reduction, not optical EPS engineering.
• Tax. FACT: FY25 effective rate 19.2% ($191.0M on $996.8M pretax; EDGAR XBRL) — clean, below the 21% federal rate, no unusual benefit driving the year.
• One-time items / run-rate distortions: the FY25 income statement is unusually clean — a trivial $(0.9)M gain on prior dispositions, no impairments, no large divestiture gain (the European commercial/refrigeration exit and the related disposals were largely effected in prior years and run through discontinued ops). The real run-rate distortions are operational, not one-time accounting: (1) the FY24 R-410A pre-buy inflated FY24 resi volume (+7% HCS volume, ~+25% Q4 segment revenue per management), making FY24’s 19.5% OM a high-water comparison; and (2) FY25 absorbed the payback (−17% HCS volume) yet held margin via pricing. Normalizing for the pre-buy, the underlying organic-volume trend is flat-to-down, and the margin is being defended by price, not demand. The two acquisitions in the period — Duro Dyne (Building Climate Solutions, FY24) and a $545.0M FY25 acquisition spend that lifted goodwill from $220M to $497.2M — add inorganic revenue and explain Building Climate’s +2% acquisition volume contribution; they are bolt-ons, not transformative.

6.5 Do unit economics improve with scale?

Partially. The incremental-margin record is strong on the way up and weak on the way down. FY22→FY24 incremental operating margins were rich (operating income +$384M on +$623M revenue ≈ 62% incremental), which is the signature of operating leverage on a fixed factory base. But the FY24→FY25 reversal exposed the asymmetry: a −17% resi volume swing produced “factory inefficiencies” and $64M of higher product costs, and only aggressive price/mix held the line. So the scale economics are real but cyclical-volume-dependent — they amplify in both directions. That is consistent with Greenwald’s economies-of-scale advantage tied to the captive direct-to-dealer base (fixed distribution and factory costs spread over more units), but it is not a recession-proof margin. Marathon’s capital-cycle lens adds a caution: ROIC in the 30s is exactly the kind of return that invites capital and competitive pricing response, and the A2L price increases that lifted FY24–FY25 margins are a one-time regulatory step that competitors (Carrier, Trane, Daikin/Goodman) took in parallel — an industry-wide reset, not a Lennox-specific moat.

Verdict

High-quality, cash-generative, capital-light economics — but at a normalized, not expanding, level. The franchise earns a genuine ~30–34% ROIC well above its cost of capital, converts net income to free cash at ~95–100% over a cycle (FY25’s 79% being a transitory A2L-inventory dip, not an accrual problem), carries modest ~1.2–1.5x net leverage, immaterial SBC and dilution, a fully-funded pension, and a clean 19% tax rate. The accounting is conservative and the earnings are cash-backed — this is not a quality-of-earnings short. The legitimate skeptic’s point is narrower and correct: the 13.9%→20.0% operating-margin surge was driven more by price/mix during a forced refrigerant transition than by durable cost-out, organic volumes have been declining, incremental margins compress sharply when volume falls, and Q1-2026 shows margins flattening as revenue recovers. The headline ~76% ROE is an artifact of a buyback-hollowed equity base and should be ignored. Net: the economics are excellent and clean, but the recent margin level is best treated as a peak-of-transition figure to be stress-tested in valuation — not extrapolated.

7. Capital Allocation

Capital allocation is where Lennox earns — or forfeits — the right to its premium multiple. The headline ROE (~77% per yfinance) is an accounting artifact of a decade of aggressive buybacks that drove book equity negative, not a measure of incremental returns; it must be read with that context or it misleads. The substantive question is whether management has deployed Lennox’s prodigious free cash flow intelligently. The record is mixed-to-good: disciplined return of capital and a smart portfolio reshaping toward higher-margin North-America HVAC, but with the recurring tension that the largest buyback in years (FY25) and the largest acquisition in years (FY25) both landed near a cyclical and valuation peak.

7.1 The buyback engine and the negative-equity structure

Lennox has used buybacks as its primary capital-return tool for over fifteen years, and the cumulative effect on the share count is dramatic. Diluted shares fell from 55.8M (FY10) to 35.4M (FY25) — a ~37% reduction, roughly 2–3% per year compounded — without a major equity-funded acquisition to offset it (FACT: 10-K XBRL, WeightedAverageNumberOfDilutedSharesOutstanding). Open-market repurchase outlays over FY16–FY25 totaled ~$2.94B ($2,936.1M per XBRL PaymentsForRepurchaseOfCommonStock; a further ~$0.2B of share repurchases to satisfy employee tax withholding is a separate line):

Source: 10-K XBRL (PaymentsForRepurchaseOfCommonStock, PaymentsOfDividendsCommonStock, StockholdersEquity); FY25 10-K (2026-02-17), Liquidity & Cash Flows. (FACT.)

The negative-equity period (FY18–FY22) was a deliberate consequence of returning more than 100% of earnings via buybacks-plus-dividends while the balance sheet ran lean — a structure that flatters ROE/ROIC mechanically. (INTERPRETATION.) Equity has since recovered to $1,163.1M (FY25) as buybacks paused in FY23 (zero repurchases) to fund the AES deal and rebuild the balance sheet (FACT). The takeaway for valuation: Lennox’s ~77% ROE and the incentive-plan’s ~47% reported ROIC are not evidence of an unusually capital-light moat on their own — they are partly the residue of share shrinkage and a thin equity base. The honest moat test is gross-margin durability and pricing power (covered in the Financial Quality discussion), not the levered return ratios. (INTERPRETATION.)

Is the buyback accretive at these prices? Lennox repurchased ~0.9M shares for $482.3M in FY25, an average of ~$536/share (FACT: 10-K XBRL TreasuryStockSharesAcquired + cash-flow). With FY25 diluted EPS of $22.79, that is a ~4.3% earnings yield (~23x P/E) on capital deployed — accretive only if Lennox’s forward returns exceed that hurdle and the multiple holds. Buying ~$536 stock against an FY25 in which revenue fell 2.7% and margins look peak-cyclical (see the Financial Quality section) is the weakest link in the capital-allocation record: management leaned in on repurchases at a high multiple just as the cycle softened. (INTERPRETATION — this is the Marathon “high returns attract capital / mean-reversion” caution applied to the company’s own buyback timing.) Notably, management then pulled back sharply: only 39,000 shares ($20.0M) were repurchased in Q1-2026, after deploying $545M on the Duro Dyne/Supco acquisition (FACT: Q1-2026 10-Q, Note 5). The Board has a $5.0B cumulative authorization with $989.5M remaining at 3/31/26, so the capacity is optional, not a commitment (FACT: Q1-2026 10-Q, Note 5).

7.2 Dividends — a genuine, growing, well-covered stream

The dividend is the more conservative leg and the cleaner signal. Declared dividends per share rose from $2.43 (FY18) → $3.53 (FY21) → $4.55 (FY24) → $5.05 (FY25), and the Board raised the quarterly rate to $1.36 on May 21, 2026 (+4.6% from $1.30) — the 16th consecutive annual increase, with dividends paid for 28 straight years (FACT: 10-K XBRL CommonStockDividendsPerShareDeclared; Lennox press release, 2026-05-21). At ~$517.78 the yield is only ~1.07% and the payout ratio is ~23% of FY25 EPS — a low, highly covered payout that leaves ample FCF for buybacks and M&A. (FACT/INTERPRETATION.) This is not a Dividend-Aristocrat-grade 25-year streak, but it is a reliable, fast-growing dividend that has compounded ~8%/year over a decade — a credible quality signal, though immaterial to total return relative to the buyback. (INTERPRETATION.)

7.3 M&A and divestitures — a deliberate reshaping toward NA HVAC

Management has actively reshaped the portfolio, and the strategic logic is sound even where the financial terms were unremarkable:

• European commercial HVAC & Refrigeration exit (Q4 2023). Lennox sold its German/French/Spanish unitary and applied-systems businesses, completing the divestiture in Q4-2023; these had generated ~$248M of FY23 revenue and were structurally low-margin/loss-making — the European operations sat in the “Corporate & Other” segment, which reported a $(93.9)M loss in 2023 (inclusive of unallocated corporate costs), a fraction of the ~20% group operating margin (FACT: FY25/FY23 10-K, segment notes). The exit cost more than it raised — Lennox booked a $22.6M PP&E impairment and a $2.3M goodwill impairment on the held-for-sale assets in Q3-2023, and 2023 investing cash flow shows only ~$23.2M net proceeds (FACT: FY25 10-K). Verdict on the exit: correct. Shedding a sub-scale, low-margin, sub-cyclical European unit to focus on the structurally superior North-America replacement market is exactly the right pruning, even at a near-zero net price. (INTERPRETATION.)
• AES Industries (Oct 2023, ~$95M). A light-commercial service/curb/recycling tuck-in into Building Climate Solutions, adding ~$40.4M goodwill (FACT: FY25 10-K goodwill note). Small, on-strategy (recurring service revenue).
• Duro Dyne & Supco / NSI Industries HVAC division (Oct 16, 2025, ~$545M net cash). The largest deal in years: Duro Dyne (sheet-metal ductwork components, Huntersville NC) and Supco (fans, belts, igniters, thousands of HVAC parts, Allenwood NJ), split across both segments (FACT: FY25 10-K; Lennox press release / PR Newswire, 2025-10-16). Strategic rationale is a deeper, higher-margin, more recurring parts-and-accessories stream attached to the existing equipment/distribution flywheel (INTERPRETATION). At ~$545M this is the deal to watch on returns; no purchase-price multiple was disclosed, an OPEN QUESTION on price discipline.
• Joint ventures (49.9% each). The Samsung JV (2024) distributes Samsung ductless mini-split AC/heat pumps and “Lennox powered by Samsung” products — a capital-light way to plug Lennox’s historic ductless gap rather than build it (FACT: FY25 10-K). An Ariston JV adds water heaters through Lennox Stores, and a long-standing 50% Mexico JV supplies refrigeration coils/compressors. These are sensible minority structures that add adjacencies without consolidating capital. (INTERPRETATION.)

7.4 Capex, leverage, and R&D intensity

Capex is modest and asset-light for a manufacturer: $118.8M (FY25), 2.3% of sales, down from a $250.2M (FY23) peak tied to the new Saltillo, Mexico commercial-HVAC factory; FY25 capex was deliberately reduced (FACT: 10-K XBRL PaymentsToAcquirePropertyPlantAndEquipment; FY25 10-K MD&A). The low capex intensity is a real attribute — it converts margins into FCF (FY25 FCF ~$639M) and is the engine behind the dividend-plus-buyback program. R&D is light at $103.4M (2.0% of sales, FY25) vs. SG&A of $681.4M (13.1%); R&D has grown from $66.8M (FY20) but remains a fraction of what scaled global peers like Carrier/Trane spend, consistent with Lennox’s NA-focused, distribution-led (not technology-led) model. (FACT/INTERPRETATION.)

Leverage is conservative and well inside covenant. Total debt was $1,388.4M at FY25 against $34.2M cash (cash drawn down to fund Duro Dyne), and the FY25 EBITDA proxy is ~$1,154M ($1,041.5M op income + $112.5M D&A), implying year-end net debt of ~$1,354M and net-debt/EBITDA of ~1.2x (~$1,354M / ~$1,154M EBITDA) — far below the 3.50x Total Net Leverage covenant (FACT: 10-K XBRL; FY25 10-K, Credit Facilities). Lennox has dry powder for both M&A and buybacks; the recent restraint is a choice, not a constraint. (INTERPRETATION.)

