BioMarin Pharmaceutical Inc. (NASDAQ: BMRN) — A Fortress of Un-Attackable Annuities, Betting the Balance Sheet on the One Franchise Anyone Can Attack

Report date: 1 August 2026 · Price: $60.02 (31 July 2026) · Market capitalisation: ~$11.7B Sector: Health Care · Biotechnology (Rare Disease / Orphan Drugs) · CIK: 0001048477 · Coverage: Initiation

An independent analyst’s report. Sections 1–15 below contain no investment recommendation and no price target. The single exception is the clearly-labelled Claude's Take block immediately following.

⚡ Claude’s Take

This block is the author’s own subjective opinion and general information only — not investment advice, and not a recommendation to buy or sell any security. It is deliberately the single place in this report where a position is taken; the analysis in sections 1–15 below carries no position and no price target. Do your own research.

Verdict: HOLD — and specifically, do not buy it for the cheapness, because it is not actually cheap. Accumulate below ~$50, and only with the ITC ruling in hand. Conviction: medium-low.

Tag: “The moat is on the wrong products.”

The screen says bargain: at $60.02 BioMarin trades at 1.9x book and 3.6x sales — the 7th percentile of its own ten-year range on sales, the 9th on book — 3.2x EV/sales against Neurocrine at 5.0x and Vertex at 8.9x, and roughly 16x trailing free cash flow. That reading is a trap, and disarming it is the most useful thing this note can do. The percentile benchmarks today’s sales against a decade when BioMarin was a growth stock priced at 7.5–9.1x EV/sales; it measures how far the narrative fell, not how cheap the cash flow is. Against a 22% normalized operating margin, 3.5x sales is roughly 16x EBIT — market, not bargain. The honest anchors are ~21x free cash flow net of stock compensation and ~20.9x normalized standalone earnings, rising to ~28x pro-forma if the promised synergies do not land. On a reverse DCF at an 8.5% cost of capital, today’s ~$14.5B pro-forma enterprise value already requires either a 25% sustainable operating margin on 5.4% revenue growth, or 22% on 9.1% — against organic growth of 2.8% in the most recent quarter. A defensible base-case range of roughly $49–63 brackets the current price. The stock is fairly valued, not mispriced.

There is nonetheless a good business in here, and it is worth being precise about that too: FY2025 free cash flow of $717M was real, gross margin ex-write-off rose to 81.4%, and returns on operating capital reached 19.4%. The product economics are excellent. The cost structure wrapped around them is not.

The problem is that I cannot find a mechanism by which shareholders get it. Reported ROIC has cleared an ~8.5–9.5% cost of capital exactly once in five years, because capital keeps being stacked on top faster than it earns. Ten years and $7.0B of R&D produced $72.7M of cumulative operating profit; fifteen years and $9.0B produced a cumulative operating loss. The pattern just repeated at maximum scale. Roctavian consumed at least $1.6B and returned $67.8M of lifetime revenue — perhaps 25–40 patients dosed worldwide — before being withdrawn from the market in December 2025. Four months later management closed the acquisition of Amicus — headlined at $4.8B, but $5,323.6M of total consideration once the debt repayment and even Amicus’s own $121.6M of banker and legal fees are counted — funded with ~$4.25B of debt, converting $1.45B of net cash into roughly $2.8–3.0B of net debt. It paid 7.6x trailing sales, a 33% premium, and above the top of its own banker’s DCF range, at the top of the M&A cycle, and has still not disclosed a synergy target. BioMarin’s own filed pro-forma shows the combination turning FY2025 net income of +$348.9M into a net loss of $205.9M. That the stock rose 17.7% on announcing its own leveraged deal — then round-tripped to a five-year low three weeks after closing — tells you how little the market credits either the standalone company or the transaction.

Meanwhile the asset that carries the equity has stalled. Voxzogo grew 2.9% in Q1 2026 against “more than 20%” patient growth, is guided to high-single digits after +26%, and now faces Ascendis’ once-weekly Yuviwel (approved February 2026, launched April, priced 20% above Voxzogo) and BridgeBio’s oral infigratinib (Phase 3 positive in NEJM, largest height-velocity effect ever recorded, filing now). BioMarin’s answer, BMN 333, arrives around 2029–30. The $4B 2027 revenue target set in September 2024 was withdrawn in October 2025 — thirteen months — and has not been replaced. The historical rhyme is exact: Kuvan fell 78.5% from its peak when generics arrived, and Palynziq now faces PTC’s oral Sephience. BioMarin sells injectables into markets that keep being won by pills and long-acting conjugates.

What keeps this a HOLD rather than an AVOID is that the durable part is genuinely durable, and better than the market’s caricature. Naglazyme and Aldurazyme have carried no unexpired exclusivity for over a decade and have never attracted a single biosimilar — roughly $694M a year defended not by law but by a patient population too small to repay anyone’s development programme. That is the cleanest “too small to attack” moat in healthcare, and 2026 policy quietly improved around it: the 100% Section 232 pharma tariff exempts orphan drugs at 0%, the IRA orphan carve-out was widened, and the rare-pediatric voucher programme was reauthorised with vouchers clearing near $200M. The feared gene-therapy disruption of chronic enzyme replacement has been empirically falsified.

What flips me bullish: an ITC exclusion order against Ascendis on or about 21 August 2026 — three weeks out — combined with two quarters of Voxzogo revenue actually converting the claimed patient growth. What flips me bearish: Q2 2026 (printing 6 August) showing the “order timing” explanation was in fact price and share, or evidence that Amicus is being absorbed into the same value-destroying machine. On the tape this is neither a falling knife nor momentum — it is a range-bound, sector-dependent name whose entire twelve-month return is factor carry: raw return +2.1%, cumulative idiosyncratic return −13.5%, and that already includes the Amicus pop. Owning it is a bet on the fundamentals, because the tape is contributing nothing.

📈 Stock Price Action — Five-Year Event Map

BioMarin has round-tripped a full cycle and ended it lower. From ~$77 at the start of August 2021 the stock ran to a five-year closing high of $117.27 on 27 January 2023, then de-rated for three years to a five-year closing low of $49.67 on 18 May 2026, before recovering to $60.02 on 31 July 2026. That is 48.8% below the five-year high and 59.8% below the all-time high of $149.13 (20 July 2015). The 52-week range is $49.67–$64.08; the stock sits 6.3% off its 52-week high and 20.8% above its 52-week low — up 11.3% over three months, 5.3% over six, 2.1% over twelve, and down 29.0% over two years. (FACT — AZI adjusted-close series, retrieved 2026-07-31.)

The cycle narrative. Voxzogo’s November 2021 approval was the last unambiguously good news the stock fully kept; the 2022 rate shock took back the re-rating, and the run into January 2023 was the market paying in advance for two launches (1–2). That advance payment was withdrawn when Roctavian, approved in the US in June 2023, sold almost nothing (3). Elliott’s roughly $1B stake and the replacement of a 19-year CEO produced the sharpest positive repricing of the five years — on governance, not fundamentals (4). 2024 was the year the market re-underwrote the growth engine: a 17.7% single-session fall on a competitor’s Phase 2b data is the cleanest evidence available that Voxzogo, not the enzyme base, carries this equity (5). A year of competitive grinding followed (6). That an acquirer rose 17.7% on announcing a $4.8B debt-funded acquisition is unusual, and says how little the market valued standalone BioMarin (7). The give-back came fast, on a safety signal and the first competitor actually launching (8). The recovery since May rests on real clinical wins in hypochondroplasia — though, as the factor analysis in section 11 shows, almost none of it is company-specific (9).

1. Executive Summary

BioMarin is a 30-year-old rare-disease specialist with $3.22B of FY2025 revenue, eight commercial products, 81% underlying gross margins and $717M of free cash flow. It is profitable, cash-generative, and trades near the cheapest multiples in its own history. It is also a company that has not earned its cost of capital in a decade, and that has just made the largest bet in its history at the moment its only growth asset stopped growing.

The portfolio splits cleanly in two. Roughly $2.24B is a slow-growing annuity of enzyme replacement therapies — Vimizim ($792.1M), Naglazyme ($485.4M), Palynziq ($433.3M), Aldurazyme ($208.5M), Brineura ($186.4M), Kuvan ($99.6M) — treating ultra-rare lysosomal storage disorders and PKU. Roughly $927M is Voxzogo, the once-daily injection for achondroplasia that has supplied essentially all of the growth and carries the equity story.

The moat is real but sits on the wrong assets. Naglazyme and Aldurazyme have had no unexpired US regulatory exclusivity or composition patents for over a decade, and have never attracted a biosimilar — because a few thousand patients worldwide cannot repay anyone’s development programme. That is a genuine, durable barrier protecting ~$694M a year. But it is a property of the disease, not of BioMarin; it is non-transferable; and it defends only against copies, offering nothing against a better molecule. Every franchise BioMarin has lost, it lost to a differentiated new agent: Kuvan fell 78.5% from its $463M peak; Palynziq now faces PTC’s oral Sephience; Voxzogo faces two attackers.

The growth engine has stalled. Voxzogo grew 2.9% year over year in Q1 2026 — against management’s claim of “more than 20%” patient growth — and is guided to high-single-digit growth in 2026 after 26% in 2025. Ascendis’ once-weekly Yuviwel was approved 27 February 2026 and launched in April at a 20% price premium; BridgeBio’s oral infigratinib posted the largest height-velocity effect ever recorded in the disease and is filing now. BioMarin’s long-acting answer, BMN 333, is a 2029–30 event. The $4B 2027 revenue target set in September 2024 was withdrawn in October 2025 and has not been replaced.

The reported financial collapse is mostly an illusion; the capital-allocation problem is not. FY2025 operating income fell to $409.5M on 12.9% revenue growth, and TTM operating income fell from $776.0M to $315.2M across four quarters. But $493.2M of identifiable one-time charges — the $241.3M Roctavian withdrawal, $221.0M of Inozyme acquired IPR&D, and a $31.0M manufacturing charge — more than account for it. Normalized operating margin is 21–23%, gross margin is rising, and free cash flow grew 50.8% in the year reported profit fell. Returns on operating capital reached 19.4%. The operations are sound. What is not sound is what happens to the cash: ten years and $7.0B of R&D produced $72.7M of cumulative operating profit, and reported ROIC has cleared its cost of capital once in five years.

And the balance sheet has changed category. On 27 April 2026 BioMarin closed the $4.8B all-cash acquisition of Amicus Therapeutics, funded with ~$4.25B of pro-forma gross debt — $850M of 5.500% notes due 2034, a $2.0B term loan B, an $800M term loan A and a $600M revolver, secured by a first-priority lien on substantially all assets. Net cash of $1.45B became net debt of roughly $2.8–3.0B, and annual interest goes from $9.5M toward $224M. Q2 2026, the first quarter including Amicus, reports on 6 August 2026 and is not in this analysis.

What the price implies. At 3.2x EV/sales and ~16x free cash flow the market is underwriting meaningful Voxzogo erosion and giving little credit for the pipeline or for synergy delivery. That is a defensible set of expectations rather than an obviously wrong one — which is why the valuation is interesting without being compelling. The binary that matters most is dated: the ITC’s initial determination on whether Ascendis infringes BioMarin’s patent is expected on or about 21 August 2026.

2. Business Overview

2.1 What the company actually does

BioMarin develops and sells drugs for genetic diseases so rare that, for most of them, no alternative therapy exists. Founded in 1997 and headquartered in San Rafael, California, it employs roughly 3,200 people and generated $3,221.3M of total revenue in FY2025, of which $3,167.8M was net product revenue. The commercial model is the classic orphan one: very small patient populations, very high annual prices, a concentrated prescriber base reachable with a small salesforce, and lifelong therapy.

The scale asymmetry is the point. Median orphan-drug list price runs roughly $219,000 a year against roughly $13,000 for non-orphan drugs — a ~17x gap that has widened over a decade. BioMarin sells $3.2B of product with roughly 500 sales people. A mid-cap specialist competes on equal commercial terms with Sanofi, because there is no primary-care detailing war to lose.

2.2 The portfolio, product by product

Two structural points are easy to miss. First, Aldurazyme is not a franchise BioMarin controls: Sanofi markets it worldwide, and BioMarin recognises revenue on release of product to Sanofi while taking 39.5–50% of net sales — which is why the line is lumpy and fell 25% in Q1 2026. Second, the revenue base is far more international and far more concentrated than a US-biotech framing suggests: roughly 63% of company-marketed revenue is ex-US (US revenue was $1,105.0M of $3,167.8M, ~35%), much of it sold through government tenders and named-patient programmes, and 37% of product revenue comes from three customers. That mix explains both the lumpy quarterly revenue recognition management repeatedly attributes to “order timing,” and the receivables build discussed in section 6.

2.3 How the money is actually made

Revenue is recurring in the strongest sense available in pharmaceuticals: lifelong, non-discretionary therapies for progressive genetic diseases, started young and continued indefinitely, with no therapeutic alternative for payers to negotiate against. Adherence is high because the alternative is disease progression. In the MPS disorders the patient is typically identified through newborn screening or a metabolic-genetics centre, treated at one of a few hundred specialist centres worldwide, and infused weekly for life.

That demand is not merely stable — it is mechanically expanded by state-funded newborn screening, which finds materially more patients than clinical diagnosis does. Pooled screening data across more than 11.6 million newborns put Pompe incidence at 1 in 18,711 births; MPS II, historically estimated at 1 in 100,000–150,000 male births, is found at roughly 1 in 36,000 male births on screening — two to three times the clinical estimate. Additions to the US Recommended Uniform Screening Panel (Pompe 2015, MPS I 2016, MPS II 2022) diffuse through state panels over roughly a decade, providing a slow, non-commercial tailwind to diagnosed prevalence.

Verdict: a genuinely high-quality revenue base — recurring, price-protected, non-discretionary and demographically expanding — attached to a growth asset that is now contested. The business model is sound; the question this report must answer is what the growth asset is worth.

