Alcon Inc. (NYSE: ALC) — A Quality Label the Numbers No Longer Support

Fresh initiation · Report date: 18 July 2026 · Price: $70.26 (17 July 2026 close) Sector: Health Care · Medical Devices & Vision Care · Swiss foreign private issuer (20-F/6-K), reports under IFRS

Sections 1–15 of this report carry no investment recommendation and no price target. The single, deliberate exception is the Claude's Take block immediately below, which is fenced off as the author’s own subjective view.

⚡ Claude’s Take

The author’s own independent, subjective opinion, offered as general information only — not investment advice, and not a recommendation to buy or sell any security. Sections 1–15 below carry no position, no recommendation and no price target.

HOLD — and AVOID at this price. Not a short. Accumulate only in the mid-to-high $50s (roughly 19–20× honest owner earnings).

Tag: “A falling knife that stopped falling — but cheap against itself is not the same as cheap.”

Alcon is a genuinely good business that was floated, and priced, as a great one. The correction to that label is most of what has happened to the stock, and I do not think it is finished. The bull case leans on the observation that book returns — ROE 4.45%, ROIC 4.48% — are artifacts of $18.3B of Novartis spin-goodwill (82.9% of equity) that Alcon never paid for, and that is correct: on the ~$7.4B of tangible capital actually employed, the operating asset earns ~20.5%. But that rescue only gets you to mid-band, not to wide-moat, and it is deteriorating: returns peaked in FY2024, core operating income grew +0.6% on +5% revenue in FY2025, and the incremental core margin was 2.5% against a 19.8% average. A business that grows revenue and keeps almost none of it is not compounding. Meanwhile Implantables — the highest-margin, most moat-dependent line, the one the whole installed-base story exists to pull through — was flat at +0.4% while J&J’s Surgical Vision grew +10.2% in the same market in the same year. That is share loss, quantified, and management’s own compensation disclosure concedes it missed share targets in three of four measured franchises.

The valuation is where I part company with the “cheap defensive” framing. On management’s core EPS of $3.07 the stock looks like 22.9×; on IFRS $1.98 it looks like 35.6×; the defensible number is honest owner earnings of ~$2.75, or 25.5× — because “core” strips not just spin amortization but recurring acquisition costs (including $46M on a deal that failed), recurring legal, and an $88M-a-quarter restructuring programme whose total size, duration and savings target the company has simply not disclosed. Twenty-five times for ~0% profit growth is not a discount. Alcon trades at a premium to CooperVision on adjusted earnings and a premium to ResMed on EV/EBIT, against ResMed’s 32.8% margins and 22% ROIC. The reverse-DCF says the market is underwriting only ~3.9% perpetual owner-earnings growth — below the 3–4% market growth management itself assumes — so the market is already refusing the FY2026 guide, and it is right to: that guide is internally inconsistent (the +10–13% core EPS line is only reachable at the top third of the margin range), back-end-loaded into 2H, and comes from a team that cut twice in 2025. What holds me at HOLD rather than something harsher is that the downside is genuinely defended: net debt is 1.19× core EBITDA, liquidity is $2.9B, the surgical consumables annuity is real and 83.6% recurring, and the buyback has been executed counter-cyclically.

The framing matters, so let me name it: this is a value trap, not yet a value name — and I mean that technically. The factor model zeroes Alcon’s Value and DividendYield betas in all four nested models, in a year when those were the two best-paying factors; it also assigns a Quality beta of ~0.03, declining to certify the quality label from a completely independent direction. Sharpe is negative at every horizon out to five years; the stock has returned +3.4% over five years against the S&P’s +84.4%, and -164pp since the 2019 spin. The dedicated value buyer has not arrived, the register is diffuse index money with no activist, and nobody is paid to close the gap. That is not a setup where I want to pay 25×.

Conviction: medium. Flips bullish if the efficiency programme proves genuinely one-time and Implantables reaccelerates to at least market growth for two consecutive quarters — that combination would mean the moat is intact and owner earnings step to ~$3.20+, at which point the mid-$60s is a real entry. Flips bearish if “costs associated with efficiency initiatives” reappears in the Q3-26 and Q4-26 core reconciliations above ~$50M/quarter, which would confirm it as mislabelled recurring opex and pull honest owner earnings toward $2.55 — on which today’s price is nearly 28×.

📈 Stock Price Action — Five-Year Event Map

Alcon has round-tripped a full cycle and ended where it started: from roughly $67 in July 2021 down to a $55.68 trough in November 2022, up 79% to an all-time closing high of $99.92 on 12 September 2024, and then down 30% to $70.26 (17 July 2026). The 52-week range is $62.02–$91.78, the stock sits -29.7% below its all-time high and below its 200-day EMA of $74.50, and its -37.9% peak-to-trough drawdown — the deepest of Alcon’s independent public life, deeper than COVID — bottomed only on 11 May 2026. Five-year price return: +3.4%, against +84.4% for the S&P 500.

1. The 2021–22 decline was a multiple event, not an earnings event — Alcon’s core EPS rose through it. 2. The 79% advance to September 2024 is the whole bull case in one line: procedure volumes normalised, core EPS compounded ~17%/yr, and the market paid up for a franchise it believed was a wide-moat compounder. Note that the flagship UNITY VCS FDA approval (24 June 2024) moved the stock +0.8% — the tape did not treat the platform refresh as a re-rating event. 3. The peak has an uncomfortable footnote: the CEO’s Form 144 notice of 12 September 2024 covered 112,182 shares at an implied ~$100.5, filed on the exact day of the all-time high close. 4–5. The two 2025 guidance cuts are the substance of the de-rating: between February and August management took constant-currency growth from +6–8% to +4–5% and core operating margin down ~150bp, and held the core EPS dollar range only by cutting the assumed tax rate from ~20% to ~18%. FY25 landed at $3.07 on a 19.8% core margin — below even the twice-cut range’s midpoint. 6. The winter recovery was bought, not earned: Alcon compressed a three-year buyback into ten months and its heaviest months (September–October, ~1.75M and ~1.55M shares) were the two cheapest of the year. Neither the STAAR termination (6 Jan, -0.04%) nor the TRYPTYR approval (28 May 2025, -0.4%) registered as price events. 7. Q1-26 was the capitulation: core margin was fine but IFRS operating margin fell to 10.8%, Implantables — the highest-margin, most moat-dependent line — grew 1% while Equipment grew 23%, and a new restructuring programme appeared with no disclosed size or savings target. 8. The bounce since May has recovered the short-term trend but not the intermediate one; on a five-year view the stock is unchanged and 81 percentage points behind the index.

1. Executive Summary

Alcon is the largest eye-care company in the world — $10,319M of FY2025 net sales across Surgical (55.7%) and Vision Care (44.3%), the industry’s largest installed base of phacoemulsification and vitrectomy consoles, and the leading position in premium intraocular lenses. It was spun out of Novartis in April 2019 and has since carried both the operational independence and the accounting inheritance of that separation: $18,262M of goodwill and intangibles, equal to 82.9% of shareholders’ equity, which generate roughly $700M a year of amortization the company never paid cash for.

The central analytical problem is that Alcon presents three different earnings numbers, and choosing among them is the entire investment debate. IFRS diluted EPS was $1.98 in FY2025. Management’s “core” measure was $3.07. Neither is right. Core does not merely add back spin-era amortization; it also strips recurring acquisition costs — including $46M spent on a deal that failed — recurring legal provisions, product discontinuations, and, from Q1-2026, an $88M-per-quarter “efficiency initiatives” programme whose total size, duration and savings target remain undisclosed. Rebuilt on a consistent basis, and cross-checked by an independent cash-flow route that agrees to within two cents, honest owner earnings are approximately $2.75 per share.

The quality label does not survive contact with the operating record. Book returns (ROE 4.45%, ROIC 4.48%) are genuinely artifacts of the spin; on the ~$7.4B of tangible capital actually employed, Alcon earns ~20.5%. But that is mid-band, not wide-moat, and it peaked in FY2024. Core operating income was $1,849M / $2,027M / $2,039M across FY2023–25 — +0.6% in FY2025 on +5% revenue, at a 2.5% incremental core margin against a 19.8% average. Surgical segment contribution has been flat in dollars at ~$1.46B for three consecutive years while its margin fell from 27.4% to 25.4%. Most tellingly, Implantables — the highest-margin line, the one the razor-and-blade installed base exists to pull through — grew +0.4% in FY2025 while J&J’s Surgical Vision grew +10.2% in the same market and the same year. Alcon’s own compensation disclosure concedes it missed share-gain targets in three of four measured franchises over the 2023–25 cycle.

One moat is real; the others are rented. The Surgical franchise has a genuine economies-of-scale-plus-switching-cost advantage: the largest installed base, a 7–10 year console replacement cycle, and an 83.6%-recurring consumables and implantables annuity. But captivity is workflow habit rather than contract, it is re-contested at every replacement decision, and Carl Zeiss Meditec — via its D.O.R.C. acquisition and the EVA NEXUS platform — is now building a rival ecosystem. Contact lenses are a genuine four-firm oligopoly, but Alcon is a tenant of that structure rather than its landlord. Ocular Health lost its prescription gate when Pataday went OTC in 2020 and now faces private label with no reimbursement protection.

Capital allocation is the weakest link, and the root cause is structural. Alcon went 0-for-2 on public M&A in a single year: STAAR was terminated in January 2026 after shareholders refused a raised bid, and LENSAR was terminated in March 2026 after the FTC signalled intent to enjoin — roughly $76M expensed for zero acquired assets. Neither the short-term nor the long-term incentive plan contains any return-on-capital metric, so the ~$1.04B contact-lens capacity build, the Aurion acquisition at roughly 53–71× sales for a pre-approval therapy, and R&D growing +19.6% against +0.6% core operating income are all scorecard-neutral or scorecard-positive while being return-dilutive. Insider ownership is under 0.5% and the register is diffuse index money with no activist.

Valuation. At $70.26 the enterprise is capitalised at ~$37.4B, or 3.62× sales, 14.2× core EBITDA, 20.3× honest EBIT and 25.5× honest owner earnings. A reverse DCF anchored on the company’s own impairment discount rates (8.0–8.5%) implies the market is underwriting only ~3.9% perpetual owner-earnings growth — below the 3–4% market growth management itself assumes. The market is refusing the FY2026 guide rather than underwriting it, and the guide is internally inconsistent: its +10–13% core EPS line is reachable only at the top third of the guided margin range, while the range’s floor produces +6.5%. Alcon is not obviously cheap against peers either — it trades at a premium to CooperVision on adjusted earnings and to ResMed on EV/EBIT, despite materially worse returns than the latter. Its own-history percentiles look inviting (P/S at the 4.7th percentile), but that is best read as a rational re-rating of a structurally lower-return business, not a mispricing.

What would change the view. The two variables that drive a 64% dispersion in FY2028 owner earnings are both currently undisclosed or undemonstrated: the true size and permanence of the efficiency programme, and whether Implantables can reaccelerate to market growth. Everything else in the story is second-order. The balance sheet is not a risk — net debt is 1.19× core EBITDA with $2.9B of liquidity — so this is a question about earning power and competitive position, not solvency.

The analysis below takes no position on the security and contains no price target outside the clearly-labelled Claude's Take block above.

2. Business Overview

Alcon Inc. is the largest pure-play eye-care company in the world. It sells the equipment, disposables, implants and consumer products used across the two dominant interventions in ophthalmology — cataract/vitreoretinal surgery and vision correction — through two reported segments. FY2025 net sales were $10,319M, +5% reported (+4% constant currency): Surgical $5,751M (55.7% of sales, +4%) and Vision Care $4,568M (44.3%, +6%) [ALC 20-F FY2025, Item 5 net-sales table, https://www.sec.gov/Archives/edgar/data/1167379/000116737926000014/alc-20251231.htm, accessed 2026-07-18]. The company employs 25,628 FTEs, is domiciled in Geneva, reports under IFRS as a US-listed foreign private issuer (20-F/6-K, not 10-K/10-Q), and is led by CEO David J. Endicott.

The two segments and their sub-lines (FACT, FY2025, $M):

How it makes money — the razor/razorblade structure, quantified. Surgical is an installed-base annuity. Alcon claims “the largest installed base of cataract phacoemulsification consoles, vitrectomy consoles and refractive lasers in the industry,” with “long buying cycles that last approximately seven to ten years [that] act as anchoring technologies that drive recurring sales of our consumables” [20-F FY2025, “Our Installed Base and Portfolio”]. The arithmetic supports the claim: implantables plus consumables were $4,810M of $5,751M, or 83.6% of Surgical revenue — one-time capital equipment is only 16.4%. Each console placed pulls through custom procedure packs (over 10,000 Custom Pak configurations offered globally), fluidics cassettes, viscoelastics and an IOL per eye, with planning and ordering routed through Alcon’s ADI cloud platform and SMARTCataract workflow. That is a real annuity and it is the single strongest structural feature of the business (Section 4). Vision Care is a different model entirely — a consumables business with no razor: contact lenses and eye drops are repeat purchases sold through eye-care practitioners, optical retail and e-commerce, where the prescriber chooses the brand and the patient re-orders it.

Customer types. Four distinct buyers, with different economics. (1) Surgeons — ophthalmologists choose the console, the IOL and the technique; they are the true decision-maker in Surgical and, per management, the binding capacity constraint on procedure volume (Section 3). (2) ASCs and hospitals — own the capital budget and re-tender the console every seven to ten years. There were 6,052 US ambulatory surgery centers in 2024, a sector growing 4.7–5.9%/yr [MedPAC March-2025 Report to Congress, Ch.10, accessed 2026-07-18]. (3) Optometrists and optical retail — the gatekeeper for contact lenses, where the fitted brand becomes the habitual re-order. (4) Consumers, who pay directly for ATIOL upgrades, LASIK, lenses and OTC drops. No customer accounted for 10% or more of net sales [20-F FY2025] — concentration risk is genuinely absent.

The reimbursed-versus-private-pay split — structurally the most important line in this section. The base cataract procedure and the monofocal IOL are reimbursed (US Medicare facility and physician fees). The premium layer is not: the 20-F describes “a growing private pay market for premium surgical devices… allowing patients to pay the non-reimbursable cost of a procedure associated with selecting premium devices, such as ATIOLs,” and states that Vision Care “is primarily private pay, with patients substantially paying for contact lenses and ocular health products out-of-pocket.” US ATIOL penetration is approximately 20% of cataract procedures per the filing itself, corroborated by Market Scope’s US Cataract Atlas (“Adoption Tops 20 Percent”); global penetration of ~17%, up ~130bp year-over-year, is a management estimate [Endicott, Q1-2026 call, 2026-05-06]. INTERPRETATION: Alcon’s highest-margin revenue is its least reimbursement-protected revenue. That cuts both ways — it insulates the premium pool from payor price-setting (Section 3, CMS Ruling 05-01), but it exposes it directly to consumer discretionary spending. Alcon’s own inflation risk factor concedes the point: price increases “may cause some customers, particularly in elective surgical and contact lens businesses where patients typically do not receive reimbursement, to reduce purchases or choose lower-cost alternatives.” What Alcon has is price/mix opportunity contingent on the consumer, not pricing power.

Geography. United States $4,657M / 45%; International $5,662M / 55%, of which Japan $609M / 6%, China only $570M / 6%, and Switzerland — the domicile — just $71M / 1% [20-F FY2025]. China deserves emphasis precisely because it is small: at $526M (2023) → $560M (2024) → $570M (2025) it has grown ~4% in two years against the group’s +10%, so it is a decelerating drag rather than the emerging-market engine most ophthalmic bull cases assume (Section 3).

The 2019 Novartis spin still dominates the financial statements. Nestlé owned Alcon from 1977, IPO’d ~25% in 2002; Novartis bought 100% across 2008–2011 and spun it off in full on 2019-04-09. Novartis retains no residual stake. What it left behind is purchase accounting: goodwill $9,256M plus other intangibles $9,006M = $18,262M, equal to 57.9% of total assets and 82.9% of shareholders’ equity ($22,034M). That stack generated $784M of intangible amortization in FY2025 against IFRS operating income of only $1,360M, and it is why headline returns — IFRS ROE 4.45%, book ROIC 4.48% — read like those of a capital-destroying business (Sections 4 and 6). FACT: Alcon never paid for those assets; Novartis did. The consequence for every reader of this report is that IFRS earnings and book returns are the wrong lens on Alcon, and management’s “core” measures are the wrong lens in the other direction — the honest number sits between them.

3. Industry Dynamics

Ophthalmic care is not one industry but three profit pools with materially different structures, and this section refuses to grade them together. The most important structural fact is stated up front: Alcon has marked down its own industry three separate times in eighteen months.

3.1 The market Alcon addresses, and the downgrade inside it

The 20-F sizes both segments directly: Surgical ~$14B, growing ~4–6%/yr 2025–2030; Vision Care ~$23B, growing ~4–5%/yr 2025–2030 — a combined addressable market of ~$37B, of which Alcon’s $10,319M is ~28% (Surgical ~41% share; Vision Care ~20%). The downgrade sequence (all FACT):

1. The FY2024 20-F guided the surgical market to grow “approximately 6% on average per year from 2024 to 2029.” The FY2025 20-F cut that to “approximately 4–6%… from 2025 to 2030.”
2. August 2025: management cut its assumed market growth to “low single digits versus a historical average of mid-single digits.”
3. FY2026 guidance assumes “aggregate eye care markets grow 3% to 4% for the year” — below the bottom of Alcon’s own published long-run TAM growth for either segment [CFO Stonesifer, Q1-2026 call, 2026-05-06].

INTERPRETATION: every bull case resting on “structural growth from aging demographics” must first survive the fact that the incumbent leader’s own near-term market assumption is 3–4%.

3.2 Pool (a) — cataract / IOL, and the ATIOL private-pay slice

Global cataract volume is ~28–30 million procedures a year, growing 3–4% on aging; the IOL sub-pool is ~$4.62B (2025) heading to ~$6.17B by 2030, a ~6.0% CAGR [MarketsandMarkets, accessed 2026-07-18 — SECONDARY, indicative]. Alcon’s $1,782M of implantables is ~39% of it.

The pool’s economics are a barbell: a large reimbursed commodity base (monofocal IOL, phaco procedure) plus a small, high-margin, consumer-discretionary premium slice. That slice exists because of a single administrative act. CMS Ruling 05-01, issued and effective 3 May 2005, held that no Medicare benefit category exists for the presbyopia-correcting functionality of an IOL, and therefore permits beneficiaries to pay the incremental charge out of pocket — they “may upgrade from a conventional IOL to a presbyopia-correcting IOL, as long as they’re willing to pay all charges beyond those associated with standard cataract surgery” [CMS Ruling 05-01, https://www.cms.gov/Regulations-and-Guidance/Guidance/Rulings/downloads/cmsr0501.pdf; AAO, “Premium IOLs—A Legal and Ethical Guide to Billing Medicare Beneficiaries,” accessed 2026-07-18]. CMS extended the same logic to astigmatism-correcting IOLs (Ruling 1536-R) and to laser-assisted cataract surgery.

Two consequences are worth stating plainly. First, the pool is hostage to consumers, not payors — which is why the discretionary caveat in Section 2 matters. Second, and rarely discussed: 05-01 is an administrative ruling, not a statute. It has stood twenty-one years with no live proposal to change it, but a future CMS could revisit it, and a reversal would vaporize the ATIOL profit pool outright. ASSUMPTION-level tail risk: low probability, catastrophic impact on the premium franchise. It is under-priced because it is never discussed.

Reimbursement is moving the other way at the margin. The CMS 2026 Medicare Physician Fee Schedule, effective 2026-01-01, cut physician reimbursement ~10.5% for cataract surgery and combined cataract+MIGS procedures, and 7.3–9.2% for other MIGS [AAO, “Medicare Reimbursement 2026: the Good and the Problematic,” accessed 2026-07-18]. That sits on top of the November-2024 MIGS Local Coverage Determinations, finalized by five of seven MACs, which declare MIGS “not considered a first-line treatment for mild-moderate glaucoma” and restrict coverage to cataract surgery plus one MIGS device per eye, with combination procedures “non-covered [and risking] denial of the entire claim” [CMS Medicare Coverage Database LCD L38301; Billing Articles A56866/A57864/A59431]. Precedent for the magnitude: the January-2022 MIGS coding change cut angle-based stenting volumes by ~20,960 procedures in a single year [Glaucoma Today/CRSToday]. This landed directly on Alcon’s Q1-2026 implantables — Endicott, asked why the line grew only 1%: “one of the reasons we kind of called out glaucoma implantables is because, as you know, the reimbursement changed this year.” INTERPRETATION: the LCDs close off MIGS as a second-implantable growth vector for IOL vendors industry-wide (it hits Glaukos’s iStent identically), but the ~10.5% cut to the cataract physician fee is the more consequential number — it reduces the surgeon’s economics on the base procedure at precisely the moment surgeon time is the binding constraint (Section 3.5).

China is a destroyed pool, not a competitive threat. Successive IOL VBP rounds cut prices ~20%, then 26%, 38%, 53% (2020) and 84% (2021), with a recent round averaging ~58–60% [mddionline.com; pacificbridgemedical.com; market-scope.com, accessed 2026-07-18]; a further round lands “in the middle part of the year” 2026 per Endicott. The naive read is that Chinese domestic makers are taking Alcon’s share. The evidence says something worse and more interesting: Eyebright Medical (SSE: 688050), the domestic champion, reported FY2025 revenue of RMB1.48B (~US$205M, +5.15%) with net profit depressed by a goodwill impairment — and IOL revenue up just +1.43% [Eyebright FY2025 annual report release, accessed 2026-07-18]. In its own protected home market, with explicit “Made in China 2025” procurement preference, the champion is growing IOLs at 1.4% and writing off goodwill. VBP destroyed the pool for everyone; Chinese makers are co-victims, not a global threat. Marathon is explicit that the capital cycle does not clear normally where the state administers prices — this is the textbook case. The correct framing: China is a ~6%-of-revenue drag with an unquantified VBP round landing now, not an existential competitive risk. It is also the entire explanation for the ATIOL penetration drag management flagged (+130bp global including China, +220bp excluding it).

3.3 Pool (b) — contact lenses

FY2025 branded manufacturer revenue, from the four issuers’ own filings: J&J $3,910M (+4.8%), Alcon $2,770M (+6%), CooperVision $2,743.8M (+5% cc), B+L ~$1.0–1.2B implied (B+L does not disclose a standalone lens line). On a ~$9.8B manufacturer-level pool, the big four are effectively the entire branded market — CR4 ~95%+. This is one of the better consumer-medtech structures in existence: no price war, high-fixed-cost silicone-hydrogel manufacturing, thousands of power/base-curve/cylinder/axis SKUs, Class II/III regulation, and an eye-care practitioner acting as agency gatekeeper. Mix-up economics (reusable→daily, hydrogel→SiHy, sphere→toric/multifocal at a 20–40% price premium) lift revenue per wearer without requiring headline price increases.

