Daikin Industries, Ltd. (TSE: 6367) — The World’s #1 Air-Conditioner Maker, Priced for a Margin Recovery It Hasn’t Yet Earned

Independent fundamental equity research. As-of date: 2026-06-17.

⚡ Claude’s Take

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

HOLD — a world-class franchise at a full-to-fair price; ACCUMULATE ON WEAKNESS, not a short. Conviction: MEDIUM. At ~¥23,680 (~25x trailing P/E on a trough 8.3% operating margin / ~10.3x EV/EBITDA), Daikin already prices in most of the FUSION30 self-help recovery before management has delivered a single quarter of it. The franchise is genuinely the best in cooling and the stock is the cheapest major HVAC name on EV/EBITDA — but that discount is largely deserved (lower margin/ROE, residential mix, China/Chemicals drag, a Japan-domicile multiple), so “cheap vs. Trane” is not the same as “cheap.” A more attractive accumulation zone sits around ~¥18,000–20,000 (≈ ~16–18x normalized EPS at a recovered ~10–11% margin / ~8.5–9x EV/EBITDA), where you pay less for the self-help optionality and the activist catalyst that are currently in the price. This is a zone, not a target: it is roughly where the stock traded as recently as late-2025/early-2026, so it is a realistic re-entry, not a fantasy.

The market is pricing correctly that Daikin is the genuine global #1 at a cyclical/transitional earnings trough — hence cheapest-on-EV/EBITDA-but-deservedly-so. What it may be pricing too optimistically is a clean margin recovery to 10%+ / ROE 12% that management promised under FUSION25 and missed outright (landed 8.3% margin / 9.1% ROE against an 11% / 11% target), and that the FY ending March 2027 guide — ~8.5% margin, +1% net income — does not yet show. The framing is therefore explicit: quality-compounder-at-a-full-price, with an Elliott activist special-situation overlay, on a recovering-from-trough tape — NOT a falling knife, NOT deep value. That framing is grounded in the price action: +~50% off the February-2025 trough, a +14% pop on Elliott’s mid-April 2026 disclosure (the biggest one-day gain since 2009), and a final leg lit by a ¥350B buyback announcement rather than an earnings beat — the FY27 net-profit guide actually came in below consensus. You are buying momentum and optics layered on a real but under-monetized franchise; the discipline is to wait for either a cheaper entry or proof the margin has turned. Conviction MEDIUM. The single bullish flip: operating margin demonstrably inflects over the next 1–2 prints — North American AC from ~8% toward 10%+, consolidated tracking above 9% toward 10% — vindicating the cyclical-trough read. The single bearish flip: margin stays ~8–8.5%, NA stays stuck ~8%, and Elliott settles for the ¥350B ASR alone with no structural reform — confirming structural erosion, in which case the ~25x P/E is real, not a trough artifact. Tag: “Scale without operating leverage — the biggest name in cooling, earning like an also-ran.”

📈 Stock Price Action — Five-Year Event Map

Over the trailing ~60 months Daikin round-tripped almost the full cycle: from ~¥21,200 in mid-2021, up to a marginal all-time monthly-close high of ~¥27,893 in June 2023, then a deep ~45% de-rate to a ~¥15,281 monthly-close trough in February 2025 (52-week low ¥15,960), and back up ~50% to ~¥23,680 today. That leaves it ~8.6% below its 52-week high (¥25,900) and ~15% below its all-time peak; the 52-week range is ¥15,960–25,900. It is the HVAC cohort’s 2nd-strongest 12-month performer (+49.8%, behind only AAON +86.6%) and trades above both its 50-day (~¥22,868) and 200-day (~¥20,035) averages — a confirmed recovery, but a recovery toward the prior cycle high, not a fresh breakout. (Price moves are FACT from monthly/daily closes; factor-model data is unavailable for this Tokyo-listed ticker, so this is a price/trend read only.)

1. The post-COVID melt-up (Aug-2021) re-rated Daikin as a decarbonization/heat-pump growth play (INTERPRETATION of cause); the +21% move is FACT.
2. The 2022 pullback paired global multiple compression on long-duration names with real cost inflation in copper, steel and logistics squeezing margins (cause INTERPRETATION).
3. The 2022–2023 melt-up to the all-time high rode yen weakness (translation tailwind to ~84%-overseas earnings) plus the broad Japan-large-cap rally (cause INTERPRETATION).
4. The grinding 2023–2025 de-rate is the bear case in price form: a protracted China property slump, share-loss fears vs. Gree/Midea, and tariff/rate overhang (cause INTERPRETATION; the ~45% drawdown is FACT).
5. The February-2025 trough coincided with peak China/tariff pessimism and a weak result, resetting the multiple to a multi-year low (context INTERPRETATION).
6. The 2025–2026 recovery leg was cohort-wide (JCI and AAON rallied hard too) on stabilizing fundamentals and refrigerant-ASP optimism (cause INTERPRETATION).
7. Elliott’s mid-April 2026 disclosure (FACT) triggered a +14% single-day gain (FACT) — the largest since 2009 — capitalizing the activist value-creation case in one session.
8. The ¥350B ASR + dividend raise (FACT events) drove a +5.4% intraday spike to a YTD high, but the stock then gave back ~7% as the below-consensus FY27 net-profit guide weighed — the move FACT, the relative weight of “optics vs. fundamentals” INTERPRETATION.

1. Executive Summary

Daikin Industries is the world’s #1 HVAC&R company by revenue — ~$36B, an estimated ~15% of the global air-conditioning market (the share figure is a third-party INTERPRETATION; Gree and Midea lead China residential by unit volume). In the year ended March 2026 it crossed ¥5T in net sales for the first time (¥5,015.0B, +5.5%), with operating profit ¥414,991M, net income ¥275,229M, and EPS ¥939.92 (FACTs, Daikin J-GAAP Brief Report, 2026-05-12). Air Conditioning & Refrigeration is ~92% of sales (¥4,621.1B); the rest is Chemicals/fluorochemicals (¥281.5B) and a small “Others” stub of oil-hydraulics, defense and electronics. The business is ~84% overseas, with the US its largest market (~¥1,786.5B, ~36% of sales, +9.2%).

The central tension is that record sales coexist with a multi-year decline in profitability. Operating margin has fallen every single year — 11.1% (FY ended March 2019) → 10.2% → 9.5% → 8.9% → 8.5% → 8.3% (FY ended March 2026) — while revenue roughly doubled (FACT, Daikin 10-year financial data). ROE slid 12.0% → 12.3% → 10.7% → 9.7% → 9.1% (and was ~15.7% back in March 2019); ROA fell from ~6.2% to 5.0%; computed ROIC is ~8.0%, barely above the cost of capital. In plain terms, scale produced negative operating leverage — a Greenwald profitability-erosion / Marathon mean-reversion signature, and the single most important fact in the file (INTERPRETATION of the cause; the monotonic decline is FACT). Daikin earns roughly half the returns of US commercial-applied peers: Trane runs an ~18.6% operating margin and ~36% ROE; Lennox ~20% margin; Carrier ~16% adjusted; JCI ~15.5% adjusted. Note that cross-border comparison must use EV/EBITDA, not P/E or headline operating margin: Daikin’s J-GAAP goodwill amortization (~¥51B/yr, ~1pp of margin drag) depresses its margin and P/E versus IFRS peers who do not amortize goodwill.

The moat is genuine but narrow. Two real advantages: (1) a supply/cost edge from vertical integration into compressors, inverters and its own fluorochemicals — Daikin makes its own R-32/A2L refrigerants, unique among pure-OEM peers; and (2) demand-captivity in applied/commercial VRV/VRF, a category Daikin invented in 1982 and whose “VRV” trademark it owns. Pricing power is evidenced: Daikin absorbed a ~¥41B US tariff hit in FY26 through pass-through pricing plus cost-down, and recovered US ducted-unitary share to ~24%. But the moat is under-monetized — an 8.3% margin / 9.1% ROE is not what a durable advantage should produce. North American AC (DNA+DAA) earns only ~8% (management concedes it has “struggled to improve”), and the service/aftermarket annuity that anchors peer economics is still aspirational (Solutions 28% → 40% target) versus Trane’s ~43% service gross margin.

The decisive question — and the entire valuation hinges on it — is whether the margin erosion is cyclical or structural. The bull (and management) frame it as a cyclical air-pocket: China real estate, a Chemicals/semiconductor down-cycle (segment profit -28% YoY), tariffs, and a US refrigerant pre-buy payback. The FUSION30 plan (unveiled May 12, 2026) targets operating margin 8.3% → 10% (FY2028) → 12% (FY2030) and ROE 9.1% → 12% → 15%, funded by ¥1.3T capex + ¥500B M&A and ≥¥700B of buybacks/M&A capacity. The bear notes that FUSION25’s identical-shape targets (11% margin / 11% ROE) were missed outright, that its North American unit-demand forecast (+3% CAGR) came in at -9% and Europe at -11% vs. +24% forecast (capacity built into an air-pocket — a textbook capital-cycle error), and that a five-to-seven-year monotonic decline through a doubling of revenue looks like erosion, not a dip. The FY ending March 2027 guide — sales ¥5,150B (+2.7%), operating profit ¥436B (~8.5% margin, +5.1%), net income ¥278B (+1.0%) — does not yet embed the recovery; its ¥436B operating-profit figure is essentially flat against the ¥435B originally planned for FY26, i.e., a lost growth year.

The catalyst is Elliott Management, which disclosed a ~3% / ~$1B stake in mid-April 2026, demanding ~14% margin, doubled EPS, a buyback of up to ¥1T, integration of Daikin’s six fragmented North American units, and a portfolio review (the stock jumped +14%). Roughly four weeks later Daikin responded with FUSION30 and its first major buyback in a decade — a ¥350B Accelerated Share Repurchase (~5% of cap, ~one-third of Elliott’s ask), a new CFO role, a 50%-outside-director target, D-ROIC + relative-TSR executive comp (stock options abolished), and a “stable & growing” dividend policy (DPS ¥340 → ¥360 guided). The governance pivot is real, but only partially confirmed — and the market may be pricing the full Elliott program when only a partial response is on the table.

