Amgen Inc. (AMGN): Deep Dive Research Report

Research date: 1 November 2026

Which results period is the latest. Amgen's fiscal year is the calendar year, and it reports quarterly. Q3 results usually come out in the first week of November (Q3 2025 was released on 4 November 2025). Amgen has announced that Q3 2026 results will be released after the US market close on Tuesday 3 November 2026. That is a future date, so it is positive evidence that Q3 2026 is not yet out. The latest released period is therefore Q2 2026, reported 4 August 2026. The six results calls used in this report are:

1. What the Company Does

Amgen invents, makes and sells prescription medicines, mostly biologics. Biologics are large, complicated protein molecules grown in living cells rather than mixed from chemicals. Doctors prescribe them, pharmacies or clinics dispense them, and insurers or governments pay most of the bill. Amgen covers the whole chain itself. It finds the molecule, runs the clinical trials that prove it works, gets regulators to approve it, makes it in its own plants, and runs the sales force that persuades cardiologists, oncologists, rheumatologists and endocrinologists to use it. The medicines treat high cholesterol, osteoporosis, severe asthma, psoriasis, rheumatoid arthritis, several cancers, and a group of rare immune and inflammatory diseases. Amgen also makes biosimilars, which are close copies of other companies' biologics once their patents expire.

Where it came from. Amgen was founded in April 1980 in Thousand Oaks, California, as Applied Molecular Genetics. "Amgen" is a contraction of that name, which the company used until 1983. George Rathmann became its first president and CEO in October 1980, and the venture capitalist William K. Bowes was among the founding backers. The company's first big bet was recombinant DNA. A team led by Fu-Kuen Lin cloned the gene for erythropoietin, the hormone that tells bone marrow to make red blood cells. That work became Epogen, approved in 1989 for anaemia in kidney-dialysis patients. A team under Larry Souza then cloned G-CSF, which became Neupogen (approved February 1991) and protects chemotherapy patients from infection. Those two drugs turned a start-up into one of the first profitable biotechnology companies. They also gave Amgen what still defines it: the know-how to make large proteins reliably at industrial scale.

The milestones that explain today's business are mostly acquisitions, made because the company's own labs could not refill the pipeline fast enough:

• Immunex (2002) brought Enbrel, the anti-TNF drug that anchored Amgen's inflammation business for two decades.
• Onyx (2013) brought Kyprolis for multiple myeloma.
• Otezla (2019) was bought from Celgene when Celgene had to sell it to complete its merger with Bristol Myers Squibb.
• A minority stake in BeiGene (2020), now BeOne Medicines, became Amgen's route into China.
• ChemoCentryx (2022) brought Tavneos.
• Horizon Therapeutics (closed 6 October 2023, per the FY2025 10-K) was Amgen's largest deal ever. It added a rare-disease franchise: Tepezza for thyroid eye disease, Krystexxa for uncontrolled gout, and Uplizna for several antibody-driven autoimmune diseases.

The result is a company that used to be "the Epogen and Neupogen company" and is now a diversified portfolio of about two dozen meaningful brands. The founding products survive only as a shrinking "established products" tail.

The value proposition. Amgen solves one problem: it turns a biological insight into a medicine that a regulator approves, a payer reimburses and a doctor trusts, and then makes enough of it, consistently, for millions of patients. The customer is ultimately a patient with a chronic or life-threatening disease, but the people Amgen has to convince are physicians and the payers who control access (Section 4).

Why it is hard to copy. A biologic is defined as much by how it is made as by its formula. Two plants using the same cell line can produce subtly different proteins, so the manufacturing process is part of what the regulator approves. Amgen runs a global network of biologics plants (Section 3) and has four decades of process data. Its newer products add another layer: BiTE bispecific antibodies that physically tether a patient's T-cells to a tumour cell, an siRNA that silences a cholesterol-related gene, and an antibody-peptide conjugate for obesity. Each is a different engineering problem.

A concrete example. Take a 62-year-old man two years after a heart attack, on a high-dose statin, whose LDL ("bad") cholesterol is still well above the target his cardiologist wants.

1. The cardiologist prescribes Repatha, an antibody that blocks PCSK9. PCSK9 is a protein that destroys the liver's LDL receptors, so blocking it lets the liver pull more LDL out of the blood.
2. The insurer's pharmacy-benefit manager asks for proof that statins alone were not enough. Once approved, a specialty or retail pharmacy ships him a pre-filled pen.
3. He injects himself under the skin every two weeks or once a month. His LDL falls sharply within weeks.
4. Amgen's large outcomes trials are what the cardiologist points to when she tells him this lowers his chance of another heart attack or stroke.
5. If he is a cash-pay patient, since 2026 he can also buy Repatha directly through the government's TrumpRx portal at a reduced list price (Section 6).

At the other end of the portfolio, a patient with small-cell lung cancer whose disease has come back after chemotherapy might receive Imdelltra, a BiTE given by intravenous infusion. Because it makes the immune system attack the tumour, the first doses have meant hours of observation for a dangerous immune over-reaction called cytokine release syndrome. Amgen is working to shorten that monitoring (Section 3).