But the cycle is turning. Growth has stepped down from the ~6–8% the category delivered in recent years to Alcon’s own Q1-2026 read of “the low end of mid-single digits.” Against that markdown, Alcon has committed ~$1.04B across ~16 new production lines — Grosswallstadt +3 ($162M, to 2027); Singapore +4 ($189M, complete), +3 plus a building ($314M, to 2027), +3 more ($157M, to 2030); Johns Creek +2 ($148M, to 2028) and +1 ($73M, to 2029) [20-F FY2025, Item 4.D]. CooperVision has spent 9–11% of sales for years on the same build; Bausch + Lomb grew revenue from $4,146M (2023) to $5,101M (2025) while earning a 3.7% operating margin and a -$360M FY2025 net loss. Three of four incumbents adding high-fixed-cost capacity into a category whose growth estimate is being cut is Marathon’s canonical late-cycle signature. THE HONEST MITIGANT: the capacity is overwhelmingly daily-SiHy and premium reusables — a mix-up, not raw unit expansion — and Alcon’s Vision Care gross margin has actually risen 175bp in two years, so the build is delivering manufacturing efficiency so far. This is a supply-side warning, not a bust. The direction of capital, however, is unambiguous: in, at the point the growth estimate is coming down.

3.4 Pool © — ocular health / OTC

Alcon’s $1,798M (+5%) sits inside the ~$23B Vision Care TAM. The demand base is unimpeachable — the 20-F cites ~1.6 billion people with dry-eye symptoms without clinical signs, 750M with both, 650M at risk. Sub-pool sizing is not: two secondary sources put the OTC dry-eye/artificial-tears category at $1.85B and $4.32B respectively for the same year, a 2.3x spread, which is itself the finding. Build no argument on it. Structurally this is the weakest pool: no reimbursement gate, no professional gatekeeper for OTC drops, low capital intensity, continuous entry, and direct private-label substitution (store-brand olopatadine and lubricant drops). Systane is category leader at ~29% of the US OTC eye-drop market, yet Alcon conceded missing its share target here on “competitive launches and overall pricing pressure,” while the #4 industry player grew the category faster on LUMIFY and OTC dry-eye. Brand buys a price premium, not a barrier. The Rx sub-segment (Tryptyr, FDA-approved 2025-05-28) is structurally better because a prescription is a gate — but it is small and the gate is a patent clock.

3.5 The demand read — “capacity, not demand,” pressure-tested

Management’s claim: cataract volumes have grown low single digits for several quarters, but “market growth will return to historical levels as health care systems adapt to increasing demand.” The mechanism, when pressed: “it really is freeing up time to do more surgery… you’re competing with, for example, a hospital OPD who wants to give that time to a more productive, more economically valuable procedure… so certain parts of the market are shrinking for available time” [Q1-2026 call].

The claim survives. The peer-reviewed workforce projection finds US ophthalmologist supply falling 2,650 FTE (-12%) while demand rises 5,150 FTE (+24%) from 2020 to 2035 — a ~30% workforce inadequacy [Berkowitz et al., Ophthalmology (AAO Journal), https://www.aaojournal.org/article/S0161-6420(23)00677-2/fulltext]. Meanwhile facility capacity is expanding: 6,052 ASCs in 2024, growing 4.7–5.9%/yr, with cataract extraction among the top projected ASC growth procedures for 2026. FACT: the bottleneck is surgeon hours, which no amount of capex relieves. Management’s corollary — that systems will adapt — is an ASSUMPTION with a decade-long time constant and no evidence yet of having started.

But the reframing is bad for Alcon, and this is the section’s most important interpretation. If the scarce asset in the value chain is the surgeon’s time, then economic rent accrues to whoever controls it — PE-backed practice roll-ups, ASC owners, and surgeons — not to the device supplier. That is the cleanest available explanation for the otherwise puzzling FY2025 result: a #1-share incumbent in a 72%-CR4 surgical market, explicitly taking price (consumables growth “reflects softer than historical market conditions as well as price increases”), and still seeing both segment contribution margins decline (Surgical 26.6%→25.4%, Vision Care 22.3%→21.5%). In a supply-capped procedure market, the supplier’s price gains are absorbed downstream. Management’s own optimistic framing — office-based surgery “is very popular right now and gaining momentum” — is a description of customers building capacity they will control, with Alcon selling equipment into it. That is a decent equipment cycle and a poor rent position.

Two further industry-structural costs. EU MDR: certification runs from ~EUR8,000 for a simple Class I device to over EUR600,000 for a Class III device with clinical investigations, with the European Commission acknowledging implementation costs up to 5% of revenue for some organizations [EC SWD(2025) 1050 final, 2025-12-16]. This cuts both ways — a genuine scale-favoring barrier against sub-scale entrants, but a large fixed compliance cost that forces SKU pruning even on incumbents (Alcon’s FY2025 $44M Vision Care product discontinuation is consistent with exactly this). The EC’s December-2025 Simplification Package would lower the barrier modestly if adopted. Tariffs: ~$100M of FY2025 cost of sales, guided to ~$100–150M for FY2026, costing 120bp of core gross margin in Q1-2026, with management explicitly choosing to reinvest the ~$25M of relief rather than let it reach earnings. A Swiss manufacturer with 45% US sales and plants in Germany, Singapore and Switzerland is structurally exposed; this is a permanent industry cost that did not exist three years ago.

3.6 Verdict — deliberately split four ways

A single grade would be dishonest here.

(1) Surgical consumables and equipment — GOOD. Concentrated (CR4 ~72%, and understated because Hoya’s IOL revenue cannot be sized), rising regulatory barriers, an installed base anchoring an 83.6%-recurring annuity, long replacement cycles, unimpeachable aging demand. This favors scale players decisively and Alcon is the scale player. The disconfirming evidence, weighed: the barrier is re-tendered every seven to ten years, and Alcon’s 25.4% contribution margin is falling, not rising.

(2) IOL / ATIOL — MIXED and deteriorating. The pool is real, patient-pay, payor-protected by CMS 05-01, and has a genuine decade-long runway (US ~20–21% against management’s stated ~35–38% ceiling). Against that: it is the most contestable sub-pool in the industry — a better optic wins share inside a single product cycle, which J&J demonstrated in FY2025; volume is capped by a surgeon shortage worsening through 2035; China administers prices to zero; and the adjacent MIGS attach has been capped by LCD and cut by fee schedule. This pool rewards the best innovator of the moment, not the largest incumbent.

(3) Contact lenses — GOOD structure, LATE in the capital cycle. Everything Greenwald asks of an industry-level barrier is present, and no meaningful entrant has established itself in decades. But growth has stepped down, ~$1.04B of Alcon capacity is going in alongside Cooper’s and B+L’s, and private label is taking international share by the incumbent’s own admission — a headwind that is qualitatively confirmed and, honestly, cannot be sized from public sources. OPEN QUESTION, and it should stay open.

(4) Ocular health / OTC — MEDIOCRE. Huge demand, no gate, private-label attack, brand equity that buys price premium rather than protection.

NET: an above-average industry — better than most of medtech, worse than its reputation, and not early-cycle. Demand is excellent, concentration is high, regulatory and manufacturing barriers are real and rising, and the patient-pay overlay is something most of medtech would envy. But the incumbent is marking its own market down, volume is supply-capped, capacity is being added into a decelerating category, the highest-value sub-pool just proved how contestable it is, and rent is drifting toward the provider. The decisive point for this memo: the two best-structured pools are precisely where Alcon is merely co-equal (lenses) or defending rather than expanding (surgical consumables), while the pool with the most attractive economics (ATIOL) is where it is visibly losing. A good industry does not rescue a firm positioned on the wrong side of its own best pools.

4. Competitive Position

4.1 The moat, named by segment

Surgical — economies of scale plus customer captivity. This is the genuine one. The mechanism is specific and checkable: the industry’s largest phaco, vitrectomy and refractive-laser installed base anchors a consumables and implantables stream that is 83.6% of segment revenue, earning a 25.4% segment contribution margin, with the anchor asset replaced only every seven to ten years. The captivity is real — surgeon muscle memory, fluidics “feel,” ORA/Verion-integrated planning, >10,000 Custom Pak configurations, and workflow locked into the ADI cloud and SMARTCataract.

Two honest limits. The captivity is workflow habit, not contract — competitor cassettes and packs exist; nothing is physically keyed the way surgical-robotics instruments are. And the barrier is re-contested at every console replacement: a moat that goes back out to tender every seven to ten years is narrower than a true switching-cost franchise. INTERPRETATION: Unity CS/VCS is best read as a defensive necessity dressed as a moat refresh — Centurion dates to 2013 and Constellation to 2008, so replacing them was mandatory to hold the base at the renewal decision, not an expansion of the franchise.

Vision Care / contact lenses — shared economies of scale that Alcon rents rather than owns. The industry barrier is genuine (Section 3.3). The firm-level advantage is not. Greenwald is explicit that scale functions as a barrier only where share is dominant and defended move-for-move; every point lost narrows the cost gap. Alcon’s own lens revenue of $2,770M is ~1% above CooperVision’s $2,743.8M — a photo finish for #2, far behind J&J at ~$3.9B. Alcon’s “second-largest in branded contact lenses” claim is defensible on revenue; CooperVision ranks #2 on measures including its large private-label volume (Costco Kirkland, Walmart Equate and similar store brands are manufactured by the majors). Both are true, and the qualifier does real work. A ~1% gap is not a position — and that is the analytical point. Alcon is the #2/#3 tenant of a four-firm oligopoly, not its landlord.

Ocular Health — demand/habit captivity via brand, the weakest form, and eroding. Artificial tears are exactly the frequent automatic repurchase habit captivity is built for. But Pataday’s 2020 Rx-to-OTC switch — the event that created the consumer franchise — simultaneously removed the prescription gate, leaving a branded drop facing store-brand olopatadine with no reimbursement moat and no professional gatekeeper.

4.2 Greenwald test (a) — market share stability. The most damning evidence in the file.

Greenwald’s observable signature of a barrier is stable-to-rising share. Alcon measures its own share — its LTI plan pays management on “Share of Peers” using third-party syndicated data (Vision Care 45% weight, two categories via GfK/IRI/Amazon/Nielsen/IQVIA; Surgical 55%, four categories via MarketScope) — and discloses the result. The verbatim FY2025 disclosure for the full 2023–2025 cycle [20-F FY2025, Compensation Report]:

• Contact lenses: “we fell short of our share-expansion target in Contact Lens primarily due to competitive pressure in the Dailies category, as well as growth of private label, particularly in the international market.”
• Ocular health: “due to competitive launches and overall pricing pressure we didn’t hit our share gain targets.”
• ATIOLs: “In ATIOLs, where Alcon already has leading share positions in most markets, Alcon trailed market growth, but stabilized US trifocal market share with the recent launch of PanOptix Pro.”
• Surgical consumables: share gained in phaco and vitreoretinal cassette packs, “driven by our recent launches of Unity VCS equipment.”

FACT: three misses out of four measured franchises over a full three-year cycle, and the single win is in the lowest-value category — fluidics cassettes — bought with a new equipment launch. “Trailed market growth” is management’s own euphemism for losing share in its highest-margin, most moat-critical line.

And the hard numbers corroborate it exactly. Lead with this. Alcon Implantables revenue: $1,703M (2023) → $1,775M (2024) → $1,782M (2025), i.e. +0.4% in FY2025. In the same market, in the same year, J&J’s Surgical Vision franchise grew +10.2% to $1,558M, with “newly launched TECNIS Odyssey [becoming] the fastest-growing intra-ocular lens in the U.S.” [J&J Form 10-K FY2025, https://www.sec.gov/Archives/edgar/data/200406/000020040626000016/jnj-20251228.htm, accessed 2026-07-18]. Flat versus +10.2% is share loss, quantified and independently corroborated. Nor is it a channel artifact — the working-capital data show no channel-stuffing or inventory tell behind the flat line (no inventory drawdown, no receivables ballooning, no DIO build), so the stall is genuine end-market and share weakness. Widen the frame and it is worse: in FY2025, every named competitor grew its comparable line faster than Alcon — J&J Surgical Vision +10.2%, Zeiss Ophthalmology +8.5%, B+L Vision Care +7%, B+L Surgical +6%, Glaukos +32%, against Alcon Surgical +4% and Implantables +0.4%. The largest firm in the industry was, in FY2025, its slowest-growing major.

4.3 Greenwald test (b) — the ROIC test

Book returns put Alcon in Greenwald’s “advantages absent” zone: IFRS ROE 4.45%, book ROIC 4.48%, against management’s own impairment-test discount rates of 8.5% (Surgical) and 8.0% (Vision Care) — i.e. roughly half the company’s own cost of capital. Those numbers are artifacts of the $18,262M spin-goodwill stack Alcon never paid for (Section 2).

The correct adjustment removes the acquired intangibles from the denominator and their amortization from the numerator. On that consistent basis Alcon earns ~20.5% on the ~$7.4B of tangible capital actually employed — the middle of Greenwald’s 15–25% “advantages present” band, not its floor. (Using core NOPAT gives 22.7%, which is too generous because core also strips recurring costs; using IFRS NOPAT over a tangible base gives ~15%, which is internally inconsistent. ~20.5% is the defensible figure.)

So the moat passes — but only after an adjustment argued this hard, and only in the past tense. The trend is the finding: honest tangible ROIC peaked at 21.8% in FY2024 and fell to 20.5% in FY2025; core ROE peaked at 7.03% and fell to 6.90%; book ROIC peaked at 6.60% and fell to 6.55%. FY2024 was the top. A widening moat does not roll over while the company spends more to stand still.

4.4 The head-to-head

The single cleanest refutation of scale as a firm-specific advantage sits in that table: Alcon’s consolidated 13.2% IFRS operating margin is barely above Carl Zeiss Meditec’s 10.9% Ophthalmology EBITA margin — for a competitor roughly one-third its size. If Alcon’s scale were a genuine cost barrier, it would show up here, and it does not. Nor does it show in lenses, where CooperVision earns a 26.6% segment operating margin against Alcon’s entire Vision Care at 21.5%. The accurate framing is “biggest, but not best” — Alcon is out-grown by J&J where it matters most, out-margined by Cooper in lenses, and only modestly out-margined by a Zeiss unit a third its size. Its clear wins are relative: it is vastly more profitable than B+L and it did not blow itself up in China like STAAR.

4.5 The threats to the ecosystem claim

The strongest challenges are to the ecosystem story, not the product line. Carl Zeiss Meditec has acquired D.O.R.C. (Dutch Ophthalmic Research Center) and now fields the EVA NEXUS vitreoretinal platform with dual-pump fluidics — closing the retina gap that was one of Alcon’s structural advantages — while bundling IOLMaster 700 biometry into a rival planning-to-OR ecosystem. That is directly material to whether the Unity refresh is a moat refresh or a defensive necessity: Alcon is not replacing 2008/2013 platforms into an uncontested field. J&J’s ELITA femtosecond laser running the SILK lenticule procedure, alongside Zeiss’s SMILE/SMILE Pro, attacks Alcon’s WaveLight LASIK franchise. Note the resonance with the failed STAAR bid: Endicott’s stated fallback after walking away was that “our new wavelight plus offering remains our focus” for refractive — Alcon is defending a laser franchise that is itself under lenticule attack, having just failed to buy the phakic-IOL alternative. (Alcon’s own Unity time-and-motion claims — 16% vitreoretinal workflow efficiency, 6% cataract turnover reduction — and PanOptix Pro’s 94% light utilization are company marketing claims, not independent data, and are labeled as such.)

4.6 The R&D treadmill

R&D: $828M (2023) → $876M (2024) → $990M (2025), +19.6% over two years against revenue +10.1% — now 9.6% of sales, with >2,100 R&D staff, 84 pipeline projects and 21 BD&L transactions in 2025. The 20-F states the economics itself: newly launched offerings “are typically introduced at a price premium… As these products age and/or competitive products advance, prices typically trend downward, requiring continuous innovation cycles to maintain and/or grow our margins.”

INTERPRETATION: that sentence is management describing a treadmill, not a moat. It concedes that pricing decays absent continuous reinvestment — which is Greenwald’s point that proprietary technology is the weakest and most transient barrier. The supporting evidence is consistent: R&D compounding at roughly twice revenue; both segment contribution margins falling in the same year; flagship launches that defend (Unity replacing 2008/2013 platforms; PanOptix Pro “stabilized” — management’s word — US trifocal share; Precision7) rather than open new ground; and the genuinely new category assets being bought, not invented (BELKIN/Voyager DSLT, Aerie/Tryptyr, LumiThera, Cylite, Aurion, plus the pending LENSAR and the failed STAAR).

A genuine disclosure gap worth naming: Alcon claims the “largest installed base” but never sizes it in units. Without console counts, “largest installed base” cannot be converted into revenue-per-console — the cleanest available test of the razor/razorblade moat. Similarly, whether Unity renews the base at higher, equal or lower ASP is unknowable from +6% equipment revenue. Both are OPEN QUESTIONS.

4.7 Verdict — weak/narrow advantage, and narrowing

Alcon holds one genuine, nameable moat and it is narrower than the bulls claim. The Surgical installed base converts into an 83.6%-recurring annuity at a 25.4% contribution margin; that is real, defensible for the length of a console cycle, and it is why this is not a bad business. But captivity is habit rather than contract, the barrier is re-tendered every seven to ten years, Zeiss has just closed the retina gap and built a rival ecosystem, and the profit-pool end of the franchise — ATIOLs — is where Alcon is visibly losing.

The disconfirming evidence, weighed honestly, does not rescue the wide-moat label. In its favor: ~20.5% on tangible capital is mid-band “advantages present”; the consumables annuity passes its own falsification test (consumables +6%, not decoupling from equipment); Vision Care’s gross margin is genuinely improving; and no competitor has broken the four-firm structure in either category. Against it: management’s own compensation disclosure concedes three misses in four measured franchises; implantables were flat against J&J’s +10.2% with no channel explanation; consolidated margin barely exceeds a competitor one-third the size; returns peaked in FY2024 and have rolled over; R&D grows at twice revenue; and ~$1.04B of new lens capacity goes into the ground alongside Cooper’s and B+L’s.

The moat test in one line: if the metric that would deteriorate absent the moat has deteriorated, the moat is failing. Consumables per console has not deteriorated — Surgical passes. Implantables growth versus market has deteriorated (+0.4% vs +10.2%) — IOL/ATIOL fails. Lens share and margin both went the wrong way — lens scale is an industry barrier Alcon rents, not a firm advantage it owns. Ocular-health share was lost to “competitive launches and overall pricing pressure” — brand there is a price premium, not a barrier. One pass, three fails.

VERDICT: a weak-to-narrow competitive advantage that is narrowing, not widening. Alcon earns a real spread on the capital it actually uses, holds one defensible franchise, and is paying a rising toll in R&D and capacity spend simply to hold position in the categories that generate its premium. That is a good business. It is not a great one, and it is not the quality compounder the market floated it as.

5. Growth History and Forward Opportunities

5.1 The three-year record, by sub-line

Alcon grew net sales from $9,370M in FY2023 to $10,319M in FY2025 — 10.1% over two years, roughly 4.9% a year, and 4% at constant currency in FY2025. That is the headline. The composition underneath it is the finding.

(FACT. Source: Alcon Form 20-F FY2025, net-sales table, https://www.sec.gov/Archives/edgar/data/1167379/000116737926000014/alc-20251231.htm, accessed 2026-07-18.)

Two things stand out. First, Vision Care — the segment with no firm-specific moat — grew faster over two years (+12.6%) than Surgical (+8.2%), the segment that carries the franchise. Second, within Surgical, Implantables grew 4.6% in two years and 0.4% in FY2025. That is the highest-margin, most moat-dependent line in the company, and it has stopped.

5.2 Organic versus acquired — the growth is organic, and that is not a compliment here

Alcon closed four bolt-ons in twenty-four months (BELKIN, Cylite, Aurion, LumiThera) for $753M of cash, and none of them moved the top line. Aurion Biotech — an $856M total-consideration transaction, 96% of it allocated to acquired in-process R&D — contributed $12M of net sales and a $37M net loss from its 2025-03-24 close to year-end (FACT, 20-F FY2025 Note 4). LumiThera closed 2025-09-02 with the purchase-price allocation still provisional and no disclosed revenue contribution; it is de minimis. BELKIN contributed $1M of sales in its 2024 stub. So essentially all of the reported 4.9% FY2025 growth is organic — while the acquisitions added $310M of goodwill and $419M of other intangibles, rebuilding the very amortization block that management’s “core” measure asks investors to disregard (Section 6).

Management’s own scorecard confirms the framing. The 2023–2025 long-term incentive plan measures results “at constant exchange rates” and “exclude[s] the impact of acquisitions, divestitures and certain non-recurring items” (FACT, 20-F FY2025 Compensation Report, Exhibit 24) — Alcon itself strips M&A out when judging whether the business grew. INTERPRETATION: the acquisition program is buying category optionality, not revenue, and it should be assessed as R&D-by-other-means rather than as growth.

5.3 The quality-of-growth red flag

Q1-2026 is where the mix problem becomes impossible to argue with. Sales of $2,706M grew 6% at constant currency, and the composition was this: Equipment +23% ($253M), Ocular health +10% ($487M), Consumables +4% ($769M), Contact lenses +4% ($738M), and Implantables +1% ($438M) (FACT, Q1-2026 press release, 6-K 2026-05-05).

Read that ordering carefully. The fastest-growing line is capital equipment — the Unity CS/VCS console replacement cycle. A console sale is a one-time, finite, lower-gross-margin transaction that pulls forward a decision customers make once every seven to ten years. The slowest-growing line is Implantables, the patient-pay premium IOL business that carries the segment’s highest gross margin and is the entire economic point of owning the installed base. And Consumables — the razor-blade annuity, the thing an installed-base moat is supposed to produce — grew 4%, below the 6% company average. INTERPRETATION: this is the inverse of the mix a franchise business wants. Alcon is converting durable, high-margin annuity growth into finite, low-margin hardware growth, and calling the aggregate +6%.

FY2025 tells the same story more slowly: Implantables +0.4% against Consumables +5.8% and Equipment +6.2%. This is not a one-quarter artifact.

5.4 The forward slate, assessed rather than listed

The Unity console cycle. Unity VCS was FDA-approved 2024-06-24 and Unity CS launched in late 2025; they replace Constellation (2008) and Centurion (2013). The refresh is real and it is driving the Equipment line. But the question that decides whether it is worth anything is unresolved: does Unity convert the installed base at a higher price, or does it merely hold units at flat ASP? Equipment revenue of +6.2% in FY2025 is consistent with either — and Alcon discloses no installed-base unit count anywhere, in any year’s 20-F, despite claiming the industry’s largest. Without units, “largest installed base” cannot be converted into revenue-per-console, which is the only clean test of the razor/razorblade moat. This is a genuine disclosure gap and it should be named as one (OPEN QUESTION).

PanOptix Pro. Management’s own word for what it achieved is that it “stabilized US trifocal market share” (FACT, 20-F FY2025 Compensation Report). Stabilized is not gained. In the same period J&J’s Surgical Vision franchise grew 10.2% on the back of TECNIS Odyssey.

TRYPTYR (acoltremon). FDA-approved 2025-05-28 for dry eye; roughly four share points captured in eight months; management guides peak sales of $250–400M. This is the single most credible new organic growth vector in the portfolio — a prescription product, meaning there is an actual gate, unlike the OTC ocular-health franchise. Size it honestly: $250–400M of peak revenue against a $10,319M base is 2.4%–3.9% of one year’s sales, arriving over several years. It is a good product. It is not a growth algorithm.

The rest of the slate — Precision7, Total30 multifocal toric (Feb-2026), Clareon TruPlus (ASCRS Apr-2026), Voyager DSLT (from BELKIN), Valeda for dry AMD (from LumiThera) — is a dense, simultaneous launch calendar. INTERPRETATION: density is itself a risk. A company whose forward growth requires six launches to execute at once, into a guide management has already back-end-loaded into 2H-2026, has concentrated its execution risk rather than diversified it.