7.5 Management incentives — well-aligned, profit-and-capital-tilted

The proxy reads cleanly and aligns pay with the metrics an owner cares about (FACT: DEF 14A, 2026-04-08):

• Short-term incentive (STI): Company Core Net Income (50%), Free Cash Flow (30%), Adjusted Revenue (20%). FY25 performance came in below target on cash flow and revenue, producing below-target STI payouts (CEO 90% of target / ~$1.33M) — evidence the plan actually flexes down when results soften. (FACT.)
• Long-term incentive (PSU, 3-year): ROIC (50%) + Core Net Income growth CAGR (50%). The ROIC design is rigorous — no payout unless ROIC exceeds estimated cost of capital, with target at ~3x cost of capital. The 2023–2025 PSUs paid ~196% of target (47.0% weighted-average ROIC blew through the max; 17.2% net-income CAGR near max), with three-year TSR of 109% (FACT: DEF 14A). That payout is defensible given the actual returns delivered, but it underscores that ROIC is partly inflated by the same thin capital base that inflates ROE — a metric the comp plan rewards but that an analyst should normalize. (INTERPRETATION.)
• CEO pay & alignment: Alok Maskara’s total compensation was $8.43M (FY25), down from $9.91M (FY24) — heavily equity-weighted (PSUs/SARs/RSUs), with 6x-base-salary stock-ownership guidelines for the CEO, anti-hedging/pledging, and clawbacks (FACT: DEF 14A Summary Comp Table; governance section). No relative-TSR metric in the plan — a mild weakness, since absolute-profit metrics can reward a rising HVAC tide rather than outperformance. (INTERPRETATION/OPEN QUESTION.) Detailed Form-4 insider activity is reviewed separately; the proxy structure itself shows incentives pointed at profit, capital efficiency, and cash — the right targets.

Verdict

Has management allocated capital intelligently? Largely yes — with one real blemish. The strategic moves are sound: exiting sub-scale, low-margin Europe; tucking in recurring-revenue service (AES) and parts/distribution (Duro Dyne/Supco); using capital-light 49.9% JVs (Samsung, Ariston) to close product gaps; running modest capex and conservative ~1.2x net leverage; and funding a low-payout, 16-year-growing dividend. The incentive plan is well-designed and demonstrably pays for ROIC and cash, and flexed down in a soft FY25. The blemish is timing and reflexivity: the buyback is the dominant value lever, and management deployed its largest repurchase in years (~$536/share) and largest acquisition in years into a cyclical/valuation peak, just as revenue turned negative — a Marathon-style warning that high reported returns were attracting the company’s own capital at a rich price. The sharp Q1-2026 pullback to $20M of buybacks suggests management recognized the same thing. Net: a genuinely disciplined allocator whose worst habit is paying up at the top of the cycle — accretive if the margin step-up proves structural and the multiple holds, financial engineering if it doesn’t. The verdict is favorable but conditional on the margin-durability question, not on the optically huge but context-dependent return ratios.

8. Changes and Headwinds — Last Two Years

The story of the last two years at Lennox is the story of one regulatory windfall (the 2024 R-410A pre-buy), its 2025–26 hangover (a volume air-pocket), and a portfolio that management quietly simplified to North America before bolting small parts/distribution businesses back on. The headline that organic demand is soft while reported revenue and margins look resilient is the single most important tension in the thesis — and it is almost entirely explained by the changes below.

8.1 Portfolio reshaping: exit Europe, simplify to two segments, then re-expand

Segment reorganization (FACT). Effective January 1, 2023, Lennox collapsed from three reportable segments to two — Home Comfort Solutions (residential) and Building/Business Climate Solutions (commercial + refrigeration) — with a residual “Corporate and Other.” Crucially, the legacy Heatcraft Worldwide Refrigeration business was retained and folded into Building Climate Solutions to “leverage synergies,” not divested (FY2025 10-K, Item 1 / Note 3). In Q1-2026 the commercial segment was renamed “Business Climate Solutions” (Q1-2026 10-Q, segment note) — a cosmetic relabel, not a structural change.

European divestiture (FACT). In Q4-2023 Lennox completed the sale of its European commercial HVAC/refrigeration operations (plants in Germany, France, Spain), the only meaningful divestiture of the period. The financial footprint was small: $23M net proceeds, against a $63.2M impairment on assets held for sale in 2023 partly offset by a $14.1M gain on sale of businesses, and a subsequent ~$62M SG&A reduction flowing through 2024 (FY2023 10-K MD&A; FY2024 10-K MD&A). INTERPRETATION: the de minimis proceeds and the held-for-sale impairment confirm Europe was a structurally inferior, sub-scale business — exiting it mechanically lifted the consolidated margin (lower-margin revenue removed) without management having “earned” the improvement operationally. This is a real but one-time, mix-driven tailwind to the margin story, and analysts should not extrapolate it.

Bolt-on M&A re-expands the parts/distribution footprint (FACT). In October 2025 Lennox acquired Duro Dyne (air-distribution accessories, controls; into Building Climate Solutions) and Supco / Sealed Unit Parts (HVAC parts; into Home Comfort Solutions) (FY2025 10-K; ~$277.2M of deal goodwill recorded — $216.3M to Building Climate Solutions and $60.9M to Home Comfort Solutions — lifting total company goodwill to $497.2M; later revised +$6.5M per Q1-2026 10-Q Note 13). These are higher-margin parts-and-supplies businesses that fit the “attachment rate” strategy; they contributed ~+1% to FY2025 consolidated revenue and a chunky +38% to Building Climate Solutions Q1-2026 net sales (Q1-2026 10-Q MD&A). INTERPRETATION: this is the principal reason FY2026 guided revenue growth of “~6–7% includes ~4% from completed acquisitions” — i.e., organic growth is guided to only ~2–3% (Q4-2025 earnings release, 2026-01-28). The acquisitions are flattering an otherwise flat organic top line.

8.2 Samsung JV and Saltillo greenfield — new growth optionality, not yet material

• Samsung JV (FACT). Announced 2024-05-28, the Samsung Lennox HVAC North America JV (Samsung 50.1% / Lennox 49.9%, HQ Roanoke TX) distributes Samsung ductless mini-split/VRF and “Lennox powered by Samsung” products through Lennox Stores and the direct-to-dealer network; operational H2-2024 (Lennox/PR Newswire, 2024-05-28; FY2025 10-K). It plugs Lennox’s biggest product gap (ductless/heat-pump), a category where U.S. share is dominated by Daikin/Mitsubishi/LG. INTERPRETATION/OPEN QUESTION: equity-method accounting keeps it off the revenue line and its contribution is not separately quantified — execution evidence is still thin two years in.
• Saltillo, Mexico commercial greenfield (FACT). A new ~800k-sq-ft commercial HVAC plant in Saltillo (~$150M, opened 2024 — Lennox’s largest-ever manufacturing investment) added emergency-replacement light-commercial capacity (company releases; FY2025 10-K property table lists two Saltillo plants, one per segment). INTERPRETATION: this is the capacity behind Building Climate Solutions’ steady share gains, but it also concentrates ~3,200 employees and key capacity in Mexico — directly in the tariff/USMCA crosshairs (discussed below).

8.3 The R-410A pre-buy and its 2025–26 payback — the core distortion

This is the puzzle that explains the optics. The EPA AIM-Act mandate to transition to ≤700-GWP refrigerants (A2L/R-454B) took effect January 1, 2025 (FY2025 10-K, Refrigerants). Ahead of it, dealers pre-bought R-410A equipment in 2024, which Lennox explicitly estimated added ~$125M to revenue (FY2024 10-K MD&A — management’s own number, FACT). The payback is now visible and severe in the residential segment:

Source: FY2024 10-K, FY2025 10-K, Q1-2026 10-Q segment MD&A.

INTERPRETATION: the −17% FY25 / −21% Q1-26 residential volume collapse is the pre-buy hangover compounded by genuine end-market softness and channel de-stocking — not a small wobble. It has been masked at the revenue line by +9–10% price/mix (A2L equipment carries higher list prices), and masked at the segment-margin line in FY25 by the favorable mix. But Q1-2026 is the tell: with the price/mix cushion no longer enough to offset −21% volume, HCS segment margin cratered ~390bps to 13.3% (Q1-2026 10-Q). INTERPRETATION: this directly stresses the durability of the 13.9%→20.0% consolidated margin expansion — a meaningful slice was price (refrigerant-transition list increases) and favorable mix, both of which fade as volumes normalize and the A2L novelty premium compresses.

8.4 Pricing actions and tariff response — the rising-cost headwind

Lennox has leaned on price/mix as the primary growth lever (+9% group, +10% HCS in FY2025; +9% HCS in Q1-2026), pairing refrigerant-transition list increases with productivity (FY2025 10-K MD&A). The offsetting headwind is escalating trade friction (FACT, Q1-2026 10-Q “Recent Developments”): 2025 IEEPA + Section 232 tariffs; a Feb-2026 Supreme Court ruling against IEEPA tariffs (without addressing refunds or repealing Section 232 steel/aluminum tariffs); a subsequent temporary Section 122 ~10% global surcharge; and April-2026 modifications to Section 232 steel/aluminum tariffs. With major capacity in Saltillo, Mexico and steel/aluminum/component inputs exposed, the company states tariffs “could directly impact” segment gross margins. OPEN QUESTION: Lennox has not disclosed a quantified tariff dollar exposure — a material gap given the Mexico manufacturing concentration and the USMCA review due in 2026.

8.5 Leadership/board, guidance trajectory, and the drawdown

• Leadership (FACT): CEO Alok Maskara (since 2018) and CFO Michael Quenzer (CFO effective Jan-1-2024) remain in place — no C-suite turnover, a stability positive. At the board level, director Sivasankaran Somasundaram resigned 2026-05-23 (“other professional responsibilities,” no disagreement), shrinking the board from 9 to 8 (8-K, 2026-05-28).
• Guidance trajectory (FACT): FY2026 guide (set 2026-01-28): revenue +6–7% (incl. ~4% from acquisitions), adjusted EPS $23.50–$25.00, capex ~$250M, FCF $750–850M. After Q1-2026 (2026-04-29), management reaffirmed the full-year EPS range — no cut, but no raise despite a Q1 beat ($3.35 adj. vs ~$3.20 consensus). INTERPRETATION: the reaffirm-after-beat signals management is bracing for a soft balance of year (the residential volume/payback drag), not banking the upside.
• Drawdown timeline (FACT/INTERPRETATION): the stock fell ~25% from a 52-wk high of $689 to ~$518 (2026-06-08). Catalysts were the 2025 quarterly prints flagging residential sell-through weakness and extended channel de-stocking (Q3-2025 commentary), and a string of 2025 sell-side target cuts (e.g., Goldman, UBS to ~$560–615, KeyBanc flagging residential de-stocking into 2026; Investing.com, 2025). The market is de-rating peak-cycle margins and pricing in the volume air-pocket.

Verdict

Net, the last two years are roughly thesis-neutral with a negative skew on quality-of-earnings. The genuine positives are real but second-order: the Europe exit and Heatcraft consolidation cleaned up the portfolio; the Duro Dyne/Supco bolt-ons add higher-margin parts/attachment revenue; the Samsung JV and Saltillo plant create optionality in ductless and light-commercial. But the dominant fact is that the 2024 pre-buy (~$125M) borrowed demand that 2025–26 is now repaying via a −17% to −21% residential volume decline, and that the resilient reported revenue/margin is being held up by price/mix and acquisitions, not organic volume — guidance itself concedes organic growth of only ~2–3%. Layer on un-quantified, escalating tariff exposure concentrated in Mexican manufacturing and a 2026 USMCA review, and the changes do not strengthen the structural-improvement narrative; they reframe a chunk of the margin expansion as cyclical/regulatory rather than durable. The board/management continuity is a modest positive; the reaffirm-don’t-raise guidance posture and the analyst de-rating corroborate the cautious read. On balance: weaken, pending evidence that volumes trough and the A2L price premium sticks.