2. Business Segments

Amgen reports one operating segment, human therapeutics (FY2025 10-K). Management describes the portfolio by therapeutic area, and that is the useful way to see where the business's weight sits. Below is an approximate split of Q2 2026 product sales using Amgen's own product groupings. The percentages are this report's calculation from the Q2 2026 product-sales table; Amgen does not publish them.

General Medicine covers common chronic conditions with very large patient populations: cardiovascular disease (Repatha), osteoporosis (Prolia, Evenity) and migraine prevention (Aimovig). What Amgen does well here is run very large outcomes trials and then sell into primary-care-adjacent specialties (cardiology, endocrinology, rheumatology) where volume, not price, decides the economics. Management treats it as both a cash engine and the launch platform for its two biggest pipeline bets, MariTide for obesity and olpasiran for lipoprotein(a), which target the same cardiologists and primary-care doctors Repatha already calls on. Main rivals: Regeneron/Sanofi, Novartis and Merck in cholesterol; biosimilar makers in osteoporosis.

Oncology sells to oncologists and hospital cancer centres. What sets it apart is the BiTE platform. Blincyto (for a form of acute leukaemia) was the first commercial BiTE, and Imdelltra extends the concept to solid tumours. Supportive-care antibodies (Xgeva, Nplate) and targeted drugs (Vectibix for colorectal cancer, Lumakras for KRAS-mutant lung cancer, Kyprolis for myeloma) round it out. It exists as its own unit because the customers, the hospital buy-and-bill economics and the regulatory pathway differ from chronic-care medicine. Management presents it as the growth bet within the existing business. Main rivals: Roche, Johnson & Johnson, AstraZeneca and others in each tumour type.

Rare Disease is the Horizon franchise. Its medicines treat thyroid eye disease (Tepezza), chronic gout that resists other treatment (Krystexxa), and antibody-driven autoimmune diseases: neuromyelitis optica, IgG4-related disease and generalised myasthenia gravis (Uplizna). It also includes ANCA-associated vasculitis (Tavneos). Patients are few and identified one by one, often by specialists who see a handful a year. Infusion logistics and patient-support services matter as much as a large sales force. Management treats it as a growth engine with long exclusivity. Main rivals: Viridian in thyroid eye disease; argenx, UCB and Johnson & Johnson in myasthenia gravis.

Inflammation pairs a growth asset with two declining ones. Tezspire is an anti-TSLP antibody for severe asthma, co-developed with AstraZeneca; Amgen books US sales and AstraZeneca leads elsewhere. Enbrel and Otezla are mature brands now under Medicare price setting (Section 6). The skill is selling to rheumatologists, dermatologists and pulmonologists against entrenched competitors. Management treats the older brands as cash cows to manage down while Tezspire expands into new diseases. Main rivals: Regeneron/Sanofi (Dupixent), GSK, AstraZeneca and AbbVie.

Biosimilars is a distinct business. Amgen copies other companies' biologics when their patents expire: Eylea (Pavblu), Humira (Amjevita), Stelara (Wezlana), Avastin (Mvasi), Herceptin (Kanjinti), Soliris (Bkemv) and Remicade (Avsola). It exists separately because it is effectively the reverse of Amgen's core business. It competes on manufacturing cost, legal strategy and launch timing rather than on new science. Management said cumulative biosimilar sales passed US$15 billion since 2018 (Q2 2026 call). It has named Eylea HD, Opdivo, Keytruda and Ocrevus as the next wave. Main rivals: Sandoz, Celltrion, Samsung Bioepis and others.

Established products are the founding franchise (Epogen, Aranesp, Neulasta) plus Parsabiv for dialysis patients. Biosimilars eroded these years ago. They get little investment and throw off cash.

Geographic mix. Amgen is a US-weighted company. The United States was 73% of product sales in 2025 and about 72% in the first half of 2026, with the rest of the world at 27-28% (FY2025 10-K; Q2 2026 10-Q). US weighting is typical for a US biotech, because US prices are a multiple of European ones. That ties Amgen's fortunes closely to US drug-pricing policy (Sections 6 and 8).

Outside the US, the business divides as follows:

• Europe is the largest international market. Products launch there later and at lower prices, and denosumab biosimilars arrived there in late 2025.
• Japan is a long-standing market. Amgen's presence there goes back to the Kirin-Amgen joint venture of the 1980s. Evenity has taken category leadership there: management cited 55% volume share (Q2 2026 call). Tepezza launched in Japan in 2025.
• China is reached mainly through BeOne Medicines. Amgen holds about 17% of BeOne (Q2 2026 10-Q). BeOne holds the right to sell Xgeva, Blincyto and Kyprolis in China under a November 2025 amendment to their collaboration.

Some products are almost entirely American. Krystexxa, Tezspire (in Amgen's books) and Pavblu have little or no booked revenue abroad. Vectibix and Aranesp, by contrast, sell more abroad than at home.