On valuation, Daikin is the cheapest major HVAC name on EV/EBITDA (~10.3x vs. Trane 25.7x, JCI 22.6x, Carrier 22.5x, Lennox 17.6x, AAON 44.4x) and on EV/sales (~1.4x) — but the discount is largely deserved (lower margin/ROE, residential mix, China/Chemicals cyclicality, Japan-domicile multiple). The ~25x trailing P/E looks rich but sits on a trough 8.3% margin; normalizing toward 10–11% compresses the implied P/E to ~16–19x, so the headline overstates expensiveness if margins recover. A reverse-DCF (Ke ~7.5%) embeds only ~4–7% long-run FCF CAGR — undemanding for a global #1 in a ~6–7%-growth market — but it quietly assumes the margin/ROE downtrend halts and reverses. In short: a high-quality global leader at a genuine trough, where the recovery is already largely capitalized in the price (+~50% off the trough, last leg buyback-driven into a below-consensus guide). The upside lives in margin DELIVERY, not in the entry multiple — and that delivery is, for now, promised rather than proven.

2. Business Overview

Daikin Industries is the world’s largest heating, ventilation, air-conditioning and refrigeration (HVAC&R) company by revenue, and in the year ended March 2026 it became the first HVAC company to cross ¥5 trillion in sales: ¥5,015,036M (+5.5% YoY), with operating profit of ¥414,991M (Fact; Daikin Brief Report on the Settlement of Accounts, J-GAAP, p.1, 2026-05-12). It is, at heart, a single-product industrial — an air-conditioner manufacturer with a fluorochemicals arm bolted on — and it makes money the way capital-goods companies do: by designing, manufacturing and selling physical equipment, recognizing revenue at the point of sale, and earning a margin on the spread between selling price and the cost of materials, labor and factory overhead. Roughly 84% of sales are overseas (Fact; Daikin 10-year Financial Data summary, 2026-05-12), making this a Japanese-domiciled but genuinely global manufacturer.

Segment structure: two reportable segments, but really one business

Daikin reports only two segments — Air-Conditioning & Refrigeration Equipment (“AC&R”) and Chemicals — with everything else swept into a non-reportable “Others” bucket (Fact; Brief Report, Segment Information note, pp.15–17). The granular product/region detail that investors care about lives in the results presentation, not the statutory segment note, so the reportable structure understates how concentrated the business actually is.

(Fact; Brief Report segment note, p.17, 2026-05-12.)

AC&R (~92% of sales) is the company. Within it sit several sub-pieces that the two-segment structure deliberately hides: (1) residential ducted unitary and ductless/split systems — the US Goodman/Amana platform, Japanese room ACs, and ductless across Asia/Europe; (2) applied/commercial — VRV/VRF systems (Daikin invented variable refrigerant flow in 1982 and owns the “VRV” trademark, forcing every competitor to use the generic “VRF”; Fact, en.wikipedia.org/wiki/Variable_refrigerant_flow, 2026-06-17), chillers, and custom air-handling units via Alliance Air; (3) light-commercial rooftop and packaged units; (4) refrigeration — commercial/industrial cold-chain showcases (the AHT Austria business acquired in 2019); and (5) air filtration/air-quality via AAF/Flanders (AAF International, the global filtration brand acquired in 2016). Daikin does not disclose product-line margins, so whether the advantaged applied/VRF pocket actually earns more than the commoditizing residential core is an open question (Open Question) — the consolidated 8.2% AC&R margin blends both.

Chemicals (~5.6%) is the fluorochemicals arm — fluoropolymers, fluoroelastomers, and, critically, refrigerants. This is the segment that makes Daikin different from every pure-OEM HVAC peer: it manufactures its own R-32 and A2L low-GWP refrigerants and holds the underlying IP, a genuine supply/cost vertical-integration edge no Carrier, Trane, or Lennox replicates. But Chemicals is itself cyclical and just delivered an ugly year — segment profit fell 28% (¥46,119M → ¥33,089M; margin 17.5% → 11.8%) on weak semiconductor-related fluoropolymer demand and distribution-inventory destocking (Fact; Daikin FY results presentation, p.4/p.21, 2026-05-12). The “vertical integration is a moat” narrative has to reckon with the fact that the integrated input business shrank in profit this year, and faces a structural PFAS/“forever chemicals” regulatory overhang globally.

Others (~2.2%) — oil-hydraulic equipment, defense systems, and electronics — is a low-margin (4.4%), legacy domestic-Japan rump with no strategic role in the thesis.

Geographic map: an American company that happens to be Japanese

By customer geography, the US is now Daikin’s single largest market at ¥1,786.5B — ~36% of group sales and growing +9.2% YoY (Fact; Brief Report disaggregated revenue note, p.17). The full split:

(Fact; Brief Report, p.17, 2026-05-12.) On the AC-only regional cut (which adds filtration), the Americas are even more dominant at ¥1,898.6B, ~41% of AC sales (Fact; FY results presentation, p.13).

The regional cycles are diverging sharply, which matters for the thesis. The US (Goodman residential unitary + applied/data-center via DNA/DAA) is the growth engine and where the refrigerant transition (R-410A → R-32/A2L) handed Daikin a share recovery to ~24% of US ducted unitary. Europe is heat-pump heating — structurally attractive in theory but subsidy-dependent and volatile (EU heat-pump sales fell ~23% in 2024 before recovering). Japan is a mature, GDP-like home market. China is the problem child: ¥403.5B of AC sales, down for a third consecutive year (¥458.8B → ¥428.1B → ¥403.5B) in a Gree/Midea/Xiaomi price war compounded by the property bust (Fact; FY results presentation, China section p.16). Daikin has no scale advantage in Chinese residential and is retreating to higher-value products to defend margin.

Revenue model: overwhelmingly one-time equipment, recurring revenue still nascent

This is the most important structural fact about the business model. Daikin’s revenue is overwhelmingly non-recurring — it is recognized when a piece of equipment ships, and the cash flow stops there. The company does not disclose a standalone service, parts, or aftermarket revenue or margin line (Open Question; Brief Report). “Expanding the Service Solutions business globally” is listed as one of six top-management priority themes and a FUSION30 pillar — i.e., it is an aspiration, framed as “generate revenue from maintenance, repair services, and parts sales,” not an established profit pool (Fact; FY results presentation p.5, 2026-05-12). Management targets lifting the “Solutions” mix from 28% to 40%, which is itself an admission that the recurring annuity does not yet anchor the franchise.

That is the gap to the best-monetized HVAC peers. Trane runs a roughly 43% gross-margin service annuity on its installed commercial base; Watsco-style distribution captures recurring replacement flow. Daikin’s economics, by contrast, are still those of a cyclical equipment OEM exposed to new-construction and replacement cycles, refrigerant pre-buy/payback swings (the 2024 US A2L pre-buy then 2025 payback slump), and FX. The installed base is enormous and the replacement tail is real, but Daikin has not yet built the captive, high-margin service layer that converts an installed base into a financial annuity (Interpretation).

Manufacturing and vertical integration

Daikin’s operating model is deep vertical integration. It manufactures its own compressors and inverters — the highest-value, most differentiating components in an air conditioner — and, uniquely, its own refrigerants through the Chemicals segment. Management explicitly names “inverters, heat pumps, and refrigerants” as proprietary strengths “not things that other companies can imitate so easily” (management commentary — hypothesis, not evidence; FY results-briefing transcript, 2026-05-12). The corroborating data point is concrete: Daikin absorbed ~¥41B of direct US tariff impact in FY ended March 2026 entirely through pass-through pricing plus cost-down (copper-to-aluminum substitution), evidence of real pricing power in its differentiated lines (Fact; FY results presentation p.4 and OP bridge p.8 — selling price +¥120.5B, cost reductions +¥81.0B). The model spans a global manufacturing footprint (~104,000 employees) and supports rising R&D (¥150.7B, up from ¥81.5B five years ago) that funds the refrigerant/inverter integration edge.

End markets

The business serves four end markets: residential (homeowner HVAC replacement and new build — the largest and most commoditized pool, the Goodman/Amana and ductless core); commercial (offices, retail, hospitality — VRV/VRF, chillers, applied, the stickiest engineering-intensive pocket); data-center/AI cooling (the management-favored growth pillar, with a new “Dynamic Data Centers Solutions, Inc.” entity and a “total solution” of heat source + air-handling + server + liquid cooling — but only ~¥100B of sales today, ~2% of group, targeting ¥300B+ by FY2030, and a crowded capital-cycle danger zone with Vertiv/Schneider/JCI/Carrier racing in); and industrial refrigeration / cold chain (AHT, food retail — the segment that just took an ¥11,849M impairment).

Verdict: a good industrial with a narrow moat, not yet a recurring-revenue compounder

Daikin is a high-quality, world-leading equipment manufacturer whose business model is structurally inferior to the best HVAC franchises on the one axis that matters most — recurring economics. The argument for quality is real: global #1 scale, genuine vertical integration into compressors/inverters/refrigerants, applied/VRF leadership, demonstrated pricing power (the ¥41B tariff pass-through), and an enormous, growing US installed base. But ~92% of revenue is HVAC equipment, the overwhelming majority of it recognized one-time at sale, and the high-margin service/aftermarket annuity that anchors Trane’s and Watsco’s quality is — on Daikin’s own disclosure — still an ambition (28%→40% Solutions target), not a built business. The disconfirming evidence for an outright negative read is equally real: pricing power exists, the installed base is durable, and the replacement tail provides a quasi-recurring volume floor even without a formal service line. But until the recurring layer is disclosed and monetized, Daikin’s economics remain those of a cyclical capital-goods company exposed to construction cycles, refrigerant pre-buys, China’s price war, and FX — which is precisely why scale has doubled revenue since FY2019 without lifting margins. The model is good-industrial-with-a-narrow-moat; the bull case rests on FUSION30 converting equipment leadership into a recurring annuity it does not yet have.