8A. SEC Filing Sweep & Insider Transaction Read

Scope: full trailing-36-month SEC corpus for Lennox International (NYSE: LII; CIK 0001069202), enumerated from EDGAR (filings since 2023-06-01). Form breakdown over the window: 153 Form 4, 51 Form 144, 27 8-K, 9 10-Q, 6 Form 3, 3 10-K, 3 DEF 14A/DEFA14A, 2 424B5, 1 S-3ASR, 2 Form 5, 8 Schedule 13D/G. Code-noise forms (424B/FWP/144) excluded from the event read except where load-bearing.

1. Material-Event Timeline (8-K corpus, FY2023 → 2026-05)

Bolt-on M&A also disclosed in 10-Ks (not standalone 8-Ks): AES (Oct 2023), plus Duro Dyne (Building Climate Solutions) and Supco (Home Comfort Solutions), and the Ariston water-heater JV (49.9%). All small, tuck-in, North-America-focused — consistent with a disciplined, low-multiple bolt-on cadence rather than transformational M&A. (Hand to Capital Allocation for price/multiple scrutiny — deal economics are not disclosed at the 8-K level. OPEN QUESTION.)

2. Insider-Transaction Read (153 Form 4s, 467 line-items)

I parsed all 153 Form 4 XMLs programmatically. Transaction-code distribution across 467 line-items:

Headline finding (FACT): zero discretionary open-market purchases (code P) by any insider in the trailing 36 months. Every share that moved into insider hands did so via equity comp (A/M), not cash purchases. There is no “skin-in-the-game buy” signal at any price — including across the ~25% drawdown from the $689 high. (Source: parsed Form 4 corpus, EDGAR, accessed 2026-06-08.)

Sales were ~$60M gross, concentrated in departing/peripheral insiders, and routine. ~145 open-market sale transactions (≈210 reported line-items) totaled ~105,700 shares / ~$59.7M gross proceeds. By insider:

The Form 4s carry the Rule 10b5-1 plan indicator on the selling filings (e.g., Norris’s July-2025 Form 4), i.e., pre-scheduled diversification, not discretionary conviction selling. INTERPRETATION. The largest seller, Norris (founding family), is trimming a legacy stake on a plan — a wealth-diversification signal, not a thesis signal.

The two principals who matter sold nothing into the market. CEO Alok Maskara: across the window, only grants (A) and tax-withholding (F) — zero code-S sales; he has retained every net share. CFO Michael Quenzer: same pattern — grants, tax-withholding, one option exercise (M) in Oct-2024 with no associated open-market sale — zero code-S sales. FACT (parsed Form 4 corpus). For a stock down ~25% from its high, the absence of CEO/CFO selling is the most constructive insider data point available, but retention of granted stock is a weak positive — it is not the same as an open-market purchase with personal cash. INTERPRETATION.

Form 144 (51 filings): these are notices of intent to sell restricted stock; they track the same routine 10b5-1 selling above and add no independent signal beyond confirming sale volume is planned and modest relative to a ~$18B cap. The 51-to-153 ratio (notices to executed Form 4s) is consistent with ordinary plan administration, not a wave of urgent disposals.

3. One-Time / Run-Rate-Distorting Items (36-month set)

• FY23 held-for-sale impairment of $63.2M and the European divestiture (net +$10.3M gain, +$14.1M in operating income) — both below the headline margin trend; FY23 operating income of $791.5M absorbed them. The bigger distortion is on the revenue base: Europe’s removal makes FY24-onward an apples-to-oranges top-line comparison (Puzzle 3 — divestiture is a real chunk of the FY25 “decline,” alongside pre-buy payback and resi-volume softness; quantify with Financials).
• FY24 R-410A pre-buy is the single largest run-rate distortion (Puzzle 1/2): it pulled residential demand into 2024 and flattered FY24 volume/margin; FY25’s -2.7% revenue is partly this payback, not pure demand erosion. The 8-Ks confirm the timing (earnings cadence) but the magnitude must come from the MD&A/transcripts — hand to Financials.
• Restructuring is immaterial: $6.8M FY25, $0 FY24, $3.1M FY23 — not a cost-out lever; margin expansion is not restructuring-driven. This supports the price/mix-and-operating-discipline (not one-time) read of margin gains.
• Pension settlements trivial ($0.4M FY25/FY24, $0.8M FY23). No litigation/insurance one-timers of size surfaced (asbestos/environmental reserves are ordinary-course and disclosed in Note 5).
• Capital structure event: the Sept-2023 $500M notes due 2033 is the only new debt issuance in the window; the 2025 S-3ASR is undrawn capacity. Net debt ~$1.9B is term-funded, not revolver-dependent (constructive for leverage durability).

4. Short-Interest Context

Short interest is elevated at ~8.4% of float, short ratio ~4.6 (third-party, accessed 2026-06-08) — high for a quality industrial compounder (peers TT/CARR typically run 1-2%). Reconciled against this sweep, the short case is not an accounting-fraud or insider-dumping thesis (insiders aren’t selling discretionarily, accounting one-timers are immaterial, debt is termed out). The crowded short is therefore most consistent with a peak-margin / pre-buy-payback / multiple-de-rate bet (Puzzle 1, 2, 6) — a fundamental valuation short, not a forensic one. INTERPRETATION; hand to Variant Perception.

Verdict

The 36-month filing record is clean and undramatic — no forensic red flags, no transformational/empire-building M&A, no debt-funded financial engineering. The only structural events are a sensible North-America refocus (European exit, Samsung/Ariston JVs, small tuck-ins), a routine 2023 bond, and a 2025 capital-return step-up (+$1.0B buyback, 13% dividend hike) executed into a soft revenue year — which Capital Allocation should weigh as either confident counter-cyclical buying or pro-cyclical buyback timing near a peak multiple.

On insiders, the signal is mildly constructive but not a conviction tell. The decisive facts are negative-of-a-negative: zero open-market purchases by anyone (so no bullish skin-in-the-game even after a 25% drawdown), but also zero discretionary CEO/CFO selling — Maskara and Quenzer have retained all net granted stock, and the ~$61M of insider sales is routine, 10b5-1-planned, concentrated in a founding-family director (Norris) and departing/peripheral officers. Net read: insiders are net sellers only in the mechanical comp sense; on a signal basis the activity is neutral-to-slightly-positive (no one is bailing), but it offers no affirmative bottom-fishing endorsement. The elevated ~8.4% short interest is best explained as a peak-margin/valuation short, which this clean filing record neither confirms nor refutes — it must be adjudicated on the margin-durability and pre-buy-normalization analysis, not on filing-corpus forensics.

9. Risk Analysis

Lennox is a high-quality, cash-generative business, but it enters 2026 carrying a stack of risks that are unusually correlated with one another — peak margins, a demand air-pocket, a refrigerant transition still in its tail, and a tariff regime in flux all push the same direction (down) in a soft-demand scenario. The matrix below scores each risk by likelihood and impact, then the prose walks the mechanism and the disconfirming evidence. The single most important framing point, in Marathon “Capital Returns” terms: operating margin has expanded ~610 bps (13.9% FY22 → 20.0% FY25) into a price-led HVAC up-cycle (FACT — MD&A, FY25 10-K p.21, filed 2026-02-17). High returns attract capital and mean-revert; the burden of proof is on durability, not reversion.

Risk Matrix

Peak-margin mean-reversion (Likelihood M / Impact H)

This is the central risk. Operating margin moved from 13.9% (FY22) to 20.0% (FY25), and FY25 gross margin gained 290 bps purely from higher price and favorable mix, against which higher product cost (−160 bps) and freight/distribution (−120 bps) only partly offset (FACT — FY25 10-K MD&A p.21–22). Roughly 9 points of the FY25 price/mix benefit came while volume fell 13% (FACT — same). Price-led margin in a self-described competitive market (Item 1A p.8) is the textbook Marathon set-up: returns this far above the historical band invite capital and competitive response, and they mean-revert when the pricing window closes. Disconfirming evidence does exist — SG&A fell 60 bps on genuine cost-out and lower incentive comp, and the direct-to-dealer model creates structural pull-through (Item 1A p.8) — so some of the gain is real operating leverage, not just price. But until margin holds through a down-volume, normal-price year, durability is unproven. INTERPRETATION: treat ~20% as a cyclical peak that the market may be capitalizing as a permanent base.

Residential-volume / housing-turnover / interest-rate sensitivity (Likelihood H / Impact H)

The largest segment, Home Comfort Solutions, is residential replacement and new-construction equipment, both correlated to rates, financing availability, and housing turnover (FACT — Item 1A p.9–10). The risk is already crystallizing: HCS net sales fell 7% with segment profit down $32M in FY25, and the weakness accelerated into Q1-26 (HCS sales −10%, profit −$37M) on “market softness” (FACT — FY25 10-K p.21; Q1-26 10-Q p.20). With rates still elevated in 2026, replacement deferral and weak new-construction are a live, not hypothetical, drag. This risk is high-likelihood and high-impact because it hits the highest-margin, highest-volume segment.

Pre-buy payback air-pocket (Likelihood H / Impact M)

The FY24 transition to low-GWP refrigerants triggered customers to pre-purchase R-410A equipment, which Lennox estimated added ~$125M to FY24 revenue (FACT — FY24 10-K MD&A). That pulled demand forward, mechanically depressing FY25/26 volume — consistent with the −13% FY25 volume print and the continued HCS decline in Q1-26. The payback is partly self-inflicted and partly indistinguishable from organic softness, which is the analytical hazard: management can attribute weak volume to “normalization” indefinitely. OPEN QUESTION: how much of the FY25–26 volume decline is pre-buy payback (transitory) versus genuine end-demand softness (structural to the cycle)? The two have very different multiple implications.

Refrigerant-transition execution / A2L field reliability (Likelihood M / Impact M)

The shift to A2L refrigerants (R-454B) introduces mildly flammable refrigerant into a residential installed base for the first time. Lennox provides warranties of 1–20 years on components like compressors and heat exchangers, and warranty liability stood at $167.2M at FY25 (FACT — Item 1A p.11). New chemistry plus long warranty tails plus field-installation variability is precisely the profile that produces deferred warranty/recall surprises (Item 1A explicitly flags recall and field-corrective-action risk). No incident is disclosed yet, so this is a tail risk, not a base case — but it is asymmetric and would hit both the P&L and the brand reputation the moat partly rests on.

Tariffs and commodity input cost (Likelihood H / Impact M)

Lennox depends on steel, copper, and aluminum and a concentrated supplier base (FACT — Item 1A p.10). The 2026 tariff backdrop is unusually fluid: after the Supreme Court struck IEEPA tariffs (Feb-26), the administration imposed a new 10% global Section 122 surcharge, and Section 232 steel/aluminum tariffs were modified again in April-26 (FACT — Q1-26 10-Q p.20). FY25 already absorbed +160 bps of higher product cost (FY25 10-K p.22). The mitigant is pricing power and commodity hedging (Item 1A p.10–11; derivatives note), but pushing price into a soft-volume residential market is exactly where pricing power gets tested. High likelihood (tariffs are here), medium impact (historically passed through, with a lag).

Competitive / pricing pressure (Likelihood M / Impact M)

Lennox concedes “substantially all” its markets are competitive and that barriers are low in parts, distribution, and commercial service (FACT — Item 1A p.8), and that low-cost-country manufacturing could hand rivals a cost edge. The direct-to-dealer model is a real demand-side advantage (captive installed base, retrofit pull-through), but it is not insurmountable, and value-tier competition from Daikin/Goodman and the scale of Carrier and Trane caps pricing over a full cycle. The risk bites specifically because margin is currently price-led: a competitor willing to trade margin for share in a down-volume year would expose how much of the 290 bps is durable.

Concentration in North America (Likelihood M / Impact M)

The Q4-23 European divestiture and refrigeration exit left non-US sales at ~7% of the total (FACT — Item 1A p.13). The strategic logic (focus, higher margins) is defensible, but it removes geographic diversification: Lennox is now a near-pure bet on one North American HVAC cycle, with no offsetting region when that cycle softens — which it currently is.