3. Products and Operations

The products that matter to the investment case

Repatha (evolocumab) is a fully human antibody against PCSK9, injected every two weeks or monthly, that lowers LDL cholesterol. Two features set it apart:

• Outcomes data in both prevention settings. It is the only PCSK9 inhibitor with Phase 3 outcomes data in both secondary prevention (FOURIER, in patients who have already had an event) and primary prevention. The primary-prevention trial, VESALIUS-CV, was presented at the American Heart Association meeting in November 2025. It showed that intensive LDL lowering before a first heart attack or stroke reduces risk. A positive CHMP opinion for a broader EU label followed (Q2 2026 call).
• Momentum after years of weak uptake. For a long time payer hurdles held Repatha back. Management said new-to-brand prescriptions grew more than 50% year on year (Q2 2026 call).

Evenity (romosozumab) is a monthly injection for 12 months that blocks sclerostin. That makes it a bone builder, not just a brake on bone loss. It is aimed at women at very high fracture risk, who then usually switch to an anti-resorptive drug. Its label carries a cardiovascular warning, which limits use in some patients.

Prolia and Xgeva (denosumab) are the same RANKL antibody at different doses: Prolia for osteoporosis, Xgeva to prevent fractures from bone metastases. Patents expired in the US in February 2025 and in Europe in November 2025 (FY2025 10-K). Multiple biosimilars have launched, and management has warned of "accelerated sales erosion" (Q4 2025 and Q1 2026 calls). Both brands fell by roughly a third year on year in Q2 2026. This is the largest drag on the portfolio today.

Tezspire (tezepelumab) blocks TSLP, a signal released by the airway lining that sits "upstream" of several inflammatory pathways. That lets it work in severe asthma whether or not the patient's disease is driven by eosinophils (a type of white blood cell), which is a point of difference from rival biologics. It gained a nasal-polyps indication in 2025. Phase 3 trials are running in COPD and eosinophilic esophagitis; the latter is due to complete in the second half of 2026 (Q2 2026 release).

The Horizon rare-disease drugs

• Tepezza is an IGF-1R antibody given as eight IV infusions for thyroid eye disease, the bulging, painful eyes that come with autoimmune thyroid disease. A subcutaneous version delivered by an on-body injector matched the IV version in Phase 3 (Q1 2026 call).
• Krystexxa is an enzyme that breaks down uric acid in gout that has not responded to other treatment. It is usually given with methotrexate to prevent the body making antibodies against it.
• Uplizna is a CD19 antibody that depletes B-cells, dosed twice a year after loading doses. Its labels cover neuromyelitis optica, IgG4-related disease and, since December 2025, antibody-positive generalised myasthenia gravis. It is the first CD19-directed therapy in that disease.

The BiTEs (Blincyto and Imdelltra). A BiTE is a small two-armed antibody fragment. One arm grabs CD3 on a T-cell; the other grabs a marker on a cancer cell (CD19 for Blincyto, DLL3 for Imdelltra). The patient's own T-cell then kills the cancer cell. Imdelltra received full US approval on 19 November 2025 after the DeLLphi-304 trial showed a 40% reduction in the risk of death compared with chemotherapy in second-line extensive-stage small-cell lung cancer. EU and Chinese approvals followed in May 2026. The practical constraint is the cytokine-release risk. Management is working to cut first-dose monitoring from 16 hours to 1-2 hours, and said more than 2,000 US accounts already use Imdelltra (Q2 2026 call).

Otezla and Enbrel. Otezla is an oral PDE4 inhibitor for psoriasis and psoriatic arthritis. Enbrel is the original anti-TNF biologic. Both are mature and now subject to Medicare price setting (Section 6).

Biosimilars. Pavblu, the Eylea biosimilar for retinal disease, is now the largest. Amgen launched it in the US in late 2024 after prevailing in patent litigation with Regeneron.

The pipeline that matters

• MariTide (maridebart cafraglutide) for obesity is the single most consequential asset (Section 7). It is an antibody that blocks the GIP receptor, chemically fused to two GLP-1-activating peptides. The antibody backbone gives it a long half-life, which is why it can be dosed monthly, and possibly every two or three months for maintenance. Weekly peptides like semaglutide and tirzepatide cannot do this.
  • Phase 2 results: up to about 20% average weight loss at 52 weeks in people without type 2 diabetes, with no plateau. In year two (presented January 2026), a large majority of patients kept their weight off on a lower monthly or a quarterly dose.
  • The weakness: nausea and vomiting at the start of treatment in Phase 2. Phase 3 uses a three-step dose escalation, which management says further reduced those rates (Q1 2026 call).
  • Programme scale: nine Phase 3 studies running and three more planned (Q2 2026 call). They cover obesity with and without diabetes, cardiovascular outcomes, heart failure, sleep apnoea, and switching from weekly drugs.
• Olpasiran is an siRNA that silences the gene for lipoprotein(a), an inherited cholesterol particle that independently raises heart risk. It cuts Lp(a) by more than 95% with dosing every 12 weeks (Q1 2026 call). OCEAN(a)-Outcomes enrolled 7,297 patients (Q2 2026 call).
• Others: xaluritamig, a STEAP1 BiTE in prostate cancer (the XALute Phase 3 has completed enrolment), and dazodalibep for Sjögren's disease (two Phase 3s due to complete in the second half of 2026).