3. Industry Dynamics

Global heating, ventilation, air-conditioning and refrigeration (HVAC&R) is an above-average industrial end-market — consolidated, barrier-protected, with a non-discretionary replacement annuity and regulation-driven price uplift — but its attractiveness is concentrated, not uniform. The durable money sits in two pools: North American applied/commercial/engineered systems (chillers, controls, VRF for commercial), and the replacement/service aftermarket on a vast installed base. The structurally weaker pools — residential/light-commercial unitary, European residential heat-pump heating, and Chinese residential AC — are exactly where Daikin is over-indexed. The central question this section answers is not whether the industry is good (it is), but whether Daikin sits in the parts of it that earn premium returns. The evidence says: partly, but more in the weaker pools than the “global #1 compounder” narrative implies.

3.1 Market size, growth, and the demand floor

The global HVAC equipment market is roughly $300B in 2025, projected to ~$380–410B by 2030 at a ~6.4–7.5% CAGR (Fact; MarketsandMarkets puts it at $299.28B → $407.77B, 6.4% CAGR; Grand View 7.5%, accessed 2026-06-17). Air-conditioning specifically is ~$130–160B. Asia-Pacific is ~50% of the market by revenue, and heat pumps are the fastest-growing sub-segment (~8.5% CAGR). Method dispersion is wide — equipment-only estimates run $175–320B and broad (services-inclusive) figures reach ~$525B — so treat any single TAM as a directional anchor, not precision (Interpretation). The defensible read is mid-single-digit blended growth: faster in commercial-applied (~7.5%) and in penetration-driven emerging markets, slower and more cyclical in mature residential.

The most reliable feature of the industry is its demand floor. ~63% of US installs (some estimates 75–85%) are retrofit/replacement on a ~130M-unit US residential installed base replacing at ~6%/yr (~8M units) versus only ~1–1.5M new-construction units (Fact; GMInsights 2024). Replacement is weather- and failure-driven, not credit- or sentiment-driven — the structurally best part of the industry because it is recurring, higher-margin, and stickier than the equipment box. This annuity is why the moat in HVAC lives in service and the installed base, not in the unit sale. It is also where Daikin is comparatively thin (Section 3.3).

3.2 Structure: a consolidated, rational oligopoly

HVAC is a consolidated oligopoly with high barriers and rational pricing. The top-5 hold ~32% of global revenue and ~84% of North American unitary share (Fact). 2024 NA share: Trane 23.0%, Daikin (incl. Goodman/Amana) ~19.2%, Carrier 17.7%, JCI 15.2%, Lennox 8.9% (Fact; deallab NA HVAC Market 2024). Globally, Daikin is #1 (~$36.3B revenue, ~15% share — the 15% is a third-party estimate, Interpretation), ahead of Gree (~$29.2B), Carrier (~$20.4B) and Midea (~$18.9B), with Trane, Johnson Controls, LG, Samsung, Mitsubishi Electric and Lennox rounding out the field (Fact; GMInsights/Mordor aggregation, accessed 2026-06-17). Regional leadership diverges sharply: Carrier ~17.7% in North America, Daikin ~18–20% in Europe, and Gree/Midea dominate China by residential unit volume — so Daikin’s global #1 is a revenue crown, not a unit-volume one (Daikin trails Gree and Midea badly in Chinese residential units).

By Greenwald’s tests the industry passes as barrier-protected: durable share leadership (Daikin has held #1 globally for years; the US top-5 are gaining not losing share, with no new entrant cracking the oligopoly in decades), high incumbent ROIC in the best pools (~20%+ segment EBITDA at the US applied/commercial players; Trane ROIC ~28% incl. goodwill), and rational competition among scaled majors that compete on M&A and efficiency rather than greenfield capacity wars. The barrier type is economies of scale plus customer captivity in engineered systems, controls, and the service aftermarket — reinforced by dealer/contractor networks, code/efficiency-regulation R&D, and installed-base service density (Interpretation). Internal rivalry is high but rational; the entrant threat is low.

3.3 The bifurcated profit pool — and where Daikin sits

The industry’s profitability is bifurcated, and this is the structural crux for Daikin.

The durable money is in the top two rows: Carrier’s own segment economics prove the point — Climate Solutions Americas earns ~20.5% operating margin versus Climate Solutions Europe (Viessmann-heavy heating) ~8.8% (Fact; Carrier FY2025 10-K). Trane converts a 43%-gross-margin service annuity (~34% of revenue, ~12% CAGR) into ~28% ROIC. Daikin is structurally lighter in both. Its Solutions/service mix is 28% of sales, with a target of 40% by FY2030 — i.e., management is explicitly pivoting toward the annuity precisely because equipment economics are deteriorating, which is itself a tell that the equipment moat is weakening (Interpretation; Daikin FUSION30 plan p.8, p.13). Whether Daikin can reach a Trane-class service mix is an open question; today it is structurally behind on the highest-quality profit pool.

Conversely, Daikin is heavy in the weak pools. In AC&R, the Americas (~¥1,898.6B) carry substantial residential/unitary exposure via Goodman/Daikin Comfort; China is ¥403.5B and falling; Europe (¥784.2B) leans on subsidy-dependent heat-pump heating. The financial fingerprint is unambiguous: consolidated operating margin fell every year from 10.2% (FY2021) to 9.5% / 8.9% / 8.5% / 8.3% (FY ended March 2026), and ROE from ~12% to 9.1%, ROA from 6.2% to 5.0% — even as revenue roughly doubled since the year ended March 2019 (Fact; Daikin Financial Data, IR). Scale produced negative operating leverage. By Greenwald’s lens, this is the “are advantages present” test flashing erosion at the equipment layer — and management itself names the cause: “selling prices declined due to accelerated equipment commoditization and intensified competition,” citing “the emergence of Chinese manufacturers” (Fact; FUSION30 plan p.8). That is management corroborating, not contesting, the bear read.

3.4 The capital cycle: capacity built into demand that collapsed

Apply the Marathon supply-side lens directly. The textbook capital-cycle hazard is high returns drawing in capacity ahead of demand that then disappoints — and Daikin’s own FUSION30 documents lay out exactly this setup. During FUSION25, Daikin forecast North American AC unit-sales at +3% CAGR (FY23→FY25); actual came in at -9%. It forecast European AC units at +24%; actual was -11% — a ~35-point miss (Fact; FUSION30 plan p.9). Capex ran ¥311.5B / ¥324.6B / ¥300.0B in FY23/24/25 against depreciation rising from ¥170B → ¥197B → ¥225B (capex/depreciation >1.3x in the peak years) — capacity added into a demand air-pocket (Fact; Daikin FY2025 Results Presentation p.25). The predictable result is the five-year margin and return compression documented above. This is the Marathon asset-growth anomaly playing out in real time: Daikin grew assets into forecasts that the supply side and the cycle did not validate.

The peer cohort confirms the read is industry-wide for the weak pools but disciplined for the strong ones. Commercial-applied is mid-cycle with disciplined supply — engineered-systems complexity and contractor lock-in keep entrants out, so incumbent returns are not being competed away. Residential is later-cycle and channel-oversupplied (it absorbed the 2025 ~20% US shipment air-pocket). European heat pumps are in a post-boom rationalization (demand down 20–48%). The one zone where capital is actively flooding in is data-center cooling (Section 3.6).

3.5 Refrigerant transition and the electrification policy reversal

The R-410A → A2L low-GWP transition is a regulatory ASP tailwind layered on a transient volume distortion — not a clean super-cycle. Under the AIM Act (US, implementing the Kigali HFC phase-down), manufacture of new R-410A residential/light-commercial equipment was banned from Jan 1, 2025 (Fact; EPA AIM Act, accessed 2026-06-17). The industry split on the replacement: peers standardized on R-454B (Trane, Carrier, JCI/York, Lennox), while Daikin is the R-32 standard-bearer — a meaningful differentiation point given Daikin makes its own refrigerants (Section 4). A2L systems run +10–30% more expensive (R-454B base ~10% above R-410A, escalating to 15–20% over two years), a durable, pro-incumbent price uplift (Fact; ACHR News).

But the volume signal is corrupted by a pre-buy. Channels over-ordered cheaper R-410A in 2024, then de-stocked in 2025: US a/c + heat-pump shipments fell ~20% YoY in 2025 on pre-buy payback, destock, high rates, low used-home turnover, tariffs, and tax-credit termination (Fact; AHRI via ACHR News). Lennox’s residential volume swung +7% (2024) to -17% (2025) while price/mix +10% offset it — a clean illustration. Daikin is in the same air-pocket and management confirms its US distribution and DNA inventory “has not changed… not able to significantly reduce it” as of March 2026 (Fact; Daikin FY2025 transcript). Daikin did recover US residential unitary share to ~24% in FY2025 (above its FY2023 pre-loss level) via R-32 units and the “Fit” inverter product — but the fact that share was lost then regained is itself a Greenwald-weak signal: a >2-point swing means the residential position is contestable, not a fortress (Interpretation), and there is an open question whether the share was “bought” back with price during the air-pocket.

