Air Products & Chemicals

Phase A — Understand the business

Lens 1 · Company Overview

Air Products is a ~$12B-revenue, ~$63B-market-cap industrial-gases major — one of three Western "majors" (with Linde and Air Liquide) that, plus China's regional players, effectively constitute a global oligopoly in atmospheric and process gases. Founded 1940, Delaware-incorporated, NYSE: APD, operating in ~50 countries. The business is not a chemicals company in the cyclical sense — it sells oxygen, nitrogen, argon, hydrogen, helium, CO and syngas, and the production equipment to make them.

The economic engine is the on-site supply model: APD builds, owns and operates an air-separation unit (ASU) or hydrogen/syngas plant on or adjacent to a large customer's facility under 10–20-year take-or-pay contracts, with cost pass-through for energy and price escalators. Around 90% of sales are on stable, long-term take-or-pay contracts. The remainder is merchant (liquid/bulk and packaged gas sold off a regional grid at spot/contract) and a small sale-of-equipment book (cryogenic/gasification plants, sits in Corporate & other).

End markets: refining (hydrogen for desulphurisation/upgrading), chemicals (H2/O2/N2/CO/syngas feedstock), metals, glass, cement, electronics/semis (helium, specialty gases), food (N2 freezing), medical, and energy production (enhanced oil recovery, gasification). Customer base is broad and investment-grade-skewed; no single customer is disclosed as material to the whole. Equity affiliates (JVs in Algeria, China, India, Italy, Mexico, Saudi Arabia) contribute ~$648M/yr of after-tax income — a large, often-overlooked earnings stream (~24% of FY25 adjusted net income).

The differentiator vs. peers since ~2018 was an outsized clean-hydrogen / energy-transition growth bet (NEOM, Louisiana blue H2, Alberta, World Energy SAF). That bet is exactly what the current management is unwinding — see Lens 9.

Lens 2 · Supply Chain

Upstream inputs → APD → end customer, with named stakeholders along the chain:

• Upstream / inputs. (1) Air — free feedstock for ASUs (O2/N2/Ar); the cost is power. Electricity and natural gas are the dominant input costs and are largely passed through to on-site customers. (2) Natural gas — feedstock for steam-methane-reforming hydrogen and syngas (HyCO). (3) Helium — APD is a major refiner/distributor; sourced from natural-gas processing (US LaBarge/Cliffside legacy, Qatar via JV, plus new entrants — see Lens 13). (4) Electrolysers + renewable power for green H2 (NEOM uses thyssenkrupp/Haldor-class electrolysers; solar+wind on-site).
• Equipment. APD designs and builds its own ASUs and cryogenic equipment (vertical integration — a genuine moat input), and procures large rotating equipment (compressors: e.g. Atlas Copco / MAN / Siemens-Energy class) from OEMs.
• The company itself — ~25.3B of net plant & equipment; capital-intensity is the business.
• Downstream / customers. Refiners and chemical producers (the HyCO offtakers — e.g. Gulf-Coast refining complex), steel/glass/electronics manufacturers, merchant distributors, and — for the energy-transition assets — TotalEnergies (15-yr green-H2 supply deal; full 600 t/day NEOM hydrogen offtake) and Yara International (renewable-ammonia marketing, up to ~1.2 Mt/yr). Saudi partners ACWA Power and NEOM/PIF are co-owners (33% each) of the NGHC JV.

Chokepoints / single-source dependencies: (1) Power price/availability is the master variable — a 10% adverse move in CNY or EUR alone cuts ~$50/$34M of operating income. (2) Helium supply is a concentrated, geopolitically-exposed input (Qatar, US, now Russia/Algeria/South Africa) — and APD is on the wrong side of it right now (oversupply crushing its merchant helium pricing; Lens 5/13). (3) NEOM concentrates execution risk in one $8.4B Saudi asset with offtake still being assembled (Lens 13).