3. Industry Dynamics

3.1 The structure, and why it worked

The orphan-drug model is one of the better neighbourhoods in healthcare, and the numbers support that plainly. Evaluate forecasts orphan drugs will generate more than $400B of annual sales by 2032, over 21% of the global prescription market, up from roughly 15% a decade earlier, on a 2025–30 orphan CAGR above 10% against roughly 7.5% for non-orphan. Fifty-four percent of 2025’s novel FDA approvals were orphan drugs.

The economic logic is straightforward. A tiny population justifies a very high price because total budget impact stays small. The prescriber base is concentrated, so commercial cost is low. Payers rarely restrict access to the only differentiated therapy for a devastating rare disease. Statutory incentives stack on top: seven years of US orphan exclusivity, a 25% §45C tax credit, waived user fees, and — for rare pediatric approvals — a Priority Review Voucher.

3.2 The 2026 policy environment is better for BioMarin than the headlines suggest

This is where most analysis of BioMarin goes wrong, and it is worth being precise.

Section 232 tariffs — orphan drugs are exempt at 0%. Proclamation 11020 (2 April 2026) imposed a 100% ad valorem tariff on patented pharmaceuticals, biologics and APIs, effective 31 July 2026. But Annex IV exempts orphan drugs at a 0% rate, alongside cell and gene therapies, plasma-derived therapies, antibody-drug conjugates and several other classes. Every BioMarin commercial product is orphan-designated. Even absent the exemption, the two Ireland-manufactured products (Vimizim and Brineura, made at Shanbally, Cork) would fall under the 15% EU-origin rate, and only the US-destined share would be affected — on a revenue base that is only ~35% US. The generic claim that a 100% pharma tariff is a major BioMarin risk is wrong on the current text of the proclamation. At the margin the tariff regime is a relative competitive advantage for an orphan-only manufacturer against large-cap pharma. (One caveat: the Annex IV scope should be verified line-by-line, and the exemption is subject to a one-year reassessment.)

IRA Medicare negotiation — close to immaterial. The ORPHAN Cures Act, enacted in the reconciliation bill signed 4 July 2025, widened the IRA’s orphan exclusion from drugs designated for only one rare disease to drugs designated for one or more, provided all approved indications are orphan — effective from the 2028 price-applicability year, as BioMarin’s own 10-K confirms. Three facts stack: only ~35% of revenue is US; BioMarin’s US patients are overwhelmingly paediatric and young-adult inherited-disease patients on commercial insurance and Medicaid, not Medicare; and the entire marketed portfolio consists of drugs whose only approved indications are orphan conditions. The genuinely relevant IRA mechanics are second-order — the Part D redesign’s manufacturer discount and Part B/D inflation rebates capping US list-price increases at CPI. Margin nibbles, not thesis-breakers. (BioMarin does not disclose its Medicare payer mix; this conclusion is an indication-based inference.)

Priority Review Vouchers — restored, and valuable. Authority to grant new rare-pediatric vouchers sunset in December 2024. On 3 February 2026 the Consolidated Appropriations Act, 2026 — incorporating the Mikaela Naylon Give Kids a Chance Act — reinstated it through 30 September 2029. Secondary-market prices are at record highs of roughly $180–205M per voucher (Fortress $205M; Jazz $200M; Denali $195M; Rocket $180M), against ~$100–110M before the lapse. For a rare-pediatric developer that is a ~$200M non-dilutive cash option per qualifying approval, and the clearest positive policy development of the period.

Where policy genuinely cuts against BioMarin. Most-favoured-nation pricing is the live risk. CMS/CMMI proposed two mandatory MFN demonstration models in December 2025 — GLOBE (Part B, proposed from October 2026) and GUARD (Part D, from January 2027) — neither finalised as of August 2026. Critically, Proclamation 11020 grants a 0%-through-January-2029 safe harbour to companies in Annex II that signed an MFN pricing agreement with HHS before 2 April 2026. BioMarin refused, and is instead a founding member of the Midsized Biotech Alliance of America, a ten-company coalition formed in February 2026 with Alnylam, Incyte, Neurocrine, Exelixis and others to contest MFN. That is a defensible strategic choice with an asymmetric downside: if exemptions are re-cut to condition orphan relief on MFN participation, BioMarin is on the wrong side of it.

Two further anti-incumbent changes landed in the same twelve months. Section 6605 of the February 2026 Consolidated Appropriations Act narrowed US orphan exclusivity from “same disease or condition” to “same approved use or indication” — overturning the Eleventh Circuit’s Catalyst decision and requiring FDA to revisit active exclusivities. In Europe, the agreed pharma package (trilogue deal December 2025, adoption slated for autumn 2026) cuts baseline orphan market exclusivity from 10 years to a 9/11/4-year tiered structure, introduces a “Global Orphan Marketing Authorisation” capping indication-stacking at +2 years, permits biosimilar applications during the final two years of exclusivity, and extends the Bolar exemption to procurement tender submissions. EU HTA Regulation joint clinical assessments hit orphan products on 13 January 2028. Both regimes are anti-incumbent by design.

3.3 The capital cycle — and why the timing of the Amicus deal matters

Marathon’s framework asks a single question: is capital entering or leaving? The honest answer is a two-stage one, and it does not flatter BioMarin.

Between 2021 and 2024 capital genuinely exited. Rates rose, the XBI collapsed, the IPO window shut, and a generation of single-asset rare-disease biotechs was starved or wound up. That supply destruction should, with a lag, improve returns for survivors — and it is the window in which BioMarin’s competitive set thinned.

Since 2025 capital has returned hard: venture funding of $38B in 2025 (+28%), the XBI up 40.8%, M&A of $209B in 2025 and $123B by June 2026, and an IPO window already beating four consecutive full years. By Marathon’s logic, capital re-entering a sector is a leading indicator of deteriorating forward returns, not improving ones. One refinement partly saves the picture: rare disease is only ~11% of venture allocation (against 32% for oncology), and capital entering rare disease arrives largely as M&A for commercial assets, which removes competitive supply rather than adding it. The net read is mid-cycle — no capital-cycle windfall, and the only favourable framing is BioMarin as prey rather than as a survivor earning excess returns.

There is a second, sharper observation. A 2011 Morgan Stanley healthcare industry primer argues the orphan exception exists because “these diseases are small enough that the first to market drug with a meaningful disease impact is often the only player.” The same document supplies its own counter-evidence: “once a company is seen as having potential in an untapped disease, several other biotech companies usually follow its lead” — hepatitis C direct-acting antivirals went from zero programmes in 2005 to roughly thirty by 2010. Achondroplasia going from one approved drug to three credible entrants in five years is that mechanism operating on what was supposed to be an orphan exception. (Morgan Stanley Research, Blue Paper: The US Healthcare Formula, 16 June 2011 — a fifteen-year-old document, cited for structural framing only, not as current data.)

3.4 The disruption that did not happen

The 2020-vintage bear case on BioMarin was that one-time gene therapy would obsolete chronic enzyme replacement. In 2024–2026 that thesis died in public: BioMarin wrote off Roctavian and pulled it from the market; Pfizer withdrew Beqvez; bluebird bio — three approved gene therapies, 57 patients ever treated — sold for roughly $30M; and Sarepta’s Elevidys was hit with patient deaths, a clinical hold and a Boxed Warning. The chronic infusion annuity is more durable than anyone believed five years ago, and that durability is now evidenced rather than assumed. This is a real and under-appreciated positive for the $2.24B enzyme base. The longer-horizon threat is in vivo gene editing — Intellia’s lonvo-z hit its Phase 3 endpoint in April 2026 — but BioMarin’s diseases require gene addition plus blood-brain-barrier crossing, which puts that threat in the 2032-and-beyond bracket.

Verdict: rare disease remains a structurally GOOD industry, but it is getting worse at the margin, and the deterioration is concentrated precisely where BioMarin lives. The orphan growth premium is forecast to compress from a ~2.5–3 point spread over non-orphan to roughly one point by 2030 — the excess-return signature of a niche that has been arbitraged. Legal protection is being trimmed on both sides of the Atlantic in the same year, and the IRA carve-out is a political artefact rather than a right. But the counter-argument is strong and must be stated: this describes a maturing industry, not a broken one; the most-feared disruption has been falsified; and 2026 policy — tariff exemption, widened IRA carve-out, restored vouchers — has on net been better than feared for this niche. A rare-disease pure-play that is ~65% ex-US, essentially unexposed to Medicare negotiation and tariff-exempt by product class is structurally less policy-exposed than large-cap pharma. The bear case here is not that the model is broken. It is the narrower claim that the excess returns have been competed away, and what remains accrues to the most defended, least glamorous assets.

4. Competitive Position

4.1 Naming the moat, in Greenwald’s taxonomy

Greenwald recognises three genuine competitive advantages: supply/cost advantages, demand-side customer captivity, and economies of scale reinforced by captivity. Applied honestly to BioMarin, the answer is uncomfortable.

There is no economies-of-scale advantage — BioMarin’s cost base is higher per dollar of revenue than the industry needs, with SG&A at 35.8% and R&D at 28.6% consuming 64% of revenue. There is modest customer captivity — a patient stable on a lifelong weekly infusion is genuinely reluctant to switch, and the treating-centre relationship is sticky — but it is defensive only: it protects the installed base and does nothing to stop a rival winning new patients. What BioMarin actually has is closest to a supply-side advantage of an unusual kind: a minimum efficient scale that exceeds the size of the market.

Recombinant enzymes are complex glycoproteins requiring specific glycosylation — mannose-6-phosphate for lysosomal targeting — to function at all. Building comparable manufacturing capacity and obtaining regulatory approval is a multi-year, multi-hundred-million-dollar undertaking. Against a population of a few thousand patients worldwide, no rational competitor can earn that back.

The decisive evidence is the absence of an attack. Naglazyme’s last composition patent expired in November 2023 and Aldurazyme’s in November 2020; neither has unexpired US or EU orphan exclusivity. Vimizim’s US biologic exclusivity expires in 2026 and both its orphan exclusivities are already gone. All three have continued to grow, and not one biosimilar has ever been developed. Roughly $1.49B of revenue — 46% of the total — is defended by nothing but the unattractiveness of the prize.

4.2 Running the two Greenwald tests

The market-share-stability test produces a split verdict that is itself the finding. Where nobody attacked — MPS VI, MPS IVA, MPS I — share has been stable at essentially 100% for two decades. Where anyone attacked, share collapsed: Kuvan fell 78.5% from its FY2019 peak of $463M to $99.6M once generic sapropterin arrived. Stability in the absence of an attacker is uninformative; the one clean natural experiment available is damning.

The ROIC test fails outright. A company with six essentially uncontested monopolies, ~$415,000 annual pricing and 81% gross margins produced returns on invested capital of 2.75%, 2.81%, 6.19% and 4.57% across FY2022–FY2025 — never clearing its cost of capital. Greenwald’s test is that a genuine barrier shows up as sustained excess returns. Here it does not, because the operating advantage is consumed by the cost base before it reaches capital. (The important qualification developed in section 6: on operating capital, normalized returns reached 19.4% in 2025. The franchise economics are fine; the enterprise economics are not.)

4.3 Voxzogo — the contested asset, and the crux of the thesis

Achondroplasia affects roughly 1 in 25,000 live births, some 250,000 people worldwide, about 80% from de novo FGFR3 mutations. The historic alternative to drug therapy is limb-lengthening surgery, with complications in up to 70% of patients. Voxzogo, approved in the US in November 2021 and extended to all ages in October 2023, reached $926.9M in FY2025 with more than 5,000 children treated across 55 countries. It is BioMarin’s only genuine growth asset, and it now faces two credible attackers.

Ascendis Pharma — Yuviwel (navepegritide). FDA-approved 27 February 2026 under accelerated approval, launched 6 April 2026. Once-weekly versus Voxzogo’s once-daily injection, in children aged 2 and over. Efficacy is essentially at parity (+1.49 cm/yr annualised growth velocity against Voxzogo’s +1.57), and it is priced at a 20% premium — $498,225 per year. It carries its own orphan exclusivity to 2033 and enrolled more than 60 children across more than 35 prescribers in its first four weeks. Channel work suggests roughly 40% of injected patients are switch candidates.

BridgeBio — infigratinib. An oral, once-daily FGFR3 inhibitor. Phase 3 PROPEL 3, published in NEJM on 29 June 2026, reported +1.74 cm/yr LS mean (+2.10 observed) — the largest height-velocity effect ever recorded in the disease — plus the only statistically significant body-proportionality and arm-span results in the field. NDA filing is guided for Q3 2026 with launch in early-to-mid 2027, backed by roughly $1B raised in July 2026.

BioMarin’s answer arrives too late. BMN 333, a long-acting CNP, has only just entered Phase 2/3, with an annualised-growth-velocity readout in 2027 and approval targeted around 2029–30 — and it must then win a fresh Phase 3 superiority trial against BioMarin’s own Voxzogo. A failure there would simultaneously sink BMN 333 and publicly validate competitor parity. That is a four-year defensive gap during which Ascendis is already selling and BridgeBio will launch.

Management’s defence, offered across four earnings calls, rests on four planks: long-term safety-data depth, the infant segment, the ex-US footprint, and patient inertia. Only the infant label is a hard barrier — Voxzogo is approved from four months of age while Yuviwel starts at two years — and it holds only until competitors secure infant labels of their own. Management has itself conceded competitor efficacy is comparable, telling analysts it assumed “two competitors and comparable data to Voxzogo” in its scenarios. The switching defence rests on proprietary market research that ranks “convenience is the third” priority — against which one analyst noted that a competitor expects about half of the patients in its early achondroplasia trials to be Voxzogo-experienced. A rival’s own trial-enrolment mix is harder evidence of switching appetite than the incumbent’s market research.