Electrification policy has turned from tailwind to headwind on both continents simultaneously. The US IRA Section 25C heat-pump credit (~$2,000/unit) was repealed effective Dec 31, 2025 (Fact; P.L. 119-21). Europe was a boom-and-bust: EU heat-pump sales fell ~22–23% in 2024 (Germany -48%) after gas-price normalization and German subsidy cuts, recovering only ~+11% in 2025 (Fact; EHPA / pv-magazine). Daikin’s FUSION25 review concedes “subsidies for heat pump heating were scaled back” and, under the Trump administration, “some energy-saving incentives were suspended or reconsidered.” Electrification is a genuine multi-decade tailwind but a poor near-term demand engine — and Daikin’s EU and US residential heat-pump bets were sized against subsidy assumptions that reversed.

3.6 Data-center cooling: the bull pillar and the danger zone

Data-center/AI cooling is the dominant bull pillar across the entire HVAC cohort — and simultaneously the clearest Marathon capital-cycle danger zone. The cohort narrative is a ~$26B (2025) → ~$128B (2033) TAM at ~22% CAGR, driven by ~$413B of 2025 hyperscaler capex (+84% YoY) and the air→liquid transition at ~250kW/rack (Fact; industry research). Every incumbent is racing capital in — Vertiv (+45% liquid-cooling capacity), Schneider (acquired Motivair), Carrier, JCI (Pune expansion), AAON, Modine, Munters — and the top-5 in liquid cooling hold only ~35% share, i.e., the supply side is fragmenting as capital floods a hot demand forecast. That is the textbook late-boom warning: treat data-center cooling as high-growth optionality likely to mean-revert over 3–5 years, not a durable moat.

Two Daikin-specific points sharpen the caveat. First, Daikin’s own DC-cooling TAM CAGR estimate is ~11% (¥1.4T → ¥2.7T) — roughly half the cohort’s ~22% narrative, a notable disagreement that argues either Daikin is conservative or the cohort is over-extrapolating. Second, Daikin is a late, sub-scale entrant: NA data-center sales were only ~¥100B in FY ended March 2026 (~2% of group), targeting ¥300B+ by FY2030 (Fact; Daikin transcript). It is chasing this with ¥1.3T of FUSION30 capex + ¥500B of M&A — deploying capital into the part of the industry where the capital cycle is most hostile, while DNA+DAA (its North American HVAC) earn only ~8% OP margin today against a 13% FY2028 target. The “premium applied” economics are an aspiration, and Daikin is entering the frothiest pocket of the industry years after Vertiv, Schneider, Carrier and JCI built positions — a 2027 order air-pocket is a real, named risk.

3.7 China, India, and the PFAS overhang

China is the industry’s weakest regional pool and Daikin’s clearest structural drag: a Gree/Midea/Haier price war intensified by Xiaomi (share to ~15% in H1-2025) and Aux, atop a property collapse. Daikin’s China AC&R sales fell ¥458.8B → ¥428.1B → ¥403.5B (FY23→FY25), the third straight annual decline, with management citing the “deteriorating real estate market and sluggish consumer spending” (Fact; Daikin Results Presentation p.13). Daikin holds margin only by retreating to premium multi-split — a defensible but shrinking position. India is the inverse: the best long-duration secular growth, with AC household penetration only ~8% (vs ~90% US, ~100% China), a market scaling ~$6.2B (2025) → ~$21.6B (2034) at ~14.8% CAGR (Fact; IMARC/Zerodha). But Daikin is not #1 — Voltas holds ~19.5% room-AC share versus Daikin ~18% — so management’s “No. 1 share” ambition is an aspiration, not a current fact (a flagged contradiction: management implies leadership the external data does not support).

Finally, PFAS / “forever chemicals” regulation is a genuine Daikin-specific overhang on the very fluorochemical vertical integration that is its differentiation. Potential restrictions cover the exact refrigerants Daikin makes (R-125, R-134a, R-1234yf, R-410A, R-454B blends) and its fluoropolymers; Daikin America faces EPA test orders and is spending $300M+ on PFAS capture (Fact; HPAC Magazine / EPA, accessed 2026-06-17). The Chemicals segment is already a drag — operating profit fell 28% (¥46.1B → ¥33.1B) in FY ended March 2026 on weak semiconductor/fluoropolymer demand — so the “vertical-integration moat” is currently a regulatory liability and an earnings headwind, not a buffer (Interpretation). The AHT refrigeration acquisition (Austria, 2019) took a ¥11.8B impairment in FY26, a concrete data point that the commercial-refrigeration sub-segment is structurally weak for Daikin and prior M&A is underperforming.

3.8 Peer EV/EBITDA comp

Daikin is the cheapest major HVAC name on EV/EBITDA — but the cross-sectional multiple is not a quality ranking; it reflects mix and domicile (Interpretation). The US cohort is rich-to-full on its own history (Trane 91st percentile, JCI 98th P/S, AAON 96th, FIX 99th), with only Lennox screening “cheap” — and explicitly because the market doubts its 20% margin is durable.

The ~10x discount is largely deserved: Daikin earns roughly half the operating margin and ROE of the premium US applied players, carries a residential/China/chemicals cyclical mix, and bears a Japan-domicile multiple. The legitimate offsetting question — left to Section 10 — is that the ~25x trailing P/E sits on a trough 8.3% margin (depressed ~1pp further by J-GAAP goodwill amortization of ~¥51B/yr, which IFRS peers do not carry; this is why EV/EBITDA, not P/E or operating margin, is the honest cross-border comparator), so the headline multiple overstates expensiveness if margins recover — but that recovery is unproven.

Verdict

Structurally above-average but sharply bifurcated — and Daikin sits more in the weak pools than its “premium compounder” narrative implies. The case for a good industry is strong: a consolidated oligopoly (top-5 ~32% global / ~84% NA), durable share leadership with no new entrant in decades, a non-discretionary ~130M-unit replacement annuity, regulatory ASP tailwinds from the A2L transition, and ~20%+ segment EBITDA in the best pools. The disconfirming evidence — weighed and not dismissed — is that the attractiveness is concentrated: the durable returns live in NA applied/commercial and the service aftermarket (where Trane and JCI over-index), while residential unitary, EU heat-pump heating, and Chinese residential AC are commoditizing, subsidy-dependent, or in an outright price war. Daikin is over-indexed to all three weak pools and underweight the annuity (Solutions 28% vs Trane’s 43%-GM, ~34%-mix service moat). Its own numbers corroborate the bear read: five straight years of margin decline (10.2% → 8.3%), ROE 12% → 9.1%, and a textbook Marathon capital-cycle misstep — capacity built for +3% / +24% demand that delivered -9% / -11%. The bull pillar everyone underwrites (data-center cooling) is the part of the industry where the capital cycle is most dangerous, and Daikin is a late, ~2%-of-sales entrant chasing it with ¥1.3T of capex into a fragmenting, capital-flooded segment. The PFAS overhang sits on the very fluorochemical integration that is supposed to be its edge. Net: the industry deserves an above-average structural grade; Daikin’s position within it is good-but-not-premium, currently contested at exactly the residential/Chinese/European seams where structure is weakest — and being squeezed there now, not protected from it.

4. Competitive Position

The right question for Daikin is not “does it have a moat?” — the global #1 HVAC&R franchise plainly has something — but “where exactly is the moat, and why is it not showing up in the returns a real moat is supposed to produce?” The answer is that Daikin holds two genuine, Greenwald-codable advantages concentrated in a minority of its revenue, surrounded by a much larger residential/unitary/China business that competes on price against scaled, hungry rivals. The franchise is broad; the advantage is narrow. And the financial test of a moat — peer-superior, durable returns on capital — Daikin currently fails versus its US commercial-applied peers, earning roughly half their margins and ROE. This section names each putative advantage in Greenwald’s taxonomy and pressure-tests it against the specific financial outcome that would deteriorate without it.

Moat 1 — Global scale and #1 share (economies of scale: real, under-monetized)

Daikin is the world’s largest HVAC&R company by revenue: ~$36B (¥5,015.0B) of sales and an estimated ~15% global AC share, ahead of Gree (~$29B), Carrier (~$20B) and Midea (~$19B), and #1 in every major region except China (Interpretation — the “~15% share” figure is third-party, TheChillBrothers/OpenPR synthesis, 2026-06-17, not primary; treat as approximate). In Greenwald’s framework, economies of scale are only a moat when paired with customer captivity that keeps the scale leader’s market share from being competed away. Daikin’s scale is genuine in absolute terms — it funds ¥150.7B of R&D (up from ¥81.5B five years ago) and ¥300B of capex, and lets it amortize refrigerant-transition and tariff costs across a global base no single-region peer can match (Fact, Daikin Financial Data summary, 2026-05-12). The clearest financial evidence the scale is real: in the year ended March 2026 Daikin fully absorbed ~¥41B of direct US tariff impact through pass-through pricing plus copper→aluminum/stainless substitution, with a selling-price benefit of +¥120.5B and cost reductions of +¥81.0B in the operating-profit bridge offsetting -¥88.0B of raw materials (Fact, Results Presentation p.4/p.8, 2026-05-12). A sub-scale OEM could not have priced through that.

But the test the scale fails is monetization. Scale that produced a moat would show up as peer-leading margins; Daikin’s consolidated operating margin is 8.3% and falling — it has declined every single year from 11.1% (FYE March 2019) to 8.3% (FYE March 2026), while revenue roughly doubled. That is negative operating leverage: scale went up, unit economics went down (Fact, Daikin 10-year summary). Under Greenwald’s logic the scale advantage is real but is being spent — on refrigerant transitions, on defending residential share, on building data-center capacity — rather than captured. Verdict on Moat 1: a real economies-of-scale advantage that is genuinely load-bearing (refrigerant leadership, tariff absorption) but under-monetized to the point that, on returns alone, it is invisible.