ATIOL penetration is the one genuinely large runway. US penetration is approximately 20% of cataract procedures per the 20-F itself, corroborated independently by Market Scope; global is approximately 17% on management’s estimate, and management puts the long-run ceiling at “mid-30s to upper level high 30s” with a historical gain rate of 50–100bp a year (FACT/MANAGEMENT ESTIMATE, 20-F FY2025 Item 4.B; Q1-2026 call, 2026-05-06). A decade of mix-up is available. The problem is that Alcon is losing share inside it.

And there is now no inorganic lever. STAAR was terminated 2026-01-06 after its shareholders declined to approve the raised $30.75 bid; LENSAR was terminated 2026-03-16 by a Termination and Mutual Release Agreement after the FTC signalled it would seek to enjoin the deal (FACT, ALC 6-K 2026-01-06; LENSAR Form 8-K, Item 1.02, filed 2026-03-17). Both public-company transactions failed in a single year, and the FTC’s stated theory — that Alcon’s share in an adjacent surgical category makes consolidation anticompetitive — is a forward constraint, not just a past event. The bull case must now be carried organically, by exactly the product slate above.

5.5 Verdict — high- or low-quality growth?

Low quality, and deteriorating. The disconfirming evidence deserves weighing first: 4.9% growth is respectable for a $10B medtech business, it is almost entirely organic, it is broad-based across five sub-lines with no line declining, Vision Care compounded at 12.6% over two years, and TRYPTYR plus a 20%-penetrated ATIOL market are real runways. This is not a shrinking company.

But growth quality is about composition, and Alcon’s composition is moving the wrong way on every axis that matters. The highest-margin line is flat while the lowest-margin line is fastest. The recurring annuity is growing below the company average while one-time hardware grows at 23%. Growth is fastest in the segment (Vision Care) where the company concedes it missed its share target to private label, and slowest in the segment (Surgical) where the moat is supposed to live. The acquisitions bought no revenue but did rebuild the amortization block. And the one growth mechanism management could have bought — phakic IOLs, or the FLACS consolidation — is now foreclosed twice over, once by a shareholder vote and once by a regulator.

The financial signature of this appears in Section 6, and it is unambiguous: Alcon grew revenue $483M in FY2025 and kept $12M of it.

6. Financial Quality

6.1 Three earnings numbers, and why the choice is the entire debate

Alcon reports under IFRS. For FY2025 it published IFRS diluted EPS of $1.98 and a non-IFRS “core” diluted EPS of $3.07 — the second number is 55% larger than the first, and the gap has widened every year: $0.78 (2023), $1.00 (2024), $1.09 (2025), and $0.46 in Q1-2026 alone. Nothing in this memo matters more than which number an owner should use.

The bull’s case for core is strong on its best component. Of FY2025’s $784M of total intangible amortization, “Marketing know-how” ($238M, gross cost $5,960M unchanged for years, 100% Surgical) and “Technologies” ($22M, net book value down to $6M) are pure Nestlé/Novartis-era step-ups from the 2019 spin. Alcon paid nothing for them, they require no cash replacement, and adding them back is correct (FACT, 20-F FY2025 Note 9).

But core is not merely IFRS-plus-amortization, and this is the part the “just use core” framing obscures. Core also removes:

• $58M of acquisition and integration costs, of which $46M is direct banker/legal/diligence fees — spent by a company that did 21 BD&L transactions in 2025. A cost incurred every year is not non-recurring. And a large share of it was spent on STAAR, a deal that failed: cash out, nothing in. Q1-2026 booked another $20M.
• $21M of legal provisions — chronic, not episodic: $201M paid to J&J Surgical Vision in 2023; the Sight Sciences willful-infringement loss (>$34M of past damages plus royalties to November 2028); and an open DOJ False Claims Act Civil Investigative Demand from July 2024 with no provision taken against total litigation provisions of $18M.
• $44M of product discontinuation in Vision Care — the innovation treadmill writing off a failed asset, in a business whose own 20-F says prices “trend downward, requiring continuous innovation cycles.”
• Self-replacing software amortization — the “other intangibles including software” category amortized $122M in FY2025 against $122M of additions in the same category in the same year.
• And, new in 2026, “$88 million of costs associated with efficiency initiatives.” Note 12 of the Q1-2026 interim report confirms what this is: accrued severance (provision movement: $0 at 1 January, $70M of additions, $7M paid, $63M balance at 31 March). This is unambiguous cash cost, merely accrued now and disbursed later, and core excludes 100% of it. No total program size, savings target, or duration has been disclosed anywhere. For scale, the 2019–2023 transformation program cost $425M all-in.

Strip the illegitimate add-backs and you get honest owner earnings. Two independent routes:

The two routes converge within two cents. That convergence is what makes the number credible — an accrual bridge and a cash bridge built from different statements agreeing to within 0.7% is not a coincidence. Call it ~$2.75 per share, or ~$1,364M. It sits at roughly 72% of the distance from IFRS to core.

At $70.26 the three multiples are:

INTERPRETATION: 35.5x is an artifact of purchase accounting and should be ignored. 22.9x is the company’s own preferred frame and it is roughly 12% too flattering. 25.6x is the defensible number.

6.2 Returns — a real artifact, and an incomplete rescue

On the reported balance sheet, Alcon looks like a value-destroying business. FY2025 net income of $980M on shareholders’ equity of $22,034M is ROE of 4.45%. NOPAT on invested capital of $25,673M (equity $22,034M + financial debt $4,737M − cash and time deposits $1,607M + leases $509M) is ROIC of 4.48%. Alcon’s own impairment-testing discount rates, disclosed in Note 9, are 8.5% for Surgical and 8.0% for Vision Care — so by the company’s own cost-of-capital proxy, book ROIC is roughly half its hurdle.

That is an artifact, and the reason is on the balance sheet in plain sight: goodwill $9,256M plus other intangibles $9,006M = $18,262M, equal to 82.9% of shareholders’ equity and 57.9% of total assets. This is Novartis purchase accounting stepped up at the 2019 spin. Alcon never paid for it; Novartis did. Strip it and tangible invested capital is $7,411M.

The adjustment must be made consistently. Removing $18.3B of acquired intangibles from the denominator requires removing the amortization of those same intangibles from the numerator — otherwise the numerator is charged for an asset the denominator no longer holds. Done consistently, using honest NOPAT, Alcon earns approximately 20.5% on the tangible capital actually employed — the middle of Greenwald’s 15–25% “advantages present” band, not its floor.

But the trend is the finding, not the level:

Returns peaked in FY2024 and have rolled over on every measure — core ROE 7.03% → 6.90%, book ROIC 6.60% → 6.55%, honest tangible ROIC 21.8% → 20.5%. And there is no compounding mechanism: equity grows faster than earnings because $18.3B of inert spin goodwill sits on the balance sheet and Alcon keeps adding to it ($310M of goodwill and $419M of intangibles in FY2025 alone). INTERPRETATION: the adjustment rescues the level of returns from an accounting distortion. It does not rescue the direction, and direction is what a quality label depends on.

6.3 The absence of operating leverage — the single most important finding

Core operating income: $1,849M (2023) → $2,027M (2024) → $2,039M (2025). That is +9.6%, then +0.6%.

FY2025 revenue rose $483M. Core operating income rose $12M. Incremental core operating margin: 2.5%, against a 19.8% average margin. A business with operating leverage converts incremental revenue at above its average margin. Alcon converted at one-eighth of it.

Core diluted EPS went $3.05 → $3.07, +0.7% — and every cent came from below the operating line. In August 2025 management cut its assumed core effective tax rate from ~20% to ~18%, worth roughly $0.07–0.08 of EPS, in the same guidance revision that cut constant-currency sales growth to +4–5% and core operating margin to 19.5–20.5%. The operating deterioration was masked by a tax assumption.

The FY2025 core margin bridge decomposes the 80bp decline: underlying gross margin ex-tariff +0.88pp; tariffs (~$100M to cost of sales) −0.97pp; R&D intensity (9.59% of sales vs 8.91%, +$114M) −0.69pp; SG&A intensity −0.07pp. Absent tariffs and the R&D step-up, core margin would have expanded ~85bp — manufacturing efficiency is genuinely working. But neither offset is a one-off: tariffs are guided at $100–150M for FY2026, and management said it would reinvest the ~$25M of tariff relief rather than drop it through. R&D at 9.6% of sales is a stated strategic level.

Say it plainly: Alcon grew revenue and kept almost none of it.

6.4 The segment finding that inverts the conventional narrative

(FACT. Source: 20-F FY2025, Note 3 segment P&L.)

Surgical’s entire 197bp two-year contribution-margin decline is 100% gross margin — the 196bp fall in cost of sales performance. Operating-expense discipline was fine: SG&A leverage of +45bp roughly offset the R&D step-up of −46bp. The problem is adverse mix: Surgical revenue shifted from Implantables (32.0% → 31.0% of segment sales) toward Consumables (51.2% → 52.7%) and Equipment, as the flat, highest-gross-margin line was out-grown by lower-margin ones. Add tariffs and Chinese VBP price erosion on IOLs and the gross margin has to fall. Surgical grew $437M of revenue over two years and produced $6M of incremental contribution — a 1.4% incremental margin. Contribution dollars have been flat at roughly $1.46B for three consecutive years.

Vision Care ran the opposite way: gross margin +175bp over two years and a 39.8% incremental contribution margin on $512M of revenue growth. Its FY2025 contribution-margin dip of 82bp is entirely a 30% R&D step-up (+$85M, $284M → $369M, for Aurion and TRYPTYR), not operational deterioration.

INTERPRETATION: the segment hierarchy is inverted relative to how the company is narrated. Alcon is described — including by its own investor materials and by the late-2025 consensus view — as a surgical franchise with a contact-lens business attached. On unit economics the opposite is true right now. The segment carrying the one genuine, nameable moat is the one whose gross margin is decaying and whose profit dollars have not moved in three years; the segment with no firm-specific moat is the one delivering manufacturing efficiency and a 40% incremental margin. This is not an argument that Vision Care is the better business — its structural position is weaker (Section 4) — but it is a fact about where the economics currently sit, and any thesis built on “the surgical annuity compounds” has to confront it.

6.5 Cash flow and the capex holiday

First, resolve the capex discrepancy explicitly, because both circulating figures are right and they are different definitions. The FY2025 cash flow statement shows “Purchase of property, plant & equipment” of $543M and “Purchase of intangible assets” of $120M; $543M + $120M = the $663M total-capex figure. Management’s published FCF of $1,733M is CFO $2,271M − PP&E $543M + $5M of PP&E disposals — it excludes the $120M of software/intangible capex. Deducting all capex, honest FCF is $1,613M.

FCF nearly tripled from $537M to $1,613M in two years, and that is widely read as operating improvement. It is only about four-fifths that. Of the $1,076M swing, roughly $883M came from CFO — much of it the non-repeat of the $201M J&J settlement paid in 2023, plus a cash-tax tailwind — and roughly $188M came purely from capex falling. PP&E intensity dropped from 8.5% of sales in 2021 to 5.3% in 2025.

And it reverses. Alcon has disclosed approximately $1.04B of contact-lens capacity expansion — roughly 16 production lines — running to 2030 (Grosswallstadt, Singapore, Johns Creek; 20-F Item 4.D). PP&E purchase commitments were $276M at year-end. Q1-2026 PP&E capex was $139M vs $106M, +31% YoY, and FY2026 is guided to “mid-single-digit % of sales.” Depreciation of $417M against $543M of PP&E capex is a 1.30x ratio, compressed from ~1.8x in 2021. The FY2024–25 FCF level is a capex trough being harvested, not a structural step-change.

Net it down and the headline yield deflates:

Two further reversals sit in FY2026: cash taxes paid were only $107M in FY2025 against $326M (2024) and $255M (2023) — a ~$150M one-year tailwind — and the core tax rate is guided back up to ~20%. The “5% FCF yield” is really about 4.0%: a normal yield for a mid-single-digit grower, not a cheap one.

6.6 Balance sheet — strong, and under-appreciated

This is the part of the story that is better than the reputation. Total financial debt $4,737M against cash and time deposits of $1,607M gives net debt of $3,130M, or 1.19x core EBITDA (~$2,633M); 1.38x including leases. Interest coverage is 10.0x (core operating income $2,039M / $204M of interest expense). The blended interest rate is 3.6%, termed out through 2052, with the maturity ladder spread $576M inside one year, $2,389M in 1–5 years, and $1,800M beyond. Total liquidity is $2,927M, including a $1.32B unsecured multicurrency revolver, undrawn and extended in Q3-2025 to October 2030. There is roughly $3.45B of incremental net-debt headroom before a still-conservative 2.5x.

That last figure carries a conclusion: FY2026’s known calls are approximately $500M of buyback, ~$170M of dividends, and — before it was terminated — ~$430M for LENSAR, against $1.6–1.7B of FCF. Alcon could comfortably have funded STAAR (~$1.8B) and the buyback and LENSAR simultaneously. The STAAR walk-away was a price decision and a failed shareholder vote, not a financing constraint.

One item requires explicit handling so it is not misread: the Notes carry a fair value of $4,466M against $4,608M of carrying value — approximately 3.1% below par. This is a rate artifact, not credit stress. Alcon’s near-dated coupons are 2.375%–3.000%, well below current market yields, so the bonds mark below par. Economically that is a benefit — cheaply locked-in, long-dated debt. There is no credit stress anywhere on this balance sheet.

6.7 Stock-based compensation and dilution

FY2025 equity-based compensation was $171M (Surgical $81M, Vision Care $68M, unallocated $22M). In context: 1.66% of revenue, 8.4% of core operating income, 10.6% of honest FCF, 12.5% of honest owner earnings. INTERPRETATION: a modest, well-controlled load. This is not a software company, and SBC is charged in full in the owner-earnings bridge above rather than being a hidden distortion.

The buyback is real. Alcon repurchased 8,456,204 shares for $682M at an average $80.71 in FY2025, completing a three-year $750M authorization in ten months (finished 2026-01-20, 9.3M shares) and then announcing a new $1.5B / three-year program on 2026-05-05. Net of the $171M SBC expense, $511M of the $682M genuinely retired stock — roughly three-quarters of the dollars, and about 77% on a share-count basis (gross 9.3M repurchased less 7.19M net decline in shares outstanding implies ~2.1M shares of employee issuance, which cross-checks to $171M / $80.71 = 2.12M within 0.5%).

An earlier read that the buyback “only funds SBC” rested on the weighted-average diluted share count falling just 1.3M (497.5M → 496.2M) in FY2025. That is a timing artifact — purchases spread through a year count only fractionally in a weighted average by construction. The real effect shows in Q1-2026: 490.2M diluted shares vs 498.0M, −1.6% YoY, and in shares outstanding falling 494,616,324 → 487,427,920, −1.45%.

The mark, however, is poor: average repurchase cost $80.71 against $70.26 today — approximately 13% underwater, an unrealized loss of roughly $88M on the completed program. Buying more as the price fell (1.75M shares in September at $77.41 against 327,500 in April at $90.49) is genuinely counter-cyclical and rarer than it should be. It has simply not been vindicated yet.

6.8 Working capital and earnings-quality tells

Inventories rose 5.4% against sales +4.9%; DIO was essentially flat at ~188 days. Receivables rose 11.9% and DSO rose 4.3 days, but only back to the 2023 level — 2024’s 63.9 days was the outlier — and payables stretched in parallel, so the cash conversion cycle improved to 183 days. Receivables aging is stable: past due more than three months was 6.2% of gross in 2025 versus 6.1% in 2024, with the provision at $50M on a 12% larger book. No customer is 10% or more of net sales.

INTERPRETATION, and this is an important negative finding: there is no channel-stuffing or inventory-build signal behind the flat implantables line. A channel stuff would show inventory drawn down at Alcon with receivables ballooning and aging deteriorating. A demand air-pocket masked by a build would show rising DIO. Neither is present. The implantables stall is therefore genuine end-market and share weakness, not an accounting or channel artifact — which removes the benign explanation and makes the competitive read harder to dismiss, not easier.

One residual question: not-overdue receivables grew 12.8% against 4.9% sales growth. The most plausible explanations are a Q4-weighted revenue skew or longer payment terms on capital-equipment placements as the Unity refresh accelerates (Equipment +23% in Q1-2026). Benign either way, but unresolved from the filings (OPEN QUESTION).

Finally, Alcon has built working capital every single year — $(314)M, $(522)M, $(404)M, $(189)M, $(240)M across 2021–2025. That is structural, and it should be treated as a permanent drag on FCF rather than normalized away.

6.9 Verdict — do economics improve with scale?

No. They are deteriorating at the margin.

The disconfirming evidence first, because it is substantial and should not be waved off. Net leverage of 1.19x, 10x interest coverage, a 3.6% blended cost of debt termed to 2052, $2.9B of liquidity with an undrawn revolver to 2030, no customer concentration, stable receivables aging, an improving cash conversion cycle, an SBC load of 1.66% of revenue, a buyback that is roughly three-quarters genuine shrink and executed counter-cyclically, and — corrected for the numerator/denominator inconsistency — approximately 20.5% return on the tangible capital actually employed. That is a business that will not get into trouble, and 20.5% sits mid-band in Greenwald’s “advantages present” range, not at its floor.

But the section’s question is about incremental economics, and the answer is unambiguous. FY2025 revenue grew $483M and core operating income grew $12M — a 2.5% incremental margin against a 19.8% average. Core operating income went $1,849M → $2,027M → $2,039M. Core diluted EPS went $3.05 → $3.07 with every cent coming from a tax-rate cut. Core ROE and honest tangible ROIC both peaked in FY2024 and have rolled over. Segment-level, the decay is concentrated precisely where the moat is supposed to be: Surgical’s 197bp two-year contribution-margin decline is 100% gross margin, driven by a flat, highest-margin implantables line being out-grown by lower-margin consumables and equipment — a 1.4% incremental margin — while Vision Care, the segment with no firm-specific moat, delivered +175bp of gross margin and a 39.8% incremental margin. Scale is not lowering unit costs faster than mix and price are eroding them.

The earnings quality is the second problem and it is widening rather than narrowing. The IFRS-to-core wedge went $0.78 → $1.00 → $1.09 and hit $0.46 in Q1-2026 alone. Part is legitimate; the rest is $58M of acquisition costs from a strategy that runs 21 deals a year (with $46M spent on a failed one), chronic legal provisions against an unprovisioned DOJ investigation, $44M of treadmill write-offs, self-replacing software amortization, and an $88M-per-quarter severance program of undisclosed total size that management excludes from its own headline metric. Every bolt-on mechanically rebuilds the amortization block core asks investors to ignore — unallocated cost of net sales rose $67M in FY2025 on the Aurion, LumiThera and Cylite closings. Honest owner earnings are ~$2.75 per share, confirmed to within two cents by an independent cash route — 25.6x at $70.26, not the 22.9x the core framing implies.

And the cash flow is flattered on two axes that reverse: a capex holiday worth roughly $188M of the FCF improvement, against $1.04B of announced lens capacity still to be spent and Q1-2026 capex already +31%; and a ~$150M one-year cash-tax tailwind against a FY2026 core tax rate guided back to 20%.

The verdict on the memo’s spine: Alcon is a good business, not a great one. It earns a real spread on the capital it actually uses, it is financed conservatively, and it is competently run. But the operating leverage is absent, the returns peaked in FY2024, the gap between what it reports and what an owner earns is widening in step with the acquisition programme, and the segment carrying its only nameable moat is the one whose unit economics are decaying. That is a quality-of-earnings caution, not a solvency one — but it is precisely the evidence that the quality label no longer fits.

7. Capital Allocation

Alcon’s capital allocation is conservative in form and expansionary in substance. It takes no balance-sheet risk, structures its deals better than most acquirers its size, and pays its executives on a governance framework cleaner than most US large-caps. And it is pointed at the wrong target. The verdict is BELOW AVERAGE for a business of this quality, and deteriorating — not because management is reckless or dishonest, but because nobody in the incentive plan, on the board, or on the share register is measuring return on capital.

7.1 The 0-for-2 M&A Year

In a single twelve-month window Alcon attempted two public-company acquisitions and lost both, for two entirely different and entirely foreseeable reasons. Neither failure was a price decision.

STAAR Surgical. Alcon signed a definitive merger agreement on 2025-08-05 at $28.00/share cash (~$1.5B equity value; a 59% premium to the 90-day VWAP), targeting STAAR’s EVO ICL phakic implantable collamer lens as a complement to its LASIK franchise (6-K, 2025-08-05). A $1.9B Morgan Stanley bridge was executed 2025-08-20. Broadwood Partners, holding ~31% of STAAR, opposed. In November 2025 Alcon — by mutual agreement — waived both its break-up fee and its matching rights and permitted an unencumbered go-shop, which expired 2025-12-06 with no superior proposal. On 2025-12-09 Alcon raised the bid ~10% to $30.75/share, which Endicott called “best and final”; 20-F Note 26 puts total consideration at ~$1.8B. STAAR’s shareholders declined to approve it, and Alcon terminated on 2026-01-06 (6-K, 2026-01-06). No termination fees were paid by either side; the bridge was cancelled undrawn.

The management framing — Endicott: “Throughout this process we remained disciplined with our views on price and risk” — does not survive the arithmetic. STAAR carried ~$187.5M of net cash and no debt (STAAR FY2025 10-K), so the implied EV at $28.00 was ~$1.31B against FY2025 sales of $239.4M: 5.5x EV/sales, rising to 5.9x at $30.75. Alcon’s own equity trades at 3.82x EV/TTM sales (ROIC.ai, TTM 2026-03-31). Alcon therefore bid a 45–55% premium to its own multiple for a business running a -19.2% operating margin, with revenue down 25.7% from its 2023 peak and ~32% of sales flowing through two Chinese distributors into a VBP regime that has cut IOL prices by 20%/26%/38%/53%/84% across successive rounds. There were no earnings, so no earnings multiple existed. INTERPRETATION: the original bid was a full price for a hoped-for revenue base, not a disciplined one for the actual base — and the “accretive in year two” claim required a China recovery Alcon neither controlled nor underwrote.

The escalation is the damning part. Having signed at a full price, Alcon raised it 10% while surrendering the two protections that exist precisely to compensate an acquirer for a failed process. That is the textbook signature of escalation of commitment. A retained break-up fee would have covered the entire ~$46M of FY2025 direct acquisition costs. And the “we dodged a bullet” reading is not available either: STAAR’s Q1-2026 preliminary net sales exceeded $90M against $239.4M for all of FY2025, implying a run-rate far above the trough Alcon underwrote. On the evidence in hand, Alcon did not escape overpaying — it lost the asset, and Broadwood’s holders were right that Alcon was buying a China-VBP casualty at the bottom of its cycle while paying for none of the recovery.

LENSAR. Agreed 2025-03-24 at $14.00/share cash plus a non-tradeable CVR of up to $2.75 contingent on 614,000 cumulative procedures over 2026–2027 (~$430M all-in). Both parties received an HSR Second Request on 2025-05-21 — roughly eight weeks after signing. Ten months later the parties abandoned: LENSAR disclosed that “the Federal Trade Commission intends to seek to enjoin the acquisition,” and a Termination and Mutual Release Agreement was executed 2026-03-16 (LENSAR 8-K, 2026-03-17). LENSAR retained a $10.0M deposit. The FTC’s theory was that the deal combined the #1 and #2 players in femtosecond laser-assisted cataract surgery — an objection visible on the face of the transaction the day it was signed.