There is also an unresolved legal binary. BioMarin is litigating US Patent RE48,267 against Ascendis at the International Trade Commission; the administrative law judge’s initial determination is expected on or about 21 August 2026, with a full-Commission decision around 21 December if review is granted. BioMarin has parallel actions in the European Unified Patent Court and won a procedural victory at the Federal Circuit on 26 March 2026, which preserved its district-court case. Management’s own “high case” 2027 scenario is explicitly predicated on “successful intellectual property events.”

4.4 The pattern that should worry a long-term owner

BioMarin sells injectables into markets that keep being won by pills and long-acting conjugates. This is not bad luck three times; it is a repeated failure to anticipate the direction of modality competition in its own indications.

Verdict: a weak and narrow competitive advantage. BioMarin owns a collection of small, genuinely durable, low-return annuities — defended by a market too small to attack rather than by anything the company built — stapled to one growth asset now contested by a better-dosed injectable and a coming oral. The barriers that exist protect the old, patent-free products beautifully and the new, growing ones not at all. That inversion is the central structural fact about this company. It is not a moat worth paying a premium for; it is a moat worth paying a discount to own, which is roughly what the market is doing.

5. Growth History and Forward Opportunities

5.1 The record

Revenue compounded at 15.0% a year over four years — a genuinely good record. But the composition is the story: Voxzogo went from nothing to $926.9M and supplied roughly two-thirds of the entire increase. The enzyme base grew in the mid single digits, and Kuvan actively shrank. Growth has been almost entirely one product, and that product’s incremental contribution has now gone to approximately zero.

5.2 Where the growth was supposed to come from, and what happened to it

Management’s September 2024 Investor Day set a $4B revenue target for 2027 and a >$1.25B operating cash flow target. The CFO reaffirmed the $4B figure in May 2025, hedged it in August, and withdrew it in October 2025, replacing a point target with an unspecified range whose low end — roughly $3.65B — was simply then-current sell-side consensus, and whose high end required “a significant delay in the competition, for example, successful intellectual property events.” The cash-flow target died with it: “that $1.25 billion was tied to the $4 billion.” Management has said it does “not plan to provide additional estimates of 2027 revenues.” A flagship target set, reaffirmed and abandoned inside thirteen months is a material mark against forecasting credibility, and it removed the anchor investors had.

FY2026 guidance now stands at $3.825–3.925B of revenue and $4.85–5.05 of non-GAAP EPS. The headline is roughly 20% growth — but $500M of that is acquired Amicus revenue. Organic growth is roughly 3–6%. Tellingly, when the $500M of acquired revenue was added, the EPS guide was cut by $0.10 at both ends. Two-thirds of the year’s EPS is guided into the second half off a Q1 base of $0.76 — a substantial execution risk in itself.

5.3 What is actually left

Voxzogo label expansion is the most credible remaining organic driver, and it is genuinely alive: the Phase 3 in hypochondroplasia met its primary endpoint in May 2026, and the FDA accepted an sNDA for full approval in achondroplasia with a PDUFA date of 28 February 2027. But the CANOPY programme has narrowed — trials in Turner syndrome, SHOX deficiency and ACAN deficiency were discontinued in March 2026 after slipped-capital-femoral-epiphysis events in investigator-sponsored studies, and Turner and SHOX were quietly dropped from the strategy language between the 10-K and the 10-Q.

The acquired Amicus franchises (Galafold in Fabry, Pombiliti + Opfolda in late-onset Pompe) add roughly $500M of 2026 revenue and a new therapeutic area. The pipeline — BMN 333, BMN 351 in Duchenne (5.0% mean absolute dystrophin at week 25 in the 9 mg/kg cohort), and BMN 401 for ENPP1 deficiency from the Inozyme acquisition — is real but distant. Newborn screening provides a slow, reliable tailwind to diagnosed prevalence across the MPS franchises.

Verdict: the historical growth was high quality but is now substantially spent. It came from one genuinely innovative product entering an untreated disease — the best kind of growth there is. What replaces it is $500M of acquired revenue bought with $5.2B of shareholders’ money, plus roughly 3–6% organic growth, plus a pipeline whose most important asset arrives after the competitive battle it is meant to fight has been decided. The company has moved from earning its growth to buying it, and the price paid is the subject of section 7.

6. Financial Quality

6.1 The reported numbers say the business broke. They are wrong.

On the face of the income statement FY2025 looks like a business losing control of its costs. Revenue grew 12.9% to $3,221.3M while operating income fell 13.7% to $409.5M, and the operating margin compressed from 16.6% to 12.7%. The trailing-twelve-month series is worse: TTM operating income fell from $776.0M at the Q2 2025 balance date to $315.2M at Q1 2026 — a $460.8M collapse — while TTM revenue rose from $3,063.6M to $3,242.3M. A 78%-gross-margin business whose profit halves while sales grow is, on its face, a broken one.

It is not. Of that swing, $493.2M is identifiable non-recurring charges — more than the entire decline:

Normalized for these, TTM operating margin is 24.9%, not 9.7%.

But the naive add-back is also misleading. Normalized quarterly operating margins run 30.0% and 33.5% in H1 2025, then 22.5%, 22.5% and 21.0% in Q3 2025, Q4 2025 and Q1 2026. H1 2025 was the anomaly, not the norm — the trough between a roughly 395-person reduction in force in 2024 and the cost re-inflation that followed. The defensible normalized run-rate is 21–23%, neither 33% nor 12.7%. That estimate is corroborated independently: BioMarin’s own reported non-GAAP operating margin was 23.3% in FY2025 (down from 28.6%) and 24.3% in Q1 2026 (against 35.7% a year earlier).

Two things follow. Anyone reading the GAAP optics as operational decay is wrong. And anyone accepting management’s “approximately 40%” non-GAAP target at face value is also wrong: that measure excludes $181.4M of stock-based compensation — 5.6% of revenue and an entirely real cost of employing people. The truth sits between, and the evidence puts it at 21–23%.

6.2 Gross margin is rising, not falling

The reported decline from 79.7% to 77.7% is an artifact of the $119.2M Roctavian inventory write-off sitting in cost of sales. Excluding it, FY2025 gross margin was 81.4% — up 170 basis points and the highest of the five-year period (74.5%, 76.0%, 78.0%, 79.7%, 81.4% normalized). Voxzogo is margin-accretive, and no Voxzogo royalty obligation is disclosed anywhere in five years of 10-K filings. The genuine swing factors in cost of sales are idle-plant absorption at Novato and Shanbally and periodic write-offs — lumpy rather than trending.

6.3 The growth engine has stalled

Strip Q4 2025 — best explained by a large contracted government order and US stocking that did not recur — and the franchise has been flat at roughly $218–221M a quarter for five consecutive quarters.

Management’s explanation for the gap between patients and revenue — patients on therapy grew “more than 20%” year over year in Q1 2026 while revenue grew 2.9% — is “entirely order timing.” That explanation has now been offered for Q4 2024, Q1 2025, Q2 2025, Q3 2025 and Q1 2026. It is plausible for a tender-driven business selling in 55 countries, and it is also unfalsifiable from outside. It sits awkwardly beside CEO Alexander Hardy’s separate disclosure that two national reimbursement renegotiations land in 2026, each producing “a price reset on your entire population” — which is, definitionally, a price driver, in the same period the CFO described as having “no significant price drivers.” We flag the inconsistency rather than resolve it. Q2 2026 reports on 6 August 2026 and will test it directly.

6.4 Cash generation is the genuinely good news

In the very year reported operating income fell 15.4%, operating cash flow rose 44.5% and free cash flow rose 50.8%. Free cash flow of $717.0M against an ~$11.7B market capitalisation is roughly 16x price-to-free-cash-flow — a materially cheaper picture than the 44x trailing GAAP P/E. Capital intensity is low at 3.2% of revenue; stock-based compensation has fallen 12.4% in dollars over two years; dilution runs ~1.2% a year with no primary equity issuance in five years.

6.5 Quality of earnings — where the bodies may be buried

Inventory. Days sales of inventory stand at 644 days as reported and 772 days against cost of sales excluding the Roctavian write-off, up from 572 days in 2021. BioMarin’s disclosed policy is to capitalise pre-launch and pre-qualification manufacturing costs where recoverability is judged “probable” — precisely the policy that produced the $119.2M Roctavian inventory write-off, product manufactured for a launch that never came. With more than two years of inventory on hand and the growth product decelerating, this is the most likely location of the next write-off.

Receivables. Accounts receivable grew 37.5% in FY2025 against 12.9% revenue growth; days sales outstanding rose from 83 to 89. With roughly 63% of company-marketed revenue ex-US and ~30.9% of revenue ($980M: Latin America $435.5M, rest-of-world $544.5M) sold into lumpy government tenders, lengthening collection is a real economic cost and a live sovereign-credit and FX exposure. In Brazil, a slice of that revenue rests on judicialização — court-ordered access — rather than payer decisions, given CONITEC’s rare-disease cost-effectiveness threshold. FY2026 guidance explicitly “reflects an allowance for a modest amount of disruption” in the Middle East, unquantified.

An unexplained reclassification. In Q1 2026 finished goods fell $270.5M while work-in-process rose $253.6M, in a quarter with only $195.0M of total cost of sales. The offsetting magnitudes suggest a reclassification rather than production activity; the filing does not explain it.

Comparability. FY2020 carried a $903M deferred-tax valuation-allowance release against a $43M operating loss, rendering any five-year EPS growth rate meaningless. The FY2024 10-K also recast FY2023 and FY2022, moving ~$27.7M from SG&A below the operating line. Build the trend from operating cash flow, not reported EPS.

6.6 Returns on capital — the finding that reframes the company

Against a cost of capital of roughly 8.5–9.5%, BioMarin has cleared its hurdle once in five years. On that evidence alone it looks like a chronic value destroyer.

But the reported figures carry a denominator stuffed with capital that is not working: a large cash balance, goodwill and intangibles from prior deals, and a ~$1.5B deferred tax asset — 24% of book equity — that is a legacy of the 2020 valuation-allowance release rather than an operating asset. Measured on operating capital and normalized earnings, returns are 7.2% (2023), 14.6% (2024) and 19.4% (2025).

That distinction is the most important analytical finding in this report, and it changes the diagnosis. BioMarin’s underlying operating business earns respectable returns. The reason shareholders have not seen them is what management does with the capital stacked on top. This is not a bad business with a good story; it is a decent business with a poor capital-allocation record — a different, and more actionable, conclusion. Note also that book value of $32.14 a share becomes roughly 2.7x on tangible book once the deferred tax asset is excluded, against the headline 1.87x P/B.

Verdict: economics do improve with scale at the operating level — gross margin has risen every year for five years, normalized operating margin is 21–23%, cash conversion is strong, and returns on operating capital reached 19.4%. But the reported return on total capital has cleared the cost of capital once in five years, because capital keeps being added faster than it earns. The financial quality of the operations is good; the financial quality of the enterprise is mediocre, and the gap between them is management’s doing.

7. Capital Allocation

This is the section on which the investment case turns, and the evidence is unusually clear.

7.1 The twenty-year scoreboard

Ten years and $7.0B of research spending produced $72.7M of cumulative operating profit. Fifteen years and $9.0B produced a cumulative operating loss. Cumulative stock-based compensation is 24 times cumulative operating profit. The $1,027M of cumulative net income over the decade is largely illusory: $854M of it is 2020 alone, a year in which a $903M deferred-tax valuation-allowance release was booked against a $43M operating loss.

These are not the numbers of a company that has converted scientific capability into shareholder value.

7.2 Roctavian — the largest destruction of capital in company history

BioMarin invested at least $1.6B — the true figure is higher, since the program-level R&D disclosure was discontinued after 2022 — and recovered roughly four cents on the dollar. At a $2.9M list price, $67.8M of lifetime revenue implies roughly 25 to 40 patients dosed worldwide across the product’s entire commercial life.

The sequence is worse than the arithmetic. A Complete Response Letter in 2020; US approval in 2023; commercial retrenchment to three countries in 2024 while management was still guiding the program to profitability by end-2025; an announced divestiture search in October 2025; no buyer found; and outright market withdrawal committed on 17 December 2025. The $241.3M of charges is the residue, not the loss.

Nor was it the first. Prosensa, acquired in 2015 for $680M, produced $797.8M of IPR&D impairments in 2015–16 — more than the purchase price.

7.3 Amicus — the largest bet, made at the worst moment

On 19 December 2025 BioMarin agreed to acquire Amicus Therapeutics for $14.50 a share in cash, a 33% premium to the last close and 58% to the 60-day VWAP. The deal closed 27 April 2026.

The headline “$4.8 billion” understates what was actually paid. The 8-K/A filed 13 July 2026 discloses total consideration of $5,323.6M: $4,563.5M to shareholders, $142.6M to cash out restricted and performance share units, $62.9M for pre-combination options, $432.9M to repay Amicus’s own debt (including a $12.0M prepayment penalty), and — notably — $121.6M of Amicus’s own banker and legal fees paid by BioMarin, including a roughly $50.7M success fee to Centerview. Net of $293.5M of cash acquired, the enterprise value paid was approximately $5.03B for $634.2M of FY2025 revenue — about 7.6x trailing and 6.5x forward sales, roughly 2.4x BioMarin’s own 3.1x multiple, entirely in cash and debt.

BioMarin’s own filed pro-forma is the most damning document in the file. Under Article 11, on a FY2025 pro-forma basis the combined company swings from BioMarin’s actual +$348.9M of net income and $1.80 of diluted EPS to a net loss of $205.9M, or $(1.07) a share — with operating income going from +$409.5M to negative $64.1M. The bridge is $390.3M of incremental intangible amortisation, $180.6M of incremental interest and $70.1M of inventory step-up. Pro-forma SG&A runs at 41.2% of revenue. Management markets the transaction as accretive to non-GAAP earnings; its own filed statements show it turning a profitable company into a loss-making one on a GAAP basis.