Lens 3 · Competitive Advantages (moats)

APD sits inside one of the best moats in the industrial economy — and it is structural, not brand:

1. On-site contract lock-in (switching costs + local monopoly). Once APD builds an ASU on a customer's fence line under a 15-year take-or-pay deal, the customer cannot switch without re-permitting, re-piping and stranding APD's asset — and APD has no incentive to leave because the asset is depreciated to match the contract. Each plant is effectively a local monopoly with a contractual annuity. ~90% take-or-pay.
2. Density economics in merchant. Liquid/bulk gas is uneconomic to truck far; whoever owns the nearest production node owns the region. The three majors have effectively carved the map — this is why pricing is rational and returns are high (Americas FY25 adj-EBITDA margin ~47% ).
3. Scale + vertical equipment integration. APD builds its own ASUs; it does not pay an OEM margin and controls plant design/efficiency — a cost-and-speed edge over a would-be entrant.
4. Capital + balance-sheet scale. Funding 10–20-year, multi-hundred-million-dollar projects (and a $6.1B project financing at NEOM) is a barrier no start-up clears.

Bargaining power: strong over merchant customers (local monopoly) and moderate-to-strong over on-site customers (locked in mid-contract, but competitively bid at renewal/new-plant award). Over suppliers (power, gas) APD is largely a pass-through taker, not a price-setter — the moat is in the contract structure, not input cost.

The moat is intact and arguably strengthening under the new regime — because management is redirecting capital away from the lower-moat, merchant-priced, demand-unproven energy-transition projects and back to the contracted on-site core. Vs. peers, APD's moat is the same kind as Linde's (the gold-standard operator) but APD has historically earned a slightly lower return on it — the entire bull case is that the Praxair-pedigree management closes that gap (Lens 9, 12).

Lens 4 · Segments

APD reports five segments: three regional industrial-gas segments (Americas, Asia, Europe), Middle East & India (mostly equity-affiliate income), and Corporate & other (sale-of-equipment + central costs). FY2025 vs FY2024, all ``:

*MEI EBITDA is almost entirely equity-affiliate income ($340.9M) — the segment is a JV holding (JIGPC/Jazan with Aramco+ACWA), not an operating P&L.

Reading the trend: the regional core is healthy and stable — Americas/Asia/Europe combined op-income ~$3.2B, mid-to-high-20s% margins, with Europe accelerating (+8% EBITDA on volume+price+FX) and Asia improving on productivity. The drag is entirely (a) Corporate & other, gutted by the September-2024 sale of the LNG business to Honeywell (−$358M sales, ~$135M lost operating income), and (b) helium weakness depressing merchant pricing in Americas/Asia. Geography: APD is majority ex-US in earnings — "The Chinese Renminbi and the Euro represent the largest exposures in terms of our foreign earnings". In fiscal Q2 FY2026 the segments inflected up: consolidated sales +9%, adj. operating margin to 23.7% from 21.6%, on +4% volume (HyCO on-sites in Americas) and FX tailwind. The segment story is "core gases were always fine; the reported P&L was wrecked by the energy-transition write-downs sitting in Corporate."

Phase B — Measure performance

Lens 5 · Earnings Result

Two prints matter: the FY2025 annual (the "kitchen-sink" year) and fiscal Q2 FY2026 (the recovery).

FY2025 (year ended 30 Sep 2025):

• Sales $12,037.3M, −1% YoY (volume −4%, price +1%, energy pass-through +2%, FX flat). Volume decline = LNG-sale loss, weak helium, project exits.
• GAAP operating loss ($877.0M) and GAAP net loss ($354.4M), GAAP EPS ($1.74) — wholly driven by $3.7B pre-tax business-and-asset-action charges ($3.0B after-tax, $13.68/share), of which ~$2.5B was plant-impairment on the cancelled clean-energy projects.
• Adjusted EPS $12.03, −3% YoY (from $12.43). Adjusted operating income $2,857.7M (23.7% margin, −70bp). Adjusted EBITDA $5,076.4M, +1%. Gross margin 31.4% (−110bp), depressed ~50bp by energy pass-through optics.
• Below the line: equity-affiliate income flat at $647.7M; interest expense $214.0M (low — $401.8M of interest capitalised into projects); income-tax benefit $94.3M on the pre-tax loss; adjusted effective tax rate 18.2%.

Fiscal Q2 FY2026 (quarter ended 31 Mar 2026) — the turn:

• Sales $3,171.8M, +9% (volume +4%, FX +4%, energy +2%, price −1% on helium). Volume led by HyCO on-sites in Americas.
• GAAP EPS $3.19 (vs ($7.77) PY); adjusted EPS $3.20, +19% YoY (vs $2.69). Adjusted operating margin 23.7% (+210bp). Equity-affiliate income $179.4M, +23% (Mexico).
• First-half FY26: sales $6.3B (+7%), adjusted EPS $6.37 (+13%), adj. operating margin 24.1%.