Moat 2 — Fluorochemical + compressor/inverter vertical integration (supply/cost: genuine and unique)

This is Daikin’s strongest and most differentiated claim, and the one with no clean analog among peers. Daikin is the only major AC OEM that manufactures its own fluorochemical refrigerants — it makes R-32, A2L blends and R-290, owns the associated IP, and pairs that with in-house compressors and inverters (Interpretation/Fact, FY-ended-Mar-2026 transcript, 2026-05-12). Management names “inverters, heat pumps, and refrigerants” as proprietary strengths “not easily imitated” — commentary that should be treated as a hypothesis, but here corroborated by an external datum: Daikin lost and then recovered US Ducted Unitary share to ~24% (above its pre-loss FYE-March-2023 level) on the back of leading the R-410A→R-32/A2L transition (Fact, transcript Q&A). The financial outcome that would deteriorate without this integration is precisely first-mover margin and share through forced regulatory refrigerant changeovers — and the A2L transition is exactly such a forced changeover (R-410A US manufacturing ban Jan-1-2025). This is a real supply/cost advantage that pure-OEM peers (Carrier, Trane, Lennox, the old standalone Goodman) cannot replicate.

Two caveats stop this from being a wide moat. First, the Chemicals segment that is the integration is itself cyclical and currently shrinking: segment profit fell ~28% YoY (¥46,119M → ¥33,089M, margin 17.5% → 11.8%) on weak semiconductor-related fluoropolymer demand (Fact, Brief Report segment note). The asset underpinning the moat narrative is having a bad year. Second, the advantage carries a tail risk no peer shares to the same degree: PFAS / “forever chemicals” regulation. Tightening global restrictions on fluoropolymers and certain refrigerants are a Daikin-specific overhang on the exact edge being cited as the moat (Open Question — the durability of the fluorochemical advantage under PFAS regulation is unresolved). Verdict on Moat 2: genuine, unique supply/cost advantage — the single best moat element Daikin owns — but cyclically depressed today and exposed to a regulatory tail the peers do not carry.

Moat 3 — VRV/VRF applied leadership (demand-captivity: the stickiest pocket)

Daikin invented variable refrigerant flow in 1982 and owns the “VRV” trademark — every competitor is legally forced to market the identical technology as the generic “VRF” (Fact). Applied/commercial — chillers, VRV, custom air-handling via Alliance Air, increasingly data-center cooling — is the most engineering-intensive, specifier-driven, least-commoditized pocket of HVAC, and is where switching costs are real: a building specified around Daikin VRV carries installed-base, controls and service captivity that does not transfer to a rival. In Greenwald’s taxonomy this is demand/customer-captivity, and it is the part of Daikin’s franchise most directly comparable to the premium applied moats of Trane and Johnson Controls. The financial outcome that would erode without it is applied pricing power and specifier captivity — the highest-quality revenue Daikin earns. Verdict on Moat 3: the most defensible advantage Daikin has on returns quality — but it is a minority of revenue, and the two-segment disclosure hides whether applied/VRF actually earns materially above the commoditizing residential core (Open Question — no product-line gross margins disclosed).

Moat 4 — US ducted/residential platform via Goodman (mixed; Greenwald-weak/contestable)

The 2012 Goodman acquisition built genuine US distribution scale — but the returns expose the contestability. Daikin’s North American AC business (DNA residential + DAA applied) earned only an ~8% operating margin in the year ended March 2026, and management openly concedes both “have struggled to improve profitability” (Fact, transcript Q&A). FUSION30 targets lifting NA AC profitability from 8% to 13% by FY2028 — an explicit admission the platform under-earns. Most damning for the moat case is the share-stability test, Greenwald’s sharpest tool: Daikin lost US Ducted Unitary share around FYE-March-2023 during the refrigerant-transition disruption and had to claw it back to ~24% by FYE-March-2026. A wide-moat franchise does not lose and regain residential share on a product-transition stumble; that volatility is the signature of a strong scale player in a contestable market, not an entrenched captive franchise. The recovery does evidence the real refrigerant/product edge (Moat 2 doing the work), but the loss proves the residential side is contestable on price. Verdict on Moat 4: scale yes, durable moat no — Greenwald-weak; a low-return, contestable platform whose share moves with product cycles.

Moat 5 — Service/aftermarket attach (aspirational, not yet an annuity)

The highest-quality HVAC moats are anchored by recurring service/parts annuities — Trane runs an estimated ~43% service gross margin. Daikin does not separately disclose a service/aftermarket revenue or margin line. “Expanding the Service Solutions business globally” is a FUSION30 priority theme with a 28%→40% “Solutions” mix target — i.e. it is an ambition, not an established profit pool (Fact, Results Presentation p.5; Interpretation on its absence as a moat). Verdict on Moat 5: aspirational — the annuity that anchors peer moats is a goal Daikin is chasing, not an advantage it already owns.

Moat 6 — Brand / dealer networks (real, regionally)

Daikin’s regional brand and dealer relationships are real and were central to the US R-32 share recovery (the dealer-development push behind the Fit platform). This is local-distribution/habit captivity in Greenwald’s terms — genuine but regional and not a group-wide moat. Verdict: real where it exists, but a supporting advantage, not the thesis.

Head-to-head: Daikin earns ~half the commercial-applied peers

The ROIC and market-share tests both point the same way: the advantages are real in kind but absent in returns. Cross-border comparison must use EV/EBITDA, not P/E or headline operating margin, because Daikin’s J-GAAP goodwill amortization (~¥51.4B/yr, plus an ¥11,849M FY26 AC&R impairment) depresses reported margin ~1pp versus IFRS/US-GAAP peers (Carrier, Trane, JCI) who do not amortize goodwill. Even normalizing for that ~1pp, the gap is structural, not accounting:

Peer margins/ROE are from public filings and market data; Daikin figures are primary (Brief Report, 2026-05-12). The pattern is unambiguous: Daikin’s edge over peers is breadth and vertical integration, not per-unit profitability. The US-listed commercial-applied players monetize a narrower but stickier profit pool — applied equipment plus service annuity — far better than Daikin monetizes its broad-but-residential-heavy one. Management knows it: FUSION30 explicitly aims to reach ROE/margins “in line with our US rivals” by 2028-2030, which is a written admission Daikin currently trails Trane/Lennox on the exact financial outcomes a moat should produce. Meanwhile the bottom of the competitive set — Gree and Midea dominating China residential by volume, with China AC sales down a third straight year (¥458.8B → ¥428.1B → ¥403.5B) — sits squarely on Daikin’s weakest, most contestable turf, where it has no scale advantage at all.

Verdict: a partial / narrow moat that is not (yet) in the returns

Daikin possesses genuine, Greenwald-codable advantages — a unique supply/cost edge from fluorochemical + compressor/inverter integration (Moat 2), demand-captivity in VRV/VRF applied (Moat 3), and a real-but-under-monetized economies-of-scale lead (Moat 1). The pricing-power evidence is concrete: ~¥41B of tariffs passed through, US share clawed back to 24%. But a moat is only a moat if it shows up in a financial outcome that would deteriorate without it, and Daikin’s returns do not pass that test against peers. Operating margin has fallen for seven straight years; ROE is 9.1% versus Trane’s ~36%; ROIC (~8.0%) barely clears WACC; the share-stability test is failed on the residential side (lost-then-regained US share); the aftermarket annuity that anchors the best HVAC moats does not yet exist as a disclosed profit pool. The disconfirming evidence weighed honestly: the refrigerant and VRF advantages are real and the US share recovery shows them converting to share — so this is not “no moat,” and the franchise is not commoditizing wholesale. The accurate characterization is a partial, narrow moat concentrated in fluorochemical integration and applied/VRF engineering, surrounded by a large, contestable, price-competitive residential/unitary/China business — structured and currently earning more like a good global industrial than a wide-moat compounder. Whether the genuine advantages ever surface in peer-competitive returns is the FUSION30 self-help question — and it is unproven, given FUSION25 missed the identical 10%-margin / 12%-ROE targets outright.

5. Growth History and Forward Opportunities

Five-year revenue history: ¥3.1T → ¥5.0T, but the trail flatters the engine

Daikin grew net sales from ¥3,109.1B (year ended Mar-2022) to ¥5,015.0B (year ended Mar-2026) — +61% over four years, ~¥2x since FY-ended-Mar-2019 — and crossed ¥5T for the first time, +5.5% YoY (Fact; Daikin Brief Report on Settlement of Accounts (J-GAAP), 2026-05-12). On the headline this is a growth franchise. Underneath, three things degrade the quality of that growth: it was heavily aided by a weakening yen, it was disproportionately acquired/price-driven rather than volume-driven, and it diluted rather than expanded margins (see Section 6).

FX-flattered top line (Interpretation). ~84% of sales are overseas (Fact; Brief Report 2026-05-12), so the reported yen revenue line is highly sensitive to translation. The yen weakened materially across the window — Daikin’s own FY-ending-Mar-2027 guidance assumes USD/JPY ¥145 against a ¥151 actual realized in the just-ended year, and explicitly quantifies a ~¥20B operating-profit FX headwind for the coming year on that mean-reversion assumption (Fact; Results Presentation, 2026-05-12). The corollary is that a meaningful slice of the ¥3.1T→¥5.0T rise is translation, not real volume — a constant-currency top line would show distinctly slower growth. Treat the +61% as partly an FX artifact (Interpretation), and weight the underlying unit/price story below.

Segment composition (year ended Mar-2026):

AC&R is the franchise; Chemicals is a high-margin but cyclical satellite whose sales grew +7.0% even as profit fell -28.3% on weak semiconductor and fluoropolymer demand (Fact; Brief Report 2026-05-12) — a margin-collapse story, not a demand-collapse one, and the single biggest YoY profit drag.