The ledger of the year: ~$76M expensed for zero acquired assets — $46M of direct acquisition costs in FY2025, $20M in Q1-2026, plus the $10M forfeited deposit. That is 3.7% of FY2025 core operating income and 5.6% of IFRS operating income, and it understates the cost, which also includes thirteen months of CEO/CFO/corp-dev attention in a year requiring two guidance cuts. There is a durable consequence too: a regulator has now asserted on the record that Alcon’s share in an adjacent surgical category makes consolidation anticompetitive. Core-adjacent M&A is now antitrust-constrained; growth must come organically or from genuinely new categories. (FACT: the failures. INTERPRETATION: the constraint.)

7.2 The Bolt-On Ledger, Graded

BELKIN is the model and proves Alcon can structure. $61M of cash up front against up to $385M of commercial milestones (contingent consideration carried at just $6M at acquisition) puts ~86% of the potential price on the seller’s ability to deliver. The LENSAR CVR carried the same instinct. Deal structuring is a genuine Alcon competence; the failures are in target selection and process, not terms.

Aurion is the verdict-driving deal. $856M of acquisition-date consideration, of which the final PPA allocated $820M — 96% — to acquired IPR&D for a corneal cell therapy approved in Japan only, plus $175M of goodwill. The asset produced $12M of sales and a $37M net loss in the nine months from close to year-end: roughly 53x annualized sales, 71x on the reported stub. It is PwC’s sole critical audit matter in the FY2025 20-F, and no impairment has been taken. This is a binary biotechnology bet booked inside a medtech balance sheet; it is the largest capital commitment Alcon has actually completed since the spin — larger than the rest of the closed ledger combined — and it received a fraction of the scrutiny the failed STAAR bid attracted. An $820M write-down would be ~40% of a year’s core operating income.

Ocumension: read the structure, not the gain. Alcon exited China rights to Bion Tears, Tears Naturale and procedural drops and took its $116M as ~16.7% of Ocumension’s ordinary shares plus royalties and AR-15512 milestones — converting a direct China exposure into an equity exposure to a Chinese ophthalmic pharma rather than de-risking it. The $57M net gain flattered FY2024 operating income, and its absence is one identified cause of the FY2025 IFRS margin step-down (14.4% → 13.2%).

Set against a Surgical segment whose contribution has been flat at ~$1.46B for three consecutive years with margin down 27.4% → 25.4%, the pattern is Marathon’s canonical one: capital flowing into growth optionality exactly as returns in the core decelerate. The 20-F records 21 BD&L transactions in 2025 and credits management with having “overachieved our business development and licensing deal flow targets” — deal count is an explicitly rewarded objective, which is a vanity metric measuring activity rather than return. Aggregate BD&L spend is not disclosed; the $692M “acquisitions of businesses” line captures only the four business combinations. Alcon is a systematic net buyer of assets more expensive than itself — 5.5–5.9x (STAAR), 6.1–7.4x (LENSAR), 53–71x (Aurion) — against its own 3.8x.

7.3 Buyback and Dividend

The $750M repurchase authorized 2025-02-25 as a three-year program was completed in ten months, on 2026-01-20, at 9.3M shares. FY2025 activity (20-F Item 16E): 8,456,204 shares at an average $80.71, $682M of consideration. Monthly average prices ran $90.49 (April) → $75.33 (October), and volume accelerated as the price fell — 327,500 shares in April against 1,750,000 in September and 1,547,394 in October. A new $1.5B / three-year authorization followed on 2026-05-05.

Both readings are worth carrying. Positive: this is genuine counter-cyclical execution, rarer than it should be, and it is not the anti-dilution treadmill it superficially resembles. Against $171M of FY2025 equity-based compensation, net buyback was ~$511M — ~75–77% of the spend genuinely retired stock. Two independent methods agree: $171M / $80.71 = 2.12M shares of dilution absorbed, while actual shares outstanding fell 494,616,324 → 487,427,920 (-7.19M) against ~9.3M repurchased, implying ~2.11M of net employee issuance. (The oft-quoted “-1.3M diluted shares on $682M” is a weighted-average artifact and materially overstates the treadmill; FY2026 guided diluted shares of ~492M vs 496.2M is the fuller effect.) Cautionary: management describes the program’s purpose as “intended to offset the dilutive effect of shares vesting” — SBC is a permanent ~25% toll on every buyback dollar — and the FY2025 tranche is ~13% underwater at $70.26, an unrealized mark of ~-$88M. The process was right; the outcome is not yet vindicated.

The dividend was frozen. CHF 0.21 (2023) → 0.24 (2024) → 0.28 (2025) → CHF 0.28 proposed 2026-02-24 — flat, the first non-raise since the 2019 spin. (The “~10% increase” in management’s own scorecard refers to the USD amount paid during 2025 and is an FX effect.) At $166M paid in FY2025 that is 10.9% of core net income, 10.3% of FCF, and a 0.48% yield. A business earning ~20.5% on tangible capital, generating $1.6B of FCF at 1.2x net leverage, paying out ~11% of core earnings, is not running a capital-return policy — it is paying a token. Freezing the committed instrument in the same year the discretionary one was doubled is a defensible preference; it is not a shareholder-friendly one, and it removes the single hard commitment that would discipline the capital budget.

7.4 The Capex Nobody Is Scrutinizing, and R&D That Is Not Converting

The largest capital-allocation decision on the table is not a deal. It is the ~$1.04B, ~16-line contact-lens capacity build running to 2030 (20-F Item 4.D): Grosswallstadt +3 lines ($162M, to 2027); Singapore +4 lines ($189M, completed 2024), +3 lines and a new building ($314M, to 2027), +3 more ($157M, to 2030); Johns Creek +2 lines ($148M, to 2028) and +1 line ($73M, to 2029). For scale, $1.04B is 106% of one full year of Vision Care segment contribution and larger than every completed acquisition combined ($753M of cash M&A across two years) — and it has attracted a fraction of the attention the two failed deals did.

The honest split: contact lenses grew +6% in FY2025 and Vision Care gross margin rose 175bp over two years on a 39.8% incremental contribution margin, so the build is delivering manufacturing efficiency, and daily-disposable silicone-hydrogel is genuinely under-penetrated. Against that, Alcon’s own compensation report concedes it missed its contact-lens share target “primarily due to competitive pressure in the Dailies category, as well as growth of private label, particularly in the international market” — it is adding capacity in the exact category where it is losing share, and losing it partly to private label, the buyer least willing to pay for a brand and most willing to fill someone’s spare line. CooperVision has spent 9–11% of sales for years on the same build; Bausch + Lomb grew revenue to $5,101M while earning a 3.7% operating margin and losing $360M. Three of four incumbents are adding high-fixed-cost capacity simultaneously into a category whose growth the leader itself now describes as “the low end of mid-single digits.” Each build is individually defensible; the aggregate is the problem. Fixed capacity must be filled, the marginal fill is private-label volume, and that is precisely the volume that erodes the branded price umbrella Vision Care’s 21.5% contribution margin depends on. INTERPRETATION: rational as a defence, dangerous as an industry event. Alcon discloses no target utilization, no incremental unit capacity and no hurdle rate — a genuine disclosure gap that makes “demand-led vs. share-defensive” unsettleable from the filings.

R&D tells the same story one line up. $828M (2023) → $876M (2024) → $990M (2025): +19.6% over two years against revenue +10.1%, now 9.6% of sales. Over the identical window core operating income moved $1,849M → $2,039M, +3.3%/yr; in FY2025 alone R&D rose 13.0% while core operating income rose 0.6%. Alcon spends 2.3x CooperVision’s R&D intensity (4.2%) and 1.3x Bausch + Lomb’s (7.3%) — and is losing lens share to Cooper, which earns a 26.6% segment operating margin against Alcon’s 21.5%, and IOL share to J&J (Implantables +0.4% vs J&J Surgical Vision +10.2%). The 20-F states the mechanism in its own words: “As these products age and/or competitive products advance, prices typically trend downward, requiring continuous innovation cycles to maintain and/or grow our margins.” That is management describing R&D as the maintenance cost of the existing price level, not as an investment that widens a moat.

7.5 The Root Cause: No Return-on-Capital Metric Exists Anywhere in the Pay Plan

Everything above resolves into one structural fact, and it is the punchline of this section.

Short-term incentive (annual, cash, 0–200%): 40% Net Sales | 40% Core Operating Income | 20% Free Cash Flow, times an Individual Performance Factor. Long-term incentive (three-year, 100% PSUs, 0–200%): 25% Net Sales CAGR | 25% Core Diluted EPS CAGR | 25% Share of Peers | 25% Innovation — the same four metrics carry unchanged into the live 2025–2027 cycle.

There is no ROIC, no ROE, no return on tangible capital, no economic profit, and no capital-efficiency test of any kind — in either plan, in either cycle. A management team scored this way is structurally indifferent to the return on the marginal dollar. It can commit $1.04B of contact-lens lines into a category it is losing, pay 53–71x sales for Aurion, bid 5.5–5.9x for a loss-making STAAR, and grow R&D four times faster than core operating profit — and every one of those decisions is scorecard-neutral or scorecard-positive while being return-dilutive.

The calibration failed in the one cycle now auditable. The 2025 STI worked: BPF 80% (Net Sales 75%, Core Operating Income 59%, Free Cash Flow 135%), CEO payout 72% of target, total ECA compensation -12% in CHF. Credit it plainly. But the 2023–2025 PSU paid 137%:

Over that cycle shareholders received a 58th-percentile TSR and 3.3%/yr core operating income growth. The one metric measuring competitive outcome — market share — paid 35%, on management’s own admission of misses in three of four franchises: contact lenses (“we fell short of our share-expansion target”), ocular health (“we didn’t hit our share gain targets”), and ATIOLs (“Alcon trailed market growth”). The plan diagnosed the problem correctly and then paid 137% anyway.

The Core Diluted EPS CAGR metric is a live quality-of-incentives question. Alcon’s reported core diluted EPS runs $2.24 (2022) → $2.74 → $3.05 → $3.07 (2025) — a three-year CAGR of 11.1%, or 5.9% measured 2023→2025. The plan reports an actual of 21.1%, roughly 1.9x anything a shareholder can compute from the audited core series, and it paid the capped maximum, contributing 50 of the 137 points. Even the plan’s target of 14.1% exceeds the reported 11.1% outturn. The footnote explains the mechanism — results are “measured at constant exchange rates” and “exclude the impact of acquisitions, divestitures and certain non-recurring items” — but Alcon discloses neither a base year nor any reconciliation. That is a core number, adjusted again for FX, again for M&A, and again for unspecified non-recurring items. Had it paid 100% rather than 200%, the PSU would have paid 112%, not 137%. OPEN QUESTION, and a genuine disclosure gap. A related one: the plan says nothing about neutralizing buybacks, so a repurchase that shrinks the share count mechanically raises the CEO’s PSU payout.

Nor is there anyone to object. CEO 2025 total compensation was $11,602,744 (LTI 65% of it), with the STI paid below target. Endicott’s vested stake — 269,775 shares plus 287,898 target PSUs — is ~0.11% of shares outstanding; total Board and executive ownership is under 0.5%, and no individual holds 1%. The register is diffuse index money (BlackRock 6.05%, UBS entities ~5.3–5.8%), Novartis exited fully in 2019, there is no activist and no strategic block, and Alcon has not opted out of the Swiss mandatory-offer regime. Alignment runs through annual grants, not through an owner’s stake — a single year’s pay is worth ~60% of Endicott’s entire accumulated vested holding.

Credit where due: the governance scaffolding is better than most US large-caps — no severance agreements, no single-trigger change-of-control, no excise-tax gross-ups, no stock options, 100% PSU long-term incentive, clawbacks, hedging and pledging prohibited, binding Swiss say-on-pay, and target pay set near median despite above-median revenue and market cap. The defect is entirely in the metrics inside the plan, not the architecture around it.

VERDICT — BELOW AVERAGE. Alcon is not destroying capital: it earns ~20.5% on the tangible capital actually employed, funds everything internally (FY2025 deployment of $2,077M against $2,271M of operating cash flow), carries 1.2x net leverage at a 3.6% blended rate with an undrawn revolver to 2030, and returns ~2.0% in shareholder yield. But in one year it went 0-for-2 on public M&A at a cost of ~$76M and thirteen months of management attention; the largest deal it did close put 96% of $856M into unapproved IPR&D; it is building $1.04B of capacity into a category it concedes it is losing; it spends 2.3x a peer’s R&D intensity to grow slower and earn less; it froze its dividend at an ~11% payout while doubling a buyback authorization; and it is paid on a scorecard that cannot see any of it. Prudent, well-structured, honestly governed — and pointed at the wrong target. Until a return metric enters the incentive plan or a holder large enough to insist appears on the register, the base case is that this pattern continues.

8. Changes and Headwinds — Last Two Years

The two-year record reads as a company defending an ageing installed base with a mandatory platform refresh, discovering mid-2025 that its growth algorithm no longer worked, and then spending the following twelve months absorbing the consequences — two failed acquisitions, a reimbursement cut, a lost patent case, a tariff regime, and an undisclosed restructuring program.

Mid-2024 — the platform refresh, which was necessary rather than expansionary. The FDA approved the UNITY Vitreoretinal Cataract System on 2024-06-24, followed by UNITY CS in late 2025 (Canada launch June 2026). The commercial logic is unambiguous: Centurion dates to 2013 and Constellation to 2008 — twelve and seventeen years old respectively — against a console replacement cycle of seven to ten years. INTERPRETATION: this was a defensive necessity dressed as a moat refresh. The corroborating evidence sits in Alcon’s own compensation report: the only category where it recorded a share gain across the 2023–2025 LTI cycle was “phaco and vitreoretinal cassette packs… driven by our recent launches of Unity VCS equipment” — the new platform bought back consumables share it needed to defend, while ATIOLs, the profit pool, lost ground. The tape agreed with that reading: the UNITY VCS approval moved the stock +0.8% across the surrounding three sessions.

July 2024 — the DOJ Civil Investigative Demand, still open and still unquantified. Alcon received a CID under the False Claims Act relating to discounts on surgical-equipment servicing contracts. It is cooperating. As of the 2026-02-24 20-F it remains open with no quantified provision, against total litigation provisions of only $18M at 2025-12-31. FCA exposure carries treble damages and per-claim penalties, and Alcon’s US surgical installed base is the largest in the industry. This is an unquantified tail risk sitting on the balance sheet with no accrual against it.

February 2025 — the Simbrinza loss. On 2025-02-05 the District of Delaware ruled Alcon had not proven infringement in its Hatch-Waxman suit against a generic filer (the court also held the asserted claims not proven invalid). Both sides appealed; appellate briefing concluded January 2026 and the outcome is unknown. Generic exposure on Simbrinza is live.

May 2025 — TRYPTYR. FDA approval of acoltremon (AR-15512) for dry eye on 2025-05-28; launched July 2025, taking ~4 share points in eight months with peak sales guided at $250–400M. Commercially real, and a structurally better business than OTC drops because a prescription is a genuine gate. It was also a price non-event (-0.4%).

August 2025 — the guidance reset, and the substance of the de-rating. This is the most important sequence in the two-year window. Across four updates management dismantled its own algorithm:

Read the columns together. Between February and August constant-currency growth was cut from +6–8% to +4–5% and core operating margin by ~150bp cumulatively. The core EPS dollar range was held across both cuts — but the constant-currency growth rate was cut from +8–11% to 0–+2%, and in August the core tax rate was simultaneously dropped from ~20% to ~18%, worth roughly $0.07–$0.08. Operating deterioration was therefore partially masked below the operating line. FY2025 landed at $3.07 on a 19.8% core operating margin — below even the twice-cut range’s midpoint. The 2025-08-20 print produced a -10.1% single-day fall, and the stock fell 18% from there to 31 October. INTERPRETATION: this was not a one-quarter miss. A “5–7% grower with operating leverage” was repriced as a “4–5% grower with flat earnings,” and the market was right to do it — core operating income grew 0.6% on 5% revenue.

August 2025 to March 2026 — the M&A failures. The STAAR saga (signed 2025-08-05, escalated 2025-12-09, terminated 2026-01-06) and the LENSAR termination (2026-03-16, FTC intent to enjoin, $10M deposit forfeited) are analysed in Section 7. For the changes narrative the relevant points are three: Alcon’s stated refractive strategy now rests entirely on organic wavelight plus/LASIK, the high-myope phakic-IOL category is left to a recovering independent STAAR, and the FTC has placed on the record a theory that constrains future core-adjacent M&A. The STAAR termination was a price non-event (-0.04%) — the tape had already re-rated the strategy.

March 2026 — the Sight Sciences loss. The court preserved a jury finding of willful infringement by Alcon on the Hydrus Microstent, carrying >$34M of past damages plus ongoing royalties through November 2028 (GlobeNewswire, 2026-03-30). A recurring cash toll on a glaucoma implantable already under reimbursement pressure.

January 2026 — reimbursement. The CMS 2026 Medicare Physician Fee Schedule, effective 2026-01-01, cut physician reimbursement ~10.5% for cataract and combined cataract+MIGS procedures and 7.3–9.2% for other MIGS. That sits on top of the November-2024 MIGS Local Coverage Determinations finalized by five of seven MACs, which demoted MIGS below drops and laser as first-line therapy and restricted coverage to one MIGS device per eye alongside cataract surgery. Endicott, asked directly why Implantables grew only 1% in Q1-2026: “one of the reasons we kind of called out glaucoma implantables is because, as you know, the reimbursement changed this year.” The precedent for magnitude is the January-2022 coding change, which cut angle-based stenting volumes by ~20,960 procedures year-on-year. This closes off an adjacency the industry had been counting on, and the cataract fee cut itself discourages allocating scarce surgeon time to the base procedure.

Tariffs — a new structural cost. ~$100M to cost of sales in FY2025; ~$100–150M guided for FY2026 on an assumed ~10% average US import tariff (cut from 15%), with the ~$25M of relief explicitly reinvested rather than dropped through. Q1-2026 carried $33M, costing 120bp of core gross margin. A Swiss-domiciled manufacturer with 45% of sales in the US and plants in Germany, Singapore and Switzerland is structurally exposed. This is a permanent industry cost, not an add-back.

February 2026 — the undisclosed efficiency program, and the conspicuous silence. Announced 2026-02-24 and charged $88M in Q1-2026 alone, of which Note 12 confirms $70M was accrued severance — unambiguous cash cost, merely accrued now and disbursed later. Core excludes 100% of it. No total program size, savings target or duration has been disclosed anywhere in the corpus. For scale, the 2019–2023 transformation program was announced with a $300M cost target and $200–225M of savings, later expanded, and cost $425M all-in for $300–325M of run-rate savings — announced, sized and tracked publicly. INTERPRETATION: the silence is the finding. A company that sized its last restructuring in advance and is not sizing this one, in the same quarter it guided core EPS up 10–13%, is asking investors to accept a reacceleration whose offsetting cost it will not quantify.

Governance. Scott H. Maw, a founding director and Chair of the Audit & Risk Committee, is not standing for re-election at the 2026-04-30 AGM. Alcon loses its Audit Chair at the precise moment three unquantified exposures are live: Aurion’s $820M of un-impaired IPR&D (PwC’s sole critical audit matter), the undisclosed-scope efficiency program, and the open DOJ False Claims Act investigation against $18M of total litigation provisions. His successor and that person’s accounting background are an open question.

VERDICT: these developments materially WEAKEN the thesis, and the weakening is broad rather than concentrated. One item would be noise. What the two-year record shows instead is deterioration on every axis at once — the demand algorithm (two guidance cuts, market growth marked down three times in eighteen months to a FY2026 assumption of 3–4%), the growth strategy (0-for-2 on public M&A, with the antitrust menu now narrowed), the reimbursement environment (MPFS cut plus MIGS LCDs), the cost structure (tariffs at 120bp of gross margin, permanent), the legal tail (an open FCA investigation with no provision, a willful-infringement royalty to 2028, a lost Hatch-Waxman case on appeal), and the disclosure regime (an unsized restructuring, a departing Audit Chair). The single genuine positive — the UNITY platform refresh — was mandatory to defend an installed base built on 2008 and 2013 hardware, and the market treated it as such. Nothing in the last two years supports the quality-compounder label; the period is better read as the market correcting it.

9. Risk Analysis

9.1 Risk Matrix

9.2 The Risks That Matter Most

1. Implantables share loss is the thesis risk, not a sector risk. Implantables is the highest-margin, most moat-dependent line Alcon has, and it has been effectively flat for three years — $1,703M → $1,775M → $1,782M — while J&J’s Surgical Vision franchise grew 10.2% in FY2025 on the back of TECNIS Odyssey, and Bausch + Lomb grew Surgical 6% despite the enVista voluntary recall. In a market Alcon leads with 41% share, every named competitor grew its comparable line faster than Alcon did in FY2025. The mechanism is structural, not cyclical: the IOL is a product sold on optical performance into a procedure the surgeon can change at will, and Greenwald is explicit that proprietary product technology is the weakest and most transient barrier. Alcon’s own compensation report concedes ATIOLs “trailed market growth.” The financial signature is already visible — Surgical contribution margin fell 197bp over two years, entirely in gross margin, because the high-margin implantables line is being out-grown by lower-margin consumables and capital equipment. Likelihood high, impact high: the ATIOL pool is where the premium economics live, and share moves within a single product cycle.

2. The efficiency program proving recurring is the single most important open question in the file. Alcon charged $88M in Q1-2026 to “costs associated with efficiency initiatives,” of which $70M is accrued severance per Note 12 — real cash, merely deferred in timing. Core excludes all of it, and no total size, savings target or duration exists anywhere in the corpus. That single undisclosed number swings the forward picture entirely: against honest owner earnings of ~$2.75/share in FY2025, a $150M program implies ~$2.92 (+6%), a $250M program implies ~$2.75 (flat), and four times the Q1 run-rate ($352M) implies ~$2.59 (-6%) — all while core EPS is guided up 10–13% to ~$3.42. The entire guided reacceleration can be absorbed by a restructuring management has excluded from its own headline metric and declines to size. Likelihood high — the prior transformation program ran four years and $425M — impact high, because it determines whether owner earnings grow at all in FY2026.

3. The capacity build landing into a price war. ~$1.04B and sixteen lines to 2030 is larger than every completed acquisition combined, and it is going into the ground alongside CooperVision’s and Bausch + Lomb’s simultaneous builds, in a category whose growth Alcon itself has marked down from ~6–8% to “the low end of mid-single digits,” and in which its own proxy concedes it lost share to private label internationally. The individual logic is sound — daily silicone-hydrogel is under-penetrated, high-fixed-cost lines are the barrier that keeps entrants out, and Vision Care gross margin has genuinely improved 175bp on a 39.8% incremental contribution margin. The aggregate is the danger: fixed capacity must be filled, the marginal fill in a lens plant is private-label volume, and that volume erodes the branded price umbrella Vision Care’s 21.5% contribution margin depends on. Likelihood medium — the mix-up to daily-SiHy is real and may absorb the lines — impact high, because a broken pricing equilibrium in a 95%-CR4 oligopoly is not quickly repaired.

4. Two unquantified legal and accounting exposures with no provision against either. The DOJ Civil Investigative Demand under the False Claims Act, received July 2024 on surgical-equipment servicing discounts, remains open at the 2026-02-24 20-F with no quantified provision against total litigation provisions of $18M. FCA matters carry treble damages and per-claim penalties, and Alcon’s US surgical installed base is the industry’s largest — the theoretical exposure is unbounded and the accrual is zero. Separately, Aurion’s $820M of acquired IPR&D rests on a corneal cell therapy approved in Japan only, generating $12M of sales and a $37M loss; it is PwC’s sole critical audit matter and no impairment has been taken. A full write-down would be ~40% of a year’s core operating income, and Alcon discloses no milestone timeline, regulatory pathway date or impairment trigger. Both are non-cash-or-unknown today and both are live. Likelihood medium each, impact medium-to-high.