Balance-sheet optics, quality-of-earnings, and litigation (Likelihood L–M / Impact L–M)

Years of buybacks left treasury stock at a cumulative −$4,923.4M and equity historically near/below zero; equity has since recovered to $1,163.1M (FY25) from $962.1M (FY24) (FACT — BS p.39). The optical “deficit” is a capital-return artifact, not distress — but the leverage underneath is real: cash fell to $34.2M (from $415.1M) against $1,388.4M total debt, partly commercial-paper funded (FACT — BS p.39; debt note). In a sharp downturn with CP markets tightening (a risk Item 1A names, p.13), refinancing flexibility is thinner than the headline. On quality-of-earnings, FY25 CFO of $758M fell below net income of $806M and dropped from $946M in FY24, while inventories sat at $1,152.6M — consistent with an A2L/working-capital build (FACT — p.20; BS p.39). Not alarming, but a watch-item: it flatters reported earnings relative to cash. Litigation (asbestos reserve $18.6M; FY25 environmental/litigation charge $10.9M) is routine for an industrial of this vintage and low-impact (FACT — accrued-expense note; p.22).

Key-person and multiple de-rating (Likelihood L–M / Impact M)

The margin transformation is closely identified with CEO Alok Maskara (since 2018); no public succession plan is disclosed (INTERPRETATION — proxy). Finally, the market has already begun pricing some of these risks: the stock sits ~25% below its $689 52-week high at ~22.7x trailing P/E with short interest of ~8.4% of float, and P/S at the 83rd own-history percentile (FACT — market data 2026-06-08). The de-rating risk is that consensus is still capitalizing peak margin; if FY26 confirms the air-pocket, the multiple compresses on lower earnings — a double hit.

Verdict

The risk profile is asymmetric to the downside in 2026, and the risks are correlated, not independent — peak margin, the volume air-pocket, the refrigerant payback, and tariff cost all worsen together in a soft-demand year, which is the scenario the data already shows beginning (FY25 volume −13%; Q1-26 HCS −10%). The catastrophic-loss risk is low: this is a debt-serviceable, FCF-generative, brand-and-distribution business, not a fragile balance sheet despite the optical equity history. But the thesis risk is high — the entire bull case rests on ~20% operating margin being a durable structural base rather than a price-led cyclical peak, and the burden of proof (Marathon’s mean-reversion prior) is unmet. The most underrated risk is the one the matrix scores M/H: that consensus has extrapolated peak unit economics into perpetuity just as volume rolls over. The most overrated is the balance-sheet “deficit,” which is a buyback artifact, not distress. Net: a good business at a point in the cycle where the risk/reward is dominated by margin-durability uncertainty.

10. Valuation Discussion

Embedded-expectations framing. No price target, no recommendation. The question is not “what is LII worth” but “what does $517.78 already require to be true,” and whether that required future is the bull’s durable-20%-margin compounder or the bear’s peak-of-transition mean-reverter.

10.1 Where the stock trades, and against what

As of the 2026-06-08 close of $517.78 (FACT, market data baseline), LII carries ~34.8M shares, a ~$18.0B market cap, ~$1.35B net debt (total debt $1,388.4M − cash $34.2M, per the FY25 10-K) and an enterprise value of ~$19.4B. On the verified FY25 base (revenue $5,195.3M; EBIT $1,041.5M; D&A $112.5M → EBITDA ~$1,154M; net income $805.8M; diluted EPS $22.79 — all reconciled to the FY25 10-K filed 2026-02-17, lii-20251231.htm), that prices LII at:

The stock sits ~25% below its 52-week high of $689.44 (FACT, market baseline). That drawdown, on a business whose FY25 EPS was up (+0.6%) and whose operating margin hit a record 20.0%, is itself the central valuation fact: the market is de-rating the multiple, not the current earnings. The de-rate is the embedded judgment that FY25 earnings are nearer a peak than a base. (INTERPRETATION.)

10.2 Peer comps — LII screens cheap to quality peers, near Carrier

Live multiples (yfinance, 2026-06-08; reconcile trailing P/E to filings — yfinance’s 14.3% operating margin and 23.0x P/E for LII are TTM/aggregator artifacts that mix periods, so anchor LII’s operating margin to the verified 20.0% 10-K figure and use the table for relative positioning, not absolute precision):

Source: yfinance peer comps (LII, CARR, TT, AAON, JCI, WSO), 2026-06-08.

LII is the cheapest of the group on trailing P/E (~22.7x) and EV/EBITDA (~17.0x) — a discount of roughly 20-30% to Carrier/JCI and ~30%+ to Trane on EBITDA, and a fraction of AAON’s nosebleed ~45x (FACT). The caveats that defang the “LII is cheap” reflex:

• AAON’s 94x P/E / 45x EBITDA is not a clean comp — its +54% revenue growth is acquisition- and data-center-cooling-driven (BasX), a different (and more speculative) growth profile; it deserves a premium and is best treated as an outlier, not a benchmark. (INTERPRETATION.)
• Trane is the quality benchmark — the share-leader (~23% N.A.) with the broadest commercial/applied mix and the most diversified geography; its ~25x EBITDA reflects that the market pays up for diversification and applied-systems durability LII lacks as the smallest, N.A.-only major. (INTERPRETATION, industry section.)
• Carrier is the nearest peer on multiple (~21x EBITDA, ~19-21x fwd P/E) and the most apt read-across: similar pure-play HVAC tilt post-Viessmann/Fire&Security reshuffle. LII trades a modest discount to Carrier — consistent with LII’s smaller scale and zero geographic diversification, partially offset by LII’s structurally higher margins (20.0% EBIT vs Carrier’s mid-teens). (INTERPRETATION.)

Read-through: LII is not mispriced as deep value; it is priced as the lowest-multiple name in a high-multiple group precisely because the market is unwilling to pay a Trane/AAON multiple for margins it suspects are peak-of-transition (the price-led 13.9%→20.0% surge documented in the Financial Quality and Capital Allocation sections). The discount is the market’s skepticism, not a free lunch. (INTERPRETATION — this is the variant-perception crux.)

10.3 Versus its own history — mid-range, not washed-out

LII’s own-history valuation percentiles (third-party valuation-percentile data; 0 = cheapest in ~10y, 100 = most expensive) place the stock mid-pack against itself, not at a trough:

Source: third-party own-history valuation-percentile data, 2026 baseline.

The honest reading: discard P/B (negative-then-tiny equity makes it noise), discount nothing from the 30th-percentile P/E but note the 83rd-percentile P/S as the tell — investors are paying near-peak per dollar of revenue because they have re-rated to the new margin structure, even as they de-rate the earnings multiple on doubt that those margins hold. The two facts are reconciled by exactly the peak-margin debate: the market half-believes the margins (P/S high) and half-doubts them (P/E below median). Composite 39th ≈ “fairly valued vs its own past, not cheap, not dear.” (INTERPRETATION.)

10.4 Embedded expectations — what $517.78 demands

Reverse-engineering the multiple (EPV vs. growth split). A no-growth earnings-power value (EPV) anchors what you pay for the current business with zero future growth credited. Using normalized FY25 NOPAT ≈ $841.5M (EBIT $1,041.5M × (1−19.2% tax); see Financial Quality) and a ~9% pre-tax-equivalent cost of capital for a low-beta, lightly-levered industrial, an unlevered EPV ≈ $841.5M / 0.09 ≈ ~$9.35B, less ~$1.35B net debt ≈ ~$8.0B equity EPV, or ~$230/share of pure no-growth earnings power. Against the $18.0B / $517.78 market, that means ~$10.0B of the ~$18.0B market cap (≈ 56% of equity value, ≈ $288 of the $518 price) is being paid for growth and/or margin-durability beyond the current run-rate. (ASSUMPTION-heavy: EPV is sensitive to the discount rate — at 8% EPV-equity ≈ $9.2B/~$263, at 10% ≈ $7.1B/~$203 — but the order of magnitude is robust: a clear majority of the price is expectations, not capitalized current earnings.)

Cross-check via a simple reverse-DCF on FCF. Capitalize the ~$639M FY25 FCF (FCF $757.6M CFO − $118.8M capex; see Financial Quality). To support a $19.4B EV at a 9% WACC and ~3% terminal growth requires a perpetual FCF stream consistent with a starting FCF near ~$1.06B (a Gordon EV of $639M × growth/discount math falls well short of $19.4B at 3% terminal), i.e., the market is underwriting meaningful FCF growth off the FY25 base, not flat FCF. Management’s own FY26 guide of $750-850M FCF and EPS $23.50-25.00 on ~8% revenue growth (Lennox Q1 2026 release/guidance, StockTitan/Investing.com, 2026-04-29) is the bridge: the price needs that FCF step-up to recur and compound, not be a one-year A2L-inventory-unwind bounce. (INTERPRETATION.)

What must be true at $517.78 (the embedded bet):

1. Margins hold near 20% operating — i.e., the price/mix-driven expansion is structural, not transition-peak. If margins mean-revert to high-teens (industry section’s base risk as the A2L price-up annualizes), trailing EPS power is ~10-15% lower and the multiple is really ~25-26x, not 22.7x. (The Q1’26 print — EPS −8% to $3.35, HCS profit −30%, ~17% incremental gross margin on the way down — is early evidence the market is right to doubt.) (FACT on Q1; INTERPRETATION on extrapolation.)
2. Volume troughs and turns — the −13% FY25 / −8% Q1’26 organic-volume decline (pre-buy payback) bottoms in H2’26 and resumes growth, so price/mix is no longer offsetting shrinking units. BCS (+38% Q1’26 revenue, margin 19.7%) must keep carrying the mix.
3. MSD revenue + buyback compounding delivers high-single/low-double-digit EPS growth for several years — the guided ~8% FY26 growth is largely price; durable double-digit EPS needs volume to return and the ~$300-500M/yr buyback to continue shrinking the count (~37% reduction since FY10; see Capital Allocation).

If those hold, ~22.7x trailing / ~19.4x forward is undemanding for a 30%±ROIC compounder. If margins are peak and volume stays negative, the same price is ~mid-20s× on normalized earnings — i.e., fully valued to expensive. The stock is the embodied disagreement on item (1). (INTERPRETATION.)

10.5 Scenario analysis (framework, NOT a target)

Explicit, falsifiable scenarios. Exit multiple applied to the implied EBITDA; equity = EV − ~$1.35B net debt; per-share on ~34.5M shares (declining with buyback). These are conditional ranges to stress-test the thesis, not a valuation opinion or recommendation.

(ASSUMPTION: each row’s exit multiple, margin and growth are stated, not derived from a model the reader can’t see; D&A held ~flat near $115-125M; net debt roughly stable net of buyback. The bull-case ~$775-800 zone is notable because it approximately recovers a prior, more-optimistic ~$737 fair-value framing circulated in 2025 — useful as a calibration: that level is only reachable on the bull path of durable structural margins + volume recovery + a re-rate, which FY25’s revenue decline and the Q1’26 margin slip have not yet validated. The base case sits modestly above the current price; the spread between bear (~low-$300s) and bull (~$775-800) is almost entirely the margin-durability question.) (INTERPRETATION.)

The asymmetry the table reveals: at $518 the downside-to-bear (~38-40%) is comparable to (slightly smaller than) the upside-to-bull (~+50-55%), with the base case modestly above spot — i.e., the market has priced LII close to a probability-weighted midpoint, and the payoff is governed almost entirely by which way the margin question resolves. This is a coin-flip-on-margin-durability setup, not an obvious bargain. (INTERPRETATION.)