On price, there is no margin of safety. Using Amicus management’s own risk-adjusted projections and its banker’s terminal assumptions, BioMarin’s internal rate of return with no synergies is roughly 9.9% — at or below Amicus’s own standalone weighted average cost of capital of 10.25–12.00%. BioMarin paid above the top of Centerview’s discounted-cash-flow range of $11.75–$14.35 and well above the top of the trading-comparables range of $7.25–$11.95. The deal requires $100–200M of run-rate synergies simply to clear the cost of capital — and BioMarin has never disclosed a synergy number, including affirmatively electing not to present one in the 8-K/A where Article 11 permitted it.

The purchase-price allocation reveals where the bet actually sits. Of the consideration, $3.0B was allocated to Galafold on an eleven-year life — exactly the January 2037 US loss-of-exclusivity date — and $1.45B to Pombiliti + Opfolda, which is 12.9x its $112.5M of FY2025 sales. That single assumption carries roughly 29% of the deal cost, and it is placed on the number-three product in a market Sanofi has dominated for two decades.

The financing:

A $3.7B bridge facility was never drawn but still cost $22.8M in fees. Run-rate interest expense goes from $9.5M in FY2025 to roughly $224M a year, and BioMarin simultaneously forgoes the ~$74.9M of interest income its net cash was earning — a pre-tax swing approaching $290M. Net cash of $1.45B at the end of 2025 becomes net debt of roughly $2.8–3.0B. The facilities are secured by a first-priority lien on substantially all assets and carry maintenance covenants of total net leverage ≤3.50x and interest coverage ≥3.00x. The $600M convertible is deeply out of the money and is therefore a cash obligation falling due in May 2027.

The leverage itself is manageable, and this report declines to make it the headline risk. At roughly 2.9x gross and 2.25x net leverage against a 3.50x covenant that binds only the $1.4B Term Loan A and the revolver — the $2.0B Term Loan B is covenant-lite — with interest coverage near 7.2x, the balance sheet has capacity. Even a severe stress case stays inside the covenants, and the stated sub-2.5x target becomes close to automatic once the $600M convertible is repaid in May 2027. The risk in this transaction is the price paid and the Pompe ramp, not the debt.

The strategic logic — diversifying away from a contested Voxzogo by adding Fabry and Pompe franchises — is coherent on its face. The objection is not to the direction but to the price, the funding and the timing: 7.6x trailing sales, all in cash and debt, at a 33% premium and above the top of its own banker’s valuation range, at the top of a biotech M&A cycle, by a management team whose two previous large capital commitments returned four cents on the dollar and less than nothing respectively. Marathon’s asset-growth anomaly warns precisely against this pattern. One telling market verdict: the +17.7% announcement pop fully round-tripped to a five-year low of $49.67 on 18 May 2026, three weeks after closing — below the pre-announcement price.

What BioMarin actually bought. Galafold (migalastat, oral, Fabry) generated $521.7M in FY2025, growing 34.5% over two years while Sanofi’s Fabrazyme and Takeda’s Replagal grew roughly 2–3% — it is the only Fabry asset growing double digits, at roughly 22% of a ~$2.2–2.4B global market. Its competitive position has improved since the deal was struck: Sanofi’s oral venglustat failed both Phase 3 trials (February and July 2026), Idorsia’s lucerastat failed in 2021, and Sangamo — whose Fabry gene therapy was the most advanced in the world with a nearly complete BLA — filed Chapter 11 in June 2026 and is selling that programme to Astellas for $25M upfront. Nothing credible is coming at Galafold before 2030. Its real constraints are different and less discussed: only 35–50% of Fabry patients are amenable to migalastat, which implies roughly half of its structural runway is already consumed; its EU orphan exclusivity lapsed in May 2026, leaving 59% of the franchise dependent on country-by-country patents and SPCs; and its US approval remains an accelerated approval that has not been converted to full approval in eight years. The US 2037 date was locked only when Aurobindo and Lupin settled — on 19 December 2025, the day the merger was announced.

Pombiliti + Opfolda (late-onset Pompe) generated $112.5M, growing 56% at constant currency — genuinely taking share, corroborated by Sanofi’s Pompe franchise decelerating from +9.7% to −0.6% by Q2 2026. But it holds only ~7.1% of a flat ~$1.59B market, in a disease with just 5,000–10,000 patients worldwide, which makes it a zero-sum switching contest against an entrenched incumbent. Critically, the US label restricts it to patients “not improving on their current enzyme replacement therapy” — it cannot be prescribed to a treatment-naive patient — while the EU label carries no such restriction, which is why 57% of its revenue is ex-US. No filing to remove that US restriction has been identified. Paying $1.45B, or 12.9x sales, for this asset is the single most aggressive assumption in the transaction.

And the synergy pool is structurally small. Amicus ran a $634M revenue business with 511 employees — more than $1.2M of revenue per head. There is little overhead to remove: the announced reduction is 58 New Jersey positions, about 11% of Amicus headcount, effective from 7 August 2026, with $72.0M of lease obligations running to 2034. This deal has to be justified on revenue, not on cost.

7.4 No buybacks, no dividend

BioMarin has repurchased zero shares between 2021 and Q1 2026 (a single $50M convert-linked purchase in 2020) and has never paid a dividend. Both new debt instruments restrict dividends. Share count has risen from 172.6M in 2016 to 192.3M, roughly 1.2% a year. Equity-plan overhang stands at roughly 38.9M shares, about 20% of shares outstanding, after shareholders approved 8.0M new shares in 2025 and a further 7.65M in June 2026.

Management has been explicit that this is deliberate. The CEO has argued that business development is “the most highly correlated with stock appreciation” and that assets “are worth more in our hands.” The stated $4–5B of firepower has now been spent — on Inozyme and Amicus.

7.5 The incentive plan — what management is actually paid to do

This is the sharpest evidence in the file, and it comes straight from the proxy.

The 2025 annual bonus paid on a non-GAAP EPS “result” of $4.83. The figure BioMarin reported to shareholders was $3.15 — below the plan’s own $3.67 threshold, at which the metric pays zero. The Compensation Committee bridged the gap by adding back the Roctavian write-off ($0.46), the Inozyme acquired-IPR&D charge ($1.12), and — remarkably — the cost of its own above-target bonus ($0.10). Revenue missed its target ($3,221M against $3,223M) and was paid at 100%. The pool funded at 130%, as it had in 2024.

The long-term plan moved the same direction. For the 2025 grant, Core Operating Margin was deleted as a performance metric and replaced with Revenue CAGR and an “Innovation” measure defined as incremental revenue from new products — in the same year management borrowed $4.25B to buy revenue. A plan that pays for revenue growth and explicitly stops paying for margin, at a company whose problem is that revenue growth has never converted into returns on capital, is not aligned with shareholders; it is aligned with the acquisition.

Ownership reinforces the point. Officers and directors together hold fewer than 1% of shares — 512,000 shares outright against 193.3M outstanding. Five-year total shareholder return took $100 to $67.77, against $124.75 for the Nasdaq Biotechnology Index. CEO pay for 2025 was $21.9M, up 48% year over year.

7.6 Insiders and the activist

There have been no open-market purchases (transaction code P) by any insider in 24 months. Against that, $4.38M of sales across five individuals, of which four of the five 2026 sales were not made under 10b5-1 plans — including the Chief Legal Officer and the Chief R&D Officer, the latter selling on 26 February 2026, the day the 10-K disclosing the Roctavian write-off was filed. The CEO and CFO have not sold. The overall read is not a scandal, but it is a tilt: no conviction buying, and discretionary selling by senior officers around a bad disclosure.

Elliott Management, whose ~$1B stake in November 2023 drove the sharpest positive repricing of the five-year period and produced a cooperation agreement and a Strategic & Operating Review Committee, fully exited in Q2 2025 — verified from its own 13F filings, which show 3,500,338 shares from Q1 2024 through Q1 2025 and nothing thereafter. It left before the Roctavian withdrawal and before the Amicus deal. Its three settlement directors remain on the board. Investors who view the Elliott involvement as ongoing governance protection should note that the activist is gone.

7.7 Scoring management against its own targets

At the September 2024 Investor Day, reaffirmed by 8-K on 24 September 2024, management committed to $4B of revenue in 2027, a 40% non-GAAP operating margin starting in 2026, more than $1.25B of operating cash flow from 2027, and a mid-teens revenue CAGR to 2034.

Verdict: management has not allocated capital intelligently, and the incentive structure explains why. The operating business is run competently — margins normalize in the low twenties, gross margin rises, cash converts. But $9.0B of research spending over fifteen years has produced a cumulative operating loss; the flagship gene-therapy programme returned four cents on the dollar; the prior large acquisition was impaired for more than its purchase price; there have been no buybacks and no dividend; insiders own less than 1% and have bought nothing in two years; the activist who forced the reset has sold and gone; every headline target set in September 2024 has been missed or withdrawn; and the response has been to lever the balance sheet with $4.25B of secured debt to buy $740M of revenue at 7.1x sales, while deleting operating margin from the long-term incentive plan and replacing it with revenue growth. On the single question this report exists to answer — has management earned the right to be trusted with $5.2B of shareholders’ capital — the honest answer is no.

8. Changes and Headwinds — Last Two Years

The past twenty-four months contain more strategic change than the prior decade, and almost all of it is a reaction to a growth engine coming under attack.

The headwinds, ranked. First and largest, the Voxzogo competitive assault — two entrants, one launched and one filing, against a franchise that has gone flat. Second, the credibility deficit: a flagship target set and withdrawn in thirteen months, an EPS guide missed on a discretionary write-off, a margin target that went backwards in its ramp year and was then dropped from the guidance table. Third, the leverage, which converts a formerly forgiving balance sheet into one that amplifies operational disappointment. Fourth, the unresolved DOJ subpoena over sponsored genetic testing for Vimizim and Naglazyme. Fifth, MFN pricing, where BioMarin has chosen to fight rather than settle.

The tailwinds are real and should not be dismissed. The gene-therapy threat to enzyme replacement has been empirically falsified. The 2026 policy package — orphan tariff exemption, widened IRA carve-out, restored $200M priority review vouchers — is net favourable. Voxzogo won its hypochondroplasia Phase 3 and has a PDUFA date of 28 February 2027 for full approval. And the ITC ruling due around 21 August 2026 is a genuine, dated call option on excluding the lead competitor from the US market.

Verdict: on balance these developments weaken the thesis. The single asset that justified a growth multiple is now contested by better-dosed and better-tolerated alternatives, management’s forecasting credibility is impaired, and the balance-sheet cushion has been spent. The offsetting positives are real but mostly defensive — they protect the annuity, not the growth.

9. Risk Analysis

The two risks that dominate are (1) and (2), and they are correlated: if Voxzogo erodes, the revenue base that the $500M cost programme is meant to leverage shrinks, and the margin target becomes unreachable. Risk (4) is the amplifier — with net debt of ~$3B against an ~$11.7B market capitalisation, the equity’s sensitivity to enterprise value rose roughly 42%, from 0.88x to 1.25x. A company that previously absorbed disappointment with net cash now transmits it to shareholders.

Catastrophic-loss risk is low. This is a profitable, cash-generative business with a diversified, durable annuity base, no covenant pressure at current earnings, and no single product representing more than 29% of revenue. A total loss is very hard to construct. But the downside is real: the bear case in section 10 implies roughly $23–33 a share, and the comparable nobody wants to discuss is Sarepta, whose market capitalisation fell from $11.8B to $2.26B in a single year and which now trades at 1.23x EV/sales.

10. Valuation Discussion — Embedded Expectations

10.1 Which multiple is the right one

BioMarin presents three multiples that tell three different stories, and adjudicating between them is the central valuation task.

The percentile trap deserves emphasis. A 7th-percentile price-to-sales ratio looks like a screaming signal. It is not. The denominator of that percentile is a decade in which BioMarin was valued as a growth company at 7.5–9.1x EV/sales. What has changed is that the market no longer believes the growth story — which is a re-rating to a correct multiple, not an overshoot. Against a 22% normalized operating margin, 3.5x sales equates to roughly 16x EBIT. That is an ordinary multiple for an ordinary-growth business.

10.2 Normalized earnings power

Standalone normalized operating margin is 21–23%, centred on 22%, as derived in section 6. Adding Amicus pre-synergy takes the combined figure to roughly 18.9% — the acquisition is margin-dilutive before cost savings. Management’s target is “approximately 40%,” from a $500M cost programme announced in September 2024.

That target should be scored low credibility for 2026, on management’s own numbers: the non-GAAP operating margin went backwards in the ramp year, from 28.6% in FY2024 to 23.3% in FY2025, and stood at 24.3% in Q1 2026 against 35.7% a year earlier. Q1 2026 delivered $0.76 of non-GAAP EPS against a full-year guide of $4.85–5.05, requiring roughly 5.5x the Q1 run-rate across the remaining three quarters — in a business whose margin seasonality historically runs the wrong way, with Q1 strongest. And the operating-margin line was dropped from the May 2026 guidance table altogether.

10.3 Comparables

BioMarin sits at the bottom of the profitable cohort alongside Incyte and Jazz, far below Neurocrine, United Therapeutics, Alnylam and Vertex. The discount is mostly earned: the lowest ROIC in the group by a wide margin, a normalized operating margin of 22% against peers at 32–46%, a stalled growth asset, zero buybacks in five years, and an exclusivity expiry in 2026. Where the discount may be too large is that BioMarin’s problem is a bad cost structure attached to excellent product economics — which is, in principle, fixable. The reverse condition would not be.

Sarepta is the comparable nobody wants to name, and it is the floor case: a rare-disease company whose lead franchise broke, now at 1.23x EV/sales with a market capitalisation that fell from $11.8B to $2.26B in a year.