How the products are made

Amgen's biologics are made in two stages, and it owns most of both:

1. Drug substance. Engineered mammalian cells (typically Chinese hamster ovary cells) are grown in large bioreactors. The secreted protein is harvested, purified through chromatography and filtration steps, and tested batch by batch. This is the hard, regulated, capital-intensive step.
2. Drug product. The purified protein is formulated, filled into vials, syringes or autoinjector pens under sterile conditions, then labelled, packaged and shipped through a cold chain.

The network listed in the FY2025 10-K:

Management says the 2026 capital plan is "focused on expanding manufacturing capacity for weight management products" in North Carolina, Ohio and Puerto Rico (Q2 2026 call, Peter Griffith). That means Amgen is building MariTide capacity before the Phase 3 data are in. The approach makes sense because obesity is a volume market measured in millions of patients, and the incumbents' supply shortages in 2022-24 showed what being capacity-constrained costs. It also means Amgen is committing capital to a launch that depends on a trial result. The US build-out has a policy dimension too, covered in Section 6.

Two other operational facts matter. First, the Puerto Rico manufacturing base is also the centre of Amgen's largest tax dispute (Section 8.4). Second, MariTide's monthly-or-less dosing changes the supply equation relative to weekly drugs: fewer pens per patient per year, but a larger and more complex molecule per dose.

4. Customers

Amgen has three layers of "customer", and confusing them is a common analytical mistake.

1. The legal buyers: three wholesalers. McKesson, Cencora and Cardinal Health each accounted for more than 10% of revenue, and together 77% of worldwide gross revenues in 2025 (FY2025 10-K). This concentration is a feature of how US drugs are distributed, not a risk in the usual sense. Wholesalers take title and redistribute to pharmacies, hospitals and clinics on thin margins. They do not choose which drug a patient gets, and they cannot credibly refuse to stock a product doctors prescribe. The exposure is credit and logistics, not demand.

2. The prescribers. These are the people who make the clinical choice:

• Cardiologists and lipid specialists decide on Repatha and, in future, olpasiran. They want hard-outcomes evidence (fewer heart attacks and deaths, not just lower LDL) and guideline backing. Their decision cycle is slow and evidence-led: practice shifts after a landmark trial and a guideline update, over years.
• Endocrinologists, rheumatologists and gynaecologists decide on osteoporosis drugs. For denosumab, the choice is increasingly driven by the clinic's purchasing economics rather than the doctor (see below).
• Oncologists choose Imdelltra or Blincyto on survival data, NCCN guideline listing and the practicality of administration. Monitoring burden matters to a cancer centre's capacity, which is why cutting Imdelltra's observation time is a commercial lever, not only a clinical one.
• Rare-disease specialists (neuro-ophthalmologists, neurologists, rheumatologists) see few patients and value Amgen's patient-identification and reimbursement-support services nearly as much as the drug.

3. The payers. They decide whether a prescription is filled:

• Pharmacy-benefit managers (CVS Caremark, Express Scripts, Optum Rx) run formularies for self-administered drugs: Repatha, Otezla, Enbrel, Evenity at the pharmacy, and in future MariTide. They negotiate rebates in exchange for formulary position. Annual contract cycles mean list price and net price can diverge sharply.
• Medicare is now a direct price-setter for Enbrel (from 2026) and Otezla (from 2027) under the Inflation Reduction Act.
• Hospitals and clinics buy infused or office-administered drugs (Prolia, Xgeva, Tepezza, Krystexxa, Uplizna, the BiTEs) under "buy-and-bill": they buy the drug, administer it, and bill the insurer. This setting is where denosumab biosimilars do damage fastest. A clinic earns a margin on reimbursement, so a cheaper biosimilar reimbursed at a similar rate is directly in the clinic's financial interest.
• The 340B programme lets qualifying hospitals buy at deep discounts. Management cited it as a source of price pressure on Otezla (Q2 2026 call).
• Since 2026, direct-to-patient channels. These are TrumpRx and Amgen's own AmgenNow programme (Q4 2025 call), which sell to cash-pay patients at a reduced price.

Why customers choose Amgen. It is product by product, not brand loyalty:

• Evidence depth. Repatha has two outcomes trials. Rival PCSK9 drugs and the new oral entrant do not yet have that for both prevention settings (Section 5).
• First-in-class positions: Tepezza, Imdelltra, Tezspire's mechanism, and Uplizna's CD19 claim in myasthenia gravis.
• Dosing convenience as a design goal. Uplizna is dosed twice a year, olpasiran quarterly, and MariTide monthly or less.
• Reliable supply. It is rarely discussed, but Amgen has not had the shortages that hurt some rivals.