10.6 FCF yield, shareholder yield, and the short signal

• FCF yield ~3.5% (~$639M / $18.0B). Thin in absolute terms (below a ~4.2% 10-year-equivalent risk-free hurdle, so the equity is priced for growth, not income) but understated as a forward figure: the FY25 FCF was depressed by a deliberate working-capital build (a ~$300M gross inventory build for the A2L transition; ~$228M net working-capital cash drag) (see Financial Quality). Normalizing that working-capital drag, “clean” FCF is nearer ~$800M+ (≈ the FY26 guide), an ~4.4-4.7% normalized FCF yield — middling for the quality. (FACT/INTERPRETATION.)
• Shareholder yield ~3.5-3.8%: ~1.07% dividend + ~2.5-2.7% buyback (FY25 repurchases ~$482M / $18.0B ≈ 2.7%; see Capital Allocation). Respectable, buyback-tilted, but the FY25 buyback was executed at ~$536/share / ~23x P/E — i.e., near the cyclical/valuation peak (Marathon: high reported returns drew the company’s own capital in at a rich price), and Q1’26 repurchases collapsed to ~$20M, suggesting management itself recognized the price. (FACT; INTERPRETATION.)
• Elevated short interest — sources disperse from ~5.9% of float (Fintel/MarketBeat, ~Mar-2026) to ~8.4%/short-ratio-4.6 on another measure; either way it is high for a quality industrial (Fintel LII short-interest, accessed 2026-06-08). The SEC sweep found a clean filing record with zero insider open-market buys and zero discretionary CEO/CFO sales — so the short is best read as a valuation/peak-margin/pre-buy-payback short, not a forensic one. It is a coherent variant-perception signal that institutional capital is positioned for exactly the bear-case margin reversion above; it is not a balance-sheet or accounting bet. A short that size also creates squeeze risk on any clean H2’26 volume-recovery print. (INTERPRETATION.)

10.7 Verdict — what the price demands

At ~$517.78 (~22.7x trailing EPS, ~17x EV/EBITDA, ~3.5% FCF yield), the market is not underwriting LII as a cheap stock and is not capitulating to a cyclical bottom. It is pricing LII as the lowest-multiple name in a premium HVAC group specifically because it doubts the durability of the record 20% operating margin. Decomposed: roughly ~$230/share is no-growth earnings power (EPV) and the remaining ~$288 — the majority of the price — is paying for margins-hold-plus-MSD-growth-plus-buyback-compounding. The single load-bearing embedded assumption is that the 13.9%→20.0% margin surge was structural rather than a forced-refrigerant-transition price-spike now annualizing into mean-reversion. The 83rd-percentile P/S (market has re-rated to the new margins) sitting beside the 30th-percentile P/E (market discounts the earnings multiple on doubt) is the quantitative signature of that unresolved bet, and the elevated short interest plus the −8% Q1’26 EPS print are the bears pressing it. The scenario spread is near-symmetric around spot (~low-$300s bear / ~$540-570 base / ~$775-800 bull), so the embedded expectation is “prove the margins are real, and the modest multiple is too low; fail to, and the modest multiple is correct or generous.” The price demands durable ~20% margins and a return to positive unit volume — neither yet confirmed in the data. No price target or recommendation is expressed; this is the conditional structure the current price already contains.

11. Variant Perception

Variant perception is the gap between what the market believes and what the evidence supports. For Lennox the debate is unusually crisp: it reduces to a single, falsifiable question — is the ~20% operating margin a durable structural base or a price-led, refrigerant-transition cyclical peak? Everything else (the buyback compounding, the Samsung optionality, the tariff overhang, the 8.4% short) is downstream of that one question. This section lays out the consensus, the strongest bull and bear cases on the evidence, the three to five assumptions that decide the outcome with the specific data that would falsify each, and where a skeptical reading lands.

11.1 The Consensus View

The Street’s posture is Hold/Neutral with a modestly positive price skew but no conviction. As of June 8, 2026 the consensus rating is a Hold built from roughly 5 Buy / 10 Hold / 2 Sell across ~17 covering analysts, with an average price target around $575 (range $450–$691) against a $517.78 quote — implied upside ~11% (FACT — MarketBeat LII forecast, accessed 2026-06-08; third-party aggregator mean target $570, accessed 2026-06-08). The dispersion is itself a finding: a $450–$691 spread (low ~13% below, high ~33% above spot) is a ±20-point band that signals genuine disagreement, not a settled view.

What does consensus actually believe? Three things, inferred from estimates and post-Q1 actions (INTERPRETATION):

1. Margins are near a structural plateau, not extending. FY26 EPS estimates cluster around ~$24.33 — squarely inside management’s reaffirmed $23.50–$25.00 guide (FACT — Q1-2026 call, reported 2026-04-29; FY26 EPS consensus per stockanalysis.com / Investing.com, accessed 2026-06-08). The Street is underwriting roughly flat earnings on ~8% revenue growth — i.e., it has stopped paying for further margin expansion and is bracing for slight margin give-back, exactly what CFO Quenzer guided (“a slight decline in the margin”). The ~22.7x trailing / ~20.8x forward P/E (FACT — third-party aggregator snapshot, accessed 2026-06-08) sits at the low end of the HVAC group (Trane, Carrier, AAON trade ~25–45x on various measures) — a relative de-rate that says the market is no longer capitalizing peak unit economics as permanent.
2. The bear seat is occupied but not dominant. JPMorgan and Morgan Stanley both carried Underweight ratings into the period (FACT — maintained 2025-10-23, per Benzinga/MarketBeat, accessed 2026-06-08), and Northcoast and Wolfe were upgrading off Sell/Underperform (to Neutral / Peer Perform) only in late 2025 — i.e., the most negative analysts were closing shorts into a stabilizing tape, not adding. Post-Q1-2026, RBC raised its target $485→$579 and Wells Fargo $480→$555, Vertical Research upgraded Hold→Buy, while Barclays trimmed $600→$597 (FACT — MarketBeat, dated 4/30/2026). Net: the Q1 beat-and-reaffirm pulled the low targets up toward the pack, compressing the bear tail without producing a bull stampede.
3. The model is “fine, fairly valued, cyclically de-risking.” Consensus treats Lennox as a quality industrial whose 2025–26 air-pocket is largely behind it — hence Hold at a market-to-slight-premium multiple rather than the growth-premium multiple peers command.

11.2 The Strongest Bull Case

The bull case is coherent and rests on real evidence, not narrative:

• A genuine, management-led margin step-change. Operating margin went 13.9%→15.9%→19.5%→20.0% over FY22–FY25 (FACT — verified to EDGAR XBRL OperatingIncomeLoss). The bull argues a durable slice is structural: SG&A fell ~60 bps on real cost-out and the direct-to-dealer model (200+ company-owned “Lennox Stores,” 76% of HCS revenue sold direct) creates pull-through and pricing control that an indirect competitor lacks (FACT — FY25 10-K MD&A; Item 1). ROIC of ~33% (NOPAT $841.5M / invested capital ~$2,517M) is the honest profitability signal and it is excellent — well above cost of capital and above most peers (FACT — Financials section, computed from FY25 10-K).
• Replacement-demand resilience. ~63%+ of HVAC installs are non-discretionary replacement, weather- and failure-driven rather than credit/sentiment-driven, which underpins the industry’s resilient high-teens-to-low-20s margins through cycles (FACT — Industry section). The installed base is an annuity.
• A2L/R-454B pricing is a regulation-enabled, durable price-up. The Jan-1-2025 EPA AIM-Act transition raised equipment prices ~10%+ industry-wide; the bull contends this is a permanent step in the price level, not a one-time spike, and Quenzer’s “for the balance of the year, it is essentially all price” supports that pricing is holding without needing mix help (FACT — Q1-2026 call).
• Buyback compounding. Diluted share count fell ~37% (55.8M FY10 → 35.4M FY25) on ~$2.94B of FY16–FY25 open-market repurchases; even flat net income compounds EPS at mid-single digits via shrink (FACT — Capital Allocation section).
• Optionality, cheap. The 49.9% Samsung ductless JV addresses Lennox’s historical under-indexing in mini-splits (a structurally growing N.A. category) at capital-light cost, and BCS (commercial/national-accounts) is genuinely growing — +26% organic sales in Q1-2026 with +300 bps margin (FACT — Q1-2026 call). Bull view: HCS volume troughs in H2-2026 and BCS carries the company through.

11.3 The Strongest Bear Case

The bear case is, in this analyst’s read, the better-supported one on current evidence:

• Peak margins mean-revert (Marathon). The margin surge is price-led, not cost-led. Management’s own FY24 segment bridge shows HCS profit gains of $122M from price/mix versus only $9M from productivity; FY25 GM gained ~290 bps purely from price/mix while volume fell 13% (FACT — FY24/FY25 10-K MD&A). Returns this far above the historical band (op margin was 13.9% as recently as FY22) invite competitive response and mean-revert when the pricing window closes — the textbook capital-cycle set-up. The first reversion signal has already printed: HCS segment margin cratered ~390 bps to 13.3% in Q1-2026 as ~$50M of factory under-absorption hit on −21% organic volume (FACT — Q1-2026 10-Q / call).
• The FY24 pre-buy air-pocket is still venting. The R-410A pre-buy added ~$125M to FY24 revenue (management’s own estimate, FY24 10-K MD&A); the payback drove HCS volume −17% in FY25 and −21% organic in Q1-2026 (FACT). Reported revenue resilience is price + acquisitions, not units — Q1-2026 consolidated was +9% price/mix, −8% organic volume, +5% M&A; FY26’s ~8% revenue guide embeds ~4% from the Duro Dyne/Supco bolt-ons, implying only ~2–3% organic (FACT — Q1-2026 call; Growth section).
• The reaffirm-don’t-raise tell. Despite raising the FY26 revenue guide (~6–7% → ~8%) after a Q1 EPS beat, management held EPS at $23.50–$25.00 because the margin outlook turned to “a slight decline” (FACT — Quenzer, Q1-2026 call). Higher revenue, same earnings = the incremental dollars are coming in at compressing margin. That is the structural-improvement thesis weakening in real time.
• Tariff and concentration risk. Section 232 steel/aluminum tariffs “caught management by surprise” in April 2026 (Maskara: aluminum +25%, steel +20–25%, diesel +50%); Lennox is the smallest major with zero geographic diversification post the Europe exit and a large Mexican (Saltillo) plant exposed to USMCA-2026 review — less scale than Daikin/Carrier/Trane to absorb commodity/tariff shocks (FACT — Q1-2026 call; Industry/Changes sections).
• Paying ~22.7x for a cyclical with declining revenue. FY25 revenue fell −2.7%; the stock trades ~22.7x trailing on what may be peak earnings, with P/S at the 83rd own-history percentile — i.e., cheap on P/E only if the “E” is durable (FACT — third-party own-history valuation-percentile data; market data 2026-06-08). If FY26 confirms the air-pocket, the double-hit is lower earnings and multiple compression.
• The 8.4% short — who and why. Short interest is elevated at ~8.4% of float / ~4.6 days to cover (third-party data, 2026-06-08), up from exchange-reported ~5.8% / ~3.9 days in Sep-2025 — high for a quality compounder (Trane/Carrier run ~1–2%). The SEC-sweep adjudication is decisive: the filing record is clean (no accounting one-timers, no discretionary insider selling, debt termed out), so this is not a forensic/fraud short — it is a fundamental peak-margin / pre-buy-payback / multiple-de-rate short placed by funds betting the ~20% margin is a transition peak (INTERPRETATION — based on the SEC-filing review and financial analysis). The shorts and this analyst share a thesis.

11.4 The 3–5 Assumptions That Matter Most (and What Falsifies Each)

Assumptions 1 and 2 are the master variables — 3, 4, and 5 are the mechanisms through which they resolve. The single cleanest falsification for the whole bear thesis is Assumption 1’s bear-falsifier: if Lennox holds ~20% operating margin in a year of positive HCS unit volume, the margin is structural and the bears (and this analyst) are wrong.