Guidance & tone: management raised FY2026 adjusted-EPS guidance to $13.00–$13.25 (from $12.85–$13.15 set in Nov-2025), Q3 guide $3.25–$3.35, capex held ~$4.0B. Tone is confident-but-cautious ("benefits in 2H from non-helium pricing, productivity, new assets ramping").

Balance-sheet flags: the worry is cash conversion, not solvency. FY25 operating cash flow $3.26B against $7.0B gross plant capex (~$3B to NEOM) — a large negative FCF year funded by $4.4B of new debt; total debt rose to $17.7B (Sep-25) and $17.8B (Mar-26). Working capital was a $851.6M use in FY25 (incl. ~$395M LNG-sale tax payment + contract-termination payouts). This is the central tension: a AA-quality franchise temporarily over-levered by an energy-transition capex bulge it is now actively deflating.

Market reaction pattern: the stock rallies on capital discipline, not on growth. FY25 Q4 results (the strategic reset) drove a ~10% surge despite an EPS dip; today's LCEC-exit announcement put it +6.8% pre-market. The market is rewarding the shrinking of the energy-transition gamble.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research-layer shelf (transcripts/ empty) — sentiment read is ``-derived from release commentary and coverage across the last ~4 prints (FY25-Q4 Nov-25, FY26-Q1 Dec-25, FY26-Q2 Apr-30-26, plus today's 8-K).

The tonal arc is one of the cleanest "regime change" signals in large-cap industrials:

• Pre-2025 (Ghasemi era): language was visionary / growth — "world's largest green hydrogen project," first-mover in the hydrogen economy, build-it-and-they-will-come offtake. Capex ran to ~$5B+/yr.
• From Feb-2025 (Menezes era): language pivots hard to discipline, returns, productivity, "core industrial gases". New recurring phrases: stringent return criteria, right-sizing, productivity, capital allocation, unlock value. Phrases that disappeared: hydrogen-economy evangelism, first-mover land-grab, growth-capex-as-virtue.
• FY26 prints + today: the walk matches the talk — guidance raised twice, capex cut, and a second round of project exits (LCEC) executed. Management is now setting quantified 2029/2030 targets (30% operating margin, mid-teens ROCE, high-single-digit EPS growth, capex to ~$2.5B in five years, $250M cost-out). Sentiment: credible, improving, with the burden of proof now on delivering the margin/ROCE ramp rather than on stopping the bleeding.

Lens 7 · Comps

The three Western majors + the key data.

Read: APD trades at a discount to both peers on forward P/E (~21–22× vs LIN 28×, AI 24×) and roughly in-line-to-slight-discount on clean EV/EBITDA (~14× vs AI ~16×, LIN ~19×). That discount is the turnaround/execution discount — the market is not yet paying APD the Linde multiple because (a) the energy-transition overhang isn't fully cleared (NEOM still ramping), (b) ROCE has lagged peers, and (c) the new strategy is ~16 months old. The bull case is multiple convergence toward Air Liquide as the discipline proves out; the bear case is the discount is deserved while NEOM offtake and Saudi execution remain open. GAAP screens (negative trailing P/E, 58× EV/EBITDA, −1.9% ROE) are all charge artifacts and will normalise mechanically as the FY25/FY26 impairments roll off — do not anchor on them.

Lens 8 · Stock-Price Catalysts (moves >5%, last ~5 years)

The pattern is unusually clean: capital-allocation/governance events move this stock far more than ordinary earnings. `` throughout.

• Oct 4 2024 — Mantle Ridge stake disclosed (~$1B+): shares surged on the activist news + Wall Street upgrades. The market immediately priced the prospect of capital discipline.
• Jan 23 2025 — Mantle Ridge wins 3 board seats; CEO Ghasemi loses his seat: decisive governance event; stock rose on the boardroom victory.
• Feb 7 2025 — Eduardo Menezes named CEO (ex-Praxair/Linde): regime change confirmed.
• Sep 30 2024 — LNG business sold to Honeywell for ~$1.8B proceeds (+$1.6B pre-tax gain): portfolio simplification.
• 2025 — the $3.7B project-exit charge / FY25 results: despite the GAAP loss, the FY25-Q4 "strategic reset" drove a ~10% surge — the market cheered the cancellation of projects.
• Jun 30 2026 (today) — Louisiana Clean Energy Complex cancelled, $2.9B charge, Yara/NEOM offtake firming: +6.8% pre-market.