10.4 Embedded expectations — what the price requires

At $60.02, with a pro-forma enterprise value of roughly $14.5B, an 8.5% weighted average cost of capital and a 2% terminal growth rate, today’s price solves for either:

• a 25% sustainable operating margin with a 5.4% revenue CAGR, or
• a 22% sustainable operating margin with a 9.1% revenue CAGR.

Organic revenue grew 2.8% in the most recent quarter, and organic FY2026 guidance implies 3–6%.

On the cost of capital: a mechanical CAPM using the observed beta of 0.618 would give roughly 7.2%, and that is rejected here. The low beta is an artifact of a decade of idiosyncratic decline (alpha −0.228), not of low business risk, and it predates a balance sheet that now carries $4.25B of secured debt. A company that has earned 2.75–6.19% returns on invested capital for seven consecutive years does not have a 7.2% cost of capital. 8.5% is used, with 7.5%/9.5% sensitivities.

What the market is underwriting: roughly a 25–28% sustainable operating margin on 2–5% growth — splitting the difference between management’s promise and management’s record, weighted toward the record. That is neither obviously too optimistic nor obviously too pessimistic, which is precisely why the stock is not a compelling value.

What the market is getting right: that Voxzogo will lose share; that the 40% margin target will not be met on schedule; that a company with this capital-allocation record should not be paid a growth multiple.

What the market may be getting wrong, in both directions: it may under-appreciate the genuine durability of the un-attacked $1.49B enzyme annuity and the favourable 2026 policy turn (tariff exemption, IRA carve-out, restored vouchers); and it may under-appreciate how much operating leverage works in reverse if Voxzogo actually declines rather than merely flattens, given $224M of fixed annual interest.

10.5 Scenarios

The current price of $60.02 sits in the upper half of the base case. The distribution is roughly symmetric in percentage terms but the bear case is better-supported by the recent evidence than the bull case is.

10.6 Sum of the parts — the most revealing framing

Valuing the pieces separately produces a total enterprise value of $8.55–16.63B, or roughly $29–71 a share. The market sits at the 74th percentile of that band. Two findings emerge that no other lens surfaces.

First, backing out the enzyme annuity, the acquired Amicus franchises and corporate overhead leaves $3.4–8.5B attributable to Voxzogo plus the pipeline — that is 3.4x to 8.5x Voxzogo’s sales. The market is paying a full multiple for the one asset that has stopped growing and faces two entrants. The common bullish framing that “you get Voxzogo for free at this price” is exactly backwards.

Second, the corporate cost layer carries a negative value of roughly $2.5–3.5B. BioMarin’s parts are worth materially more than BioMarin. That is an unusual and clarifying result, and it is precisely what a $500M cost-reduction programme is an admission of. It is also why the company is more interesting as an acquisition target than as an acquirer.

10.7 The Amicus arithmetic

BioMarin paid 7.6x trailing and 6.5x forward sales for Amicus (FY2025 revenue $634.2M) — roughly 2.4x its own 3.1x multiple, entirely in cash and debt. The financing costs $267M pre-tax annually, or $1.11 a share. For the transaction to be merely EPS-neutral, Amicus must therefore deliver an operating margin of roughly 36% — implying $230–267M of synergies, about a third of acquired revenue.

Management guided the deal as “accretive in the first 12 months” at announcement in December 2025. By May 2026 the language was “slightly dilutive.” That walk-back inside four months, on a $5.2B commitment, is the single most under-examined number in this story. No publicly quantified synergy target has been disclosed — which, for a deal of this size and this financing structure, is itself a disclosure failure.

11. Variant Perception

11.1 What consensus believes

Consensus holds that BioMarin is a cheap, defensive rare-disease compounder with a stalled growth asset, that the Amicus deal buys time and diversification, and that a new management team executing a cost programme will re-rate the shares as margins expand toward the promised 40%. Sell-side price targets cluster meaningfully above the current price, and the 2027 revenue consensus of roughly $3.65B sits at the low end of management’s own withdrawn scenario range.

11.2 The strongest bull case

BioMarin owns an irreplaceable, genuinely un-attackable annuity: roughly $1.49B of revenue from products that have had no regulatory exclusivity for a decade and have never attracted a single biosimilar, because the patient populations are too small to repay a competitor’s development cost. That annuity throws off cash at 81% gross margins, is expanding through newborn screening, and has just been vindicated by the public collapse of the gene-therapy threat. Layer on a policy environment that turned favourable in 2026 — orphan drugs exempt at 0% from a 100% pharma tariff, the IRA carve-out widened, $200M vouchers restored — and add $634M of acquired Amicus revenue with a third of it convertible to synergies. The stock trades at the 7th percentile of its own historical sales multiple with an activist-installed board, an imminent ITC ruling that could exclude the lead competitor from the US market, and a hypochondroplasia label expansion with a February 2027 PDUFA date. If the $500M cost programme lands, this is a 30%+ operating margin business at 12x forward earnings.

11.3 The strongest bear case

BioMarin is a serial destroyer of capital that has just made its largest and most leveraged bet at the worst possible moment. Fifteen years and $9.0B of research spending produced a cumulative operating loss. The flagship gene therapy returned four cents on the dollar. The prior large acquisition was impaired for more than its purchase price. Every target set at the September 2024 Investor Day has been missed or withdrawn, and the margin metric was deleted from the long-term incentive plan and replaced with revenue growth in the same year management borrowed $4.25B to buy revenue. The one asset carrying the equity has gone flat while a once-weekly competitor launches at a premium price and an oral competitor with the best efficacy data in the disease prepares to file. The company’s own answer arrives in 2029–30. Insiders own less than 1% and have not bought a share in two years; the activist who forced the reset has sold and left. And the stock is not cheap on normalized numbers — the base case brackets the current price, and the sum of the parts says the market is already paying a full multiple for the stalled growth asset.

11.4 The 3–5 assumptions that actually matter

1. Does Voxzogo hold roughly flat, or does it decline? Flat is priced. Decline is not.
2. Does the $500M cost programme convert an 81.4% gross margin into a 30%+ operating margin? The entire valuation rests here.
3. Do Amicus synergies reach ~$230–267M? Below that, the deal destroys value on day one.
4. Does the un-attacked enzyme annuity stay un-attacked? The evidence says yes, and this is the most under-appreciated positive.
5. Does the ITC exclude Ascendis from the US market on or about 21 August 2026?

11.5 The factor-positioning read

The tape offers an unusually clean piece of evidence, and it cuts against the “turn has begun” narrative. Over the trailing twelve months BioMarin’s raw return was +2.1%, but its cumulative factor-residual return was −13.5% — and that figure already includes the +15.65% Amicus announcement day. Strip that single session and the year’s idiosyncratic drag is roughly −25%. Over the most recent quarter the raw move was +11.3% and the cumulative specific return was +0.05%. The three-to-six-month “recovery” is sector carry, not company-specific improvement.

In factor space BMRN is an abandoned sector-beta name: Industry Biotechnology +0.690, Market +0.660, Health Care +0.617, with Momentum negative in all four nested models and Quality and Value essentially zeroed. It is not a deep-value name, not a quality compounder, not a momentum name. Beta is 0.618 with alpha of −0.228; ten-year Sharpe is −0.181, five-year −0.211, three-year −0.412.

The regime has been favourable to nearly everything BioMarin loads on — biotech and healthcare factors positive with z-scores near +1, Value the strongest factor on the tape — and the stock still returned +2.1% over twelve months while the XBI rose roughly 70%. It massively underperformed a raging bull market in its own sector. Short interest stands at roughly 7.4% of float and rose 14.4% month-on-month into the rally.

The honest framing: this is neither a falling knife nor a momentum name — it is range-bound and sector-dependent. Owning it underwrites the fundamental thesis; it does not ride a trend. The factor-similar peer list is dominated by healthcare ETFs rather than operating companies, which is itself evidence that the market trades this as sector beta rather than as a differentiated franchise.

Where consensus may be offsides: consensus reads the six-month bounce as the beginning of a turn. The residual-return data says nothing company-specific has improved. If Q2 2026 (6 August) confirms the Voxzogo stall, the sector carry that has supported the price has nothing underneath it.

12. Fact vs. Interpretation

13. Open Questions

1. What are the Amicus synergy targets in dollars, and by when? No figure has been publicly quantified for a $5.2B transaction. This is the most important missing disclosure.
2. What is Voxzogo’s discontinuation and persistence rate? BioMarin has never published one despite being asked directly on multiple calls. Without it, the “patients won’t switch” defence cannot be tested.
3. Is the patient/revenue gap really order timing? Patients +20% against revenue +2.9% has been explained as timing for five consecutive quarters, while management separately disclosed two national price resets in 2026. Q2 2026 will be informative.
4. What is the actual scope and status of the DOJ sponsored-testing investigation? Two years unresolved, no accrual, no quantified exposure, and it touches the demand funnel for 40% of revenue.
5. How did the FDA dispose of BioMarin’s citizen petition on Voxzogo orphan exclusivity? It evidently did not block Yuviwel’s approval; the grounds matter for the remaining IP strategy.
6. What explains the Q1 2026 inventory reclassification — finished goods down $270.5M, work-in-process up $253.6M, against $195.0M of quarterly COGS?
7. What is BioMarin’s actual Medicare payer mix? The conclusion that IRA exposure is immaterial is an indication-based inference, not a disclosed fact.
8. Does the Annex IV orphan tariff exemption survive its one-year reassessment, and does it remain unconditioned on MFN participation?
9. What is management’s plan for the $600M convertible maturing in May 2027 — refinance into an already-levered structure, or pay in cash?
10. Will the Vimizim US biologic exclusivity expiry in 2026 draw any biosimilar interest at all? A decade of evidence says no, but it has not been tested at this revenue scale.
11. What patent and SPC coverage protects Galafold outside the US now that EU orphan exclusivity lapsed in May 2026? This is arguably the most important unanswered question in the acquired portfolio: 59% of Galafold’s revenue is ex-US, the orphan wall came down three months ago, and there is no company disclosure quantifying the remaining jurisdiction-by-jurisdiction protection. No EU generic filing has surfaced either — but the absence of evidence is not evidence of absence.
12. Is there a filing to remove the “not improving on current ERT” restriction from Pombiliti’s US label? Without one, the $1.45B allocated to that asset — 12.9x sales — rests on a second-line label in a flat market.
13. Why has Galafold’s US accelerated approval not been converted to full approval after eight years?

14. What Must Be True

The bull case requires

1. Voxzogo holds roughly $950M–$1.05B through 2027 despite two competitors, on the strength of the infant label, ex-US footprint and patient inertia. Falsification test: two consecutive quarters of Voxzogo revenue below ~$210M, or any disclosure of net patient attrition to Yuviwel. Q2 2026, reporting 6 August, is the first test.
2. The $500M cost programme delivers — non-GAAP operating margin reaches the mid-30s in 2026 and approaches 40% in 2027. Falsification test: Q2 2026 non-GAAP operating margin below ~30%, or SG&A remaining above 30% of revenue. A second consecutive year of margin going backwards ends this case.
3. Amicus delivers ~$230–267M of synergies and turns accretive within 18 months. Falsification test: FY2027 guidance that does not show Amicus contributing at least ~30% incremental operating margin, or any further “dilutive” walk-back.
4. The enzyme annuity stays un-attacked and grows mid-single digits. Falsification test: a biosimilar filing against Vimizim, Naglazyme or Aldurazyme; or Palynziq declining more than 10% as Sephience scales.

The bear case requires

1. Voxzogo declines outright in 2027 as Yuviwel scales and infigratinib launches. Falsification test: Voxzogo revenue above ~$1.05B in 2027, or an ITC exclusion order removing Ascendis from the US market.
2. The margin programme continues to slip, leaving normalized operating margin in the low 20s. Falsification test: two consecutive quarters of non-GAAP operating margin above 32% with SG&A below 30% of revenue.
3. Leverage becomes a constraint — deleveraging crowds out R&D and the May 2027 convertible must be refinanced on worse terms. Falsification test: net debt/EBITDA below 2.0x by the end of 2027 with R&D maintained above 25% of revenue.
4. Capital misallocation repeats. Falsification test: the initiation of a buyback once covenants permit, a meaningful increase in insider ownership, or the restoration of an operating-margin metric to the long-term incentive plan.

15. Source Appendix

Primary — SEC filings (mirrored locally at output/BMRN/sources/)

• BioMarin Pharmaceutical Inc., Form 10-K for FY2025, filed 2026-02-26 — product revenue table, IP/exclusivity table, geographic disaggregation, Note 19 (ROCTAVIAN restructuring), Amicus and Bridge Facility disclosure, manufacturing footprint, Government Regulation, Risk Factors (DOJ sponsored-testing subpoena, tariffs).
• BioMarin Forms 10-K for FY2021–FY2024, filed 2022-02-25, 2023-02-27, 2024-02-26, 2025-02-24.
• BioMarin Form 10-Q for Q1 2026, filed 2026-05-05 — debt footnote, inventory detail, subsequent events.
• BioMarin Forms 8-K: 2024-09-24 (Investor Day targets reaffirmed); 2025-10-07 and 2025-10-27 (Roctavian divestiture, Q3 results); 2025-12-17/19/22 (Roctavian withdrawal; Amicus merger agreement); 2026-01-29 and 2026-02-12 (5.500% notes due 2034); 2026-02-23 (Q4/FY2025 results); 2026-03-16 (VOXZOGO SCFE trial discontinuations); 2026-04-23 (French FDI clearance — final condition); 2026-04-27 (completion of merger; Item 2.03 credit facilities); 2026-05-04 (Q1 2026 results); 2026-06-04 (2026 annual meeting; equity plan amendment).
• BioMarin Form 8-K/A, filed 2026-07-13 — Amicus audited financials and unaudited Article 11 pro-forma financial information; purchase price allocation.
• BioMarin DEF 14A proxy statements, filed 2022-04-12, 2023-04-11, 2024-04-09, 2025-04-08 and 2026-04-21 — incentive-plan metrics, 2025 bonus determination, LTI metric substitution, ownership, CEO compensation, TSR table.
• BioMarin Forms 3, 4 and 5, 2024–2026 — insider transaction codes, 10b5-1 status.
• Amicus Therapeutics, Inc. Form 10-K for FY2025, filed 2026-02-20 (CIK 0001178879) — Galafold and Pombiliti revenue disaggregation, amenability disclosure, Hatch-Waxman settlements, competitor table, headcount, properties.
• Amicus DEFM14A, filed 2026-02-02 — Background of the Merger, Centerview financial analyses, management projections, golden-parachute table; DEFA14A, 2026-02-23 (supplemental disclosures).
• Amicus Form 8-K, 2026-04-27 (Item 5.02 — officer and director departures); Forms 4 filed 2026-04-27.
• Elliott Investment Management Forms 13F, Q1 2024 through Q3 2025 — position established and fully exited.