11.5 Where This Variant Perception Sits — and What an Earlier Optimistic Call Got Wrong

Our reading is modestly more bearish than consensus on earnings quality, broadly in line on the multiple — i.e., we think the Street is correctly declining to pay up for further margin expansion (the relative de-rate to ~21x forward is appropriate) but is still under-appreciating how much of the FY24–FY25 margin and revenue is price/mix and pre-buy rather than durable volume-led operating leverage. The variant view is not “Lennox is a bad business” — ROIC ~33%, clean cash-backed earnings, and disciplined capital allocation are real. It is that the ~20% operating margin should be modeled as a peak-of-transition figure to stress-test, not a base to extrapolate, and that the air-pocket may run longer than the H2-2026 trough the guide implies. The Q1-2026 reaffirm-don’t-raise and the −390 bps HCS margin print are early corroboration; the 8.4% short is a like-minded constituency, not a contrarian crowd to fade.

This is also a direct lesson from an earlier, more-optimistic ~$737 fair-value framing circulated in 2025 (a Buy-equivalent view) — a call that has not played out (FY25 revenue fell −2.7%; the stock is ~$518, ~25% below its $689 high). With hindsight, that view made the canonical mistakes the variant frame is designed to catch: it extrapolated the 2024 print — citing the Q4-2024 HCS revenue +~25% and segment-margin +~550 bps lift without flagging that those were the pre-buy itself, i.e., borrowed 2025 demand, not run-rate; and it capitalized peak margin into perpetuity rather than treating ~20% as a cyclical/regulatory high. The $737 fair value implicitly required the FY24 unit economics to persist and compound — the precise assumption the FY25 −17% volume collapse and the Q1-2026 margin crater have since contradicted. The cross-check is therefore not “that analysis was careless” but “this is exactly how a high-quality compounder lures over-optimism at a refrigerant-transition peak” — a Marathon set-up, and a reminder to weight the disconfirming volume data over the flattering margin optics.

Verdict. The market and this analyst broadly agree on the multiple (~21x forward is a reasonable, de-rated price for a quality industrial whose growth is decelerating), so the variant edge is not in valuation — it is in earnings durability. Consensus is implicitly underwriting ~flat-to-slightly-down margin on a volume trough in H2-2026; the disconfirming evidence (price-led FY24–25 margin, the still-venting pre-buy, the −390 bps Q1-2026 HCS print, the reaffirm-don’t-raise posture, the elevated fundamental short) skews the risk to the downside of even those modest expectations. The honest position: this is a genuinely good business at a fair multiple where the entire debate hinges on one unproven assumption — margin durability — and the burden of proof (Marathon’s mean-reversion prior) is, on today’s evidence, unmet. The two pieces of evidence to watch are mechanical: positive HCS organic volume would vindicate the bull; another quarter of HCS margin compression on negative volume would vindicate the bear.

12. Fact vs. Interpretation

The thesis rests on a small number of load-bearing facts and a larger set of interpretations built on them. Separating the two is the discipline that keeps a “quality compounder” narrative honest.

13. Open Questions

These are the unresolved items most likely to move the thesis; each is a research follow-up, not a settled conclusion.

1. Margin durability (the master question). Is 20% operating margin a structural floor or a refrigerant-transition peak? The only clean test is a full year of positive HCS unit volume with margin sustained ≥18–19%. Q1-2026 (HCS margin 13.3%) is early evidence for the bear; one print is not a trend.
2. Replacement vs new-construction split. The 10-K says only “a substantial portion” is replacement and does not quantify it; the precise residential split (industry estimates range ~63% to ~75–85%) materially affects how cyclical the demand base really is. Needs the investor deck/transcript.
3. How much of the volume decline is pre-buy payback (transitory) vs genuine end-demand softness (cyclical)? Management calls destocking “largely concluded,” but new-construction (~25%+ of HCS) is down sharply on rates. The two have very different multiple implications and are not separable from one improving quarter.
4. A2L pricing persistence. Does the ~10%+ refrigerant-transition list price hold as it laps, or does a competitor (Daikin/Goodman, Carrier) break discipline to take share? Net industry pricing through 2026 is the tell.
5. Tariff dollar exposure. Lennox has not quantified its Section 232/Section 122 tariff exposure despite concentrated Mexican (Saltillo) manufacturing and a 2026 USMCA review — a material disclosure gap.
6. Duro Dyne/Supco price discipline. No purchase-price multiple was disclosed on the ~$545M FY25 acquisition (the largest in years); returns on that capital are unverifiable.
7. Samsung ductless JV scale. As a 49.9% equity-method JV, its contribution sits in equity income (~$6–8M to date) and is not separately quantified — promising adjacency, not yet a numbers story.
8. Data-center cooling. Frequently cited by sell-side, but management gave no data-center revenue commentary on the Q1-2026 call and the 10-K does not size it; exposure appears far smaller than AAON’s. Do not credit it in a base case.
9. EPA R-410A sell-through timeline. A potential extension of the R-410A installation deadline (HARDI flagged ~$8B stranded-inventory concern) could prolong the transition or strand inventory — directly affecting the 2026 demand bridge.
10. European divestiture exact economics. Cross-section figures for the held-for-sale impairment (~$63M FY23) and net gain (~$10–14M) vary slightly by disclosure period; immaterial to the thesis but worth a clean reconciliation to the FY23 10-K.

14. What Must Be True

The position reduces to one variable — margin durability — expressed through volume. Each side has a clean falsification test.

Bull case — what must be true

• The 20% operating margin is structural, not a price spike. A durable slice (cost-out, owned-distribution pull-through, mix) survives once the A2L list-price step laps — i.e., margin holds ≥18–19% in a normal-price year.
• Residential unit volume troughs in H2-2026 and inflects positive, so price/mix stops papering over shrinking units, while BCS (emergency-replacement, national accounts) and the Samsung ductless JV add genuine volume-led growth.
• Buyback compounding continues below intrinsic value, shrinking the count ~2–3%/year and turning even flat net income into mid-single-digit EPS growth; the ~8.4% short becomes squeeze fuel on any clean volume-recovery print.
• Falsification test (kills the bull): another two to three quarters of HCS margin compression (sub-18%) on still-negative organic volume, or a major competitor breaking price to take share. If margins fall with volume, the expansion was cyclical/price, and ~22.7x is on peak earnings.

Bear case — what must be true

• The margin mean-reverts toward the high-teens as the refrigerant-transition price-up annualizes and “factory inefficiencies” recur on low volume (the FY25→Q1-2026 pattern extends).
• The pre-buy air-pocket runs longer than guided, with HCS organic volume negative into 2027 and reported revenue growth carried by acquisitions (~4% of the ~8% FY26 guide) and price, not units.
• The multiple de-rates further as the market marks earnings down to normalized levels — a double hit of lower EPS and lower multiple — with tariffs/commodities compressing gross margin >150 bps un-recovered.
• Falsification test (kills the bear): Lennox holds ~20% consolidated operating margin in a year of positive HCS unit volume. If margin is defended on returning volume rather than price, the step-change is structural, the bears (and this analyst’s cautious lean) are wrong, and the modest multiple is too low.

The honest synthesis: this is a high-quality business at a fair price where the bull and bear scenarios are nearly symmetric around spot and the outcome is governed almost entirely by whether the record margin is real. On today’s evidence — price-led FY24–25 margins, a still-venting pre-buy, the −390 bps Q1-2026 HCS print, a reaffirm-don’t-raise guidance posture, and an elevated fundamental short — the burden of proof is unmet, and the risk skews modestly to the downside of even consensus’s already-cautious expectations. The two things to watch are mechanical and public: positive HCS organic volume vindicates the bull; another quarter of HCS margin compression on negative volume vindicates the bear.

APPENDIX A — Standard Diligence Questionnaire

Lennox International Inc. (NYSE: LII) — supplemental to the research memo. Answers are grounded in the FY2025 10-K (filed 2026-02-17), the FY2023–FY2024 10-Ks, the Q1-2026 10-Q (filed 2026-04-29), the 2026 DEF 14A (filed 2026-04-08), EDGAR XBRL facts, and the memo workstreams. Claims are labeled FACT / INTERPRETATION / ASSUMPTION / OPEN QUESTION. No price target or buy/sell recommendation appears anywhere in this appendix; valuation is discussed only as embedded expectations.

General — What thoughtful questions have other investors asked about this company?

The serious bull/bear debate on Lennox has collapsed onto a single, falsifiable question, and most thoughtful diligence orbits it:

• Is the ~20% operating margin (13.9% FY22 → 15.9% FY23 → 19.5% FY24 → 20.0% FY25) a durable structural base, or a price-led refrigerant-transition peak that mean-reverts? [FACT — margin trajectory, FY2025 10-K statement of operations, verified to EDGAR XBRL OperatingIncomeLoss]. This is the question every other question feeds into.
• How much of FY24 was a borrowed-from-2025 R-410A pre-buy? Management estimated ~$125M of FY24 revenue was pulled forward ahead of the Jan-1-2025 low-GWP refrigerant cut-over [FACT — FY2024 10-K MD&A]. Investors ask how much demand is still “venting” — Home Comfort Solutions (HCS) volume fell −17% in FY25 and −21% organic in Q1-2026 [FACT — FY2025 10-K MD&A; Q1-2026 10-Q MD&A].
• Why is the cheapest HVAC name (~22.7x P/E, ~17x EV/EBITDA) cheap? Peers trade richer (TT ~25x, CARR/JCI ~21–22x, AAON ~45x EV/EBITDA) [FACT — yfinance peer comps, 2026-06-08]. The market is de-rating the multiple, not the current earnings — FY25 EPS actually rose +0.6% to $22.79 [FACT].
• What is the ~8.4%-of-float short interest betting on? The SEC sweep confirmed it is a peak-margin/pre-buy-payback short, not a forensic one (clean filings, zero discretionary CEO/CFO selling) [INTERPRETATION grounded in parsed Form 4 corpus, EDGAR].
• Was the earlier optimism (a ~$737 fair-value framing circulated in 2025) the right read? That call extrapolated the FY24 pre-buy print as run-rate; it has not played out (FY25 revenue −2.7%; stock ~$518 on 2026-06-08) [INTERPRETATION].
• Does the owned-store distribution moat still matter when growth has shifted to commercial? Building Climate Solutions (BCS), which does not run on the residential owned-store channel, is now the higher-margin (23.4%) and only-growing segment [FACT — FY2025 10-K MD&A].

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? At or near a cyclical/structural high (INTERPRETATION, well-evidenced). Operating margin is at an all-time high (20.0%), up ~610 bps in three years; FY25 gross margin expanded +290 bps almost entirely on price/mix [FACT — FY2025 10-K MD&A gross-margin bridge]. The Marathon “high returns mean-revert” prior is unmet, and the first reversion signal has already printed: Q1-2026 HCS segment margin compressed ~390 bps to 13.3% on −21% organic volume and ~$50M of factory under-absorption [FACT — Q1-2026 10-Q / earnings call, 2026-04-29]. Unit volume, however, looks closer to a cyclical low — the residential air-pocket (FY25 volume −13% consolidated, HCS −17%) is the payback of the FY24 pre-buy plus soft housing turnover [INTERPRETATION]. So margin is high while volume is low — an unusual and unstable combination that is the crux of the thesis.

External environment or internal actions? Both, but the recent surge is more external than the CEO-transformation narrative implies (INTERPRETATION). Genuine internal drivers exist (Maskara-era footprint rationalization, the Europe exit, pricing discipline, mix toward direct/parts). But the FY24 HCS segment bridge shows $122M of the profit gain came from price/mix versus only $9M from productivity [FACT — FY2024 10-K segment MD&A], and FY25 was +$256M price/mix offsetting a −17% volume crash [FACT — FY2025 10-K]. The dominant external drivers are: (1) the regulatory mandated price-up (DOE SEER2 2023; AIM Act / EPA A2L R-454B cut-over Jan-1-2025, lifting equipment prices ~10%+); (2) the FY24 R-410A pre-buy timing distortion; and (3) tariff pass-through [FACT — FY2025 10-K Refrigerants; ACHR News 2025]. Regulation-enabled pricing is the most reversible form of margin if a competitor chooses share over price — a risk management itself flags.