5. Reimbursement — the one that landed, and the one that would end the story. The 2026 MPFS cut is already in the numbers: ~10.5% off the physician fee for cataract and cataract+MIGS, on top of the November-2024 MIGS LCDs capping coverage at one device per eye and demoting MIGS below drops and laser. Management named it directly as a cause of Implantables +1% in Q1-2026. That is a known, bounded, industry-wide hit that strikes Glaukos identically. The tail risk is different in kind. CMS Ruling 05-01 — issued 2005-05-03 — is the administrative foundation of the entire ATIOL private-pay profit pool: it holds that no Medicare benefit category exists for the presbyopia-correcting functionality of an IOL and therefore permits beneficiaries to pay the upgrade out of pocket. It is a ruling, not a statute; it has stood twenty-one years; there is no live proposal to change it. But a reversal would vaporize the premium pool that makes ~20% of US cataract procedures carry a disproportionate share of the industry’s economics. ASSUMPTION-level tail risk: low probability, severe impact — and almost never discussed.

6. The discretionary-consumer exposure of the highest-margin revenue. This is the structural awkwardness underneath several of the above. The base cataract procedure and the monofocal IOL are reimbursed; the ATIOL upgrade is patient-pay, and Vision Care “is primarily private pay, with patients substantially paying for contact lenses and ocular health products out-of-pocket.” Alcon’s growth thesis rests on trading patients up into out-of-pocket premium products — which means its most profitable revenue is also its most consumer-discretionary and the least defended by reimbursement. The 20-F concedes the consequence: price increases “may cause some customers, particularly in elective surgical and contact lens businesses where patients typically do not receive reimbursement, to reduce purchases or choose lower-cost alternatives.” INTERPRETATION: Alcon has price realization in good consumer conditions and very little in bad ones. That is price/mix opportunity contingent on the consumer, not pricing power. Likelihood medium (it requires a consumer downturn), impact high (it hits the premium slice first and hardest).

A note on what is not a top risk. Solvency is not in question: net debt of $3,210M is ~1.2x core EBITDA at a 3.6% blended rate, interest coverage is ~10x, the $1.32B revolver is undrawn and extended to October 2030, and the $1.9B STAAR bridge was cancelled without ever being drawn. Nor is China existential — it is 6% of sales and IOLs ~5%, and the domestic champion Eyebright (~US$205M of revenue, IOL sales +1.4%, taking a goodwill impairment) is a co-victim of the VBP regime rather than a credible global competitor. FX is a real quarterly distortion — Q1-2026 was +6% constant-currency against +10% reported — but it is translation noise against a diversified base, not an economic threat. The risks that actually threaten the thesis are competitive, structural and disclosure-related, not financial.

10. Valuation Discussion — Embedded Expectations

10.1 Three earnings numbers, three different companies

Alcon’s valuation debate is not really about the multiple. It is about the denominator. The same share price supports three defensible-looking multiples that differ by more than 50%, and choosing between them is the analytical work.

At $70.26 (17 July 2026) on 487.43M shares outstanding, market capitalisation is $34,247M. From the FY2025 20-F balance sheet, total financial debts of $4,737M less cash and time deposits of $1,607M give net debt of $3,130M and an enterprise value of $37,377M ($37,886M including $509M of lease liabilities). [FACT — ALC 20-F FY2025, consolidated balance sheet, https://www.sec.gov/Archives/edgar/data/1167379/000116737926000014/alc-20251231.htm, accessed 2026-07-18.] A data-quality note that materially affects anyone screening this name: ROIC.ai’s FY2025 enterprise-value row shows total debt of $429M — that is the lease liability alone and is wrong. Its TTM row (debt $5,253M, cash $1,578M) reconciles to the filing-derived figure within 1.5%. We use the filing throughout.

Why $2.75 is the defensible basis. The 35.6x IFRS multiple is an artifact of 2019 Novartis purchase accounting. Roughly $700M a year of amortisation runs against “marketing know-how” ($238M, gross cost unchanged for years, 100% Surgical) and “technologies” ($22M, net book value now $6M — effectively exhausted) step-ups that Alcon never paid for and never has to replace. Adding those back is correct and we defend it. But management’s core number makes the opposite error. Core also strips $58M of acquisition and integration costs from a company that executed 21 business-development transactions in 2025 — a cost incurred every year is not non-recurring — of which $46M was spent on the STAAR deal that failed and a further $20M in Q1-2026 on LENSAR, which also failed. Core strips $21M of legal provisions from a company carrying an open, unprovisioned DOJ False Claims Act civil investigative demand against total litigation provisions of $18M. Core strips $44M of product-discontinuation write-offs that are the innovation treadmill functioning as designed. Core adds back self-replacing software amortisation ($122M amortised in FY2025 against $122M of additions in the same category in the same year). And new in 2026, core strips “costs associated with efficiency initiatives” — $88M in Q1 alone, which Note 12 of the Q1-2026 interim report confirms is accrued severance, unambiguous cash cost, for a programme whose total size has never been disclosed.

Charge those back and honest owner earnings are $2.75 per share, or $1,364M, arrived at by two fully independent routes that converge within two cents: an accrual route through the core bridge ($2.74) and a cash route from operating cash flow less all capex, less SBC, less a cash-tax normalisation ($2.76). [FACT — computed from ALC 20-F FY2025 Notes 9, 12, 20.] That is 25.5x — roughly 12% dearer than the core framing implies and 28% cheaper than the IFRS headline.

INTERPRETATION: 25.5x honest owner earnings, a 3.91% owner-earnings yield and a 4.19% free-cash-flow yield net of stock compensation is not a distressed multiple. It is an ordinary, full multiple for a business whose core operating income grew 0.6% on 5% revenue growth in FY2025 and whose returns peaked in FY2024. The de-rating has removed the premium. It has not created a discount.

10.2 Own-history percentiles — one clean read, and it does not say what it appears to

The AZI valuation index (re-run 2026-07-18, feed updated 2026-07-17) places Alcon at a composite 22.2nd percentile of its own post-spin history, with P/E 42.58x at the 37.0th percentile, P/B 1.55x at the 24.9th, and P/S 3.28x at the 4.7th.

Two of the three are unusable, and the reason matters more than the fact.

The P/E percentile is not merely depressed — its denominator is unstable. Alcon’s IFRS diluted EPS has run $0.68 (2022) → $1.96 (2023, flattered by a $263M discrete Swiss tax benefit that turned the year’s tax line into a net credit) → $2.05 (2024, flattered by a $57M Ocumension divestment gain) → $1.98 (2025) → $1.65 TTM. A percentile rank built on an earnings series that swings threefold on one-time tax and divestment items is not measuring valuation at all. Ignore the 37th percentile entirely. P/B is compromised too, though less so: book equity of $22,034M is 82.9% composed of goodwill and other intangibles ($18,262M) that Novartis paid for, not Alcon — the ratio measures price against an accounting artifact. Note the contrast: price to tangible book is roughly 9.3x, a very different picture from 1.55x.

P/S at the 4.7th percentile is the only clean read. Revenue is stable and uninterrupted by one-time items across the full post-spin history. Alcon is within a whisker of the cheapest it has ever been on sales.

And now the pressure test, because that fact is routinely mis-read as a conclusion. A stock at the 4.7th percentile of its own price-to-sales range is cheap only if the margin and return structure that historically justified the higher multiple is intact. Be precise about what is and is not wrong. What is not wrong: consolidated core operating margin has been range-bound — 19.7% (2023), 20.6% (2024), 19.8% (2025) — and core gross margin was essentially flat at 62.71% against 62.80%. The crude “margins collapsed, hence the low multiple” argument is not supported. What is wrong is worse. A price-to-sales multiple capitalises future margin multiplied by growth, and it is the growth of profit that has gone to zero. FY2025 revenue rose $483M and core operating income rose $12M — an incremental core operating margin of 2.5% against a 19.8% average. Core operating income has gone $1,849M → $2,027M → $2,039M: +9.6%, then +0.6%. Core ROE peaked at 7.03% in FY2024 and is now 6.90%; honest tangible ROIC peaked at 21.8% in FY2024 and is now 20.5%. Segment-level, Surgical’s contribution margin fell 197bp in two years (27.4% → 25.4%), 100% of it in gross margin, because the highest-margin line — Implantables — has been flat at roughly $1.78B for three years.

VERDICT on the percentile: the 4.7th-percentile P/S is a rational re-rating, not a self-evident mispricing. The market is not paying less for the same economics; it has lowered its terminal growth and terminal-margin assumption on a revenue base that now converts incremental sales at one-eighth of its average margin. The honest formulation: Alcon has never been cheaper on sales, and it has also never converted sales into profit growth this poorly. This is own-history context only — never cross-sectional, and never a target.

10.3 Embedded expectations — what $70.26 actually requires

Cost of capital, stated and defended. The most defensible anchor is inside the filing. The FY2025 20-F Note 9 discloses management’s own post-tax discount rates for goodwill impairment testing: Surgical 8.5%, Vision Care 8.0%, with 3.0% terminal growth. Revenue-weighted (55.7% / 44.3%) that is 8.28%. This is Alcon’s own audited view of its cost of capital applied to its own assets. A CAPM cross-check on a 0.708 beta, a ~4.2% risk-free rate and a ~5.0% equity risk premium (both conventional ASSUMPTIONS, not filing-sourced) gives a 7.74% cost of equity and a 7.16% WACC. We use a cost-of-equity band of 7.5%–8.5%, base 8.0%, and lean to the upper half for two evidenced reasons: the company’s own impairment rates exceed the CAPM output, and the 0.708 beta understates the risk an owner actually bears — 55–57% of Alcon’s return variance is idiosyncratic, realised volatility has been 27.2% annualised over five years, and the deepest drawdown of Alcon’s independent public life (-37.9%) troughed on 11 May 2026. A low market beta earned by being uncorrelated while falling is not evidence of low risk.

For the avoidance of doubt, the 4.15% WACC that circulated in screener-derived commentary on Alcon is explicitly rejected here. It sits below Alcon’s own 3.6% weighted-average cost of debt before any equity risk premium, which is arithmetically indefensible for a levered equity. It appears nowhere in this report.

The reverse DCF. Solving a Gordon-growth construction at the equity level (owner earnings are already post-interest, post-tax, post-all-capex and post-SBC, so the cost of equity is the correct discount rate):

At $70.26 the market is underwriting roughly 3.1%–4.4% perpetual nominal growth in owner earnings, centred on ~3.9% at an 8% cost of equity. Against a 2–2.5% long-run inflation assumption that is 1.5–2% real growth in perpetuity, with no margin expansion beyond what is embedded and no re-rating.

The second reading is the more important one: the market is not underwriting the FY2026 guide — it is refusing it. Take management’s guided FY2026 core EPS of ~$3.42 entirely at face value and treat it as true earnings power. The current price then implies only ~3.0% perpetual growth thereafter at an 8% cost of equity, and 2.2% at the CAPM rate. The market will pay for one year of the guide and then prices sub-inflation growth forever. Read the other way, on the honest $2.75 base the price implies ~3.9% forever — below the 3–4% aggregate eye-care market growth management itself assumes for FY2026, below the 4–5% Vision Care and 4–6% Surgical long-run TAM growth published in the 20-F, and far below the +10–13% core EPS the company is guiding. The stock is priced for Alcon to grow slightly slower than its own market, in perpetuity.

That is a low bar. It is the single strongest argument against an outright negative view, and it is why this is a genuine debate rather than a rout. A useful way to hold the trade-off: at 3.0% perpetual growth and an 8% cost of equity the price supports owner earnings of roughly $2.63; to justify the price on 2.0% perpetual growth would require roughly $3.16 of owner earnings, about +15% from here, or roughly 300bp of honest operating margin (17.85% → ~20.9%). The market is underwriting either ~4% perpetual growth on today’s owner earnings, or ~2% growth plus full delivery of the guided margin expansion. It is not underwriting both.

10.4 The guidance internal-consistency test

This is an original finding and it deserves room, because it explains the refusal above better than sentiment does.

Take the FY2026 guide on its own terms: +5–7% constant-currency sales, +70–170bp of core operating margin expansion, +10–13% core diluted EPS growth, roughly 492M diluted shares, ~$210M of non-operating expense, ~20% core tax rate. On approximately $10.84B of FY2026 revenue, solve for the core operating margin required to hit each core-EPS outcome:

The two guidance lines are not independently satisfiable. Delivering the bottom of the margin guide produces core EPS growth of roughly 6.5% — below the bottom of the EPS guide. The +10–13% EPS line requires core operating margin of at least ~21.2%, the top third of the +70–170bp range, roughly +140bp or better. This is a company that delivered +0.6% core operating income growth last year on a 2.5% incremental margin. The margin guide is not a range with an EPS range attached; the EPS guide is effectively a bet on the top of the margin range, and the framing obscures that.

Three further facts compound it. First, management has said the majority of the margin expansion falls in the second half — a back-end-loaded guide is an execution-risk guide. Second, the same management team cut guidance twice in 2025: February’s +6–8% cc sales / 21–22% core margin / $3.15–3.25 core EPS became May’s +6–7% / 20–21% / $3.05–3.15, then August’s +4–5% / 19.5–20.5% / $3.05–3.15 — and the EPS dollar range was held across the August cut only by simultaneously dropping the assumed core tax rate from ~20% to ~18%, worth roughly $0.07–0.08. FY2025 landed at $3.07 on a 19.8% core margin, below even the twice-cut range’s midpoint. Third, Q1-2026’s margin expansion came entirely from SG&A leverage (+67bp) while core gross margin fell 26bp — the expansion is being delivered by cutting people, and the severance that funds it sits below the core line in a programme of undisclosed size.

INTERPRETATION: demanding delivery before paying for this guide is rational, not myopic.

10.5 Comparables — and they kill the “quality at a discount” framing

Note first that the factor-similar peer set is useless here as a comp table: the model returns ETFs (IHI, XHE, RSPH) and European dividend-growth funds, plus only SYK, AVY and ZTS, and not one ophthalmic competitor. That is a positioning fact, addressed in Section 11, not a valuation input.

The premium to Bausch + Lomb is clearly earned. Alcon at 3.62x EV/sales against BLCO’s 2.02x is a 79% premium; B+L earns a 3.7% operating margin, lost $360M in FY2025, and carries debt at 122% of total capital against Alcon’s 1.19x net-debt/core-EBITDA and 10x interest coverage. No issue.

The positioning versus CooperVision is not “in line” — Alcon trades at a premium, and it is hard to defend. On EV/sales it is 3.62x against 3.47x. On adjusted earnings the gap is far wider: COO at roughly 16x forward adjusted EPS against Alcon at 22.9x core and 25.5x honest owner earnings. Both stocks sit at roughly the same point in their own history (COO ~5th percentile P/S, ALC 4.7th). But CooperVision earns a 26.6% segment operating margin in contact lenses against Alcon’s 21.5% Vision Care contribution margin, on 4.2% R&D intensity against Alcon’s 9.6% — Alcon spends 2.3x the R&D intensity, earns roughly five points less margin in the same category, and its own compensation report concedes it lost lens share to Cooper and to private label. Paying a 40–55% premium on adjusted earnings for the lower-margin, share-losing participant in the same four-firm oligopoly is the least defensible relative position in this table.

The premium to ResMed on operating earnings is the sharpest comparison here. Alcon trades at 20.3x honest EV/EBIT — 18.3x even on the generous core basis — against ResMed at 16.7x. ResMed earns a 32.8% operating margin and ~22% ROIC against Alcon’s 19.8% core margin and 20.5% tangible ROIC (4.4% on book), and ResMed’s 21.6x P/E rests on clean earnings requiring no add-back argument at all, versus Alcon’s 25.5x on a figure that took two independent bridges to construct. Alcon is more expensive than ResMed on operating earnings and on honest net earnings, while growing slower, earning less, and requiring more accounting charity. Against STERIS — a genuine #1 sterilisation franchise at 21.1x EV/EBIT and ~19x forward adjusted — Alcon is at parity despite materially worse incremental economics. Only against Zimmer Biomet (17.6x EV/EBIT, ~10x adjusted, ROIC below WACC) does Alcon look expensive-but-better, and ZBH is the comparable for a business that does not earn its cost of capital.

Synthesis: Alcon is priced above the ophthalmic peer with better lens economics, above the quality-medtech peer whose margins and returns are 50–60% higher, at parity with a genuine wide-moat #1, and above the sector’s clearest capital-destroyer by more than the quality gap warrants. The “quality medtech at a discount” framing does not survive contact with the actual quality-medtech comp set.

10.6 Scenarios — expressed as owner earnings and the multiple $70.26 implies

Common assumptions across all three: non-operating expense $205–210M per guidance; 20% core tax rate; recurring “core” add-backs an owner must charge back of roughly $155M a year (acquisition/integration ~$58M, legal ~$21M, product discontinuation ~$44M, software and post-spin intangible amortisation ~$34M); net share shrink from the $1.5B three-year authorisation less roughly 2.1M shares a year of SBC dilution. No inorganic contribution is modelled in any case — STAAR was terminated on 6 January 2026 and LENSAR on 16 March 2026, and the FTC’s stated theory in the LENSAR matter forecloses further consolidation of surgical adjacencies. All owner-EPS figures are ASSUMPTIONS built on FACT-labelled inputs.

BEAR — what must go wrong, each item evidenced in this file rather than invented: Implantables stay flat or decline as J&J’s TECNIS Odyssey, “the fastest-growing intra-ocular lens in the U.S.,” compounds against Alcon’s +0.4% in FY2025 and +1% in Q1-2026; the ~$1.04B, sixteen-line contact-lens capacity build lands into a price war as Alcon, CooperVision and B+L all add high-fixed-cost silicone-hydrogel capacity into a category the leader now describes as growing at “the low end of mid-single digits,” with private label taking the marginal fill; the efficiency programme proves recurring rather than one-time, so owner earnings go flat-to-down while core EPS is reported up 10–13%; the mid-2026 China IOL volume-based-procurement round reaches into the premium tier on top of successive cuts of roughly 20%, 26%, 38%, 53% and 84%; and either the open, unprovisioned DOJ investigation or the $820M of Aurion acquired IPR&D — PwC’s sole critical audit matter, resting on an unapproved corneal cell therapy that produced $12M of sales and a $37M loss — crystallises. Note the asymmetry on Aurion: an impairment flows through the “impairments” core add-back and would not touch core EPS at all. The reported number would be untouched by an $820M destruction of capital. That is a feature of the core framework worth naming.

At the bear path, $70.26 is 28.7x FY2028 owner earnings — a higher multiple of terminal earnings than of today’s. The bear case is not “the multiple compresses.” It is “there is no multiple at which flat owner earnings and no growth is interesting.”

BASE: mid-single-digit revenue consistent with the 3–4% market growth management assumes plus modest price/mix; core margin reaching the floor of the FY2026 guide then building slowly as tariffs anniversary and the R&D step-up flattens; the efficiency programme totalling roughly $300M across two years and genuinely rolling off (for scale, the 2019–2023 transformation programme cost $425M all-in for $300–325M of run-rate savings); buyback executing at roughly $500M a year. Owner earnings then compound at 9–10% a year — but note carefully that FY2026 owner earnings of $2.72 are flat against FY2025’s $2.75. The base case is a lost year followed by recovery, not a clean compounding path.

BULL — and it must now be entirely organic: the Unity CS/VCS console conversion delivers consumables pull-through at price rather than merely holding units; PanOptix Pro stabilises and then regains US trifocal ATIOL share against TECNIS Odyssey, converting the ~20–21% US and ~17% global penetration runway toward management’s stated 35–38% ceiling; TRYPTYR reaches its guided $250–400M peak (it has taken roughly four share points in eight months with ~55% commercial lives covered, Medicare Part D still pending); the efficiency programme is genuinely one-time and core margin reaches 22%+; and Vision Care’s demonstrated 39.8% incremental contribution margin carries the consolidated line. Note what is absent: no M&A contribution is available to the bull case any more.

The dispersion is the finding. FY2028 owner EPS spans $2.45 to $4.01 — a 64% range, unusually wide for a defensive medtech — and essentially all of it is controlled by two variables the company will not disclose or has not yet demonstrated: the total size and duration of the efficiency programme, and whether Implantables can grow again. Neither is a macro variable. Moving the FY2026 programme from $150M to $350M alone swings owner EPS by roughly $0.33, or 12%. This is a stock whose outcome is almost entirely in management’s hands and almost entirely unauditable from outside.

10.7 Sum-of-the-parts — and the inversion it surfaces

At defensible peer multiples: Surgical at 2.6–3.2x sales (bracketing B+L Surgical’s low multiple and STERIS’s 3.94x, adjusted down for flat implantables) gives $15.0–18.4B; Vision Care at 2.8–3.4x sales (bracketing CooperVision’s 3.47x, adjusted for Alcon’s five-point-lower segment margin) gives $12.8–15.5B; less true corporate cost of $359M pre-tax, roughly $296M post-tax capitalised at 8%, or $(3.7)B. The sum is $24.0–30.2B against a traded EV of $37.4B.

A sum-of-the-parts is usually run in the hope of finding a conglomerate discount. Alcon has the reverse: there is no hidden value to unlock. The traded EV is supported only by applying multiples above the ophthalmic peer set to at least one segment — the same conclusion Section 10.5 reached from a different direction. The corporate block is real, large and permanent at roughly 15% of core operating income, and there is no break-up, spin or separation argument here. This is an order-of-magnitude test, not a valuation; Alcon reports no segment EBITDA, capital employed or capex, so no segment return can be computed.

But the exercise surfaces something that matters more than the arithmetic: the segment hierarchy is inverted. The conventional narrative — and the late-2025 consensus framing discussed in Section 11 — is that Surgical is the moat and Vision Care is the commodity. The segment P&L says the reverse is currently true. Over two years Surgical grew revenue $437M and produced $6M of incremental contribution: a 1.4% incremental margin. Over the same two years Vision Care grew $512M and produced $204M: a 39.8% incremental margin, with gross margin up 175bp. The segment narrated as the moat is the one not converting growth into profit. Two consequences: the consolidated multiple is averaging a decelerating asset with an accelerating one and hiding both, and the ~$1.04B contact-lens capacity build is capex into the better business, not the worse one — which softens, without eliminating, the capital-cycle warning. OPEN QUESTION: this inversion is only two years old and may be a tariff-and-mix artifact rather than a structural reversal. Present it as an evidenced observation, not an established new hierarchy.

10.8 What the market is pricing correctly, and incorrectly

Correctly — the label. The market is repricing a good-not-great compounder that was floated on a great-business narrative, and this is the strongest conclusion in the file because three independent methods reach it: the 20-F (share targets missed in three of four measured franchises; Implantables +0.4% against J&J Surgical Vision +10.2%), the P&L (2.5% incremental core margin; returns peaked FY2024), and a statistical factor model that assigns Alcon a Quality beta of +0.03 to +0.08 — effectively declining to certify the quality label. When the tape and the fundamentals agree, the contrarian case is weak by construction. Roughly 3.9% implied perpetual growth is about right for a business converting incremental revenue at one-eighth of its average margin.

Correctly — the refusal to underwrite the guide. Given the February→May→August 2025 cut sequence, an EPS line that requires the top third of its own margin range, 2H weighting, expansion sourced entirely from SG&A while gross margin falls, and an undisclosed-size severance programme, demanding delivery before paying is correct.