Primary — company disclosures and transcripts

• BioMarin earnings-call transcripts, Q1 2025 (2025-05-01), Q2 2025 (2025-08-04), Q3 2025 (2025-10-27), Q4/FY2025 (2026-02-23), Q1 2026 (2026-05-04) — read in full; saved locally to output/BMRN/transcripts/. Note: transcript bodies obtained through a third-party aggregator contain machine-normalisation artifacts (entity-name substitution, one speaker mis-attribution); all quotations relied upon were checked against the corresponding press release or 8-K.
• BioMarin press releases: Amicus acquisition (2025-12-19); Q1 2026 results; VOXZOGO hypochondroplasia Phase 3 (May 2026); ENDO 2026 VOXZOGO and BMN 333 data; FDA acceptance of the VOXZOGO sNDA (PDUFA 2027-02-28); legal action against Ascendis in the European Unified Patent Court.
• BioMarin September 2024 Investor Day materials — $4B 2027 revenue, 40% operating margin, >$1.25B operating cash flow targets.
• Sanofi quarterly and full-year results releases: FY2024 (2025-01-30), FY2025 (2026-01-29), Q1 2026 (2026-04-23), Q2 2026 (2026-07-30) — Fabrazyme, Nexviazyme and Myozyme/Lumizyme revenue.
• Takeda FY2025 results (2026-05-13) — Replagal.
• Protalix Biotherapeutics FY2025 results and Chiesi partnership disclosures — Elfabrio transfer-price revenue.
• Ascendis Pharma disclosures — YUVIWEL (navepegritide) FDA approval 2026-02-27, April 2026 launch, pricing; COACH combination Phase 2 data (2026-01-08).
• Sangamo Therapeutics Form 10-Q Q1 2026 (2026-05-14) and Forms 8-K 2026-06-23 and 2026-07-20 — Chapter 11 filing, DIP facility, Lilly and Astellas stalking-horse bids, auction timetable.
• 4D Molecular Therapeutics release, 2025-01-10 — 4D-310 deprioritisation.

Legal and regulatory primary sources

• Ascendis Pharma A/S v. BioMarin Pharmaceutical Inc., No. 26-1026 (Fed. Cir., decided 2026-03-26) — mandatory stay under 28 U.S.C. §1659(a)(2).
• US International Trade Commission investigation concerning US Patent RE48,267 — ALJ initial determination expected on or about 2026-08-21.
• Proclamation 11020 (2026-04-02), “Adjusting Imports of Pharmaceuticals and Pharmaceutical Ingredients Into the United States,” 91 FR 18183 — Annex I (covered goods), Annex IV (orphan-drug exemption at 0%), Annex II (MFN safe harbour).
• Consolidated Appropriations Act, 2026 (signed 2026-02-03) — Mikaela Naylon Give Kids a Chance Act (rare pediatric disease PRV reauthorisation through 2029-09-30); §6605 (narrowing orphan exclusivity to approved use or indication).
• One Big Beautiful Bill Act (signed 2025-07-04) — ORPHAN Cures Act provisions amending the IRA orphan exclusion, effective from the 2028 price-applicability year.
• CMS/CMMI proposed rules, December 2025 — GLOBE (Medicare Part B MFN model) and GUARD (Part D).
• EU pharmaceutical package — trilogue agreement December 2025; orphan market exclusivity tiering, Global Orphan Marketing Authorisation, Bolar extension. EU HTA Regulation joint clinical assessments for orphan products from 2028-01-13.
• FDA: Recommended Uniform Screening Panel additions (Pompe 2015, MPS I 2016, MPS II 2022); Novel Drug Approvals 2025.
• HHS-OIG Advisory Opinion 22-06 (April 2022) — sponsored genetic testing under the Anti-Kickback Statute.
• New Jersey Department of Labor 2026 WARN Notice Archive — Amicus Therapeutics, Princeton, 58 positions, effective 2026-08-07.
• European Commission merger case M.12314; US HSR early termination granted 2026-02-11.

Clinical and scientific literature

• PROPEL 3 (infigratinib in achondroplasia), New England Journal of Medicine, published 2026-06-29.
• Ascendis PERIDOT and Sanofi CARAT Phase 3 results in Fabry disease (2026-02-02 and 2026-07-30 disclosures); Idorsia MODIFY Phase 3 (2021).
• Pooled newborn-screening incidence data across 29 programmes and >11.6 million newborns (Pompe, MPS II).

Industry and market data

• Evaluate, 2026 Orphan Drug Report (March 2026) and Orphan Drugs at a Crossroads — orphan market size, growth premium compression.
• Priority Review Voucher secondary-market transactions: Fortress/Cyprium $205M; Jazz $200M; Denali $195M; Rocket $180M; PTC $150M.
• Biotech capital-cycle data: venture funding, XBI performance, IPO and M&A volumes 2024–2026.

Quantitative aggregators (computed data, reconciled to filings)

• ROIC.ai MCP — income statement, balance sheet, cash flow, profitability ratios, enterprise value and valuation multiples for NASDAQ:BMRN and the peer set, accessed 2026-08-01. Third-party aggregated data; every material figure reconciled to the underlying filing.
• AZI — daily adjusted OHLCV price history (6,797 rows through 2026-07-31) and valuation_index own-history percentile ranks (P/E 44.4th, P/B 9.1st, P/S 7.3rd, composite 20.3rd, dated 2026-07-31).
• FactorsToday — factor loadings across four nested models, leaderboard risk-adjusted returns by horizon, stock-specific volatility, related stocks, and factor-return regime data, accessed 2026-08-01. Third-party statistical estimates; loadings are in-sample and the 756-day window predates the Amicus close.
• SEC EDGAR XBRL company facts via scripts/edgar.sh.

Third-party research cited for framing

• Morgan Stanley Research, Blue Paper: The US Healthcare Formula — Cost Control and True Innovation, 16 June 2011. Fifteen years old; cited solely for structural and value-chain framing (orphan barrier-to-entry logic, the loss-of-exclusivity framework, capital-cycle crowding evidence, and cost-structure benchmarks), never as current data. Third-party research, not the author’s conclusion.

Material limitations of this report

1. Q2 2026 results — the first quarter including Amicus — report on 6 August 2026, five days after this report’s date. Every trailing figure here is standalone BioMarin and understates both revenue and leverage. The synergy target, accretion math and long-term Amicus outlook are all expected on that call.
2. Purchase accounting for Amicus remains preliminary under the ASC 805 twelve-month measurement period.
3. The ITC initial determination (~21 August 2026) had not issued at the time of writing.
4. BioMarin does not disclose its Medicare payer mix; the conclusion that IRA negotiation exposure is immaterial is an indication-based inference.
5. Annex IV of Proclamation 11020 should be verified line-by-line before the orphan tariff exemption is relied upon, and it carries a one-year reassessment.

APPENDIX A — Standard Diligence Questionnaire

BioMarin Pharmaceutical Inc. (NASDAQ: BMRN) · Report date 1 August 2026

Supplemental to the report above. Answers are grounded in the evidence set out in the report and labelled Fact / Interpretation / Assumption where the distinction matters.

General

What thoughtful questions have other investors asked about this company?

The sell-side dialogue over the past five earnings calls has converged on four questions, and management’s answers to three of them have been evasive in ways that are themselves informative.

First, the switching question. Christopher Raymond of Raymond James put it most directly on the Q4 2025 call: “we were kind of surprised to learn one of your competitors is anticipating about half of their patients in their early achondroplasia trials will be Voxzogo experienced.” Management’s answer rested on proprietary market research ranking “convenience is the third” priority. [Interpretation] A competitor’s own trial-enrolment mix is harder evidence than the incumbent’s unpublished market research.

Second, the demand-versus-revenue gap. An analyst on the Q1 2026 call asked management to “bridge the disconnect between the 20% year-over-year patient growth in Q1 versus a 3% revenue growth.” The answer — “entirely order timing” — has now been given for five consecutive quarters. [Open Question]

Third, why the 2027 guidance was withdrawn. Phil Nadeau of TD Cowen pressed on whether management’s view of competitive risk had changed; CFO Brian Mueller conceded, “We have taken a different view.” [Fact]

Fourth, and most pointed: Sean Laaman’s framing that “without business development, BioMarin is a capital accumulator” — which management effectively endorsed by spending its entire stated $4–5B of firepower on Inozyme and Amicus rather than on buybacks.

The question investors have not pressed hard enough: what are the Amicus synergies in dollars? No figure has been disclosed for a $5.3B transaction, and BioMarin affirmatively elected not to present one in the 8-K/A where Article 11 would have permitted it.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Driven by the external environment or internal actions?

Neither, in the conventional sense — this is not a cyclical business. [Fact] Reported FY2025 earnings are at an artificial low: $493.2M of one-time charges (the $241.3M Roctavian withdrawal, $221.0M of Inozyme acquired IPR&D, a $31.0M manufacturing charge) depressed GAAP operating income to $409.5M against a normalized 21–23% margin. [Interpretation] Almost all of the depression is internal and discretionary — management chose to withdraw Roctavian and chose to expense an acquisition. The external environment (demand for lifelong rare-disease therapy) was essentially unchanged.

Looking forward, earnings will be depressed again in FY2026 by purchase accounting: BioMarin’s own Article 11 pro-forma shows $390.3M of incremental intangible amortisation and $180.6M of incremental interest, which on a FY2025 pro-forma basis convert +$348.9M of net income into a $205.9M net loss. [Fact]

How stable are revenues?

Structurally very stable, but reported quarterly revenue is lumpy. [Fact] The underlying demand is lifelong, non-discretionary therapy for progressive genetic disease with no therapeutic alternative — about as stable as pharmaceutical demand gets. But roughly 63% of company-marketed revenue is ex-US and ~30.9% ($980M: Latin America $435.5M, rest-of-world $544.5M) is sold through government tenders and named-patient programmes, where a single order can move a quarter. Q4 2025 Voxzogo of $273.4M against $219.9M in Q1 2026 is the clearest example.

Outlook for products/services?

Bifurcated, and this is the central finding of the report. The enzyme annuity (~$2.24B) is durable and growing mid-single digits. Voxzogo ($926.9M) has stopped growing — +2.9% year over year in Q1 2026 — and is guided to high-single-digit growth in 2026 against +26% in 2025. [Fact]

How big will this market be — growing, shrinking, domestic or international?

Growing, and predominantly international. Evaluate forecasts orphan drugs at >$400B of annual sales by 2032, over 21% of the global prescription market, on a 2025–30 CAGR above 10%. [Fact] But the orphan growth premium over non-orphan is forecast to compress from ~2.5–3 points to roughly one point by 2030 — the excess returns are being arbitraged away. [Fact/Interpretation] For BioMarin specifically, only ~35% of revenue is US ($1,105.0M of $3,167.8M).

Underlying demand is also mechanically expanding through newborn screening, which finds materially more patients than clinical diagnosis: pooled data across >11.6 million newborns put Pompe incidence at 1 in 18,711 births, and MPS II at roughly 1 in 36,000 male births against a historical clinical estimate of 1 in 100,000–150,000. [Fact]

Business Quality & Competitive Moat

Is the industry getting more or less competitive?

More, and precisely where it matters. [Fact] Achondroplasia went from one approved drug to three credible entrants in five years: Voxzogo (2021), Ascendis’ once-weekly Yuviwel (approved 2026-02-27, launched April 2026 at a 20% price premium of $498,225/yr), and BridgeBio’s oral infigratinib (Phase 3 positive in NEJM, 2026-06-29, filing Q3 2026). PKU is going the same way — Kuvan already lost to generics, and Palynziq now faces PTC’s oral Sephience (1,647 patients, $151M in Q2 2026 alone).

Conversely, the ultra-rare enzyme replacement markets have become less contested, because the feared gene-therapy disruption failed: Roctavian withdrawn, Pfizer’s Beqvez withdrawn, bluebird bio sold for ~$30M after 57 patients ever treated, Elevidys carrying a Boxed Warning. [Fact]

How profitable is the business (ROIC, ROE)?

Poorly, on reported figures, and this is the crux. [Fact] ROIC was 2.75%, 2.81%, 6.19% and 4.57% across FY2022–FY2025; ROE 5.94% in FY2025. Against a cost of capital of roughly 8.5–9.5%, BioMarin has cleared its hurdle once in five years.

But the denominator contains a great deal of non-working capital — cash, goodwill and intangibles from prior deals, and a ~$1.5B deferred tax asset (24% of book equity) left over from the 2020 valuation-allowance release. [Interpretation] On operating capital and normalized earnings, returns were 7.2% (2023), 14.6% (2024) and 19.4% (2025). The operating business earns respectable returns; the enterprise does not, because capital is added faster than it earns.

How profitable is the industry — how many competitors, what barriers to entry?