How stable are revenues? Structurally resilient at the demand level, but recently volatile at the print level (FACT/INTERPRETATION). The demand base is ~63%+ replacement-driven, non-discretionary, and weather/age-of-installed-base-pulled rather than credit/sentiment-driven — the single most attractive structural feature [FACT — replacement-heavy framing, FY2025 10-K Item 1A; GMInsights 2024]. But revenue is recognized at a point in time on equipment shipment with essentially no contracted backlog [FACT — FY2025 10-K revenue recognition note], so a regulatory pre-buy can swing reported HCS volume from +7% (FY24) to −17% (FY25) in one year [FACT]. Revenue is durable across a cycle; it is not smooth quarter-to-quarter.

Outlook for products/services; market size/growth/geography? North American HVAC is a ~$55bn (2024) market growing ~4.5%, of which U.S. residential unitary is ~$4.7bn [FACT — deallab 2024; GMInsights 2025]. Lennox is the smallest of five majors (~9% N.A. share) and a pure-play North American business after the Q4-2023 Europe exit — ~92.8% U.S., ~7.2% Canada, international ~nil [FACT — FY2025 10-K revenue-by-geography footnote]. The market is growing structurally (electrification/heat-pump adoption, aging installed base, regulatory replacement cycles), domestic, and not shrinking — but Lennox has deliberately removed its only geographic-diversification lever. FY26 guidance is +~8% revenue (~4% from the Duro Dyne/Supco bolt-ons, implying only ~2–3% organic), EPS $23.50–25.00, FCF $750–850M — set in January and merely reaffirmed (not raised) after a Q1 beat, a cautious posture [FACT — Q4-2025 release 2026-01-28; Q1-2026 release 2026-04-29].

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Less competitive / consolidating (FACT). The top five — Trane (~23%), Daikin/Goodman (~19%), Carrier (~18%), JCI (~15%), Lennox (~9%) — hold ~84% of N.A. and are gaining share; each regulatory wave (SEER2, A2L) raises entry barriers and squeezes the subscale fringe [FACT — deallab 2024; FY2025 10-K Refrigerants]. It is a structurally attractive oligopoly. The caveat: the rents accrue to the oligopoly collectively (plus distributor Watsco), and Lennox is the smallest, least-diversified member.

How profitable is the business (ROIC, ROE)? Genuinely high ROIC; ROE is an artifact — use ROIC. NOPAT-ROIC ~33% (NOPAT ~$841.5M = EBIT $1,041.5M × (1−19.2% tax) on invested capital ~$2,517M = total debt $1,388.4M + equity $1,163.1M − cash $34.2M; ~29% capitalizing operating leases), comfortably above cost of capital and passing the Greenwald current-returns moat test [INTERPRETATION/FACT — computed from FY2025 10-K; EDGAR XBRL]. The headline ~76–77% ROE is economically meaningless — a treasury-stock artifact: cumulative buybacks (treasury −$4,923.4M) nearly cancel $4,891.1M of retained earnings, leaving equity at only $1,163.1M (it was negative −$203.1M at FY22) [FACT — FY2025 10-K balance sheet; EDGAR XBRL]. Capitalize ROIC ~33%, ignore ROE ~76%.

How profitable is the industry; how many competitors; barriers to entry? The oligopoly earns resilient high-teens-to-low-20s operating margins [FACT — industry workstream]. Barriers in core equipment are real and rising: refrigerant/efficiency regulation, manufacturing scale, brand, dealer relationships, and distribution density. But they are not impregnable — Lennox’s own 10-K describes its distribution as a “low-barrier business” [FACT — FY2025 10-K], and Daikin/Goodman’s scale-and-price model is a persistent threat in residential.

Can the business be easily understood? Yes (FACT). It makes and distributes furnaces, ACs, heat pumps, rooftop units and refrigeration equipment, sold mostly for replacement through dealers — a simple, durable, understandable model. The only genuine complexity is the regulatory refrigerant/efficiency cadence and segment-bridge decomposition (price/mix vs volume), which this memo unpacks.

Can it be undermined by foreign low-cost labor? Limited direct threat; an indirect one via Daikin (INTERPRETATION). HVAC equipment is bulky, freight-sensitive, regulation-specific (U.S. refrigerant/efficiency standards) and installed locally, so commodity offshoring of finished units is structurally constrained — local manufacturing and distribution density are advantages. Lennox itself manufactures partly in Mexico (Saltillo, ~$150M greenfield), exposing it to USMCA-2026 review and tariffs rather than to low-cost-labor disruption [FACT — FY2025 10-K property table; Q1-2026 10-Q Recent Developments]. The real “foreign” competitive pressure is Japan’s Daikin (the global #1, via Goodman) competing on scale and price inside the U.S., not labor arbitrage on imports.

Do brands matter? Moderately — they keep Lennox in the leadership tier but are not dominant (INTERPRETATION). The Lennox / Dave Lennox Signature / Armstrong Air / Ducane / AirEase portfolio carries pricing power and dealer loyalty, but sits behind/alongside Trane and Carrier in residential prestige; the buyer is usually the contractor, not the homeowner, so brand operates through dealer relationships more than consumer pull [INTERPRETATION — FY2025 10-K brand list; competitive-position workstream].

Customers’ switching costs? Real but channel-specific (INTERPRETATION). The moat is dealer captivity in the company-owned direct channel: 76% of HCS revenue is direct-to-dealer through 200+ owned Lennox Stores (~245 U.S. sites) backed by ~30 distribution centers [FACT — FY2025 10-K]. A dealer trained, parts-stocked and credit-tied to Lennox Stores faces real switching friction, and the installed base creates a replacement/parts annuity. But the captivity is local-economies-of-scale-based and residential-channel-specific; it does not extend to the commercial/refrigeration BCS book that now drives growth, and the brand is not strong enough to compel a homeowner to demand Lennox over Trane/Carrier. Verdict echo: a real but narrow, moderately durable moat — not a wide, Trane-class one.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Yes — the economic franchise is largely off-balance-sheet (INTERPRETATION). The most valuable assets — the Lennox brand portfolio, dealer relationships, the owned-store distribution network’s local scale, and the installed-base replacement annuity — are internally generated and not capitalized. This is why book equity ($1,163.1M) is trivial relative to ~$18B market cap and why P/B (3rd own-history percentile) is uninformative [FACT — EDGAR XBRL; third-party own-history valuation-percentile data]. Recognized intangibles are modest and acquisition-related: finite-lived intangibles jumped to $273.0M FY25 (from $45.1M) and goodwill to $497.2M (from $220.0M) on the Duro Dyne/Supco deal [FACT — EDGAR XBRL FiniteLivedIntangibleAssetsNet, Goodwill].

Off-balance-sheet liabilities? Modest and disclosed (FACT). Operating-lease liability $382.3M and finance-lease liability $68.9M (both now on-balance-sheet under ASC 842, but worth adding ~$0.4B to debt for leverage purposes — pushes net debt/EBITDA from ~1.2x to ~1.5x) [FACT — EDGAR XBRL OperatingLeaseLiability $382.3M, FinanceLeaseLiability $68.9M]. Pension is a small, defined-benefit legacy plan with plan assets ~$147.3M — not a material funding overhang [FACT — EDGAR XBRL DefinedBenefitPlanFairValueOfPlanAssets]. Other tail liabilities: a $167.2M product-warranty accrual (warranties run 1–20 years, a real A2L field-reliability tail), an ~$18.6M asbestos reserve, and routine environmental/litigation accruals (FY25 charge $10.9M) [FACT — FY2025 10-K; EDGAR XBRL ProductWarrantyAccrual]. None is distress-level.

How conservative is the accounting? Conservative-to-clean (FACT/INTERPRETATION). Earnings are cash-backed (4-yr CFO/NI ~99%); the SEC sweep found no forensic flags, no aggressive revenue recognition (point-in-time on shipment), trivial restructuring ($6.8M FY25), and immaterial one-time items [FACT — financial analysis & SEC-filing review]. The one quality-of-earnings watch-item (not a red flag): FY25 CFO ($757.6M) fell below net income ($805.8M) and dropped from $945.7M FY24, driven by a deliberate ~$300M A2L/R-454B inventory build (inventory $853.0M → $1,152.6M) ahead of the refrigerant transition — a ~$228M net working-capital cash drag, a timing item, not an accrual problem [FACT — FY2025 10-K statement of cash flows; EDGAR XBRL InventoryNet]. Tax rate is clean at 19.2% [FACT — EDGAR XBRL]. SBC is immaterial at ~0.56% of revenue and the share count is falling, so reported EPS is not flattered by buried dilution [FACT].

How CapEx-hungry is the business? Capital-light (FACT). FY25 capex was $118.8M (~2.3% of sales); the four-year range is $101–250M [FACT — EDGAR XBRL PaymentsToAcquirePropertyPlantAndEquipment]. FY26 guidance lifts capex to ~$250M (Saltillo, A2L retooling) but it remains a low-single-digit-of-sales business — most cash converts to free cash flow [FACT — Q1-2026 release]. The capital intensity sits in working-capital swings (the A2L inventory build) far more than in PP&E.

Capital Allocation & Management

How much FCF, and how is it used — the philosophy? Lennox generated ~$639M FCF in FY25 (CFO $757.6M − capex $118.8M; ~79% of NI on the transitory inventory drag, ~95–100% over a cycle) [FACT]. Philosophy: dividends-plus-aggressive-buybacks, with opportunistic tuck-in M&A and capital-light JVs — buybacks are the dominant value lever (FACT).

Significant acquisitions recently? Yes, but disciplined tuck-ins, not empire-building (FACT/INTERPRETATION). The Oct-2025 Duro Dyne + Supco parts/distribution deal (~$545M cash, the largest in years) deepens the higher-margin parts/attachment book; AES (Oct-2023, ~$95M, recurring commercial service) and capital-light 49.9% JVs (Samsung ductless 2024, Ariston water heaters) close product gaps without owning manufacturing [FACT — FY2025 10-K; PR Newswire 2024-05-28, 2025-10-16]. The portfolio direction is coherent: shed low-margin geography (Europe Q4-2023), deepen the North American installed-base annuity. The one blemish is timing — the largest acquisition in years landed at a cyclical/valuation peak just as FY25 revenue turned negative, a Marathon-style warning [INTERPRETATION].

Buying back shares? Heavily — this is the defining capital-allocation fact (FACT). Diluted shares fell ~37% (55.8M FY10 → 35.4M FY25) on ~$2.94B of FY16–FY25 open-market repurchases [FACT — EDGAR XBRL]. FY25 alone was $482.3M (~0.9M shares at ~$536, ~23x P/E). The buyback was so aggressive it drove book equity negative FY18–FY22 [FACT]. Tellingly, Q1-2026 buybacks collapsed to $20.0M (39,000 shares) — management appears to recognize it was buying at a rich multiple [INTERPRETATION; FACT on the figures — Q1-2026 10-Q Note 5]. $989.5M of the $5.0B cumulative authorization remained at 3/31/26 [FACT].

Issuing large amounts of new shares to insiders? No (FACT). SBC is ~0.56% of revenue and the net share count is falling. There is no evidence of dilutive insider issuance; equity comp is modest and the buyback overwhelms it.

Compensation policy of directors/management? Well-designed and profit/capital-tilted (FACT). STI = Core Net Income 50% / FCF 30% / Adjusted Revenue 20%; LTI PSUs = ROIC 50% + Core Net Income CAGR 50%. The 2023–2025 PSU paid ~196% of target (ROIC 47.0% > max, Core NI CAGR 17.2%, 3-yr TSR 109%) [FACT — DEF 14A 2026-04-08]. Note the comp-plan ROIC (~47%) is, like ROE, partly a buyback-hollowed-equity artifact and is not clean moat evidence [INTERPRETATION]. CEO Maskara FY25 total comp was $8.43M (down from $9.91M FY24; STI paid 90% of target), with a 6×-salary ownership guideline — pay flexed down in a soft year, a positive alignment signal [FACT — DEF 14A].