Regional divergence: the US and Europe carried a stalling Asia

Revenue by geography (year ended Mar-2026, all segments):

The growth was front-loaded onto the Americas and Europe while China and Asia-Oceania shrank (Fact; Brief Report 2026-05-12). On an AC-only basis the Americas reached ¥1,898.6B (~41% of AC&R), Europe ¥784.2B, Japan ¥675.8B, and China fell to ¥403.5B — a third consecutive annual decline (Fact; Results Presentation 2026-05-12). The +9-10% US/Europe prints are partly price and FX, not pure volume: management itself disclosed that North-American AC unit sales over FY23→FY25 came in at a -9% CAGR against a +3% forecast, and Europe at -11% against +24% (Fact; FUSION30 plan, p.9, 2026-05-12). Revenue rose while units fell — the gap is price, mix, and yen. That is the signature of a company growing the invoice faster than the box count, which is fine if pricing sticks and ugly if it is commoditization-driven discounting offset by FX (see Section 4 / Section 6).

Organic vs. acquired (Interpretation). The structural step-up in scale since 2012 is overwhelmingly the Goodman acquisition (the US residential platform, now Daikin Comfort Technologies North America / “DNA”), AAF (filtration) and AHT (Austrian refrigeration). The US — Daikin’s single largest market at ~36% of group — exists at this scale because of M&A. The most recent four-year doubling is a blend of (a) acquired base compounding, (b) price/refrigerant-mix ASP, and © FX. Pure same-store volume growth, stripped of those three, is modest — and in the two key Western markets it was negative over FY23→FY25 (Fact; FUSION30 p.9). This is not a high-volume-growth story; it is a price/FX/acquired-scale story.

Forward opportunities — honestly sized

• Data-center / AI cooling — real, but small and late. Daikin’s North-American DC sales were ~¥100B in the year ended Mar-2026, ~2% of group, targeting ¥300B+ by FY2030 (Fact; transcript 2026-05-12). The catch: Daikin’s own DC-cooling TAM CAGR (~11%, ¥1.4T→¥2.7T) is roughly half the ~22% narrative the cohort markets, and Daikin is a sub-scale, late entrant against Vertiv, Schneider, JCI, Carrier and AAON all racing capacity in (Interpretation; industry research, 2026-06-17). This is mean-reverting optionality, not a moat.
• India — best secular growth, but Daikin is not #1. AC household penetration is ~8% (vs ~90% US), the market compounds ~14.8% to ~$21.6B by 2034, and Daikin targets 5M units / #1 share by 2030 (Fact; IMARC / Daikin FUSION30 p.9). But today Voltas (~19.5%) leads Daikin (~18%) in room AC, so management’s “No.1” is an aspiration, not a fact (Open Question), and India OP margin is only ~8% on prior capacity build.
• Heat-pump electrification — subsidy-dependent and currently a value-destruction zone. EU heat-pump sales fell -23% in 2024, recovered only +11% in 2025; the US IRA 25C credit terminated 12/31/2025 (Fact; EHPA / Rewiring America). Daikin built European capacity for +24% growth that came in at -11%.
• Applied / Solutions mix-shift — the right pivot, unproven. Daikin targets Solutions (service/aftermarket annuity) from 28%→40% of sales (Fact; FUSION30 p.13). This is the durable profit pool US peers over-index to (Trane ~43% service gross margin), and Daikin is behind — the pivot is necessary precisely because equipment economics are eroding.
• Refrigerant ASP — the R-410A→A2L (R-32/R-454B) transition lifts unit ASPs and plays to Daikin’s owned-refrigerant edge, but drove a 2024 US pre-buy and a 2025 payback slump.

Verdict — high revenue growth, LOW-quality growth

The revenue line doubled, but the growth is low-quality: it diluted margins every year (Section 6), leaned on FX and acquired/price ASP more than organic volume, and — most damningly — Daikin built capacity into a demand forecast that collapsed (NA units +3% forecast vs -9% actual; Europe +24% vs -11%; Fact, FUSION30 p.9), the textbook Marathon “supply into an air-pocket.” The forward pillars are genuine but each is either small (DC ~2% of sales), contested (India #2, not #1), subsidy-cliffed (heat pumps), or a not-yet-delivered pivot (Solutions). The disconfirming evidence — US/Europe revenue +9-10%, India’s secular runway, the applied/DC optionality — is real, which is why this is low-quality growth, not no growth. But growth that arrives with falling returns is value-neutral at best.

6. Financial Quality

The central fact: scale produced negative operating leverage

Daikin’s profitability is at a multi-year low even as revenue roughly doubled. Operating margin fell every single year — 11.1% (FYE Mar-2019) → 10.2% → 9.5% → 8.9% → 8.5% → 8.3% (FYE Mar-2026) — and ROE fell from ~15.7% (Mar-2019) to 9.1%, with ROA down 6.2%→5.0% over the recent window (Fact; Daikin Financial Data IR + Brief Report, 2026-05-12). Revenue grew ~60% over FY22→FY26 (and ~2x since FY19) while operating profit grew only ~31% (¥316.4B→¥415.0B) and margin fell ~280bps. For a self-styled global #1 claiming scale advantages, scale not surfacing in the financials is a yellow flag (Interpretation). Computed ROIC is ~8.0% — barely above a ~7-8% WACC (Interpretation; NOPAT ¥294.4B on invested capital ~¥3,704.6B). The cleanest one-line summary of the bear case: Daikin got twice as big and half as profitable on ROE.

Five-year financials (year-end convention; period ending March 31)

(Source: Daikin Brief Report 2026-05-12; Financial Data IR; live market data reconciled to filing.)

Decomposing the compression: cyclical and structural, not one or the other

The ~280bps margin fall is a blend — the central QoE conclusion.

Cyclical / transient drivers (the latest-year softness):

• Chemicals -28.3% profit (¥46.1B→¥33.1B; margin 17.5%→11.8%) on semiconductor/fluoropolymer weakness — the single biggest FY-Mar2026 YoY drag (Fact; Brief Report 2026-05-12).
• China AC volume -4.5% on the property-sector recession — though the China segment margin held at ~22%, so this is a revenue/volume problem, not a profitability collapse (Fact; Results Presentation 2026-05-12).
• US residential destocking after the A2L pre-buy; management confirms DNA distribution inventory “has not changed… not able to significantly reduce it” (Fact; transcript 2026-05-12).
• ~¥88B raw-material/tariff/logistics cost inflation, including a ~¥41B direct US tariff hit “absorbed by utilizing pass-through pricing” (Fact; Results Presentation 2026-05-12).
• FX -¥11.5B YoY.

Structural drivers (the deeper 7-year erosion):

• Management-admitted commoditization — “the decline in equipment profitability due to lower selling prices caused by commoditization and intensified price competition,” naming the emergence of Chinese manufacturers (Fact, but management commentary = hypothesis; FUSION30 transcript 2026-05-12). This is a direct moat-erosion admission, concentrated in residential unitary/China.
• Under-earning growth capital — the Goodman/DNA, AAF and India buildout earns below the corporate average; DNA+DAA together earn only ~8% OP margin, with management conceding “delays in recovering capital investment” and “we lack the speed… to bounce back” (Fact/hypothesis; transcript 2026-05-12). This is the crux of the ~8% ROIC.
• Mix dilution — acquiring lower-margin residential/distribution (Goodman) and filters (AAF) against higher-margin legacy commercial/chemicals.
• J-GAAP goodwill amortization ~¥51.4B/yr (~1pp of sales) — an accounting, not economic, drag that IFRS peers (Carrier/Trane/JCI) do not bear (Fact; Brief Report 2026-05-12). This is the single most important cross-border comparability adjustment: it depresses Daikin’s reported margin and P/E vs IFRS peers, which is exactly why peer comparison must run on EV/EBITDA (which adds it back), never on headline operating margin or P/E.

The operating-profit bridge (Mar-2025 → Mar-2026)

(Source: Results Presentation p.8, 2026-05-12. OP rose only ¥401.7B→¥415.0B, +3.3%.) The read: input-cost inflation plus growth fixed-cost drag nearly swallowed pricing and volume. This is not a demand collapse — volume and price both contributed +¥80-120B — it is a cost/investment-absorption problem.

Gross margin is stable — the compression is below the GP line

Gross margin held ~34% (34.5% in the year ended Mar-2026; GP ¥1,732.5B / sales ¥5,015.0B) across the five years (Fact; Brief Report 2026-05-12). This matters: stable GM argues the product-level pricing has not collapsed — the ~280bps operating-margin erosion sits in SG&A, rising DD&A (depreciation rose 170→197→225B as capacity came online), goodwill amortization, and mix. It tempers the “commoditization is eating the moat” reading: the gross-margin floor has held, which argues commoditization is localized to residential/China, not a broad collapse (Interpretation/Open Question).

Segment economics

• AC&R (¥4,621.1B / 8.2% margin) actually improved (+7.4% profit, margin 8.0%→8.2%) — the core is fine; consolidated compression is driven by Chemicals and DD&A, not the franchise (Fact; Brief Report 2026-05-12).
• Chemicals (¥281.5B / 11.8% margin) is the swing — profit -28.3% on semiconductors, and carries a PFAS regulatory overhang on the very fluoropolymers/refrigerants that are Daikin’s differentiation. The vertical-integration “moat” is currently a drag, not a buffer.
• NA AC (DNA+DAA) ~8% margin — the under-earning growth-capital block; management targets 13% by FY2028 but admits it has “struggled to improve.”

ROE / ROIC decomposition — and the FX-translation distortion

ROE fell 12.0%→9.1% (Mar-2022→Mar-2026); ROIC ~8.0%. A critical nuance: reported equity is materially inflated by yen-weakness FX translation. The foreign-currency translation adjustment in OCI rose +¥256.0B (¥512.3B→¥768.3B) in the year ended Mar-2026 — the single largest driver of the ¥449.8B net-asset increase, and roughly half of comprehensive income (Fact; Brief Report equity statement, 2026-05-12). With ~84% overseas sales, a weak yen mechanically inflates the equity denominator and suppresses reported ROE independent of operating performance. A constant-currency ROE would be materially higher than 9.1% (Interpretation/Open Question) — so part of the apparent “ROE deterioration” is accounting optics, not economic decline. This cuts both ways: it also means the reported ~9.1% understates underlying capital efficiency, but does not rescue the absolute ~8% ROIC vs WACC.