Correctly — M&A optionality at approximately zero. Alcon went 0-for-2 on public deals in one year at a cost of roughly $76M expensed for zero assets acquired, and a regulator has now asserted on the record that consolidating surgical adjacencies is anticompetitive.

Possibly incorrectly — three candidates, carried into Section 11: the segment inversion (Vision Care’s 39.8% incremental margin against Surgical’s 1.4%) does not appear to be in the price; the medical-device industry factor is at an extreme the stock’s own fundamentals do not fully explain; and the efficiency-programme asymmetry is unpriced in both directions — a roughly 12% swing in economic earnings resting on a number the company has simply not disclosed.

The characterisation we land on is neither “value opportunity” nor “value trap” but something less comfortable: a low-return-on-price compounder. The trap evidence is genuinely strong — negative Sharpe at every horizon out to five years, 27% annualised volatility, a -37.9% drawdown that troughed nine weeks ago, no dedicated value or dividend constituency, and a 4.7th-percentile price-to-sales that is explained by zero profit growth rather than unexplained. But the classic trap signature is absent: this is not a melting asset. It earns ~20.5% on tangible capital, carries 1.19x net leverage and 10x interest coverage at a 3.6% blended cost of debt termed to 2052, converts more than 100% of core net income to cash, has no customer concentration, and shows no channel-stuffing or inventory tell behind the flat implantables line. At 25.5x honest owner earnings and ~3.9% implied perpetual growth, an owner’s expected return is roughly the earnings yield plus growth — arithmetic that is unexciting rather than dangerous. The risk here is opportunity cost and time, not permanent impairment — with the genuine tail exceptions of the unprovisioned DOJ investigation and the $820M of un-impaired Aurion IPR&D.

Section verdict. This is a debate about delivery, not about multiple. The embedded bar is low enough that credible execution against the guide would be worth a great deal. The bear case does not require a collapse; it requires only that the efficiency programme prove recurring and Implantables stay flat, at which point owner earnings go sideways and the current price is 28.7x terminal earnings that never grew.

11. Variant Perception

11.1 What consensus believed — and its own bear case came true

No compiled sell-side estimate set was obtainable for this analysis, and that gap should be stated rather than papered over. What can be reconstructed is the prevailing sell-side and market framing of Alcon as it stood in late 2025 — roughly three weeks before the STAAR termination, and before the de-rating had completed. It is referenced here for its framing only, never for a number.

That framing ran as follows: the moat was described as wide; the pitch was “quality at a reasonable price”; the Unity platform refresh was treated as a supercycle; the base case clustered around $85–95, with a bull case above $105 and a bear case below $70; returns were quoted from screener data as ROIC ~7.35% against a WACC of ~4.15% and ROE ~11.8–12%; and the advice was to hold through what would “likely be a transition year,” with success on the STAAR acquisition treated as a catalyst that would accrete immediately to Alcon’s growth profile.

Two observations, and the second is the useful one.

First, the return metrics do not reconcile to the filing. FY2025 IFRS net income of $980M on $22,034M of shareholders’ equity is an ROE of 4.45%, not 11.8–12%. The 4.15% WACC is below Alcon’s own 3.6% weighted-average cost of debt before any equity risk premium. The wide-moat label was asserted, not tested — no share-stability test was run, and Alcon’s own FY2025 compensation disclosure concedes missed share-gain targets in three of four measured franchises. None of these figures carries into this report.

Second, and more instructive: that framing’s own stated bear case is exactly what came to pass. Its bear scenario was a STAAR deal collapse leading to capital-allocation uncertainty, with a sub-$70 price. The deal collapsed on 6 January 2026 when STAAR shareholders declined to approve the raised $30.75 bid. LENSAR followed on 16 March 2026 on FTC opposition. The stock trades at $70.26. INTERPRETATION: consensus was not wrong about the risk. It identified the correct bear case, assigned it a low weight, and was overwhelmed by a “wide moat” label it had not tested and return metrics it had not reconciled. That is the most useful lesson available here — not what to think about Alcon, but how quickly a quality label decays when it is asserted rather than measured.

11.2 The factor-positioning read — where consensus is offsides, and where it is not

The empirical positioning evidence is unusually clean, and it cuts against the easy contrarian story.

The model zeroes Alcon’s Value beta in all four nested models — and its DividendYield beta too. Over the trailing year, Value returned +14.3% (z +1.70) and DividendYield +16.8% (z +1.66): the two most extreme in-favour factors in the entire model. Alcon carries neither. Momentum is negative at -0.21 while the Momentum factor itself returned +15.5% (z +0.67). The three best-paying style factors of the last twelve months were each a headwind or an absence for this stock. INTERPRETATION: cheap against its own history is not the same as being owned by the cohort that gets paid when cheapness works. The dedicated value buyer has not arrived, and that fact should discipline any bull case built on the 4.7th-percentile price-to-sales.

The model independently declines to certify the quality label. Quality beta is +0.03 in the base model, +0.08 with sector and industry, and is zeroed entirely in the full model — statistically indistinguishable from no quality signature at all. Arrived at from a purely statistical direction, that is the same verdict the fundamental work reached from the 20-F. For a contrarian process this is the uncomfortable result, not the exciting one: there is no evident gap between what the market believes and what the evidence shows.

The regime is a headwind on the axis that matters. Alcon’s second-largest non-market loading is Industry: Medical Devices at +0.27 — the worst-performing industry basket in the model, -18.0% over 252 days at z -1.22 and still deteriorating on the 126-day window — while Health Care as a sector returned +8.7% (z +0.79). Money has been in healthcare and out of devices. Roughly a quarter of Alcon’s de-rating is a cohort effect, not a company effect. A capital-cycle reading says an industry basket at that extreme is where capital exits and supply eventually tightens; the counterweight, which must be presented alongside and not resolved as a bull point, is that supply is still expanding — roughly $1B of new contact-lens capacity going in simultaneously at Alcon, CooperVision and Bausch + Lomb. Out-of-favour tape plus still-expanding supply is the worst of both. Separately, LowVolatility (+0.34, Alcon’s largest style loading) paid just +1.46% (z -0.29) over the year: being a 0.708-beta defensive in a +20% market was pure opportunity cost.

The track record is unambiguous. The Sharpe ratio is negative at every available horizon out to five years (-1.22 at three months through -0.06 at five years), on 27.2% annualised volatility. Cumulatively Alcon has returned +3.4% over five years against the S&P 500’s +84.4% — -81 percentage points — and since the April 2019 spin +23.6% against +187.9%, -164 percentage points. Maximum drawdown since the spin is -37.9%, with the trough on 11 May 2026 — the deepest of Alcon’s independent public life, deeper than COVID, and nine weeks old at this report date.

And the positioning result is a negative one, which is the most quotable finding here. Alcon’s factor-similar peers are ETFs — IHI, XHE, RSPH — plus European and international quality-dividend-growth funds, with only SYK, AVY and ZTS appearing as individual stocks. Not one of JNJ, COO, BLCO, GKOS, STAA or Carl Zeiss Meditec appears anywhere in the list. Alcon does not trade as an ophthalmic pure-play; it trades as a passive proxy for international large-cap medical devices. The ownership register corroborates: BlackRock 6.05%, UBS entities roughly 5.3%, otherwise diffuse, no activist and no strategic block, and no opt-out from the Swiss mandatory-takeover regime.

INTERPRETATION: this is what “orphaned” looks like, and it cuts both ways. There is no crowded position to unwind and no forced seller — the de-rating is not a positioning accident awaiting an unwind. But there is equally no informed marginal buyer doing the work, no dedicated value constituency, and nobody on the register paid to be the catalyst. A recovery, if it comes, will have to be earned by the P&L rather than delivered by the tape. That is consistent with the idiosyncratic-risk evidence: 55–57% of Alcon’s return variance is company-specific, and the two defining sessions of its recent history — -11.7% on 6 May 2026 (the largest single-day decline in its post-spin history) and -10.1% on 20 August 2025 — were almost purely idiosyncratic, administered by guidance.

11.3 The strongest bull case, argued in good faith

The bull case is better than the price action suggests, and it should be stated at full strength.

The amortisation is genuinely non-economic. Roughly $700M a year runs against Novartis-era step-ups Alcon never paid for and never has to replace; “technologies” is down to a $6M net book value and “marketing know-how” has an unchanged gross cost. Adding that back is not accounting charity, it is arithmetic. On the capital actually employed, the operating asset earns roughly 20.5% — the middle of the “advantages present” band, not the floor, once numerator and denominator are made consistent.

The balance sheet is a genuine asset. Net debt of 1.19x core EBITDA, 10x interest coverage, a 3.6% blended cost of debt termed out to 2052, a $1.32B revolver undrawn and extended to October 2030, and roughly $3.45B of headroom to a still-conservative 2.5x. Alcon will not be forced into anything.

The buyback is counter-cyclical and mostly real. A three-year $750M programme was completed in ten months, with volume accelerating 5.3x from April at $90.49 to September at $77.41, and roughly 77% of the spend genuinely retired stock (two independent methods agree). A new $1.5B authorisation followed.

The organic pipeline is not nothing. TRYPTYR is a real, approved, differentiated Rx dry-eye asset that took roughly four share points in eight months with Medicare Part D still pending and a $250–400M peak-sales guide. Unity CS/VCS replaces platforms dating to 2008 and 2013 and, if it converts the installed base, is a decade-long consumables annuity attached to a 7–10 year capital cycle in which 83.6% of Surgical revenue is already recurring. ATIOL penetration at roughly 20–21% in the US against a management-stated 35–38% ceiling, and 17% globally, is a genuine multi-year runway in a category protected from payors by a twenty-one-year-old CMS ruling.

And the strongest version of the bull case is about the efficiency programme. If it is genuinely one-time — the prior transformation programme cost $425M for $300–325M of run-rate savings and did roll off — then honest owner earnings rise roughly 6% in FY2026 and the reported number is roughly honest. The market is then over-extrapolating a single bad guidance year in a demographically advantaged category, at an embedded growth rate below the growth of Alcon’s own market.

11.4 The strongest bear case

The quality label was always wrong, and both the fundamentals and the factor model now say so independently. Share targets were missed in three of four measured franchises over a full three-year cycle; the one win was fluidics cassette packs, the lowest-value category, bought with a new equipment launch.

Share is being lost in the highest-margin line. Implantables ran $1,703M → $1,775M → $1,782M — +0.4% in FY2025, +1% in Q1-2026 — while J&J Surgical Vision grew +10.2% on the back of TECNIS Odyssey, “the fastest-growing intra-ocular lens in the U.S.” Every named competitor grew its comparable line faster than Alcon in FY2025. There is no channel-stuffing or inventory tell behind the stall, which makes it harder to dismiss, not easier.

Operating leverage is absent. $483M of incremental revenue produced $12M of incremental core operating profit. Surgical’s 197bp two-year contribution-margin decline is 100% gross margin, driven by the flat high-margin line. R&D has risen 19.6% in two years to 9.6% of sales — 2.3x CooperVision’s intensity — while core operating income rose 0.6%, and the 20-F itself describes the mechanism as a treadmill: “as these products age and/or competitive products advance, prices typically trend downward, requiring continuous innovation cycles to maintain and/or grow our margins.”

The capacity build is late-cycle. Roughly $1.04B and sixteen new contact-lens lines running to 2030, into a category whose growth the leader itself has marked down, alongside simultaneous builds by CooperVision and B+L, in the precise franchise where Alcon concedes it lost share to private label. No target utilisation, incremental unit capacity or hurdle rate is disclosed.

Incentives contain no return-on-capital metric of any kind. The short-term plan pays on Net Sales, Core Operating Income and Free Cash Flow; the long-term plan on Net Sales CAGR, Core Diluted EPS CAGR, Share of Peers and Innovation. No ROIC, ROE or economic profit, in either plan, in either the completed 2023–2025 cycle or the live 2025–2027 cycle. In the one cycle that can now be audited, shareholders received a 58th-percentile TSR and 3.3%/yr core operating income growth; the single metric measuring competitive outcome — market share — paid 35%; and the award nonetheless paid 137%, because a Core Diluted EPS CAGR reported at 21.1% paid the capped 200% against roughly 11.1% computable from the audited core-EPS series, with no base year and no reconciliation disclosed.

And the price is not a discount. 25.5x honest owner earnings for a business compounding profit at approximately 0%, at a premium to CooperVision on adjusted earnings and to ResMed on operating earnings, is a full multiple. Add an unprovisioned DOJ False Claims Act investigation, $820M of un-impaired Aurion IPR&D whose write-down would not touch core EPS, and a five-year record of negative Sharpe at every horizon.

11.5 The assumptions that actually matter — and what would falsify each side

The honest summary: on the label question — is Alcon a quality compounder? — the tape, the factor model and the fundamentals all agree that it is not, so there is no consensus gap to exploit there. The genuine variant-perception candidates are narrower and more specific: the segment inversion (Vision Care is currently the better business and is not priced as such), the medical-device industry factor at a -1.22 z-score while the parent sector is positive, and an efficiency programme whose size is unpriced in both directions. Each is checkable at the next print, and each is listed in Section 14.

12. Fact vs. Interpretation

13. Open Questions

1. The size, duration and savings target of the efficiency programme announced 2026-02-24 — the single most important undisclosed number in this file. $88M was charged in Q1-2026 alone, $70M of it provisioned as severance, and no total, no savings target and no duration appear anywhere in the corpus. For scale, the 2019–2023 transformation programme cost $425M all-in for $300–325M of run-rate savings. Moving the FY2026 figure from $150M to $350M swings honest owner EPS by roughly $0.33, or 12% — the difference between honest earnings rising 6% and falling 6% in a year the reported number is guided up 10–13%.

2. Does Unity convert the installed base at price, or merely hold units at flat ASP? Equipment revenue of +6% in FY2025 and +23% in Q1-2026 is consistent with either. This cannot be settled from the outside, and the reason is itself a finding: Alcon claims the industry’s largest installed base of phacoemulsification and vitrectomy consoles but has never disclosed an absolute unit count. Without units, “largest installed base” cannot be converted into revenue-per-console productivity — the cleanest available test of the razor/razorblade moat. Name this disclosure gap.

3. The DOJ Civil Investigative Demand under the False Claims Act, received July 2024 and relating to discounts on surgical equipment servicing contracts, remains open with no quantified provision — total litigation provisions are $18M at 2025-12-31. An unquantified tail risk to both earnings and cash.

4. Aurion’s $820M of acquired IPR&D — PwC’s sole critical audit matter — rests on an unapproved corneal cell therapy with no disclosed regulatory pathway date, milestone timeline or impairment trigger. No impairment has been taken. A write-down would be roughly 40% of a year’s core operating income and would bypass core EPS entirely.

5. Can the FY2025 Capital Markets Day medium-term targets be sourced? The event reset the guidance framework used from FY2026 onward but was not furnished on a 6-K and is absent from every publicly available transcript corpus. Any medium-term target quoted must come from the company’s investor site or be omitted.

6. Does Alcon return to phakic IOLs after STAAR? Endicott’s stated fallback is that “our refractive strategy is unchanged and our new wavelight plus offering remains our focus” — i.e. defending a LASIK franchise that is itself under lenticule-extraction attack from J&J’s ELITA/SILK and Zeiss’s SMILE, having just failed to buy the phakic alternative. Nothing in the filings addresses whether Alcon re-approaches the category organically or otherwise.

7. The Simbrinza Hatch-Waxman appeal. Alcon lost at trial on infringement in February 2025 (D. Del.); both sides appealed and briefing concluded in January 2026. Outcome unknown; generic exposure is live.

8. Is the ~$1.04B, sixteen-line contact-lens capacity build demand-led or share-defensive? Alcon discloses no target utilisation, no incremental unit capacity and no hurdle rate, so the question cannot be settled from the filings — in a category where the company’s own compensation report concedes it lost share “primarily due to competitive pressure in the Dailies category, as well as growth of private label.”

9. The Core Diluted EPS CAGR reconciliation. The 2023–2025 award paid the capped 200% on a reported 21.1% CAGR against roughly 11.1% computable from Alcon’s audited core-EPS series. No base year and no bridge are disclosed. Related and also undisclosed: whether the metric is neutralised for share repurchases.

10. The mid-2026 China IOL volume-based-procurement round. Management flagged that it “rolls around again in the middle part of the year.” Its terms are not in the corpus and it affects roughly 5% of sales.

11. Whether the Vision Care / Surgical incremental-margin inversion is structural or a two-year tariff-and-mix artifact. This is the most interesting unresolved question the valuation work surfaced.

12. Succession of the Audit & Risk Committee chair. The founding director chairing that committee is not standing for re-election, at the precise moment three unquantified accounting and legal exposures are live (items 1, 3 and 4 above).

14. What Must Be True

Each proposition below is paired with a specific, dated, observable falsification test — an actual line item or disclosure in a named future filing or quarter.

The bull case requires all five

BULL 1 — The efficiency programme is one-time, not recurring opex relabelled as restructuring. FALSIFICATION: if “costs associated with efficiency initiatives” appears again in the Q3-2026 and Q4-2026 core reconciliations at a run-rate above ~$50M per quarter, or if the cumulative FY2026 charge exceeds ~$300M without a disclosed total programme size, it is recurring operating cost excluded from the headline metric, and honest owner earnings are nearer $2.55 than $2.90.

BULL 2 — Implantables can return to at least market growth. FALSIFICATION: if the FY2026 20-F net-sales table (filed ~February 2027) shows Implantables growth at or below ~1% — a fourth consecutive year of a flat ~$1.78B line — while J&J’s FY2026 Surgical Vision line again compounds at double digits, the organic bull case is dead. With STAAR and LENSAR both terminated and the FTC blocking adjacency consolidation, there is no inorganic substitute available.

BULL 3 — Core operating margin reaches the top third of its guided range, and does so on gross margin rather than headcount. FALSIFICATION: if the Q3-2026 press release shows core operating margin tracking at or below ~20.8% year-to-date, or if core gross margin declines year-over-year in three of the four FY2026 quarters while core operating margin expands, then the guided +10–13% EPS is unreachable on the arithmetic in Section 10.4 and the expansion that is occurring is severance-funded rather than structural.

BULL 4 — Unity converts the installed base at price. FALSIFICATION: if Surgical segment gross margin (20-F Note 3, FY2026, filed ~February 2027) declines for a third consecutive year while Equipment revenue growth decelerates from its Q1-2026 +23%, the console refresh has been sold on units and discount rather than on price, and the consumables annuity it was supposed to renew will not carry the segment.

BULL 5 — The amortisation add-back is not being continuously rebuilt by acquisition. FALSIFICATION: if the FY2026 unallocated cost of net sales (Note 3) rises again above the FY2025 $(776)M, and goodwill plus other intangibles rise again from $18,262M, the wedge between reported and economic earnings is widening as a direct consequence of the deal programme, and the add-back is a rolling subsidy rather than a fixed spin-era artifact.

The bear case requires all four

BEAR 1 — Reported earnings growth is running ahead of economic earnings growth. FALSIFICATION: if Alcon discloses a total efficiency-programme cost of ~$200M or less (in the Q2-2026 or Q3-2026 release, or the FY2026 20-F) and core gross margin expands year-over-year in at least two of the four FY2026 quarters, the thesis breaks and honest owner earnings genuinely compound rather than merely appearing to.

BEAR 2 — The high-margin franchises are structurally losing share, not cyclically pausing. FALSIFICATION: if the FY2026 or FY2027 Compensation Report “Share of Peers” disclosure records a share gain in ATIOLs and in contact lenses — the two categories management has conceded missing — the share-loss thesis is falsified by management’s own third-party-measured metric, which is the same metric that convicted it.

BEAR 3 — Operating leverage is structurally absent. FALSIFICATION: if FY2026 incremental core operating margin exceeds ~15% (i.e. core operating income grows by more than ~$75M on ~$500M of incremental revenue), against 2.5% in FY2025 and 1.4% in Surgical over two years, the “scale does not convert” verdict is wrong and the flat-profit period was a tariff-and-launch-spend trough rather than a structural condition.

BEAR 4 — Capital allocation is pointed at the wrong target because nothing measures return on capital. FALSIFICATION: if the 2027 Compensation Report introduces a return-on-capital metric — ROIC, ROTC or economic profit — into either the short-term or long-term plan, or if a shareholder large enough to insist appears on the register (a >5% non-index holder or an activist filing), the structural indifference this memo identifies as the root cause has been addressed at source. Absent that, the base case is that the pattern continues.

A note on symmetry. BULL 1 and BEAR 1 are the same disclosure viewed from opposite sides, and that is deliberate: a single undisclosed number — the total size of the efficiency programme — controls roughly a 12% swing in FY2026 economic earnings and is the largest single driver of the 64% dispersion in the FY2028 scenario range. It is unpriced in both directions, it is not a macro variable, and it should be the first question asked of management.

APPENDIX A — Standard Diligence Questionnaire

Supplemental to the report on Alcon Inc. (NYSE/SIX: ALC), report date 2026-07-18, reference price $70.26. Alcon is a Swiss foreign private issuer reporting under IFRS as issued by the IASB — every reference to reported earnings below is IFRS, not US GAAP.

General

What thoughtful questions have other investors asked about this company?

(1) Was the STAAR bid a good use of $1.8B? Broadwood Partners, STAAR’s largest holder at ~31%, ran a public campaign (opposition filings 2025-11-04, 2025-12-09) arguing Alcon was buying a China-VBP casualty at the bottom of its cycle and paying nothing for the recovery. Broadwood won: STAAR holders declined to approve even the raised $30.75 bid, and Alcon terminated 2026-01-06. FACT: STAAR’s Q1-2026 preliminary net sales exceeded $90M against $239.4M for all of FY2025. INTERPRETATION: on the evidence, Broadwood was right and Alcon lost an asset it had correctly identified.

(2) Is the spin amortization economic? The central dispute. FACT: goodwill $9,256M plus other intangibles $9,006M = $18,262M, 82.9% of shareholders’ equity, stepped up by Novartis at the 2019 separation. Of FY2025’s $784M of intangible amortization, “marketing know-how” ($238M) and “technologies” ($22M) are pure spin-era step-ups whose gross cost has not moved in years — adding them back is correct. But ~$630M (~9%) of “currently marketed products” is post-spin, bought with Alcon’s cash; and “other intangibles including software” amortized $122M against $122M of additions in the same year — a self-replacing, cash-consuming asset. Neither extreme survives.

(3) How exposed is Alcon to China VBP? FACT: China is 6% of net sales ($570M), and management puts China IOLs at “about the same as the full business, which is about 5%.” Successive IOL VBP rounds cut prices ~20%, 26%, 38%, 53% and 84%. INTERPRETATION: the sharper point is that VBP destroyed the pool for everyone — Eyebright, the domestic champion, grew IOL revenue +1.4% and took a goodwill impairment in its own protected home market. China is a ~6% drag with another round landing mid-2026, not an existential competitive threat.

(4) Is the premium-IOL position durable against J&J? FACT: Implantables revenue was $1,703M / $1,775M / $1,782M across FY2023-25 — +0.4% in FY2025, +1% in Q1-2026 — while J&J Surgical Vision grew +10.2% to $1,558M, with TECNIS Odyssey “the fastest-growing intra-ocular lens in the U.S.” The hardest competitive datapoint in the file.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Neither — a plateau. FACT: core operating income ran $1,849M → $2,027M → $2,039M (FY2023/24/25): +9.6%, then +0.6%. Core diluted EPS $3.05 → $3.07. Core ROE peaked at 7.03% in FY2024 (now 6.90%); honest tangible ROIC peaked at 21.8% (now 20.5%). INTERPRETATION: FY2024 was the return peak. A stall is not a trough — which matters, because trough earnings justify paying up and plateau earnings do not.