Switching costs. These vary a great deal by product:

• Low for clinically interchangeable molecules after patent expiry. Denosumab is the textbook case: a clinic can switch the next patient to a biosimilar with no retraining.
• Moderate for chronic self-injected drugs. A stable patient rarely switches unless a formulary forces it, but formularies do force it.
• High where the product is a treatment course with a defined end (Tepezza's eight infusions), or where clinicians have built protocols around it (BiTE monitoring workflows in cancer centres).
• Embedded in regulation for new competitors. A rival must run its own trials and win its own approval. For biosimilars the bar is lower, but it still takes years of process development and an abbreviated approval.

Contract structure and predictability. There are almost no long-term supply contracts in the usual sense. Revenue is a function of patients on therapy multiplied by net price per patient. Patient counts are sticky for chronic drugs, so volumes are fairly predictable quarter to quarter. Net price is reset each year in PBM negotiations, and policy now resets it too. The predictability problem is therefore price, not volume. There are seasonal effects as well: Q1 is usually the weakest quarter because US patients reset insurance deductibles and wholesalers reduce inventory. Management flagged this again for Q1 2026 (Q4 2025 call).

5. Competitive Landscape and Moat

Amgen does not have one competitor. Each franchise faces a different set, and the structure of competition differs by mechanism. Three types of rivalry matter:

1. Mechanism races. Several companies attack the same biology (PCSK9, incretins, Lp(a), IGF-1R). The winner is whoever has the best combination of outcomes evidence, dosing convenience and payer access, usually in that order.
2. Class competition. Different mechanisms compete for the same patient: in severe asthma, Tezspire versus Dupixent, Nucala and Fasenra.
3. Biosimilar erosion. Once a patent expires, several copies arrive and price, not science, decides share.

Cholesterol and Lp(a) (Repatha, olpasiran). Repatha's direct rival is Praluent from Regeneron and Sanofi. Amgen has won that two-horse race on evidence and contracting: Praluent never matched Repatha's outcomes programme breadth. The structural shift is Merck's Lipfendra (enlicitide), the first oral PCSK9 inhibitor, approved by the FDA in July 2026 for LDL lowering. A daily pill removes the injection barrier that has kept PCSK9 drugs a niche. Repatha's defence is that it has outcomes data and Lipfendra, at approval, did not. Novartis's Leqvio, a twice-yearly siRNA, competes on convenience. In Lp(a), Novartis's pelacarsen is expected to read out its HORIZON outcomes trial first (mid-2026 per CNBC, April 2026). Lilly's lepodisiran Phase 3 data are expected in 2029. Being second to an outcomes readout is not fatal in a class if the first readout is positive, because it validates the mechanism. A negative first readout would damage the whole class.

Obesity (MariTide). This is the hardest arena Amgen has entered. Eli Lilly (tirzepatide, Zepbound) and Novo Nordisk (semaglutide, Wegovy) dominate. Both are moving into oral pills and next-generation molecules, and both have years of manufacturing scale-up behind them. Amgen's wedge is dosing frequency (monthly, possibly quarterly for maintenance), not peak weight loss. Its Phase 2 efficacy was competitive, not superior. The bet is that a large group of patients will prefer a few injections a year for long-term maintenance, and that payers will value adherence. Amgen arrives late (Phase 3 data from 2027, Section 7) against two incumbents with far larger obesity sales forces and supply chains. Section 6 covers the industry dynamics.

Bone (Evenity, Prolia). Evenity's competition is the other bone-building drugs: teriparatide (Lilly's Forteo and its generics) and abaloparatide. Evenity's advantage is monthly dosing and a 12-month course. Prolia's competition is now its own biosimilars from Sandoz, Celltrion, Samsung Bioepis, Fresenius and others. That race is about price, and Amgen is the incumbent losing share by design.

Respiratory and inflammation (Tezspire, Otezla, Enbrel). Dupixent (Regeneron/Sanofi) is the category giant in type-2 inflammation. Tezspire wins where disease is not purely eosinophilic and has, per management, led new-to-brand share in severe uncontrolled asthma (Q2 2026 call). Otezla is squeezed between cheaper generics abroad and more potent biologics and oral TYK2 drugs from AbbVie, Bristol Myers Squibb and others.

Rare disease (Tepezza, Uplizna, Krystexxa). Tepezza faced its first approved direct rival in June 2026: Viridian's Lumvoa (veligrotug), another IGF-1R antibody. It has a shorter IV course and a label covering both active and chronic disease. Amgen's responses are the subcutaneous Tepezza (Section 3) and its installed base of infusion centres. In myasthenia gravis, Uplizna competes with FcRn blockers (argenx's Vyvgart, UCB, Johnson & Johnson) and complement inhibitors. Its edge is twice-yearly dosing.