Cash flow and reinvestment

OCF was ¥465,848M (-9.4% YoY), covering net income ¥275.2B ~1.7x — no NI/cash divergence (Fact; Brief Report 2026-05-12). FCF reads ~¥257B on an OCF-minus-PP&E-capex basis (¥465.8B - ¥209.1B) or ~¥140B on the databook’s broader-capex definition. The gap is the ~¥300B/yr total reinvestment (capex + intangibles), running at 6-7% of sales with capex/depreciation >1.3x in FY23-24 — capacity addition, the Marathon late-cycle warning, into the demand air-pocket above. The reinvestment is genuine and conservatively funded, but it has not yet earned its return (ROIC ~8%).

Quality-of-earnings flags

These are small, disclosed, and — Turkish/AHT aside — favorable to earnings quality. Conservative J-GAAP, no aggressive capitalization, zero dilution (293.1M shares flat; SBC immaterial), and the new ¥350B ASR will shrink the count.

R&D and reinvestment intensity — rising spend, not-yet-rising returns

R&D climbed steadily from ¥81.5B (FYE Mar-2022) to ¥150.7B (FYE Mar-2026) — +85% over four years, outpacing the +61% sales growth (Fact; Brief Report / Financial Data IR, 2026-05-12). On a stable-margin business that would be a tolerable mix shift toward higher-value product; here it lands on a falling margin, which means the incremental R&D and capex are being spent into a period where the franchise is monetizing less per yen of investment, not more. Stacking the two reinvestment lines together — R&D ¥150.7B plus capex ~¥300B — Daikin reinvested ~¥450B (~9% of sales) in the year ended Mar-2026 against ~¥257B of OCF-minus-PP&E FCF. The company is outspending its own franchise-level free cash flow on growth when total intangibles and M&A are included (databook FCF ¥139.6B), which is why ROIC sits at ~8% and why the FUSION30 plan’s explicit pivot — cutting the invest/OCF ratio from 58% to 43% and “focusing on profit margins and capital efficiency instead of chasing volume” (Fact/hypothesis; FUSION30 transcript 2026-05-12) — is itself an admission that the prior reinvestment cadence was destroying, not creating, marginal return. The disconfirming read: the heavy spend built the US/India/applied platforms whose harvest “is just beginning” (management’s words; hypothesis), and if those mature toward target margins the same capital base re-rates. That is the entire bull case for the financials, and it is forward-looking, not in the numbers yet.

Is the FY-ending-Mar-2027 margin inflection credible? Weigh it against FUSION25’s record

Guidance for the year ending Mar-2027 is sales ¥5,150B (+2.7%), operating profit ¥436B (+5.1%, margin up to ~8.5%), NI ¥278B (+1.0%) — and it does so despite a built-in ~¥20B FX headwind (the ¥145 USD/JPY assumption vs ¥151 realized), so ex-FX the guide implies ~+10% underlying OP growth (Fact; Brief Report / Results Presentation 2026-05-12). That is a genuine margin-inflection bet, and it is the first datapoint the recovery thesis needs. Two cautions weigh against taking it at face value. First, the ¥436B OP guide is roughly flat against the ¥435B Daikin had originally planned for the year just ended — i.e., a full year of intended growth was lost, and the company is only now guiding back to where it expected to be a year ago. Second, the credibility record is poor: FUSION25 set the same shape of targets (11% margin / 12% ROE) and missed outright (8.3% / 9.1%), and its demand forecasts (+3% NA AC units, +24% Europe) came in at -9% and -11% (Fact; FUSION30 p.9, 2026-05-12) — capacity built into an air-pocket, the Marathon capital-cycle signature. The new FUSION30 path (8.3%→10% margin by FY2028→12% by FY2030; ROE 9.1%→12%→15%) rests on the same forecasting apparatus that produced those misses. Treat the guided inflection as a hypothesis to be validated quarter-by-quarter, not a base case (Interpretation). The single most important thing to watch is whether the AC&R segment margin and the DNA+DAA ~8% margin actually move toward target over the next 2-3 quarters; the gross-margin trend and channel-inventory normalization in US residential are the leading indicators.

Fortress balance sheet

Equity ratio 55.9%, net debt ~¥388B, net debt/EBITDA ~0.56x, interest coverage ~10.6x, cash & equivalents ¥706.5B (Fact; Brief Report 2026-05-12). Ample capacity for the ¥350B ASR plus continued growth investment. Inventory is elevated (¥1,137.9B, up ¥85B) on the M&A/destock cycle — a channel overhang to watch, but not a solvency issue. The fortress sheet is what makes the self-help re-rating fundable (a ¥350B ASR plus ¥1.3T capex + ¥500B M&A over five years is well within capacity at 0.56x net leverage); it does not make the margin recovery itself any more certain.

Verdict — do economics improve with scale? No.

They have not. Over seven years revenue ~doubled while operating margin fell ~280bps and ROE ~660bps — negative operating leverage, the opposite of what a scale moat should produce. The disconfirming evidence is real and must be weighed: the AC&R core improved (+7.4% profit, 8.2% margin); gross margin held ~34%, so this is not a product-pricing collapse; roughly half the latest-year softness is genuinely cyclical (Chemicals, China, destock, tariff/raw-material inflation, FX); and a chunk of the ROE decline is yen-translation optics, not economic erosion. Earnings quality is high — cash-backed, conservatively accounted, undiluted, fortress-financed. But the bull case for a re-rating rests entirely on a self-help recovery that is management-asserted and unproven: FUSION25 targeted 11% margin / 12% ROE and missed outright (landed 8.3% / 9.1%), and its demand forecasts (+3% NA / +24% EU) collapsed (-9% / -11%). The honest verdict is a structurally good, cash-generative, conservatively-financed franchise currently earning below its own potential — where current depressed profitability is a blend of cyclical and structural, and recovery is unproven, not demonstrated.

7. Capital Allocation

Daikin’s capital-allocation record is a study in two halves: a competent, franchise-building M&A history that earned it the global #1 position, set against a multi-year deterioration in the return on retained capital that the empire-building obscured. The defining 2026 development is that this gap was forced into the open by an activist — and that Daikin, for the first time in its corporate history, is now run against capital-efficiency metrics. The honest read is a competent operator whose retained capital has been earning a falling return, now on probation with a credible-but-externally-imposed self-help plan it has not yet delivered.

M&A history — one transformational success, one clear miss, a string of bolt-ons

Goodman is the anchor of the bull case on management’s competence (Fact: $3.7B at ~1.76x sales, ACHR News, 2012-09-10). It gave Daikin the US ducted-unitary and furnace lines it could not build organically, made it the global #1, and is the platform now capturing US applied and data-center demand. AHT is the counterweight (Fact: Daikin Brief Report, 2026-05-12, p.19): an acknowledged integration failure that just took an ¥11.8B write-down, with remaining AHT intangibles (goodwill ¥13.2B + customer relationships ¥39.1B + other ¥29.9B at Mar-2025) still at risk if the business keeps missing plan (Open Question). The aggregate balance is positive on franchise-building but uneven on price discipline — and the legacy of acquisition is a goodwill problem the auditor singles out.

The goodwill overhang and the J-GAAP drag

M&A-derived goodwill stood at ¥274,767M at Mar-2026 (~9% of total assets), with customer-related intangibles of ¥228,708M — together ~¥503B, the auditor’s single Key Audit Matter (Fact: Daikin Brief Report, 2026-05-12; Deloitte FY25 KAM). Two consequences matter for the thesis. First, under J-GAAP this goodwill is amortized — ¥51,371M of expense in FY ended Mar-2026 (Fact) — a real, recurring drag on reported operating profit (~1pp of margin) that IFRS peers Carrier, Trane and JCI do not take. Cross-border margin and P/E comparisons therefore overstate Daikin’s gap to US peers; the cleaner read is on EV/EBITDA, where amortization is added back. Second, the goodwill sits on a balance sheet whose returns have been falling — the precise condition under which acquisition accounting becomes a risk rather than a footnote.

The core problem: retained capital earning a falling return

The hard evidence of capital-discipline weakness is the trend, not any single deal. Daikin retains ~64% of earnings — FY ended Mar-2026 payout was 36.2% (DPS ¥340, ¥99,565M paid on ¥275,229M of profit attributable) — yet the return on that retained capital has fallen every single year:

(Fact: Daikin Financial Data summary; ROE was ~15.7% as far back as Mar-2019.) Net assets grew from ~¥2.87T to ~¥3.32T in FY26 largely via retained earnings and FX translation, while the return on that growing base shrank — i.e. retained capital was not earning its keep, and incremental ROIC (computed ~8.0%, barely above WACC) ran below both the prior average and, arguably, the cost of equity. This is a Marathon capital-cycle signature: Daikin grew the asset base aggressively into the post-COVID HVAC/heat-pump boom — capex ran ¥156.3B → ¥250.3B → ¥311.5B → ¥324.6B (FY21–FY24) before discipline cut it to ¥300.0B in FY26 — and returns mean-reverted as residential/heat-pump demand disappointed (management itself lists “capital investment was reinforced in anticipation of future demand growth” as a margin-pressure cause; FUSION30 deck p.8). Capacity was built into an air-pocket. R&D, by contrast, is defensible reinvestment tied directly to the moat: ¥150.7B (~3.0% of sales) in FY26, up from ¥81.5B in FY21, directed at inverters and low-GWP refrigerants (R-32) — the supply/cost advantage the business actually earns excess returns on.