Motivations of management? Aligned with per-share value and returns on capital (INTERPRETATION). The incentive design rewards ROIC, cash and per-share compounding rather than empty revenue/asset growth, and the insider read corroborates alignment: zero discretionary CEO/CFO open-market selling across the trailing 36 months (only grants, exercises, tax-withholding; net stock retained) [FACT — parsed Form 4 corpus]. The honest caveat: there were zero open-market insider purchases either, so there is no affirmative bottom-fishing conviction signal even after the ~25% drawdown [FACT].

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No — none of these (FACT). Lennox International Inc. is a Delaware C-corporation, common stock listed directly on the NYSE under “LII.” It is not an ADR (it is a U.S. domestic issuer), not a master limited partnership, and issues a Form 1099-DIV, not a K-1 [FACT — FY2025 10-K cover/Item 5; it files standard 10-K/10-Q/DEF 14A as a domestic filer]. There are no pass-through tax complications for U.S. shareholders.

Dividend policy? A reliable, secondary, growing dividend (FACT). 16 consecutive annual increases; the quarterly dividend was raised to $1.36/share in May 2026 (+4.6%), having stepped up +13% to $1.30 in May 2025 [FACT — Lennox press releases 2026-05-21, 2025-05-27; EDGAR XBRL]. But the yield is only ~1.07% on a ~23% payout — deliberately subordinated to the buyback as the primary cash-return tool [FACT — market data 2026-06-08; FY25 EPS $22.79]. Total shareholder yield is ~3.5–3.8% (~1.07% dividend + ~2.7% buyback), though the buyback was executed near a peak multiple.

How profitable is the business? Among the most profitable in industrials: ~33% gross margin, 20% operating margin, ~$806M net income, ~$23 diluted EPS, ROIC ~33% [FACT]. Profitability is excellent; the open question is durability, not level.

Is net income diverging from cash from operations? Yes in FY25, but benignly (FACT/INTERPRETATION). FY25 CFO $757.6M < net income $805.8M (CFO/NI 94%) and CFO fell from $945.7M FY24 — driven entirely by the deliberate A2L inventory build (~$300M gross; ~$228M net working-capital drag), not accrual quality [FACT — FY2025 10-K cash flows]. Normalizing for that timing item, “clean” FCF approaches ~$800M, roughly the FY26 guide. The four-year CFO/NI of ~99% confirms earnings are cash-backed [FACT]. This is a watch-item, not a divergence that impeaches the earnings.

Embedded expectations (no target). At ~$517.78 (~22.7x trailing EPS, ~17x EV/EBITDA, ~3.5% FCF yield) Lennox is the cheapest name in a premium HVAC group. An EPV decomposition implies only ~$230/share of no-growth earnings power, so ~56% of the price (~$288) pays for margins-hold + MSD growth + buyback compounding [INTERPRETATION — valuation workstream]. The 83rd-percentile own-history P/S sitting beside the 30th-percentile P/E is the quantitative fingerprint: the market re-rated to the new margins but discounts the earnings multiple on durability doubt. No price target or recommendation is expressed.

Risks & Downside

What factors would cause the stock to decline? The risks are correlated, not independent, and all push the same direction in a soft-demand year (INTERPRETATION):

• Margin mean-reversion — the ~20% operating margin proving to be a price-led refrigerant-transition peak rather than a structural base; Q1-2026’s −390 bps HCS compression to 13.3% is the first signal [FACT].
• Residential-volume air-pocket — HCS volume −17% FY25, −21% organic Q1-2026, as the ~$125M FY24 R-410A pre-buy keeps venting and incremental margins compress sharply on the way down [FACT].
• A2L price give-back — if a competitor pursues share, the regulation-enabled +9–10% price/mix (the most reversible growth) erodes [INTERPRETATION; management flags pricing discipline risk].
• Tariff/commodity — un-quantified, escalating exposure concentrated in Mexican (Saltillo) manufacturing; Section 232 steel/aluminum intact, a new ~10% Section 122 surcharge, and a 2026 USMCA review [FACT — Q1-2026 10-Q Recent Developments].
• Multiple de-rate — a ~22.7x P/E for a company with −2.7% revenue and decelerating organic volume is vulnerable if peak-earnings doubt deepens [INTERPRETATION].

Risk of a catastrophic loss? Low (INTERPRETATION). Lennox is FCF-generative, investment-grade-style levered (~1.2x net-debt/EBITDA, ~1.5x capitalizing leases, vs a 3.50x covenant), with a fully-funded pension and no large contingent liabilities; the optical equity “deficit”/thin equity is a buyback artifact, not financial distress [FACT — FY2025 10-K credit facilities; balance sheet]. The credible downside is a valuation/earnings drawdown (peak-margin reversion + multiple de-rate), not insolvency. Cash did fall to $34.2M at FY25 (from $415.1M) against $1,388.4M total debt and $226.0M commercial paper — a liquidity item to monitor in a downturn, though the revolver provides backstop [FACT — FY2025 10-K balance sheet].

Chance of a total loss? Negligible (INTERPRETATION). This is a 130-year-old, cash-generative, brand-and-distribution oligopolist in a structurally good industry. A permanent capital loss would require a simultaneous demand collapse and debt crisis that the balance sheet and replacement-demand base make remote. The realistic risk is opportunity cost / drawdown, not zero.

Recent News & Events

Has the business environment changed recently? Yes — dominated by one regulatory distortion and its hangover (FACT/INTERPRETATION). The Jan-1-2025 AIM Act / EPA low-GWP (R-454B/A2L) refrigerant cut-over reshaped 2024–2026: it pulled ~$125M of demand into FY24 (pre-buy), is now being paid back via the −17% FY25 / −21% Q1-2026 HCS volume decline, and supplied the +9–10% price/mix that has masked the volume loss at the revenue line [FACT]. The tariff regime also shifted in 2026 (Feb-2026 SCOTUS struck IEEPA tariffs; Section 232 steel/aluminum modified April 2026; new ~10% Section 122 surcharge) — an escalating, un-quantified cost headwind [FACT — Q1-2026 10-Q].

Significant acquisitions? Yes — Duro Dyne + Supco (~$545M cash, closed Oct 16, 2025; ~$277M deal goodwill — lifting total company goodwill to $497.2M — plus ~$273M finite-lived intangibles), the largest in years, adding higher-margin parts/distribution (it drove BCS sales +38% in Q1-2026 and contributed ~4% of the FY26 revenue guide) [FACT — FY2025 10-K; Q1-2026 10-Q Note 13; EDGAR XBRL].

Change in accounting policies? None material (FACT). No restatements, no policy changes flagged in the SEC sweep; the FY24 reorganization into two reportable segments was a presentation change, not an accounting-policy change. Restructuring is trivial ($6.8M FY25) and one-time items are immaterial to run-rate [FACT].

Recent changes — new markets, facilities, management? New facility: the ~$150M, ~800k-sq-ft Saltillo, Mexico commercial greenfield (largest-ever Lennox manufacturing investment, opened 2024, ramping) — capacity for BCS but also the locus of tariff exposure [FACT — FY2025 10-K property table]. New markets/products: the 2024 Samsung 49.9% ductless JV (closes Lennox’s historical ductless under-indexing; equity-method, off the revenue line, ~$6–8M equity income — de minimis in scale) and the Ariston water-heater JV [FACT]. Management/board: stable — Maskara (CEO) and Quenzer (CFO) in place; the only governance change was a single director resignation (S. Somasundaram, 2026-05-23) reducing the board 9→8 [FACT — 8-K 2026-05-28]. Capital-return event: the May-2025 step-up (+$1.0B buyback authorization to $5.0B cumulative; +13% dividend) executed into a soft revenue year [FACT].

Net read. The recent news flow does not strengthen the durable-structural-improvement story; it reframes a meaningful share of the margin expansion as cyclical/regulatory rather than durable. The reaffirm-don’t-raise FY26 guidance after a Q1 beat, the collapse of buybacks to $20M, and the 2025 sell-side target cuts (Goldman, UBS, KeyBanc) all corroborate a cautious read [FACT — Q1-2026 release; Investing.com 2025]. Watch two mechanical signals: a return to positive HCS organic unit volume (vindicates the bull) versus another quarter of HCS margin compression on negative volume (vindicates the bear).

APPENDIX B — Source Appendix

Target: Lennox International Inc. (NYSE: LII) | CIK 0001069202 Compiled: 2026-06-08 | Web sources accessed 2026-06-08 unless noted.

Sources are listed primary-first. The documents below are the relied-on subset of the trailing 36-month SEC corpus. EDGAR archive base for all LII filings: https://www.sec.gov/Archives/edgar/data/1069202/. SEC filings are dated by filing date.

1. Primary — SEC Filings

1.1 Annual Reports (Form 10-K) — three fiscal years

Relied-on items: Item 1 Business (segments, distribution model, competition, refrigerants/SEER2, seasonality, strategic sourcing); Item 1A Risk Factors (low-barrier distribution, single-location plants, tariffs/USMCA, North-America concentration, warranties); Item 7 MD&A (consolidated and segment revenue/volume/price-mix bridges, gross/operating margin walk); Consolidated Statements of Operations, Balance Sheets, Cash Flows; Note 3 segment footnote; goodwill, debt, retirement-plans, business-combination, and commitments/contingencies notes. The FY2024 10-K is the source of the management’s-own ~$125M R-410A pre-buy estimate.

1.2 Quarterly Reports (Form 10-Q)

Relied-on items: Q1-2026 segment MD&A (HCS −10% sales / −21% organic volume / 13.3% segment margin; BCS +38% / +17% organic volume / 19.7% margin); “Recent Developments” tariff narrative (IEEPA/Section 232/Section 122); Note 5 share-repurchase ($20.0M / 39,000 shares, $989.5M authorization remaining); Note 13 acquisition purchase-price revision (+$6.5M).

1.3 Proxy Statement (DEF 14A) — executive compensation & governance

1.4 Current Reports (Form 8-K) — material events & earnings releases

1.5 Insider Transactions (Forms 3/4/5) and Other

• Form 4 corpus (insider transactions): all 153 Form 4 filings over the trailing 36 months were downloaded and parsed programmatically (467 line-items) for the insider read (zero open-market purchases; routine 10b5-1 sales; CEO Maskara / CFO Quenzer no discretionary sales). Index base: https://www.sec.gov/Archives/edgar/data/1069202/. Representative recent Form 4: 2026-05-07 (.../000106920226000056/).
• Form 3 (new insiders): six over the window (e.g., 2025-06-03, 2025-04-29, 2025-01-06, 2024-06-04, 2024-03-18, 2024-01-09).
• Form 144 (notice of proposed sale): 51 filings tracked as corroboration of planned-sale volume; no independent signal.
• S-3ASR (universal shelf): 2025-10-24 — https://www.sec.gov/Archives/edgar/data/1069202/000106920225000053/a2025shelfregistrationstat.htm (undrawn issuance optionality).
• 424B5 (senior notes due 2033 prospectus supplements): 2023-09-07 / 2023-09-08 — https://www.sec.gov/Archives/edgar/data/1069202/000119312523230170/d741027d424b5.htm and .../000119312523231556/d741027d424b5.htm.
• Schedule 13G/13D (institutional ownership): multiple over the window (e.g., 2026-04-30, 2026-02-13 13Gs), enumerated in the filing index.

2. Earnings Materials & Transcripts

Public transcript sources were used for the most recent quarters, validated against the 8-K earnings releases and 10-Q/10-K above.

3. Regulatory & Industry Sources

4. Market / Quantitative Data