Oncology (BiTEs). Imdelltra has no approved DLL3 rival in small-cell lung cancer. Several DLL3 bispecifics and antibody-drug conjugates are in development, including from Chinese biotechs licensing to Western partners. Blincyto competes with CAR-T therapies and inotuzumab in ALL.

Biosimilars. Amgen is one of a handful of companies with a full biosimilars pipeline. Sandoz and Celltrion are the scale players. Amgen wins on manufacturing quality, litigation capability (Pavblu) and its existing hospital contracting relationships.

Peer market caps from the Motley Fool's October 2026 healthcare ranking, ad-hoc-news (Sandoz, 23 Oct 2026) and Viridian quote data (30 Oct 2026). They are peer-size references only.

Barriers to entry. For a new company copying Amgen's approach, the barriers are very high: years of discovery, Phase 3 programmes costing billions, regulatory approval, and a biologics manufacturing network. They are lower for a well-funded large pharma attacking one franchise. They fall sharply at patent expiry for biosimilar makers, and lower still for oral small-molecule entrants into an antibody class (Lipfendra). The barrier that has held best over four decades is not any single patent. It is the combination of outcomes-trial capability, biologics manufacturing and a multi-specialty commercial organisation. That combination is what lets Amgen keep replacing products as their patents run out.

Market share structure. Share is set mechanism by mechanism. Amgen leads where it was first or has the best evidence (PCSK9, TSLP, IGF-1R, DLL3 BiTE). It is a minority player where it entered late (obesity, psoriasis), and it loses share by design after patent expiry.

Structural shifts

1. Oral entrants into injectable classes: PCSK9 now, and incretins too.
2. Chinese biotechs as a cheap source of assets for rivals, which shortens the time before Amgen's first-in-class drugs get followers.
3. US policy turning price into a negotiated, government-shaped variable (Section 6).
4. Biosimilar interchangeability becoming easier. This speeds erosion for everyone, Amgen included, while helping its own biosimilars.

Moat: Narrow - intangible assets (patents and regulatory exclusivity), scale / cost advantage in biologics manufacturing and outcomes-trial execution

The evidence is in Sections 4 and 5. Amgen keeps winning mechanism races where outcomes evidence decides (PCSK9 over Praluent) or where it was first (Tepezza, Imdelltra). It has replaced several generations of products after their patents expired. But every individual product moat has an expiry date written into the 10-K, and each franchise is now being directly attacked: Lipfendra on Repatha, Lumvoa on Tepezza, biosimilars on denosumab. The company's ability to keep renewing is an organisational capability, not a structural lock, which is why the verdict is Narrow rather than Wide. The single thing most likely to wear it down is late-stage pipeline failure, specifically MariTide, olpasiran or both. The renewal engine only works if the next wave arrives before the current wave's patents expire in 2028-29 (Section 8.3).

6. Industry

What drives demand. Branded biopharmaceutical demand is driven by:

• Disease prevalence. It rises with ageing populations and lifestyle disease: cardiovascular disease, obesity and osteoporosis are all growing in absolute numbers.
• Diagnosis and treatment rates. Many patients eligible for PCSK9 or osteoporosis drugs go untreated.
• Scientific breakthroughs that open new treatable populations: obesity as a disease category, Lp(a) as a target, bispecifics in solid tumours.
• Payer willingness to fund. This is where policy enters.

Demand for a successful chronic drug compounds. Patients stay on therapy for years, so each year's new starts add to the base.

Industry size and the obesity question. The obesity opportunity is the largest new market in pharma in a generation, and estimates vary widely:

• J.P. Morgan Global Research forecasts the broader incretin market (diabetes and obesity) reaching US$200 billion by 2030.
• Goldman Sachs has published two different views: one forecast of US$95 billion by 2030 for anti-obesity drugs (down from an earlier US$130 billion), and a later upgrade to US$114 billion as oral drugs broaden access (via J.P. Morgan, Goldman Sachs and ZeroHedge summaries, 2026).

The spread reflects real uncertainty about price per patient, payer coverage, and how long patients stay on therapy. That last factor is directly relevant to a maintenance-focused drug like MariTide. The Lp(a) class is earlier: consensus estimates cited by CNBC (April 2026) put Lp(a)-lowering drug sales at about US$5.6 billion by 2032, a figure that depends entirely on outcomes trials yet to read out.

Where Amgen sits in the supply chain. Amgen is a fully integrated innovator. It discovers, develops, manufactures and commercialises, and outsources less than many peers. Upstream suppliers are cell-culture media, chromatography resins, single-use bioreactor bags, glass and device components. Downstream are the wholesalers, PBMs and providers described in Section 4.