External environment or internal actions? Both, roughly evenly. External: ~$100M of FY2025 tariffs to cost of sales (-97bp of core margin), China VBP, and cataract volumes growing “low single digits” versus a ~4% historical average. Internal: R&D +13% to $990M (-69bp) and mix shift away from Implantables. FACT (computed): absent tariffs and the R&D step-up, core margin would have expanded ~85bp — manufacturing efficiency is genuinely working. INTERPRETATION: neither offset is a one-off; tariffs are guided at $100-150M for FY2026 with the ~$25M of relief explicitly reinvested rather than dropped through.

How stable are revenues? Very. FACT: 83.6% of Surgical revenue is implantables plus consumables — a genuine annuity anchored on a 7-10 year console cycle; Vision Care is repeat-purchase consumer. No customer is 10%+ of sales. The instability is in conversion, not revenue: FY2025 revenue rose $483M and core operating income rose $12M — an incremental core margin of 2.5% against a 19.8% average.

Outlook for products/services? Dense and simultaneous: Unity CS/VCS (replacing 2013 and 2008 platforms), PanOptix Pro, Clareon TruPlus, Total30 multifocal toric, Precision7, TRYPTYR (~4 share points in 8 months, peak guided $250-400M), Valeda, Voyager DSLT. INTERPRETATION: growth now requires a slate of launches executing at once, with FY2026 margin expansion back-end-loaded into 2H — concentrated execution risk, with both inorganic substitutes gone.

How big will this market be — growing, shrinking, domestic or international? FACT: Alcon sizes its markets at ~$14B Surgical (+4-6%/yr) and ~$23B Vision Care (+4-5%/yr) — ~$37B combined against $10,319M of sales, i.e. ~28% share. Geography: US 45%, International 55% (Japan 6%, China 6%). The key fact is the downgrade: Alcon has marked down its own market three times in eighteen months — the FY2024 20-F guided surgical to ~6%/yr and the FY2025 20-F to 4-6%; Aug-2025 cut assumed growth to “low single digits versus a historical average of mid-single digits”; FY2026 guidance assumes aggregate markets grow 3-4%, below the floor of its own published TAM growth.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More. FACT: in FY2025, in a market Alcon leads, every named competitor grew its comparable line faster — J&J Surgical Vision +10.2%, Carl Zeiss Meditec Ophthalmology +8.5%, B+L Surgical +6%, B+L Vision Care +7%, Glaukos +32% — against Alcon Surgical +4% and Implantables +0.4%. The industry’s biggest firm was its slowest-growing major.

How profitable is the business (ROIC, ROE)? The book numbers are artifacts. FACT (computed from the 20-F): net income $980M on $22,034M of equity = book ROE 4.45%; NOPAT $1,149M on $25,673M of invested capital = book ROIC 4.48% — roughly half Alcon’s own disclosed post-tax impairment discount rates (Surgical 8.5%, Vision Care 8.0%). Both are depressed by $18.3B of spin-goodwill Alcon never paid for. FACT (computed): stripping intangibles from the denominator and their amortization from the numerator — the two must move together — tangible invested capital is $7,411M and honest tangible ROIC is ~20.5%. INTERPRETATION: mid-band in Greenwald’s 15-25% “advantages present” range — good, not the 20-40% of genuine wide-moat medtech (ResMed 22.0%, IDEXX ~41%). And it peaked in FY2024.

How profitable is the industry — how many competitors, what barriers to entry? FACT (computed): Surgical CR4 ≈ 72% (upper bound; Hoya’s IOL revenue is not separately disclosed); contact lenses CR4 ≈ 95%+. INTERPRETATION — the Greenwald point: concentration is not producing pricing power for Alcon. Both segment contribution margins fell in FY2025 (Surgical 26.6% → 25.4%, Vision Care 22.3% → 21.5%) while it was taking price. CooperVision earns a 26.6% lens segment operating margin on 4.2% R&D intensity against Alcon’s 21.5% Vision Care contribution on 9.6%. High concentration plus falling incumbent margins is the signature of rent competed away at the product level or captured downstream.

Market-share stability (Greenwald’s test) — the most damning evidence. Alcon’s LTI pays partly on “Share of Peers” using third-party syndicated data, so management measures and discloses its own outcome. FACT (20-F Compensation Report, verbatim): contact lenses — “we fell short of our share-expansion target… competitive pressure in the Dailies category, as well as growth of private label”; ocular health — “we didn’t hit our share gain targets”; ATIOLs — “Alcon trailed market growth, but stabilized US trifocal market share.” The one win was phaco and vitreoretinal cassette packs. INTERPRETATION: three of four measured franchises missed over a full three-year cycle, the single gain was the lowest-value category, and the metric paid 35% — yet the overall PSU paid 137%.

Marathon capital-cycle read. FACT (20-F Item 4.D): ~$1.04B and ~16 new contact-lens production lines in flight — Grosswallstadt +3 ($162M to 2027); Singapore +4 ($189M, completed 2024), +3 plus a new building ($314M to 2027), +3 more ($157M to 2030); Johns Creek +2 ($148M to 2028), +1 ($73M to 2029). That is 106% of one year of Vision Care segment contribution and larger than every completed acquisition combined ($753M of cash M&A over two years). CooperVision has spent 9-11% of sales for years on the same build; B+L grew revenue $4,146M → $5,101M while earning a 3.7% operating margin and losing $360M in FY2025. INTERPRETATION: three of four incumbents are adding high-fixed-cost capacity into a category whose leader now calls growth “the low end of mid-single digits,” in the exact franchise where Alcon concedes it lost share to private label — the buyer least willing to pay for a brand and most willing to fill a spare line. Canonical late-cycle. Honest mitigant: the capacity is daily-disposable silicone-hydrogel and premium reusables (mix-up, not pure unit expansion), and Vision Care gross margin is up 175bp over two years. A warning, not yet a bust.

Can the business be easily understood? Yes — two segments, four revenue lines, one end organ. The complexity is entirely in the accounting.

Can it be undermined by foreign low-cost labor? Partially, and the filing says so: the 20-F flags “increased competition from manufacturers in Asia” and “private-label alternatives.” INTERPRETATION: lenses are the exposed franchise; consoles and premium IOLs far less so — regulatory approval, service networks and surgeon workflow are not low-cost-labor problems. Offsetting structural cost: FACT: EU MDR certification runs ~EUR8,000 (Class I) to over EUR600,000 (Class III with clinical investigations), with the European Commission acknowledging implementation costs up to 5% of revenue for some organizations — a genuine, scale-favoring barrier that has risen.

Do brands matter? Differentially, mapping onto moat strength. In Surgical the “brand” is really the installed base and workflow lock. In lenses it supports a 20-40% specialty price premium but is losing to private label internationally. In ocular health, FACT: Systane is “the world’s leading Artificial Tears brand” at ~29% of the US OTC eye-drops market — yet Pataday’s 2020 Rx-to-OTC switch removed the prescription gate, and management concedes it missed its share target on “competitive launches and overall pricing pressure.” INTERPRETATION: here brand buys a price premium, not a barrier. Weakest of the three.

What is the nature of competition? Per the 20-F: “technological leadership and innovation, quality and efficacy… relationships with eye care professionals and healthcare providers, breadth and depth of product offerings and pricing.” INTERPRETATION: in IOLs this collapses to one axis — optical performance — and share moves within a single product cycle, as J&J demonstrated. The 20-F concedes the treadmill: “As these products age and/or competitive products advance, prices typically trend downward, requiring continuous innovation cycles to maintain and/or grow our margins.”

Customers’ switching costs? Real in Surgical, thin elsewhere. FACT: Alcon holds “the largest installed base of cataract phacoemulsification consoles, vitrectomy consoles and refractive lasers in the industry… long buying cycles that last approximately seven to ten years and act as anchoring technologies that drive recurring sales of our consumables,” plus >10,000 Custom Pak configurations and the ADI/SMARTCataract workflow layer. INTERPRETATION: captivity is workflow habit and capital-cycle timing, not contract — a barrier re-tendered every seven to ten years. Carl Zeiss Meditec’s acquisition of D.O.R.C. (EVA NEXUS, bundled with IOLMaster diagnostics) means that renewal is now contested by a rival ecosystem, not a point product.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Yes, materially — the installed console base, surgeon relationships, the >10,000 Custom Pak configurations, the Systane brand, and the internally-generated pipeline (84 projects, 56 past proof-of-concept, >2,100 R&D staff, $990M of FY2025 R&D expensed) all carry no balance-sheet value. INTERPRETATION: this is exactly why book returns understate economics while book equity overstates capital — the balance sheet carries $18.3B of acquired intangibles Novartis paid for and none of the organic franchise Alcon built.

Off-balance-sheet liabilities? Leases ($509M) are on-balance-sheet under IFRS 16. PP&E purchase commitments were $276M against ~$1.04B of announced lens capacity; BELKIN carries up to $385M of contingent commercial milestones held at $6M of fair value. The genuinely unquantified item is legal: FACT: an open DOJ Civil Investigative Demand under the False Claims Act (received July 2024, relating to discounts on surgical equipment servicing contracts) with no quantified provision, against total litigation provisions of only $18M at year-end; plus the Sight Sciences willful-infringement loss (>$34M of past damages plus ongoing royalties through November 2028) and a live Simbrinza Hatch-Waxman appeal Alcon lost at trial.

How conservative is the accounting? Mixed — and the mix is the finding. Conservative: FACT (computed): no channel-stuffing or inventory tell behind the flat Implantables line — DIO flat at ~188 days, inventories +5.4% against sales +4.9%, receivables past due more than three months at 6.2% of gross versus 6.1% prior year, and the cash conversion cycle improved to 183.0 days from 188.0. The stall is genuine demand and share, not an artifact. Aggressive: the non-IFRS “core” framework strips items an owner must charge — $58M of acquisition costs from a company doing 21 BD&L deals a year ($46M of it on a deal that failed), $21M of chronic legal provisions, $44M of product-discontinuation write-offs, self-replacing software amortization, and — new in 2026 — an “efficiency initiatives” line that Note 12 of the Q1-2026 interim report confirms is accrued severance, i.e. unambiguous cash cost. The most exposed single position: FACT: Aurion Biotech’s purchase price allocation put $820M into acquired IPR&D (95.8% of $856M of consideration) plus $175M of goodwill, for an asset generating $12M of sales and -$37M of net income. It is PwC’s sole critical audit matter in the FY2025 20-F, rests on a corneal cell therapy not approved in the US or EU, and no impairment has been taken.

How CapEx-hungry is the business? Moderately — and the recent numbers are flattered. FACT: total capex (PP&E plus intangibles) as a share of sales ran 14.4% (2021) → 8.6% → 9.1% → 6.8% → 6.4% (2025); PP&E intensity alone fell from 8.5% to 4.8% at the FY2024 trough. INTERPRETATION: FCF nearly tripled from $537M (2023) to $1,613M (2025), and roughly $188M of that $1,076M swing came purely from capex falling — a cyclical trough harvested as “FCF growth.” It is reversing: Q1-2026 PP&E capex was $139M versus $106M (+31% YoY), FY2026 is guided to “mid-single-digit % of sales,” and ~$1.04B of lens capacity remains to be spent.

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy? FACT: management-defined FY2025 FCF was $1,733M — but that excludes $120M of intangible/software capex. Deducting all capex gives $1,613M; net of $171M of stock-based compensation, $1,442M ($2.94/share, a 4.19% yield at $70.26, versus the 5.03% headline). FACT (computed): FY2025 deployment was acquisitions $692M + PP&E capex $543M + buyback $676M + dividends $166M = $2,077M against $2,271M of operating cash flow (91%), with $990M of R&D expensed above that line. Everything is funded internally.

Significant acquisitions recently? FACT: both attempted public-company deals failed in the same year. STAAR Surgical (signed 2025-08-05 at $28.00/share, raised 2025-12-09 to $30.75, ~$1.8B total consideration) was terminated 2026-01-06 after shareholders declined to approve; no break fee either way, and the $1.9B bridge was never drawn. LENSAR (signed 2025-03-23 at $14.00 plus a CVR up to $2.75, ~$430M) was terminated 2026-03-16 on the FTC’s stated intent to seek to enjoin — the deal combined the #1 and #2 players in femtosecond laser-assisted cataract surgery — and LENSAR retained a $10.0M deposit. FACT (computed): $46M of direct acquisition costs in FY2025 + $20M in Q1-2026 + the $10M deposit = ~$76M expensed for zero assets acquired, plus roughly thirteen months of senior management attention in a year requiring two guidance cuts. INTERPRETATION: management’s “throughout this process we remained disciplined with our views on price and risk” narrates an outcome imposed by others. Neither failure was a price decision, and both causes were foreseeable at signing. The escalation is the damning part: having bid a full 5.5x EV/sales for a business whose revenue had fallen 26% from peak into a -19.2% operating margin, Alcon raised the price ~10% while waiving both its break-up fee and its matching rights. Forward-looking and material: the FTC has now asserted that Alcon’s share in an adjacent surgical category makes consolidation anticompetitive — permanently narrowing the M&A menu.

What did close: BELKIN ($61M cash against up to $385M of milestones — the best-structured deal in the ledger), Cylite ($72M), LumiThera ($124M, PPA still provisional), Aurion ($856M, ~53-71x annualized sales). INTERPRETATION: deal structuring is a genuine Alcon competence — it consistently makes the seller carry the risk. The failures are in target selection and process. FACT: Alcon paid or bid 5.5-5.9x sales (STAAR), 6.1-7.4x (LENSAR) and ~53-71x (Aurion) against its own 3.62x EV/sales — a systematic net buyer of assets more expensive than itself.

Buying back shares? Yes, and more genuinely than the sceptical read allows. FACT: the $750M program authorized 2025-02-25 as a three-year program was completed in ~10 months (2026-01-20); FY2025 activity was 8,456,204 shares at an average $80.71, with volume accelerating as the price fell (327,500 shares in April at $90.49 versus 1,750,000 in September at $77.41). FACT (computed, two independent methods agreeing): buyback net of SBC = $682M − $171M = $511M, i.e. ~75-77% genuinely retired stock; shares outstanding fell 494,616,324 → 487,427,920 (−1.45%) against ~9.3M repurchased, implying ~2.11M shares of net employee issuance versus the 2.12M derived independently from SBC expense. The often-cited “-1.3M diluted shares on $682M” figure is a weighted-average timing artifact and should not be used. A new $1.5B / 3-year authorization followed on 2026-05-05. INTERPRETATION: counter-cyclical process, unvindicated outcome — the FY2025 tranche is ~13% underwater at $70.26.

Issuing large amounts of new shares to insiders? No. FACT: SBC of $171M is 1.64% of revenue, 8.4% of core operating income and 12.5% of honest owner earnings; unvested equity instruments are ~1% of issued shares. Modest and well controlled.

Compensation policy of directors/management? FACT: CEO 2025 total compensation $11,602,744, of which LTI (100% PSUs) is 65.0%; total ECA pay fell 12% in CHF because the 2025 STI paid a Business Performance Factor of 80% (CEO 72% of target) on missed sales and core operating income. Pay is targeted near median despite above-median revenue and market cap. The architecture is strong and deserves credit: no severance agreements, no single-trigger change-of-control payments, no excise-tax gross-ups, no stock options, no guaranteed compensation, clawbacks including Dodd-Frank restatement recovery, hedging and pledging prohibited, and binding Swiss AGM votes on the compensation caps.

The defect is entirely in the metrics inside that architecture. FACT: STI = 40% Net Sales / 40% Core Operating Income / 20% Free Cash Flow. LTI = 25% Net Sales CAGR / 25% Core Diluted EPS CAGR / 25% Share of Peers / 25% Innovation — unchanged into the 2025-2027 cycle. There is no return-on-capital metric of any kind — no ROIC, ROE, ROTC or economic profit — in either plan, in either cycle. INTERPRETATION: a team scored this way is structurally indifferent to the return on the marginal dollar. The $1.04B lens build, ~53-71x sales for Aurion, 5.5x sales for a loss-making STAAR, and R&D +19.6% against flat core operating income are each scorecard-neutral or scorecard-positive while being return-dilutive. FACT (computed) — the sharpest governance finding: the 2023-2025 PSU paid 137% because Core Diluted EPS CAGR was reported at 21.1% and paid the capped 200% — against roughly 11.1% computable from Alcon’s own audited core EPS series ($2.24 → $3.07), or 5.9% measured 2023→2025. The plan’s footnote says results are measured at constant currency and exclude acquisitions, divestitures and “certain non-recurring items,” but no base year and no reconciliation are disclosed anywhere in the filing. That metric alone added 25 points; without it the payout would have been 112%. Meanwhile the one metric measuring competitive outcome paid 35%.

Motivations of management? FACT: insider ownership is thin — the CEO holds 269,775 vested shares plus 287,898 target PSUs ≈ 0.114% of shares outstanding; his vested stake at $70.26 is ~$18.95M = 13.3x base salary (above the 6x guideline) but only ~1.6x his annual total compensation. No Board or ECA member owns 1% or more; total insider ownership is well under 0.5%. INTERPRETATION: alignment runs through annual grants, not an owner’s stake. Combined with a diffuse, passive-flavoured register (BlackRock 6.05%, UBS entities ~5.3-5.8%, no activist or strategic block, Novartis fully exited in 2019, and no opt-out from the Swiss mandatory-takeover regime), no shareholder has the size or incentive to force a change in capital-allocation policy. Insider-transaction signal is structurally unavailable: as a foreign private issuer Alcon has no Form 3/4/5 corpus — Form 144 covers proposed sales only (~226,000 shares / ~$22.3M across 2023-2026, ~half of it a single 2024-09-12 notice) — and with no options and 100% PSU compensation, executives have no disclosed mechanism to accumulate. Do not over-read it in either direction.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No to all three — and this is frequently gotten wrong. FACT (verified in the FY2025 20-F): Item 12.D, “American Depositary Shares — Not Applicable.” Alcon shares are listed on both the SIX Swiss Exchange and the NYSE as global registered shares under the ticker “ALC,” and “can be traded and transferred across applicable borders, without the need for conversion, with identical shares traded on different stock exchanges in different currencies.” FY2025 average daily volume was ~1.1M shares on the SIX and ~1.8M on the NYSE; Alcon is a Swiss Market Index constituent. The correct analogs to flag instead of ADR/K-1 mechanics are: (a) Swiss withholding tax — dividends carry a Swiss federal withholding tax at a current statutory rate of 35%, reducible for eligible non-Swiss holders under treaty, so the headline yield overstates what a US holder receives absent reclaim; (b) PFIC status — the 20-F addresses the passive-foreign-investment-company rules and states Alcon believes it may qualify as a “qualified foreign corporation,” making distributions potentially eligible for the preferential rate; © IFRS reporting, so every segment definition and “core” measure must be bridged before comparison with a US-reporting peer.

Dividend policy? FACT — correcting a widely repeated error: the FY2025 dividend is FLAT at CHF 0.28, the first non-raise since the 2019 spin. The sequence is CHF 0.21 (2023) → 0.24 (+14.3%) → 0.28 (+16.7%) → 0.28 (0.0%). The “~10% increase” in management’s own scorecard refers to the USD amount paid during 2025 ($166M versus $130M); the ~$182M maximum payable in 2026 rises only because the Swiss franc appreciated. Stated policy is a dividend “based on the prior year’s core net income.” FACT (computed): the $166M paid in FY2025 is 10.9% of core net income, 16.9% of IFRS net income, 10.3% of FCF, and a 0.48% yield; total shareholder yield including the $511M net buyback is ~1.97%. INTERPRETATION: an ~11% payout from a business earning ~20.5% on tangible capital at 1.19x net leverage is a token, not a policy — and freezing it in the same year the buyback authorization was doubled signals a preference for the discretionary, reversible instrument over the committed one.

How profitable is the business? In one place: IFRS operating margin 13.2%, core 19.8%, honest 17.85%; IFRS diluted EPS $1.98, honest owner EPS $2.75, core diluted EPS $3.07; book ROE 4.45%, book ROIC 4.48%, honest tangible ROIC ~20.5%. At $70.26 that is 35.6x IFRS, 25.5x honest owner earnings (the figure to use), and 22.9x core.

Is net income diverging from cash from operations? Yes, in the favourable direction — but less than it appears. FACT: honest FCF of $1,613M is 1.06x core net income and 1.65x IFRS net income. The wedge is dominated by ~$700M/yr of non-cash spin-era amortization, which is legitimate. But two temporary tailwinds inflate FY2025 cash: FACT: cash taxes paid were only $107M against a $180M P&L charge and a ~$230M three-year average (a ~$150M one-year benefit, with the FY2026 core tax rate guided up to 20%), plus the capex trough above. FACT: Alcon has built working capital every year ($314M / $522M / $404M / $189M / $240M, 2021-25) — a structural, permanent drag, not something to normalize away.

Risks & Downside

What factors would cause the stock to decline? By evidenced likelihood: (1) Implantables staying flat while J&J compounds — FACT (computed): the entire 197bp two-year decline in Surgical contribution margin is a gross-margin story driven by that mix; (2) the efficiency program proving recurring — FACT: $88M was charged in Q1-2026 alone ($70M provisioned as severance) for a program whose total size, duration and savings target are undisclosed anywhere in the corpus; (3) missing the FY2026 guide — FACT (computed): the +10-13% core EPS guide is only reachable at roughly the top third of the +70-170bp core margin range; delivering the margin floor yields only ~+6.5% EPS growth, so the two guidance lines are not independently satisfiable; (4) the mid-2026 China IOL VBP round; (5) a contact-lens price war as ~$1B of industry capacity lands; (6) crystallization of the DOJ CID or an Aurion IPR&D write-down.

Risk of a catastrophic loss? Low. FACT: net debt $3,130M = 1.19x core EBITDA, interest coverage 10.0x, weighted-average interest rate 3.6% termed out to 2052, $1.32B revolver undrawn and extended to October 2030, total liquidity $2,927M, no customer concentration. The Notes trade ~3% below par — INTERPRETATION: a rate artifact from low coupons, economically a benefit, not a credit signal. The genuine tail is regulatory rather than financial: FACT: CMS Ruling 05-01 (May 2005) is what legally permits Medicare beneficiaries to pay out of pocket for the presbyopia-correcting functionality of an IOL — the foundation of the entire ATIOL profit pool — and it is an administrative ruling, not a statute. ASSUMPTION-level tail risk: low probability (21 years standing, no live proposal to change it), catastrophic for the premium franchise if revisited.

Chance of a total loss? Effectively nil on the evidence — a #1-share, investment-grade, cash-generative business with a 4.19% FCF yield net of SBC. INTERPRETATION: the realistic risk is opportunity cost and time, not permanent impairment — 25.5x honest owner earnings implies roughly 3.1-4.4% perpetual growth (~3.9% central at an 8% cost of equity), below the 3-4% market growth management itself assumes for 2026. A low bar, and an unexciting arithmetic. The classic value-trap signature — a melting asset — is absent. But note the empirical record: FACT: the Sharpe ratio is negative at every horizon out to five years; the deepest drawdown of Alcon’s independent public life (-37.9%, trough 2026-05-11) is deeper than COVID and only nine weeks old; and since the 2019 spin Alcon has compounded at 2.95%/yr against the S&P’s 15.65% — -164 percentage points cumulatively, at 27% annualized volatility.