What the tape reveals: for APD specifically, the market reacts to (1) capital discipline and governance above all, (2) energy-transition de-risking (exits = up), and (3) helium/volume in the core. It largely shrugs off the headline GAAP losses because they are recognised, non-cash write-offs of capital the market already wanted reallocated. This is the signature of a stock where the bet is on management behaviour, not the commodity cycle.

Phase C — Judge people & books

Lens 9 · Management

This is the crux of the whole story — APD is a management-driven re-rating.

1. Track record. CEO Eduardo Menezes (61), appointed Feb-2025: ~3 decades at Praxair (the industry's best-run operator) then EVP of Linde 2018–21 (ran EMEA, >$8B sales, 18,000 staff, global hydrogen). Vice Chairman Dennis Reilley — the architect of the Praxair best-in-class model (empowerment, cost discipline, rigorous capital allocation, risk-adjusted returns — the "gold standard"). This is, almost literally, the Praxair playbook being installed at Air Products. Praxair-pedigree management is the single highest-conviction fact in the dossier.
2. Tenure & skin in the game. New board/management installed by Mantle Ridge (Paul Hilal), which holds a $1B+ stake — a large, aligned, activist owner with board control. Management tenure is short (16 months) — the risk is execution-not-yet-proven, not misalignment. Insider-transactions.csv not on shelf; insider ownership otherwise modest (typical of a professionally-managed major). ``
3. Capital-allocation history — the entire thesis. The prior regime's allocation (huge speculative clean-H2 capex with offtake unsecured) is exactly what destroyed returns; the new regime is reversing it: capex $5.1B FY25 → ~$4.0B FY26 → ~$2.5B within 5 years, $3.6B+ of clean-energy projects exited (FY25 round + today's LCEC + Casa Grande AZ), with stated targets of 30% operating margin and mid-teens ROCE by 2030. Dividend: 44 consecutive years of increases, raised to $1.81/qtr in Jan-2026. FY25 payout was ~$1.6B.
4. Red flags. The proxy fight itself cost $86.3M (incl. a $24.7M reimbursement to Mantle Ridge and $29.7M ex-CEO separation) — a one-off governance cost, now done. The Mantle Ridge reimbursement (one director, Hilal, abstained) is a related-party item to flag but was board-approved and disclosed. No accounting red flags.
5. Founder vs professional manager. Squarely professional-manager / operator archetype (Praxair/Linde lifers) — exactly the right archetype for a mature, capital-intensive monopoly whose problem was over-ambition, not under-ambition.

Lens 10 · Forensic Red Flags

Acting as a forensic analyst — and the surprising conclusion is that the books are clean for a company that just took $3.7B+ in charges.

• The charges are recognised, not hidden. The FY25 $3.7B (and the incoming $2.9B LCEC) are non-cash impairments + contract-termination accruals on the energy-transition assets, fully disclosed in Note 5 / Business & Asset Actions and the impairment notes. Impairment of held-for-sale assets to $418.3M net, and a ~$2.1B liquidation-value write-down on equipment-market assets, are detailed with the valuation approach. This is the cleanup, not earnings management — the GAAP loss makes the company look worse, not better.
• Cash flow vs earnings. OCF $3.26B (FY25) is below adjusted net income $2.68B + D&A $1.56B because of the $851.6M working-capital use (LNG-sale tax + termination payouts) — explainable, one-off. Watch: with two charge years, monitor whether termination/severance cash outflows keep dragging OCF in FY26 (H1 FY26 OCF was $2.0B ).
• Capitalised interest. $401.8M of interest capitalised in FY25 (vs $214.0M expensed) — large, and it flatters reported interest expense. Legitimate for a heavy-construction phase, but as projects complete (NEOM 2027) capitalised interest falls and reported interest expense will step up — a known headwind to model, not a flag.
• NEOM / VIE structure. The $6.1B NEOM project financing is non-recourse and consolidated as a VIE; APD reports a non-GAAP capex that strips out the $2.47B (FY25) of NGHC spend funded by the JV's own financing/partners. This is the most aggressive presentation in the filing — the reported "$5.1B capex" excludes ~$2.5B of NEOM cash that did flow through additions-to-PP&E ($7.0B gross). Defensible (non-recourse, partner-funded) but the reader must know the gross number to judge the capex bulge. Flagged, not damning.
• SBC $76.4M (FY25) — small (~0.6% of sales); non-GAAP adjustments are dominated by the charges, not by chronic SBC add-backs. Goodwill $963.9M, intangibles small; no goodwill impairment (annual test passed).