Highly profitable at the product level: median orphan list price is roughly $219,000 a year against ~$13,000 for non-orphan, a ~17x asymmetry that has widened. Gross margins run near 80% — BioMarin’s was 81.4% in FY2025 excluding the Roctavian write-off, its highest in five years. [Fact] The commercial model needs only ~500 sales people to sell $3.2B.

Barriers are genuine but narrow: manufacturing complexity (recombinant enzymes require specific glycosylation — mannose-6-phosphate for lysosomal targeting — to function at all), a patient-identification network running through a few hundred specialist centres, and statutory exclusivity. [Fact/Interpretation]

Can the business be easily understood?

Yes at the unit level — a known number of patients, a known price, a lifelong therapy. The complexity is in the ex-US tender timing, the non-GAAP adjustments, and now the purchase accounting.

Can it be undermined by foreign low-cost labour?

No. [Fact] This is not a labour-cost business; manufacturing is capital- and know-how-intensive, at Novato, California and Shanbally, Cork. The relevant analogue is biosimilar competition, which has never materialised — see below.

Do brands matter?

Not as consumer brands. What matters is the clinical dataset and the relationship with the few hundred treating centres and key opinion leaders who define standard of care. [Interpretation] BioMarin’s stated defence of Voxzogo rests substantially on ten years of safety data and >10,000 patient-years — that is the closest thing to a brand in this market.

What is the nature of competition?

Modality substitution, not price. [Interpretation] This is the single most important pattern in the company’s history and it has now repeated three times: Kuvan (oral) lost to oral generics, falling 78.5% from a $463M peak to $99.6M; Palynziq (daily injection) is being attacked by an oral; Voxzogo (daily injection) is being attacked by a weekly injection and then an oral. BioMarin sells injectables into markets that keep being won by pills and long-acting conjugates.

Customers’ switching costs?

Real but defensive only. [Interpretation] A patient stable on a lifelong weekly infusion is genuinely reluctant to switch, and the treating-centre relationship is sticky. But switching costs protect the installed base; they do nothing to stop a rival winning new patients. In Fabry, the acquired Pombiliti’s US label actually requires prior ERT failure — a label built entirely around switching. Notably, BioMarin has never published a Voxzogo discontinuation or persistence rate despite being asked directly on multiple calls. [Open Question]

The moat, named: in Greenwald’s taxonomy this is a supply-side advantage of an unusual kind — a minimum efficient scale that exceeds the size of the market. The decisive evidence: Naglazyme’s last patent expired November 2023 and Aldurazyme’s November 2020; neither has unexpired orphan exclusivity; both have grown; and not one biosimilar has ever been developed against roughly $694M a year of combined revenue. A few thousand patients worldwide cannot repay a biosimilar development programme. [Fact] That barrier is genuine and durable — and it protects only against copies, not against better molecules. The moat protects the old, patent-free products and does nothing for the new, growing ones.

Financial Condition & Balance Sheet

Assets not fully recognised on the balance sheet?

Yes, and materially. [Interpretation] Roughly $9.0B of R&D has been expensed over fifteen years, creating regulatory approvals, manufacturing know-how and clinical datasets that carry no balance-sheet value. The Novato and Shanbally facilities are carried at depreciated cost against a replacement cost that would be far higher for a new entrant. The patient-identification network and KOL relationships built over 25 years are unrecognised.

Off-balance-sheet liabilities?

Nothing exotic. Operating leases (including $72.0M of acquired Amicus lease obligations running to 2034), ordinary purchase commitments, and contingent milestone obligations on in-licensed programmes. The one genuinely unquantified item is the DOJ subpoena over sponsored genetic testing for Vimizim and Naglazyme — disclosed in Risk Factors, not Legal Proceedings, with no accrual and no quantified exposure, unresolved since the FY2023 10-K. [Fact]

How conservative is the accounting?

Not conservative in the one place it matters most: inventory. [Interpretation] BioMarin capitalises pre-launch and pre-qualification manufacturing costs when recoverability is judged “probable.” Days sales of inventory stand at 644 as reported and 772 days excluding the Roctavian write-off, up from 572 in 2021 — more than two years of stock. That exact policy produced the $119.2M Roctavian inventory write-off: product manufactured for a launch that never came. With the growth product decelerating, this is the most likely location of the next write-off.

Two further flags: receivables grew 37.5% against 12.9% revenue growth (DSO 83 → 89); and in Q1 2026 finished goods fell $270.5M while work-in-process rose $253.6M in a quarter with only $195.0M of total COGS — offsetting magnitudes that suggest a reclassification the filing does not explain. [Open Question]

On the other side, the non-GAAP presentation is aggressive: it excludes $181.4M of stock-based compensation, 5.6% of revenue.

How CapEx-hungry is the business?

Not very. [Fact] Capital expenditure ran $97–132M a year over five years, just 3.2% of revenue in FY2025. Manufacturing capacity is largely built. The cash intensity of this business is in R&D and working capital, not fixed assets.

Capital Allocation & Management

How much FCF does the business generate, and how does management use it?

FY2025 free cash flow was $717.0M (operating cash flow $828.0M less $111.0M capex), up 50.8% in the very year reported operating income fell 15.4%. [Fact] Over five years: $185.3M, $44.4M, $51.6M, $475.4M, $717.0M.

Management’s use of it is the report’s central criticism. [Fact] Ten years (2016–2025) of R&D spending totalling $7,006M produced $72.7M of cumulative GAAP operating income. Fifteen years and $8,974M produced a cumulative operating loss of $431M. Cumulative stock-based compensation of $1,757M is 24 times cumulative operating profit.

Significant acquisitions recently?

Two, in nine months. [Fact] Inozyme (1 July 2025, ~$285M cash, ~$329.1M total consideration) for BMN 401 — expensed as a $221.0M acquired-IPR&D charge. And Amicus Therapeutics, announced 19 December 2025 and closed 27 April 2026: $14.50 a share, $5,323.6M of total consideration (of which $121.6M was Amicus’s own banker and legal fees paid by BioMarin), ~$5.03B enterprise value net of cash acquired, for $634.2M of FY2025 revenue — 7.6x trailing sales, roughly 2.4x BioMarin’s own multiple, at a 33% premium and above the top of Centerview’s own DCF range of $11.75–$14.35.

The prior large deal, Prosensa (2015, $680M), produced $797.8M of IPR&D impairments — more than the purchase price. And Roctavian, the flagship internal programme, absorbed at least $1.6B and returned $67.8M of lifetime revenue, implying roughly 25–40 patients dosed worldwide before withdrawal.

Buying back shares?

No. [Fact] Zero repurchases between 2021 and Q1 2026 (a single $50M convert-linked purchase in 2020). No dividend has ever been paid, and both new debt instruments restrict dividends. Management has been explicit that this is deliberate — the CEO has argued business development is “the most highly correlated with stock appreciation” and that assets “are worth more in our hands.”

Issuing large amounts of new shares to insiders?

Dilution is moderate but the overhang is not. [Fact] Share count rose from 172.6M (2016) to 192.3M, about 1.2% a year, with no primary equity issuance in five years. Stock-based compensation has actually fallen 12.4% in dollars over two years to $181.4M. But equity-plan overhang stands at roughly 38.9M shares, about 20% of shares outstanding, after shareholders approved 8.0M new shares in 2025 and a further 7.65M in June 2026.

Compensation policy of directors/management?

This is the sharpest single finding in the report. [Fact] The 2025 annual bonus paid on a non-GAAP EPS “result” of $4.83. The figure reported to shareholders was $3.15 — below the plan’s own $3.67 threshold, at which the metric pays zero. The Compensation Committee bridged the gap by adding back the Roctavian write-off ($0.46), the Inozyme IPR&D charge ($1.12), and the cost of its own above-target bonus ($0.10). Revenue missed its target ($3,221M against $3,223M) and was paid at 100%. The pool funded at 130%.

For the 2025 long-term incentive grant, Core Operating Margin was deleted as a metric and replaced with Revenue CAGR and an “Innovation” measure defined as incremental revenue from new products — in the same year management borrowed $4.25B to buy revenue. [Interpretation] A plan that pays for revenue growth and stops paying for margin, at a company whose problem is that revenue growth never converts into returns on capital, is aligned with the acquisition rather than with shareholders.

CEO pay for 2025 was $21.9M, up 48%, against five-year total shareholder return that took $100 to $67.77 while the Nasdaq Biotechnology Index went to $124.75.

Motivations of management?

Weakly aligned with shareholders on the evidence available. [Interpretation] Officers and directors together own fewer than 1% of shares — 512,000 shares outright against 193.3M outstanding. There have been no open-market purchases (code P) by any insider in 24 months; against that, $4.38M of sales across five individuals, of which four of the five 2026 sales were not made under 10b5-1 plans, including the Chief R&D Officer selling on 26 February 2026 — the day the 10-K disclosing the Roctavian write-off was filed. The CEO and CFO have not sold.

Elliott Management, whose ~$1B stake in November 2023 forced the CEO change and the strategic review, fully exited in Q2 2025 — before the Roctavian withdrawal and before the Amicus deal. Its three settlement directors remain. [Fact]

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer?

No. BioMarin is a Delaware corporation listed on Nasdaq, issuing ordinary common stock and a standard Form 1099. No K-1, no ADR structure, no partnership complications. [Fact]

Dividend policy?

No dividend has ever been paid, and none is contemplated. The April 2026 credit facilities contain restricted-payment covenants that constrain both dividends and buybacks. [Fact] [Interpretation] The practical effect is that BioMarin cannot repurchase stock at a 9th-percentile price-to-book even if it wished to — the capital structure now forecloses the option.

How profitable is the business?

Underlying profitability is good and improving; reported profitability is poor. Normalized operating margin is 21–23% (corroborated by BioMarin’s own reported non-GAAP margin of 23.3% in FY2025 and 24.3% in Q1 2026), on a gross margin of 81.4%. Reported GAAP operating margin was 12.7%. Management targets “approximately 40%,” a figure this report scores as low credibility given that the margin went backwards from 28.6% to 23.3% in the programme’s ramp year and was then dropped from the May 2026 guidance table. [Fact/Interpretation]

Is net income diverging from cash from operations?

Yes, sharply, and in the favourable direction. [Fact] FY2025 operating cash flow of $828.0M against GAAP net income of $348.9M is a ratio of 2.37x, driven by the $244.2M of non-cash impairment and $181.4M of stock-based compensation. [Interpretation] This divergence is benign — it reflects non-cash charges, not earnings quality problems — and it is why the trailing GAAP P/E of 44x and the ~16x price-to-free-cash-flow tell opposite stories. The honest anchor is ~21x free cash flow net of stock compensation.

Valuation context. At $60.02 the stock trades at 1.9x book (2.7x tangible book excluding the deferred tax asset), 3.6x sales, and 3.2–3.34x EV/sales. AZI’s own-history percentiles read P/B 9.1st, P/S 7.3rd, composite 20.3rd. [Fact] The P/E percentile (44.4th) should be discarded because trailing GAAP EPS is distorted by one-time charges. [Interpretation] And the low P/S percentile is not a buy signal: it benchmarks today’s sales against a decade when BioMarin was priced as a growth company at 7.5–9.1x EV/sales. Against a 22% operating margin, 3.5x sales is roughly 16x EBIT — an ordinary multiple.

Risks & Downside

What factors would cause the stock to decline?

In rough order of probability-weighted impact: (1) Voxzogo revenue declining outright rather than flattening as Yuviwel scales and infigratinib launches; (2) the margin programme missing again, leaving normalized operating margin in the low 20s against a promised 40%; (3) Amicus synergies under-delivering — the deal needs roughly $230–267M of run-rate savings simply to be EPS-neutral against $267M of annual financing cost, and no synergy target has been disclosed; (4) an adverse ITC ruling around 21 August 2026; (5) an adverse resolution of the DOJ sponsored-testing investigation constraining the patient-identification funnel for 40% of revenue; (6) another inventory write-off from a 772-day stock position.

Risk of a catastrophic loss?

Low. [Interpretation] This is a profitable, cash-generative business with a diversified annuity base, no product above 29% of revenue, comfortable covenant headroom (~2.9x gross and ~2.25x net leverage against a 3.50x covenant that binds only the Term Loan A and revolver, with ~7.2x interest coverage), and demand that is lifelong and non-discretionary. The bear case in this report implies roughly $23–33 a share — a severe outcome, not a wipeout.

The instructive comparable is Sarepta, whose market capitalisation fell from $11.8B to $2.26B in a single year when its lead franchise broke, and which now trades at 1.23x EV/sales. That is the shape of the downside if Voxzogo does not merely flatten but breaks — though BioMarin’s diversified enzyme annuity makes an equivalent outcome considerably harder to construct.

Chance of a total loss?

Negligible. [Interpretation] A total loss would require simultaneous failure of six independent, patent-free, therapeutically essential franchises with no alternative treatments, plus a covenant breach. There is no plausible path.

Recent News & Events

Has the business environment changed recently?

Profoundly, in both directions, within eighteen months. [Fact]

Against the company: Ascendis’ Yuviwel approved (27 February 2026) and launched (April 2026) as the first direct Voxzogo competitor; BridgeBio’s oral infigratinib posting the largest height-velocity effect ever recorded in achondroplasia (NEJM, 29 June 2026); the March 2026 SCFE safety signal halting three Voxzogo label-expansion trials; the October 2025 withdrawal of the $4B 2027 revenue target; and Galafold’s EU orphan exclusivity lapsing in May 2026 on 59% of that franchise.

For the company: the Section 232 pharma tariff regime taking effect 31 July 2026 with orphan drugs exempt at 0%; the ORPHAN Cures Act widening the IRA orphan carve-out; the rare-pediatric Priority Review Voucher programme reauthorised through September 2029 with vouchers clearing at $180–205M; the Voxzogo hypochondroplasia Phase 3 meeting its primary endpoint (May 2026) with a PDUFA date of 28 February 2027; and the empirical collapse of the gene-therapy threat to enzyme replacement therapy. In the acquired Fabry franchise, both oral competitors failed Phase 3 (Sanofi’s venglustat in February and July 2026, Idorsia’s lucerastat in 2021) and the leading Fabry gene therapy is being sold out of Sangamo’s Chapter 11 for $25M.