Recent News & Events

Has the business environment changed recently? Yes, on four axes within eighteen months. FACT: (i) the industry growth assumption was marked down three times (~6% → 4-6%; “low single digits” in Aug-2025; 3-4% assumed for FY2026); (ii) tariffs became a permanent structural cost — ~$100M in FY2025, $100-150M guided for FY2026, costing 120bp of core gross margin in Q1-2026; (iii) US glaucoma reimbursement changed — the CMS 2026 Medicare Physician Fee Schedule (effective 2026-01-01) cut physician reimbursement ~10.5% for cataract and combined cataract+MIGS procedures, on top of November-2024 MIGS Local Coverage Determinations from five of seven MACs capping coverage at one MIGS device per eye and demoting MIGS below drops and laser; (iv) volume is supply-capped — the peer-reviewed US ophthalmology workforce projection finds supply falling 12% while demand rises 24% from 2020 to 2035, a ~30% inadequacy. INTERPRETATION: management’s “capacity not demand” framing is corroborated and should be accepted; its optimistic corollary — that health systems will “adapt” — is an assumption with a decade-long time constant and no evidence yet of having started. Worse for Alcon specifically: if surgeon hours are the binding constraint, the rent accrues to whoever controls them — practice roll-ups, ASC owners and surgeons — not to the device supplier. That is the most parsimonious explanation for a #1-share incumbent in a 72%-CR4 market taking price and still losing margin in both segments.

Significant acquisitions? As above — 0-for-2 on public deals, ~$76M expensed for zero assets, against four completed bolt-ons and 21 disclosed BD&L transactions in 2025 whose aggregate spend is not disclosed, only the count. A non-exclusive RxSight collaboration on adjustable IOLs was announced 2026-07-06.

Change in accounting policies? None identified. But FACT: the IFRS-to-core wedge has widened — $0.78 (2023) → $1.00 (2024) → $1.09 (2025) → $0.46 in Q1-2026 alone, the widest quarterly wedge in the corpus — driven by the new efficiency add-back and by acquisition amortization that each bolt-on mechanically rebuilds (+$310M goodwill, +$419M other intangibles in FY2025; unallocated cost of net sales +$67M). INTERPRETATION: the amortization add-back is not a free perpetual pass; the wedge widens as a direct consequence of the M&A strategy.

Recent changes — new markets, facilities, management? FACT: ~16 new contact-lens production lines running to 2030; a new efficiency/restructuring program announced 2026-02-24; and a governance change worth naming — Scott H. Maw, a founding board member and Chair of the Audit & Risk Committee, did not stand for re-election at the 2026-04-30 AGM. INTERPRETATION: Alcon loses its Audit Chair at precisely the moment three unquantified exposures are live — the un-impaired $820M Aurion IPR&D, an undisclosed-scope restructuring program, and an open DOJ False Claims Act investigation against $18M of total litigation provisions. His successor is not disclosed in the corpus.

Disclosure gaps we could not close

APPENDIX B — Source Appendix

Methodology. Every non-obvious fact in this report carries a source below. Sources are ranked primary-over-secondary without exception: the company’s own filings and disclosures outrank management commentary; management commentary is treated as a hypothesis and is validated against filings, competitor disclosures, and independent data; a governing regulatory document outranks any article describing it; and third-party aggregated data services (ROIC.ai, AZI, FactorsToday) are used to accelerate and cross-check, never to replace, the filing. Where an aggregator and a filing disagreed on a material number, the filing won and the discrepancy is recorded in the final section of this appendix. Alcon is a Swiss foreign private issuer reporting under IFRS as issued by the IASB — it files Forms 20-F and 6-K, not 10-K/10-Q/8-K, and has no Form 3/4/5 insider corpus; all references to “GAAP” earnings elsewhere in the market’s commentary on this company are, properly, references to IFRS. Unless otherwise stated, all URLs were accessed 2026-07-18.

A note on link status: a sample of the URLs below was re-tested on 2026-07-18 and resolved (HTTP 200), including every SEC EDGAR document, the CMS Ruling 05-01 PDF, the CMS Medicare Coverage Database LCD, the European Commission SWD, the MedPAC chapter, and the Carl Zeiss Meditec and HOYA investor pages. Three sources return HTTP 403/405 to automated retrieval while remaining publicly readable in a browser — the American Academy of Ophthalmology (aao.org), the AAO’s journal Ophthalmology (aaojournal.org), and MDDI Online (mddionline.com); the FTC’s own site likewise refuses automated fetches (see the LENSAR entry). These are anti-bot blocks, not broken links, but the distinction is recorded here rather than assumed away.

1. Primary — Company Filings and Disclosures (Alcon Inc., CIK 0001167379)

The trailing 60-month EDGAR corpus (since 2021-07-01) comprises 80 filings: 6-K ×37, Form 144 ×20, SC 13G/13G-A ×7, SD ×5, 20-F ×5, DFAN14A ×3, DEFM14A ×1, DEFA14A ×1, IRANNOTICE ×1. There is no 424B*/FWP structured-note noise.

1.1 Annual reports (Form 20-F)

1.2 Earnings press releases and interim financial reports (Form 6-K)

1.3 The STAAR Surgical transaction (“Project Strasbourg”) and the proxy filings

Filename note for auditability: the 6-K wrappers form6-kstrasbourgaug2025.htm and form6-kstrasbourgterminati.htm refer to Alcon’s internal project codename for the STAAR acquisition, not to a French manufacturing site, restructuring, or contract termination. No charge is associated with a “Strasbourg termination.” Similarly, the 2025-05-28 6-K form6-kxarx1551230may2025.htm is unrelated — “AR-15512” is the TRYPTYR compound code.

1.4 Regulatory-approval and product 6-Ks

1.5 Form 144 filings (the only insider-transaction filing type available for an FPI)

Seventeen notices reviewed on EDGAR spanning 2023-05-11 through 2026-02-26, CIK 0001167379; filer CIKs map to the CEO (1532133, Officer + Director), the CFO (1656085), and three further executive-committee members (1967480, 1967524, 1967526). Aggregate ~226,000 shares / ~$22.3M of proposed sales, of which ~50% is a single CEO notice dated 2024-09-12 (112,182 shares / $11,273,015). Form 144 covers proposed sales only; every notice traces to restricted-stock vesting. These filings are also the source of the shares-outstanding series used in the buyback arithmetic (490,086,981 in 2023 → 493,244,479 in 2024 → 494,616,324 in Feb-2025 → 487,427,920 in Feb-2026).

Corpus enumeration: EDGAR’s full filing history for CIK 0001167379, retrieved 2026-07-18 and enumerated from 2021-07-01 forward.

2. Primary — Earnings Call Transcripts

Retrieved via the ROIC.ai transcript service (list_earnings_calls, get_earnings_call_transcript, get_latest_earnings_call), accessed 2026-07-18, and read in full.

Coverage gap, stated explicitly. ROIC’s transcript corpus is earnings-call-centric. Alcon’s FY2025 Capital Markets Day — which reset the guidance framework used from FY2026 onward — is in no corpus available to this engagement: it was not furnished on a 6-K and it is not in ROIC’s transcript set. Its medium-term targets are therefore not quoted anywhere in this report; anyone wanting them must source the deck directly from investor.alcon.com.

3. Primary — Competitor and Peer Filings

4. Primary — Regulatory and Government Sources

5. Quantitative Data Services (Third-Party Aggregated — NOT PRIMARY)

These services accelerate and cross-check the analysis; they do not replace the filing. For a US-listed foreign private issuer, the 20-F and the 6-K interim reports are primary, and every material number in this report was reconciled to them. The corruptions found in these feeds during this engagement are documented in the final section.

ROIC.ai MCP (accessed 2026-07-18) — tools used and purpose:

AZI (azitrading.com) — used for exactly two things:

• Price history CSV — https://azitrading.com/controls/download-data.php?t=ALC, accessed 2026-07-18: 1,828 sessions covering the full post-spin history (2019-04-09 through 2026-07-17), adjusted and unadjusted OHLC, dividend/split columns, and pre-computed 21/50/200-day EMAs, beta and alpha. The SPY benchmark CSV was pulled from the same endpoint. Data-integrity check performed and passed: the CSV’s monthly average closes were compared month-by-month against the FY2025 20-F Item 16E monthly buyback average prices for April–December 2025 — every month agreed within ~1% (e.g. April $90.43 vs $90.49; October $74.71 vs $75.33). The AZI series is therefore validated against a primary filing and was used with confidence.
• valuation_index own-history percentile ranks — the AZI fundamentals endpoint’s valuation_index block for ALC, retrieved twice independently (feed updated_at 2026-07-17 19:00:32) with identical results: composite 22.21st percentile; P/E 42.575× @ 37.05th; P/B 1.548× @ 24.86th; P/S 3.280× @ 4.73rd; n_components 3; latest price $70.26, TTM EPS $1.6503, BVPS $45.395, TTM sales/share $21.4191. Interpretive binds applied in the body: the P/E percentile was discarded because the IFRS EPS denominator is unstable (swinging on the FY2023 $263M discrete Swiss tax benefit and the FY2024 $57M Ocumension divestment gain); the P/B percentile was discounted because 82.9% of book equity is spin-era goodwill and intangibles; only the P/S percentile was treated as a clean own-history read, and it is context only — never cross-sectional, never a price target. No other AZI datapoint (statement arrays, snapshot EV/market cap) was used.

FactorsToday (https://www.factorstoday.com/api, no-auth; methodology at /about), all endpoints accessed 2026-07-18:

• /api/stock-loadings/ALC — loadings date 2026-07-17, 756-day window, four nested models (All Factors R² 0.447 / adj. 0.409; Base R² 0.326). Read within one model only, per the hierarchical-orthogonalisation rule.
• /api/leaderboard/ALC — risk-adjusted track record by horizon. All returns and Sharpe ratios are ANNUALIZED, including short windows; de-annualisation checks against the AZI CSV are recorded in the body (m3 −43.04% annualized → −13.13% raw vs. CSV −13.18%).
• /api/stock-info/ALC — beta 0.708, alpha −0.187, relative strength (rs_ytd/6m/12m/peak), market cap, latest OHLC (latest_data 2026-07-17).
• /api/stock-specific-vol/ALC — idiosyncratic volatility 22.24% annualized (252-day window).
• /api/related-stocks/ALC — factor-similar peers (a negative result, reported as such: the set returns ETFs and international dividend-growth funds, with only SYK, AVY and ZTS as individual names and none of the ophthalmic competitors).
• /api/factor-returns/historic — the regime read (factor returns and z-scores at 21/63/126/252 days).
• Known-bias caveats applied: R² figures are in-sample and mildly overstate fit; the SmallSize loading (+0.24) carries a known heavy-issuer bias; custom-factor names are model-generated labels on real baskets and were not relied on.

SEC EDGAR — the full filing history for CIK 0001167379 from 2021-07-01 forward, plus the underlying documents. EDGAR XBRL and filings data is authoritative primary data and is listed here only for completeness.

6. Industry Data and Trade Press (SECONDARY — labeled as such throughout the report)

A blanket caveat applies to every market-sizing figure in this group: apart from Alcon’s own ~$14B Surgical and ~$23B Vision Care TAM figures (which are primary, from the 20-F), all market sizing below is third-party estimation with wide dispersion. The OTC dry-eye category alone spans $1.85B to $4.32B across two sources for the same year — a 2.3× spread — which is itself the finding. No argument in this report is built on secondary market sizing.

Peer-reviewed literature (a distinct and higher tier than the trade press above):

• “Ophthalmology Workforce Projections in the United States, 2020 to 2035,” Ophthalmology (Journal of the American Academy of Ophthalmology) — https://www.aaojournal.org/article/S0161-6420(23)00677-2/fulltext (HTTP 403 to automated fetch; publicly readable in a browser). Total ophthalmologist supply projected to decline by 2,650 FTE (−12%) while demand rises 5,150 FTE (+24%) from 2020 to 2035 — a ~30% workforce inadequacy. This is the evidentiary basis for accepting management’s “capacity, not demand” framing as fact while rejecting its optimistic corollary that “health care systems adapt” on a guidance-relevant horizon.

Analytical frameworks: Competition Demystified (Greenwald & Kahn) and Capital Returns (Chancellor / Marathon).

7. Data-Quality Warnings and Corrections Applied

This section exists because several third-party feeds were materially wrong on this issuer, and because the analysis corrected itself in six specific, auditable places. Both categories are recorded so that any figure in this report can be traced to the version that survived scrutiny.

7.1 Third-party feed corruptions found on ALC

(a) ROIC.ai’s FY2025 balance sheet for ALC is CORRUPT and was not used at all. The returned FY2025 balance sheet shows bs_tot_liab = $31,555M — identical to total assets, which is arithmetically impossible for a solvent issuer with $22.0B of equity; bs_other_intangible_assets_detailed = −$9,256M (a negative intangibles line, and note that $9,256M is in fact the goodwill figure with the sign flipped); and no cash, receivables, or debt lines and no equity line at all. Correction applied: every FY2025 balance-sheet figure in this report — equity $22,034M, goodwill $9,256M, other intangibles $9,006M, financial debts $4,737M ($4,162M non-current + $575M current), cash $1,527M + time deposits $80M, lease liabilities $509M, inventories $2,391M, trade receivables $1,942M, total assets $31,555M — was taken directly from the FY2025 20-F consolidated balance sheet and its notes, and was independently re-verified against the same filing.

(b) ROIC.ai’s return_com_eqy for ALC is off by roughly 1000×. It returns 9800 for FY2025, 5090 for FY2024 and 4870 for FY2023 — i.e. the FY2025 figure appears to be 9.8% expressed in a broken unit, and even taken at face value it does not reconcile: $980M of IFRS net income on $22,034M of shareholders’ equity is ROE 4.45%, not 9.8%. Separately, ROIC’s return_on_inv_capital is internally inconsistent, swinging from 4.76% (FY2024) to 10.15% (FY2025) on a balance sheet that barely changed. Correction applied: every return metric in this report is hand-computed from the 20-F. The full five-year series (IFRS ROE, core ROE, book ROIC on core NOPAT, and tangible ROIC on honest NOPAT) was rebuilt from the FY2023/FY2024/FY2025 20-F balance sheets and income statements, with tax normalised at each year’s core effective rate because the IFRS rate is unusable (FY2023 was a net $142M tax benefit, driven by the $263M discrete Swiss Tax Agreement).

© ROIC.ai’s FY2025 enterprise-value row for ALC is wrong; the TTM row reconciles. The FY2025 EV row shows total debt of only $429M — that is the lease liability alone, omitting $4,737M of financial debt. Correction applied: the FY2025 EV row was discarded. The Q1-2026 TTM row (debt $5,253M, cash $1,578M) was used and independently reconciled: at $70.26 it implies EV ~$37,922M against a filing-derived, lease-inclusive $37,886M — agreement within 1.5%. The report’s headline EV of ~$37.4B (ex-leases) is computed from the 20-F balance sheet, not from the feed. A related corruption: ROIC’s pr_to_book_ratio for ALC reads 1850.69× (a broken book-value denominator); AZI’s P/B of 1.55× was used instead. ROIC’s get_credit_ratios is likewise corrupt for FY2024–FY2025 (short_and_long_term_debt = $429M, tot_debt_to_tot_cap 99.77%, net_debt_to_capital −277% for FY2024), so all leverage figures were hand-computed. Finally, ROIC’s JNJ TTM EV row omits debt and cash entirely (EV = market cap), so JNJ multiples are treated as approximate and JNJ is used only for the Vision-segment growth comparison, never as a valuation comp.

(d) The ROIC.ai news feed for “ALC” is contaminated with two unrelated issuers. The feed mixes in Algoma Central Corporation (TSX: ALC), a Canadian marine shipping company, and Alcon Silver Corp., a British Columbia mining shell being acquired by Mexican Gold — plus a large block of congressional-trade disclosure items unrelated to Alcon. Correction applied: the feed was used strictly as a triage layer; every material item was traced to the underlying filing, company release, or article before it entered the report, and the unrelated issuers’ items were discarded.

(e) Alcon reports under IFRS, not US GAAP, and “core” is a non-IFRS management measure. Any “GAAP” label attached to Alcon figures by an aggregator or by market commentary is a category error. The company’s audited statements are IFRS as issued by the IASB; “core” operating income, “core” diluted EPS and “core” gross margin are management-defined non-IFRS measures, reconciled in the earnings releases, that exclude amortisation and impairment of acquired intangibles, acquisition and integration costs, restructuring, certain legal provisions, product-discontinuation charges, and fair-value movements on associates. This report uses IFRS figures as the audited base, presents management’s core figures where they are the relevant comparison to guidance, and constructs an independent “honest owner earnings” figure between the two — with the composition of the wedge itself treated as an analytical finding rather than an accounting technicality.

7.2 Corrections applied during the analysis (recorded for auditability)

Six figures were asserted at one stage of the work and corrected at a later one. In every case the corrected figure is what appears in the report; the superseded figure is recorded here so the trail is complete.

7.3 Screener-derived return and cost-of-capital figures: rejected and excluded

The screener-derived figures circulating on Alcon in late 2025 — ROIC 7.35%, WACC 4.15%, and ROE 11.8–12% — are the numbers underpinning the then-prevailing bull framing. None of these reconcile to the 20-F, and one of them is not a defensible cost of capital. FY2025 IFRS net income of $980M on $22,034M of shareholders’ equity is ROE 4.45%, not 11.8–12%. A “WACC of roughly 4.15%” for a levered equity with a 0.708 beta is arithmetically indefensible — it sits below Alcon’s own 3.6% weighted-average cost of debt before any equity risk premium. All three figures were rejected and excluded from the report. In their place the cost-of-capital discussion is anchored on the most defensible source available: the FY2025 20-F Note 9 post-tax discount rates Alcon itself applies to its own goodwill impairment testing — Surgical 8.5%, Vision Care 8.0% (revenue-weighted 8.28%, terminal growth 3.0%) — cross-checked against a CAPM construction whose risk-free and equity-risk-premium inputs are labeled as assumptions rather than facts. The late-2025 framing is referenced in the report only as a record of what consensus believed at the time, and never for a number.

7.4 Claims that could NOT be independently verified

The following were encountered during the work and were either qualified in place or excluded entirely. They are listed so that no reader mistakes their absence for oversight.

• “>60% global PCIOL share” for Alcon — EXCLUDED. This figure appears only in secondary financial press (traced to a Nasdaq/Zacks article) and never in a filing, never in a syndicated-data disclosure, and never in management commentary on the record. No premium-IOL share figure for Alcon is stated as fact anywhere in this report. Note the tension that made the exclusion necessary: Alcon’s own compensation disclosure concedes that in ATIOLs it “trailed market growth,” and Implantables revenue was flat (+0.4%) in FY2025 while J&J’s Surgical Vision franchise grew +10.2% — a share profile difficult to reconcile with a stated 60%-plus dominant position.
• Alcon’s Unity time-and-motion efficiency claims (a 16% vitreoretinal workflow efficiency gain; a 6% cataract turnover reduction) and PanOptix Pro’s “Enlighten NXT” 94% light-utilization figure — QUALIFIED as COMPANY MARKETING CLAIMS. These originate in Alcon’s own investor and product press releases, not in independent, peer-reviewed, or third-party data. Where they appear in this report they are attributed to the company and labeled as such; no analytical conclusion rests on them.
• The “~1–2% of sales” figure for Alcon’s direct China public-tender VBP exposure — NOT VERIFIED, and not stated as fact. It appears in trade-press commentary (mddionline), not in anything sourced to Alcon. What is established from primary sources and used instead: China is 6% of Alcon net sales ($570M FY2025 vs $560M FY2024 vs $526M FY2023 — ~4% growth over two years against group +10%), and management stated on the Q1-2026 call that IOLs in China are “about the same as the full business, which is about 5%.” The bounded, filing-anchored version is what appears in the report.
• Carl Zeiss Meditec R&D intensity (~14–15%) — EXCLUDED. Flagged as an assumption when first raised and never verified to a primary source; the figure does not appear in the report. Likewise no Carl Zeiss Meditec valuation multiple is printed anywhere, because no clean enterprise-value row was obtainable and the figure would have to come from the Zeiss annual report or an exchange feed.
• Private-label share of the soft contact-lens market — NOT SIZEABLE from public sources. No manufacturer discloses private-label volume and no credible public source quantifies the segment. It is treated in the report as a qualitatively confirmed but quantitatively unmeasured share headwind — confirmed because Alcon’s own compensation disclosure blames it for a missed share target “particularly in the international market” — and no number is attached to it.
• Whether the FTC actually filed a complaint against the Alcon/LENSAR transaction, or merely signalled intent — UNCONFIRMED. Both LENSAR’s filed release and the trade coverage indicate abandonment in anticipation of suit. The FTC’s own site returns HTTP 403 to automated retrieval, so the primary regulatory document could not be obtained; the report quotes LENSAR’s filed language (“intends to seek to enjoin”), which is itself primary.
• The 21.1% “Core Diluted EPS CAGR” that paid the 2023–2025 LTI at its capped 200% — NOT RECONCILABLE. Alcon’s reported core diluted EPS series ($2.24 / $2.74 / $3.05 / $3.07 for 2022–2025) yields a three-year CAGR of ~11.1%, or +5.9% measured 2023→2025. The plan discloses neither a base year nor a bridge, stating only that results are “measured at constant exchange rates” and “exclude the impact of acquisitions, divestitures and certain non-recurring items.” The gap is reported in the body as a disclosure gap and an open question, not as an allegation.
• Two figures used with an explicitly stated non-contemporaneity assumption: the Alcon-versus-CooperVision “roughly tied for #2 in branded contact lenses” ranking rests on a ~1% revenue gap ($2,770M vs $2,743.8M) across non-contemporaneous fiscal years (Cooper’s ends 31 October, Alcon’s 31 December). The report states the ranking only in its qualified form and never as a clean “#2” or “#3.” Similarly, the Surgical CR4 of ~72% is an upper-bound estimate, because it uses Carl Zeiss Meditec’s full Ophthalmology SBU (which includes diagnostics and refractive lasers only partly overlapping Alcon’s market definition) and because HOYA Surgical Optics’ IOL revenue is not separately disclosed anywhere and is therefore excluded.
• The CMS 2026 Medicare Physician Fee Schedule cut percentages (~10.5% cataract and combined cataract+MIGS; 7.3%–9.2% other MIGS) are sourced to AAO and Reuters coverage, not to the final-rule text. The direction and the existence of the cut are corroborated by management’s own Q1-2026 call commentary (“the reimbursement changed this year”); the precise percentages retain a secondary-source caveat.
• The China IOL VBP price-cut percentages (~20% / 26% / 38% / 53% / 84%; a recent round ~58–60%) are sourced to three independent trade publications, not to an NHSA tender document. The existence, mechanism and directional severity of VBP are established primarily from Alcon’s own 20-F risk factors; the specific percentages carry a secondary-source caveat wherever they appear.
• Forward consensus estimates were not obtainable (no professional estimates feed was available). “What consensus expects” is inferred from management guidance and from the late-2025 prevailing articulation of the bull case, not from a compiled estimate set. This is a genuine gap and is labeled as such in the Variant Perception discussion rather than papered over.
• No short-interest or days-to-cover figure was obtainable from the available feeds. The positioning read therefore rests on the ownership register (BlackRock 6.05%, UBS entities ~5.3%–5.8%, Cede & Co. 17.1% nominee, otherwise diffuse; no activist or strategic block; no Swiss opt-out from the mandatory-takeover regime) rather than on short data.