Regulatory findings (required sub-section).

• SEC Litigation Releases / AAERs: None. Verified via SEC EDGAR EFTS (LR + AAER) search 2021-06-30→2026-06-30, 0 findings.
• 10-K Item 3 (Legal Proceedings): routine/incidental litigation only; 26 CERCLA/RCRA environmental sites (none expected material); the long-running Brazil CADE antitrust matter (R$179.2M ≈ $34M fine) was finally annulled — "in October 2025 the Supreme Court of Brazil rendered a judgment confirming the appellate ruling, which annulled the administrative proceeding and the fine". Company states no proceedings reasonably possible to be material.
• Non-SEC enforcement (web): no material FTC/DOJ/EU enforcement actions surfaced against Air Products as of 2026-06-30.
• Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-30. The only governance "event" was the (now-concluded, disclosed) proxy contest.

Phase D — Project & stress-test

Lens 11 · Forward Projection

Built bottom-up from FY2025 actuals + FY2026 guidance. APD's fiscal year ends 30 September, so the next three fiscal years are FY2026, FY2027, FY2028. Every input labelled; outputs ``.

Anchor: FY2025 adjusted EPS $12.03. FY2026 guidance $13.00–$13.25 (midpoint $13.13, ~+9%).

• FY2026E adj. EPS ≈ $13.13 (base = guidance midpoint). Drivers: +4% volume (HyCO on-sites ramping), non-helium pricing, $250M-run-rate cost-out beginning to land, FX tailwind; offset by continued helium-price weakness and rising reported interest as NEOM nears completion. ``
• FY2027E adj. EPS ≈ $14.4 base ($13.13 × ~1.10). The swing factor is NEOM start-up (mid-2027) — first green-ammonia revenue + the end of capitalised-interest relief (interest expense steps up as the asset goes in-service). Management's "high-single-digit EPS growth through 2029" frames the trajectory. Bull ~$14.9 (faster cost-out + NEOM accretive on the Yara/Total offtakes); bear ~$13.6 (NEOM start-up cost drag, helium still weak, soft industrial volumes). ``
• FY2028E adj. EPS ≈ $15.8 base ($14.4 × ~1.10). By FY28 the margin/ROCE program should be visibly compounding (toward the 30% op-margin / mid-teens ROCE 2030 goal), capex steps down toward ~$2.5B, and FCF inflects strongly positive — the deleveraging + buyback-optionality phase. Bull ~$16.8, bear ~$14.5. ``

The more important number than EPS is free cash flow: as capex falls $5.1B→$4.0B→~$2.5B while EBITDA grows toward ~$5.5–6.0B, APD swings from negative-FCF (FY25) to strongly FCF-positive by FY27–28, which funds the dividend (44-yr streak) with room to spare and opens buyback/deleveraging optionality. That FCF inflection, not the EPS line, is the re-rating fuel.

Brier forecast — NOT logged (per --watchlist rule: skip forecast.ts create in the breadth loop; only log on genuine committed conviction in a /thesis pass). Candidate forecast for a future thesis: "APD FY26 (ending Sep-2026) adjusted EPS ≥ $13.00, p≈0.80" — guidance midpoint $13.13, raised twice, with H1 already at $6.37 and a confident Q3 guide.

Lens 12 · Bull vs Bear

Adversarial institutional view.

Bull case. APD is a top-3 global monopoly franchise being handed to the best operators in the industry (Praxair/Linde alumni) with a $1B+ aligned activist on the board. The moat (90% take-or-pay, on-site lock-in, regional density, vertical equipment integration) was never broken — only the capital allocation was. The new regime is (a) cutting capex $5.1B→$2.5B, (b) exiting $3.6B+ of value-destructive clean-energy projects, (c) targeting 30% operating margins + mid-teens ROCE by 2030, and (d) converting the company from a negative-FCF growth-gambler into a FCF-compounding dividend aristocrat (44 yrs). NEOM — the one big surviving bet — is 95%+ built, on-track for mid-2027, and offtake is firming (TotalEnergies 15-yr deal + 600 t/day; Yara up to 1.2 Mt/yr ammonia). The stock trades at a forward-P/E discount to Linde and Air Liquide that should converge as discipline proves out. Earnings surprise vector: cost-out lands faster than modelled and NEOM signs full offtake → multiple re-rates toward 25×+ on a rising EPS base.