Significant acquisitions?

Yes — Inozyme (July 2025) and Amicus (closed 27 April 2026), covered above. [Fact]

Change in accounting policies?

No policy change, but two comparability items matter. The FY2024 10-K recast FY2023 and FY2022, moving roughly $27.7M from SG&A to below the operating line. And FY2020 carried a $903M deferred-tax valuation-allowance release against a $43M operating loss, which makes any five-year EPS growth rate meaningless. [Fact] Purchase accounting for Amicus remains preliminary under the ASC 805 twelve-month measurement period.

Recent changes — new markets, facilities, management?

Management: Alexander Hardy replaced Jean-Jacques Bienaimé as CEO on 1 December 2023, ending an 18-year tenure, following Elliott’s activist stake. The Chief Commercial Officer, Chief R&D Officer and Chief Accounting Officer have all turned over since. On the Amicus side, all five executive officers and all nine directors departed at closing, with $51.3M of disclosed golden-parachute compensation across the five named executives and $218.6M of total equity settlement. [Fact]

Facilities and markets: the acquisition adds Fabry and Pompe franchises and sites in Princeton NJ, Philadelphia PA and Marlow UK, with $72.0M of lease obligations to 2034. A 58-position reduction at Princeton was filed with New Jersey’s WARN system, effective from 7 August 2026. Voxzogo availability expanded from 49 to 55 countries with a target of more than 60 by 2027. [Fact]

The dating limitation that governs this entire appendix: BioMarin reports Q2 2026 on 6 August 2026 — the first quarter containing Amicus, and the call on which the synergy target, accretion math and long-term outlook are expected. Everything above is standalone BioMarin plus one week of disclosed integration detail. [Fact]

APPENDIX B — Source Appendix

Sources supporting the BioMarin Pharmaceutical Inc. (NASDAQ: BMRN) report dated 1 August 2026. Primary sources are listed before secondary. Every material quantitative claim in the report was reconciled to a filing; where a third-party aggregator and a filing disagreed, the filing governs and the discrepancy is noted.

Primary — SEC filings (mirrored locally at output/BMRN/sources/)

• BioMarin Pharmaceutical Inc., Form 10-K for FY2025, filed 2026-02-26 — product revenue table, IP/exclusivity table, geographic disaggregation, Note 19 (ROCTAVIAN restructuring), Amicus and Bridge Facility disclosure, manufacturing footprint, Government Regulation, Risk Factors (DOJ sponsored-testing subpoena, tariffs).
• BioMarin Forms 10-K for FY2021–FY2024, filed 2022-02-25, 2023-02-27, 2024-02-26, 2025-02-24.
• BioMarin Form 10-Q for Q1 2026, filed 2026-05-05 — debt footnote, inventory detail, subsequent events.
• BioMarin Forms 8-K: 2024-09-24 (Investor Day targets reaffirmed); 2025-10-07 and 2025-10-27 (Roctavian divestiture, Q3 results); 2025-12-17/19/22 (Roctavian withdrawal; Amicus merger agreement); 2026-01-29 and 2026-02-12 (5.500% notes due 2034); 2026-02-23 (Q4/FY2025 results); 2026-03-16 (VOXZOGO SCFE trial discontinuations); 2026-04-23 (French FDI clearance — final condition); 2026-04-27 (completion of merger; Item 2.03 credit facilities); 2026-05-04 (Q1 2026 results); 2026-06-04 (2026 annual meeting; equity plan amendment).
• BioMarin Form 8-K/A, filed 2026-07-13 — Amicus audited financials and unaudited Article 11 pro-forma financial information; purchase price allocation.
• BioMarin DEF 14A proxy statements, filed 2022-04-12, 2023-04-11, 2024-04-09, 2025-04-08 and 2026-04-21 — incentive-plan metrics, 2025 bonus determination, LTI metric substitution, ownership, CEO compensation, TSR table.
• BioMarin Forms 3, 4 and 5, 2024–2026 — insider transaction codes, 10b5-1 status.
• Amicus Therapeutics, Inc. Form 10-K for FY2025, filed 2026-02-20 (CIK 0001178879) — Galafold and Pombiliti revenue disaggregation, amenability disclosure, Hatch-Waxman settlements, competitor table, headcount, properties.
• Amicus DEFM14A, filed 2026-02-02 — Background of the Merger, Centerview financial analyses, management projections, golden-parachute table; DEFA14A, 2026-02-23 (supplemental disclosures).
• Amicus Form 8-K, 2026-04-27 (Item 5.02 — officer and director departures); Forms 4 filed 2026-04-27.
• Elliott Investment Management Forms 13F, Q1 2024 through Q3 2025 — position established and fully exited.

Primary — company disclosures and transcripts

• BioMarin earnings-call transcripts, Q1 2025 (2025-05-01), Q2 2025 (2025-08-04), Q3 2025 (2025-10-27), Q4/FY2025 (2026-02-23), Q1 2026 (2026-05-04) — read in full; saved locally to output/BMRN/transcripts/. Note: transcript bodies obtained through a third-party aggregator contain machine-normalisation artifacts (entity-name substitution, one speaker mis-attribution); all quotations relied upon were checked against the corresponding press release or 8-K.
• BioMarin press releases: Amicus acquisition (2025-12-19); Q1 2026 results; VOXZOGO hypochondroplasia Phase 3 (May 2026); ENDO 2026 VOXZOGO and BMN 333 data; FDA acceptance of the VOXZOGO sNDA (PDUFA 2027-02-28); legal action against Ascendis in the European Unified Patent Court.
• BioMarin September 2024 Investor Day materials — $4B 2027 revenue, 40% operating margin, >$1.25B operating cash flow targets.
• Sanofi quarterly and full-year results releases: FY2024 (2025-01-30), FY2025 (2026-01-29), Q1 2026 (2026-04-23), Q2 2026 (2026-07-30) — Fabrazyme, Nexviazyme and Myozyme/Lumizyme revenue.
• Takeda FY2025 results (2026-05-13) — Replagal.
• Protalix Biotherapeutics FY2025 results and Chiesi partnership disclosures — Elfabrio transfer-price revenue.
• Ascendis Pharma disclosures — YUVIWEL (navepegritide) FDA approval 2026-02-27, April 2026 launch, pricing; COACH combination Phase 2 data (2026-01-08).
• Sangamo Therapeutics Form 10-Q Q1 2026 (2026-05-14) and Forms 8-K 2026-06-23 and 2026-07-20 — Chapter 11 filing, DIP facility, Lilly and Astellas stalking-horse bids, auction timetable.
• 4D Molecular Therapeutics release, 2025-01-10 — 4D-310 deprioritisation.

Legal and regulatory primary sources

• Ascendis Pharma A/S v. BioMarin Pharmaceutical Inc., No. 26-1026 (Fed. Cir., decided 2026-03-26) — mandatory stay under 28 U.S.C. §1659(a)(2).
• US International Trade Commission investigation concerning US Patent RE48,267 — ALJ initial determination expected on or about 2026-08-21.
• Proclamation 11020 (2026-04-02), “Adjusting Imports of Pharmaceuticals and Pharmaceutical Ingredients Into the United States,” 91 FR 18183 — Annex I (covered goods), Annex IV (orphan-drug exemption at 0%), Annex II (MFN safe harbour).
• Consolidated Appropriations Act, 2026 (signed 2026-02-03) — Mikaela Naylon Give Kids a Chance Act (rare pediatric disease PRV reauthorisation through 2029-09-30); §6605 (narrowing orphan exclusivity to approved use or indication).
• One Big Beautiful Bill Act (signed 2025-07-04) — ORPHAN Cures Act provisions amending the IRA orphan exclusion, effective from the 2028 price-applicability year.
• CMS/CMMI proposed rules, December 2025 — GLOBE (Medicare Part B MFN model) and GUARD (Part D).
• EU pharmaceutical package — trilogue agreement December 2025; orphan market exclusivity tiering, Global Orphan Marketing Authorisation, Bolar extension. EU HTA Regulation joint clinical assessments for orphan products from 2028-01-13.
• FDA: Recommended Uniform Screening Panel additions (Pompe 2015, MPS I 2016, MPS II 2022); Novel Drug Approvals 2025.
• HHS-OIG Advisory Opinion 22-06 (April 2022) — sponsored genetic testing under the Anti-Kickback Statute.
• New Jersey Department of Labor 2026 WARN Notice Archive — Amicus Therapeutics, Princeton, 58 positions, effective 2026-08-07.
• European Commission merger case M.12314; US HSR early termination granted 2026-02-11.

Clinical and scientific literature

• PROPEL 3 (infigratinib in achondroplasia), New England Journal of Medicine, published 2026-06-29.
• Ascendis PERIDOT and Sanofi CARAT Phase 3 results in Fabry disease (2026-02-02 and 2026-07-30 disclosures); Idorsia MODIFY Phase 3 (2021).
• Pooled newborn-screening incidence data across 29 programmes and >11.6 million newborns (Pompe, MPS II).

Industry and market data

• Evaluate, 2026 Orphan Drug Report (March 2026) and Orphan Drugs at a Crossroads — orphan market size, growth premium compression.
• Priority Review Voucher secondary-market transactions: Fortress/Cyprium $205M; Jazz $200M; Denali $195M; Rocket $180M; PTC $150M.
• Biotech capital-cycle data: venture funding, XBI performance, IPO and M&A volumes 2024–2026.

Quantitative aggregators (computed data, reconciled to filings)

• ROIC.ai MCP — income statement, balance sheet, cash flow, profitability ratios, enterprise value and valuation multiples for NASDAQ:BMRN and the peer set, accessed 2026-08-01. Third-party aggregated data; every material figure reconciled to the underlying filing.
• AZI — daily adjusted OHLCV price history (6,797 rows through 2026-07-31) and valuation_index own-history percentile ranks (P/E 44.4th, P/B 9.1st, P/S 7.3rd, composite 20.3rd, dated 2026-07-31).
• FactorsToday — factor loadings across four nested models, leaderboard risk-adjusted returns by horizon, stock-specific volatility, related stocks, and factor-return regime data, accessed 2026-08-01. Third-party statistical estimates; loadings are in-sample and the 756-day window predates the Amicus close.
• SEC EDGAR XBRL company facts via scripts/edgar.sh.

Third-party research cited for framing

• Morgan Stanley Research, Blue Paper: The US Healthcare Formula — Cost Control and True Innovation, 16 June 2011. Fifteen years old; cited solely for structural and value-chain framing (orphan barrier-to-entry logic, the loss-of-exclusivity framework, capital-cycle crowding evidence, and cost-structure benchmarks), never as current data. Third-party research, not the author’s conclusion.

Material limitations of this report

1. Q2 2026 results — the first quarter including Amicus — report on 6 August 2026, five days after this report’s date. Every trailing figure here is standalone BioMarin and understates both revenue and leverage. The synergy target, accretion math and long-term Amicus outlook are all expected on that call.
2. Purchase accounting for Amicus remains preliminary under the ASC 805 twelve-month measurement period.
3. The ITC initial determination (~21 August 2026) had not issued at the time of writing.
4. BioMarin does not disclose its Medicare payer mix; the conclusion that IRA negotiation exposure is immaterial is an indication-based inference.
5. Annex IV of Proclamation 11020 should be verified line-by-line before the orphan tariff exemption is relied upon, and it carries a one-year reassessment.

Evidence-verification notes

• Reconciliation policy. ROIC.ai, AZI and FactorsToday are third-party aggregated or modelled data and are not primary. Every figure driving a verdict in this memo was traced to the underlying 10-K, 10-Q, 8-K or proxy. Where the ROCTAVIAN restructuring charge is quoted, this report uses the 10-K Note 19 aggregation of $241.253M; a $244.2M figure appears in the cash-flow statement’s asset-impairment line under a different aggregation. The difference is immaterial and is noted rather than averaged.
• Transcript quotations. Earnings-call bodies obtained through a third-party aggregator were found to contain machine-normalisation artifacts — entity-name substitution (one instance replaced “Inozyme” with “Enzyme Therapies”), a product-name corruption (“Pombiliti” rendered “Pombility”), and one answer mis-attributed to a speaker who does not appear on the call. Substance was reliable; exact wording was not. Every quotation relied upon in this report was checked against the corresponding press release or 8-K.
• Claims stated as absences. Several findings are framed as no evidence found rather than as confirmed negatives, and should be read that way: no filed DGCL §262 appraisal petition in respect of the Amicus merger; no EU generic migalastat filing; no sNDA to remove the “not improving on current ERT” restriction from the Pombiliti US label; no publicly quantified Amicus synergy target; no development update on Shionogi/Maze MZE001 since May 2024. Absence of evidence in a public-filing search is weaker than a confirmed negative.
• Figures this report declines to use. A “$1.2B peak sales” figure for Pombiliti + Opfolda circulates widely but could not be traced to any Amicus primary document; the only $1.2B in Amicus materials refers to the global Pompe ERT market. It is not used. Similarly, BridgeBio’s “>65% share of a ~$2.5B TAM” for infigratinib is management’s own projection and is reported as such, not adopted.
• Unattributed price moves. The +8.6% session of 29 May 2026 has no cleanly dated primary-source driver and no cause is asserted for it. The −17.2% move of 10 January 2022 and the −12.2% move of 9 November 2023 (in Amicus) coincide with sector-wide drawdowns and are likewise not attributed.
• Model-window caveat. FactorsToday factor loadings are fitted on a 756-day window that predates the Amicus close (27 April 2026), so they describe legacy BioMarin more than the combined entity. Loadings are L1-sparse and in-sample; R² is mildly overstated by construction.