Regulation shapes everything

• Approval. The FDA approves biologics through a Biologics License Application. New biologics get 12 years of US data exclusivity, which stops a biosimilar relying on the originator's data, on top of patents. Biosimilars follow the BPCIA abbreviated pathway. FDA moves to drop separate switching studies for "interchangeable" status make substitution easier.
• The Inflation Reduction Act lets Medicare negotiate prices for older high-spend drugs. Small molecules are eligible after 9 years and biologics after 13, so the law favours biologics like most of Amgen's portfolio. It has already hit Enbrel (2026) and Otezla (2027).
• Most-favoured-nation pricing. The 2025 executive order and follow-up letters pushed companies toward aligning US prices with other rich countries. In December 2025 Amgen was among the companies that signed an MFN agreement with the administration. Terms are largely confidential. Disclosed elements include lower Medicaid pricing, TrumpRx cash prices (Repatha cut from US$573 to US$239 a month for TrumpRx buyers), and a three-year suspension of threatened pharmaceutical tariffs in exchange (White House fact sheet, December 2025; Pharmacy Times; Health Affairs).
• Tariffs. The threat of tariffs on imported branded drugs is the reason the whole industry, Amgen included, has accelerated US plant investment.
• 340B discounts keep growing as a share of volume for hospital-administered and specialty drugs.

Cyclicality. Prescription drug demand barely responds to the economic cycle. Patients do not stop taking cholesterol or cancer drugs in a recession. The industry's real cycle is the patent cycle: each company faces a cliff when a cluster of major products loses exclusivity. And the policy cycle: pricing reform comes in waves, often around US elections. Both matter more than GDP.

Tailwinds:

• Obesity becoming a reimbursed chronic-disease category.
• Cardiovascular prevention moving earlier, toward primary prevention.
• Bispecifics and siRNA opening new mechanisms.
• Ageing populations in every major market.
• Japanese and European uptake of products launched first in the US.

Headwinds:

• US net prices falling under IRA, MFN, 340B and PBM pressure.
• Faster biosimilar erosion.
• Oral small molecules entering antibody classes.
• Cheaper assets from Chinese biotechs shortening the time to a second-in-class rival.
• Event-driven outcomes trials taking longer because modern background therapy (statins, SGLT2s, GLP-1s) lowers event rates, which delays every cardiovascular readout in the industry, not just Amgen's.

7. Growth Triggers

Ranked by how much each could change the business, most material first.

• MariTide Phase 3 readouts and the regulatory filing that follows. The two chronic weight-management trials, MARITIME-1 (no diabetes) and MARITIME-2 (type 2 diabetes), were fully enrolled by Q3 2025. The programme has grown to nine ongoing and three planned Phase 3 studies across cardiovascular outcomes, heart failure, sleep apnoea, three new type 2 diabetes trials, and switching from weekly injectables. Management expects a regulatory filing after the data read out (Q1 2026 concall, 30 Apr 2026; Q2 2026 concall, 4 Aug 2026). ClinicalTrials.gov lists primary completion for the main trials in January 2027. Repeated in all six concalls.
• MariTide as a switching and maintenance product. Two Phase 3 studies test moving patients from weekly semaglutide or tirzepatide onto MariTide every 8 or 12 weeks, with a 300-patient switch study using weight change at 52 weeks as its endpoint. Maintenance extension trials test "as few as 4 or 6 doses per year" (Q1 2026 concall, 30 Apr 2026; Q2 2026 concall, 4 Aug 2026).
• Manufacturing capacity for MariTide. The 2026 capital plan targets weight-management manufacturing capacity in North Carolina, Ohio and Puerto Rico ahead of launch (Q1 2026 concall, 30 Apr 2026; Q2 2026 concall, 4 Aug 2026). The Juncos expansion was announced September 2025 and the further Puerto Rico investment in May 2026. Repeated.
• Olpasiran OCEAN(a)-Outcomes. This is a 7,297-patient, event-driven trial in patients with Lp(a) above 200 nmol/L, plus OCEAN(a)-PreEvent in primary prevention and the newly started OCEAN(a)-CCTA plaque study (Q1 2026 concall, 30 Apr 2026; Q2 2026 concall, 4 Aug 2026). Management plans a timing update in early 2027 (CNBC, 27 Apr 2026). The slippage is judged in Section 9.
• Repatha label broadening to primary prevention on the strength of VESALIUS-CV. A positive CHMP opinion has been received for the EU label (Q2 2026 concall, 4 Aug 2026), and management is positioning Repatha for earlier, more intensive LDL lowering (Q4 2025 concall, 3 Feb 2026). Repeated.
• Imdelltra moving to first-line small-cell lung cancer through the DeLLphi programme (nine studies, including the DeLLphi-305 first-line maintenance Phase 3, combining with PD-L1 inhibitors and chemotherapy). Monitoring requirements are being cut from 16 hours to 1-2 hours to widen community-oncology use (Q1 2026 concall, 30 Apr 2026; Q2 2026 concall, 4 Aug 2026).
• Uplizna expansion. The myasthenia gravis launch follows the December 2025 approval. A Phase 3 in autoimmune hepatitis (MERCURY) has started, and a CIDP (a chronic nerve-inflammation disease) Phase 3 is planned for H2 2026 to H1 2027 (Q2 2026 concall, 4 Aug 2026).
• New indications and next-wave biosimilars. These are a group of smaller late-stage events:
  • Dazodalibep's two Phase 3 Sjögren's studies complete in H2 2026 (Q1 2026 concall, 30 Apr 2026).
  • Tezspire's eosinophilic esophagitis Phase 3 completes in H2 2026, with two COPD Phase 3s enrolling (Q2 2026 concall, 4 Aug 2026).
  • Subcutaneous Tepezza and further global Tepezza launches are planned for "2026 and beyond" (Q1 2026 concall, 30 Apr 2026).
  • Xaluritamig's XALute prostate-cancer Phase 3 has completed enrolment (Q2 2026 concall, 4 Aug 2026).
  • Biosimilars of Eylea HD, Opdivo, Keytruda and Ocrevus are in late-stage development (Q2 2026 concall, 4 Aug 2026).