Bear case (2–3 permanent-impairment risks).

1. NEOM becomes a stranded $8.4B asset. If green-ammonia demand/pricing doesn't materialise at scale, APD owns a first-of-a-kind Saudi mega-plant whose output it must market (it's now the marketer, not just producer). Offtake is being assembled, not locked — Yara is "finalising," and APD has openly struggled to "line up buyers." A half-empty NEOM is a multi-year ROCE drag and a credibility hit to the very discipline thesis driving the stock.
2. The re-rating is already substantially priced. At ~21–22× forward EPS and +6.8% on today's news, with consensus PT ~$311–318 (~10–13% above ~$282), the market is already paying for a successful turnaround. If margin/ROCE convergence is slower than the 2030 targets, there's little valuation cushion — this is no longer a "cheap broken company," it's a "fairly-priced quality turnaround."
3. Core-volume cyclicality + helium overhang. ~Half of earnings are ex-US (CNY/EUR exposed); industrial volumes track global manufacturing. Helium has flipped from shortage to structural oversupply (supply ~6.5 Bcf vs demand ~6.0 Bcf, new entrants Renergen/Qatar/Iran, demand growing only ~2.5%/yr ) — a persistent merchant-pricing headwind APD has flagged repeatedly.

Pre-mortem (18 months out, thesis broke): It's late 2027. NEOM started up but is running below nameplate with only partial offtake; the Yara deal slipped or priced poorly; green-ammonia spreads are uneconomic without subsidy (OBBBA trimmed IRA hydrogen credits). Industrial volumes softened in a China/Europe slowdown, helium pricing kept falling, and the $250M cost-out was offset by start-up costs — so FY27 EPS came in flat-to-down vs the high-single-digit promise. The discipline narrative cracks ("they cut capex but couldn't grow earnings"), the multiple de-rates back toward 18×, and the stock round-trips its 2025–26 activist gains.

Are multiples too high? Not egregiously — ~21–22× forward is below peers for a comparable-or-better moat, so on relative value it's defensible. But it's no longer cheap; the easy "stop the bleeding" money has been made.

Contrarian view (what the market refuses to see): The market is treating NEOM as a risk to be minimised. The non-consensus upside is that NEOM — once derisked with Total+Yara offtake — becomes a 20+ year contracted annuity that anchors APD as the only major with a built, financed, operating green-H2 platform exactly as carbon-border taxes (EU CBAM) start forcing hard-to-abate ammonia/steel/refining customers to pay for low-carbon molecules. The bear sees a stranded asset; the contrarian sees the first take-or-pay clean-molecule plant in the world, fully built while everyone else cancelled.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case.

• What structurally breaks the model? Very little on the core — the on-site take-or-pay book is genuinely durable. The break risk is concentrated in the two things bulls wave away: NEOM and helium. The short thesis isn't "the gas business is bad"; it's "you're paying a quality multiple for a turnaround whose one remaining growth asset is an un-de-risked Saudi mega-project and whose merchant cash cow (helium) is in structural deflation."
• Revenue concentration / what shifts? Geographic: majority ex-US earnings, biggest FX exposures CNY + EUR — a China industrial slowdown or a EUR reversal directly hits earnings ($50M/$34M op-income per 10% move ). NEOM concentrates a single counterparty/country (Saudi/PIF/ACWA) and a still-forming offtake book in one $8.4B asset.
• Why the moat may be weaker than bulls think: at renewal and new-plant award the on-site moat is competitively bid against Linde/Air Liquide — APD doesn't price-set, it wins on cost/relationship. And in merchant, the moat is regional density, which doesn't protect against a commodity (helium) going into oversupply. Bulls conflate the contracted-annuity moat with pricing power over commodities APD doesn't control.
• Most dangerous competitor bulls underestimate: Linde — not as a share-stealer (the map is carved) but as the benchmark that sets the multiple and the operating bar. If Menezes's Praxair playbook doesn't close the ROCE gap to Linde, the entire "convergence" thesis evaporates and APD stays a perpetual discount. Also watch new helium entrants (Renergen, Qatar expansions, Iran) structurally capping APD's highest-margin merchant line.
• Worst capital-allocation moves / incentives: the prior board's clean-H2 land-grab already destroyed billions (that's why the activist won); the residual risk is the new team now over-correcting — under-investing in genuine core-growth ASUs to hit a capex-cut headline, or the Mantle Ridge related-party dynamic (board control by a single activist) prioritising a near-term re-rate over long-duration franchise health.
• Assumptions that must hold for today's price: (1) FY26 EPS ≥ $13 (H1 supports it — low risk); (2) NEOM starts mid-2027 and secures economic offtake (medium-high risk); (3) the 30%-op-margin / mid-teens-ROCE 2030 ramp actually compounds (medium risk, unproven); (4) helium doesn't deteriorate further (medium risk, currently negative).
• Valuation if growth disappoints 20–30%: if FY27/28 EPS growth halves (high-single → low-single digit), the multiple likely compresses toward ~18× on a ~$13.5–14 EPS base → ~$245–250, roughly today's level minus the activist premium — i.e. ~12–15% downside with the dividend as the floor.
• Single scenario that permanently impairs the business: a genuinely stranded NEOM (built, financed, but chronically under-offtaken / uneconomic green-ammonia spreads) combined with a multi-year global industrial recession — turning the FY27 FCF inflection into another debt-funded drag and breaking the discipline narrative. Plausibility: moderate — NEOM is non-recourse-financed (limits the balance-sheet hit) but a flop would gut the equity story.