The most damning datapoint is the scorecard on management’s own plan. FUSION25’s profitability targets were missed outright: operating margin landed 8.3% against an 11.0% target (itself already revised down from 12%), and ROE 9.1% against a 12.0% target (Fact: FUSION30 deck p.7–8). The headline net-sales “beat” (¥5,015B vs ¥4,550B target) was substantially FX-flattered (budget ¥125/USD vs actual ¥151/USD). Management’s recent target hit-rate is therefore poor — the relevant prior for how to weight FUSION30’s promises.

The activist and the governance pivot

The under-levered balance sheet — equity ratio 55.9% (up from 51.5% in FY22), ¥706.5B cash, interest coverage 12.0x — was exactly the excess capacity an activist would target (Interpretation). Elliott Management disclosed a ~3% / ~$1B stake around April 15, 2026; shares rose ~14%, the biggest single-day gain since 2009 (Fact: Nikkei Asia, 2026-04-16). Elliott’s demands: a ~14% operating margin, doubled EPS, up to ¥1T of buybacks, integration of the six separately-managed North American units, and a portfolio review. Roughly four weeks later (May 12, 2026) Daikin’s FUSION30 plan delivered, for the first time in its history, a capital-efficiency architecture:

• A ¥350.0B Accelerated Share Repurchase — the largest buyback ever, ~5% of the ~¥6.6T cap, executed May 13 via ToSTNeT-3 (Fact: Brief Report, Subsequent Events p.21). Treasury shares were previously ~0.09%; this is roughly one-third of Elliott’s ¥1T ask.
• A newly-created CFO position responsible for financial strategy and capital policy.
• A target of 50% outside directors, including a non-Japanese independent with a financial-institution background.
• A comp overhaul: stock options abolished, replaced by performance-share units (70%, tied to D-ROIC + relative TSR vs TOPIX) and retention RSUs (30%), with malus/clawback and post-vesting holding requirements (Fact: FUSION30 deck p.36–37).
• A 5-year plan that cuts the investment/operating-cash-flow ratio from 58% to 43%, earmarks OCF ¥3,000B, capex ¥1,300B, M&A ¥500B, dividends ≥¥500B and buybacks/additional M&A ≥¥700B, with M&A steered toward Solutions/service bolt-ons rather than transformational deals (Fact: FUSION30 deck p.14–15, p.32).
• A dividend policy moved off the rigid DOE formula to “stable and continuously growing,” DPS guided ¥340 → ¥360.

The targets — operating margin 8.3% → 10% (FY28) → 12% (FY30); ROE 9.1% → 12% → 15%; invested-capital turnover →1.3x — are the explicit bull bridge. The credibility caveat is decisive: these are the same metrics FUSION25 missed, set by a management team that just conceded “our earning capacity has declined.” Treat them as aspirational, not committed.

Ownership and incentives

Ownership is diffuse with no controlling founder today (the Yamada family is historical). At Mar-31-2026 the top holders are nominee trust banks (Master Trust ~18.4%, Custody Bank ~7.3%), heavy foreign custodians (State Street ×3, Chase, JP Morgan — signalling large foreign ownership), Elliott International ~3.0% (8,828k shares), and a residual Sumitomo keiretsu remnant (SMBC ~2.4%, MUFG ~1.3%) — a likely cross-holding-unwind candidate under Prime-Market reform (Open Question). No customer exceeds 10% of sales. CEO Togawa’s total comp is ~¥389M (≈61% variable; third-party Simply Wall St, verify vs the Jun-2026 Securities Report) — modest globally and majority-variable, now tilting further to capital-efficiency metrics.

Verdict — mixed, improving under duress

Mixed; competent empire-builder, mediocre-and-declining capital returns, now on probation with a credible (forced) self-help plan. The bull evidence is real: Goodman built the franchise, R&D feeds the moat, and the governance/comp reform genuinely re-aligns incentives toward returns. But the disconfirming evidence is heavier and more recent: ROE fell every year for five years while ~64% of earnings was retained; capex ran hot into demand that did not arrive; AHT took an impairment; goodwill is the auditor’s sole flag; and FUSION25’s margin/ROE targets were missed outright. The pivot is the right one — D-ROIC pay, a 43% reinvestment ratio, the first real buyback in a decade — but it arrives only under activist duress, untested, with a multi-year execution-and-falsification gap. The capital-return discipline is, for now, a promise priced ahead of a track record.

8. Changes and Headwinds — Last Two Years

The defining feature of Daikin’s last two years is a divergence: the operating story weakened while the capital/governance response sharpened. Earning power has visibly ground lower for two years, and the company is now staking its re-rating on a self-help margin-restoration plan it has not yet delivered.

Recent-events timeline

The headwinds that bit — and persist

The year ended March 2026 was a record top line (¥5,015.0B, first time above ¥5T, +5.5%) but an operating-profit miss: OP of ¥415.0B came in ~¥20B / ~4.6% below the ¥435.0B initially guided in May 2025 (Fact: 4Q presentation p.3, 2026-05-12). The guidance arc tells the story — OP held at ¥435B through 1H, was cut to ¥413B at Q3 (Feb 4, 2026) citing demand “more severe than expected,” US tariffs, China property and semiconductor weakness, then nosed back to ¥415B on a strong Q4 self-help push. Four concrete headwinds bit simultaneously and carry into FY-ending-March-2027:

1. US tariffs — a ~¥41B direct OP hit (initially feared ~¥47B), absorbed only via price pass-through and a copper→aluminum/steel material switch (Fact: 4Q transcript). Management calls this “fully absorbed” — true for the tariff line, but it masks that overall earning power fell; absorption came alongside, not instead of, the demand and Chemicals deterioration (contradiction flag).
2. US residential air-pocket — the 2024 GWP pre-buy plus channel destock and high mortgage rates left residential “sluggish.” Daikin’s own inventory rose because it kept producing R-32 strategic product to win share — a working-capital drag it could not unwind on plan in North America and Asia/Oceania (Fact: 3Q transcript). The offset: it recovered US ducted-unitary share from ~22% to ~25% on the R-32 win-back — competitive execution, not loss.
3. China contraction — AC net sales fell to ¥403.5B (−6%), the third straight annual decline (¥458.8B → ¥428.1B → ¥403.5B), guided down a further ~6% to ¥380.0B for FY27. Driver: the property recession plus a Gree/Midea-led price war pressuring Daikin’s premium positioning; Daikin defends margin only by ceding volume to premium multi-split (Fact: 4Q presentation p.13/p.17).
4. Chemicals collapse — the highest-margin segment’s OP fell ~28% to ¥33.1B (margin 17.5% → 11.8%) on semiconductor weakness and fluoropolymer destocking; guided to recover only partially to ¥39.0B (Fact: 4Q presentation p.4). This is genuine profit-pool erosion in Daikin’s most differentiated business — and it sits behind an escalating PFAS litigation thread (Catoosa County GA, Huntsville AL May 2026, France Jan 2026), a structural overhang on the fluorochemical edge that underpins the moat.

Two new pressures were baked into FY27 guidance: an exogenous Middle East cost/demand risk (Middle East AC guided down to ¥156.0B / 88% YoY, with crude/logistics/fluoropolymer-procurement risk), and FX flipping from tailwind to headwind (a guided −¥120B to sales / −¥20B to OP at USD ¥145 vs ¥151) — so the reported +5.1% OP guide understates ~+10% ex-FX growth. Notably, the ¥436B FY27 OP guide is essentially flat versus the ¥435B originally planned for FY26 — a lost growth year.

Below the operating line, FY26 also carried the ¥11.8B AHT impairment and a +¥23.5B Turkey inflation-accounting swing — non-operating noise that distorts the bridge from OP to net income and flattered ordinary profit (+11.4%). A separate clarification: the FY26 deconsolidation of “Goodman Global Holdings, Inc.” (35 entities excluded) is a phased internal legal-entity merger into Daikin Comfort Technologies (rebrand began 2022), NOT a divestiture of the US residential business — easy to misread, and consistent with the integration Elliott is pushing.

The offsets

Against the weakening operating story, four developments preserve the long-term case: (a) the data-center applied build-out (DDC Solutions, the Tijuana ramp, a new global DC unit, a DC hub for FY27) — applied/commercial carried company growth while residential sagged, though this is also the Marathon capital-cycle danger zone every HVAC peer is racing into; (b) the US share recovery (~22% → ~25%); © FUSION30, an unusually candid returns reset explicitly admitting the gap to Trane/Carrier; and (d) the capital pivot — the ¥350B ASR, new CFO, more outside directors and ROE/D-ROIC-linked pay. Management’s own framing is the central confession: it “takes very seriously the fact that we have not achieved our operating margin; in other words, our earning capacity has declined” (Fact: FUSION30 transcript). That is an admission to weigh as evidence, not the reassurance management intends.

Verdict — net weakening of the thesis, with a genuine self-help offset

On balance these changes weaken the near-term thesis and raise execution risk, but they do not break the franchise. The disconfirming weight is real and dated: a guidance cut, three straight China declines, a Chemicals profit-pool erosion, tariffs and a new Middle East cost, FX reversing, and an admitted decline in earning power. But the franchise is intact — record revenue, recovered US share, a credible data-center growth pillar, and a decisive capital/governance reset. The deterioration is better characterized as earning-power slippage plus a richer-but-deserved valuation question than a broken business. The net effect is to shift the bull case onto an as-yet-undelivered margin-restoration plan, with the operating environment offering little tailwind — exactly the “recovering-from-trough, not yet delivered” posture the rest of this analysis describes. The changes are explicitly testable against FUSION30’s FY28/FY30 margin and ROE milestones; until those are hit, the thesis rests on promise, not proof.