Lens 14 · Management Questions (ordered by information value)

1. NEOM offtake: what % of the 600 t/day green-H2 / ~1.2 Mt/yr ammonia is contracted today (Total + Yara) vs spot, and at what clearing economics — does NEOM earn your stringent return hurdle at signed-offtake prices, or does it need subsidy/CBAM to clear?
2. Walk us from FY25's $5.1B capex to ~$2.5B in five years: how much of the cut is energy-transition exit (one-time) vs core-ASU growth you're forgoing — and what does that imply for the on-site contract-win pipeline (backlog) you're feeding for 2028+ growth?
3. What is the explicit ROCE bridge from today's GAAP-distorted low-teens to "mid-teens by 2030" — how much is denominator (asset write-downs/capex cuts) vs numerator (margin, cost-out, NEOM earnings)?
4. Quantify the FY27 step-up in reported interest expense as NEOM completes and capitalised interest ($401.8M in FY25) rolls off — what's the net EPS effect once NEOM earnings and that interest both hit?
5. Helium: at current oversupply, what is your merchant-helium volume/price assumption in the FY26 guide and the 2029 plan — and at what helium price does it become a material drag rather than a manageable headwind?
6. After LCEC and Casa Grande, is the project-exit program complete, or should we model further charges — and will the remaining held-for-sale assets ($418.3M net) clear at book in FY26?
7. With Mantle Ridge holding board control and a $1B+ stake, how do you ensure capital-allocation decisions optimise the 20-year franchise rather than the activist's exit timeline?
8. Capital returns priority once FCF inflects positive (FY27+): accelerate the 44-year dividend growth, initiate buybacks, or deleverage the $17.8B debt first — and what's your target net-debt/EBITDA?
9. On-site renewal economics: across the contract book renewing in the next 3–5 years, are you re-pricing at better, equal, or worse terms than the original 15–20-year deals, given the competitive bid against Linde/Air Liquide?
10. What is the right normalised tax rate (adjusted ETR 18.2% in FY25) given OBBBA, and how much of the clean-H2 production tax credit do you still expect to capture after the IRA modifications?
11. Equity affiliates throw off ~$648M/yr of after-tax income (JIGPC/Jazan, Mexico, etc.) — how durable is that, what's the cash-conversion (distributions vs equity income), and are any up for restructuring?
12. China specifically: how exposed is the FY26 plan to Chinese industrial demand, and what's your read on competitive intensity from local Chinese gas players on new on-site awards?
13. What organisational/cultural changes (the "Praxair model") are you installing, and what's the leading indicator we should track to know the operating transformation is working before it shows in ROCE?
14. Casa Grande, World Energy SAF, BHIG deconsolidation, Uzbekistan syngas financing — give us the post-mortem: what's the single repeatable lesson now embedded in the "stringent return criteria" gate?
15. What would have to be true for you to re-accelerate growth capex above $4B — i.e., what's the bar a new project must clear, and is there anything in the pipeline today that clears it?