8. Key Risks

8.1 MariTide disappoints, or succeeds but commercially too late. MariTide is the asset most likely to decide whether Amgen grows into the 2030s, so it is also the largest single risk. There are three ways it can go wrong:

• Efficacy or tolerability falls short. The Phase 3 efficacy or tolerability profile comes in below Phase 2, or below what payers and physicians now expect from incretins. The nausea seen early in Phase 2 forced a titration redesign, and Phase 3 is where that fix is proven or not.
• The maintenance benefit is not paid for. The data are good, but the dosing advantage does not convert into payer preference because Lilly and Novo can offer bundled rebates across a large installed base, and oral drugs reset what "convenient" means.
• Capacity is stranded. The capacity being built in North Carolina, Ohio and Puerto Rico (Section 3) is committed before the answer is known, so a negative or delayed readout would leave fixed cost behind.

Calibration: this is a binary event in early-to-mid 2027 with high impact. A clean failure is low-to-moderate probability given the Phase 2 data. A "works but under-delivers commercially" outcome is a meaningfully higher probability.

8.2 US pricing policy compounds across several mechanisms at once. About 72% of product sales are American (Section 2), so Amgen's exposure to US net price is larger than for European peers. The mechanisms are all live together:

• IRA price setting on Enbrel (from 2026) and Otezla (from 2027), with further selections possible in later rounds.
• The confidential MFN agreement (Section 6). Its future price-alignment obligations are not public, and its tariff protection lasts only three years.
• 340B leakage, which management already cites on Otezla (Q2 2026 call).
• TrumpRx and direct-to-consumer cash prices that may anchor commercial expectations lower.

The damage comes slowly, through lower net prices on the same volumes, not through a single blow. Calibration: high probability, moderate and cumulative drag. It becomes severe only if MFN obligations widen to commercial insurance or new launches such as MariTide are priced against international benchmarks from day one.

8.3 A second wave of patent expiries lands before the pipeline matures. Denosumab is already eroding (Section 3). The FY2025 10-K lists the next wave of US expiries:

The way this plays out: if MariTide and olpasiran approvals land in 2028-29 rather than 2027-28, Amgen faces a window where Repatha, Enbrel and Otezla are all declining together while new launches are still ramping. Calibration: the expiries are near-certain. The risk is the overlap, which is moderate in probability and high in impact. It ties directly to 8.1 and 8.6.

8.4 The Puerto Rico tax dispute with the IRS. The IRS challenges how Amgen allocated profits between its US entities and its Puerto Rico manufacturing operations. For 2010-12 it seeks additional federal tax of about US$3.6 billion plus interest. For 2013-15 it seeks about US$5.1 billion plus interest, plus penalties of about US$2.0 billion. A draft notice on similar issues for 2016-18 arrived in April 2026. The Tax Court trial covering 2010-2015 ended in January 2025, and a decision is expected "no earlier than late 2026 or early 2027" (Q2 2026 10-Q). Management has said the 2016-18 matter could have a "material impact" if sustained in full (Q1 2026 call).

An adverse ruling would mean a large cash payment for past years, including interest that compounds over more than a decade. It could also set a precedent that raises Amgen's tax rate going forward and undermines the economic logic of the Puerto Rico manufacturing base. Calibration: the timing is near-term. The outcome is genuinely uncertain. The impact is large and lands on capital returns and debt capacity rather than on the drug portfolio.

8.5 The in-market growth brands face their first direct challengers in the same year. Amgen's growth today comes from a handful of brands. Two of them got direct rivals in mid-2026: Merck's oral Lipfendra against Repatha (July) and Viridian's Lumvoa against Tepezza (June). Section 5 covers both. The mechanism is share loss at the margin among new patients first. A pill for the patient who refuses injections; a shorter course for the thyroid-eye patient whose doctor wants both active and chronic disease covered. Calibration: high probability of some share loss. The size depends on whether Lipfendra produces outcomes data and on payer contracting. Management's answer is evidence (VESALIUS-CV) and format (subcutaneous Tepezza). Neither can be judged until 2027.

8.6 Late-stage attrition and event-driven delay leave the pipeline thinner than it looks. In 18 months Amgen exited rocatinlimab, which was handed back to Kyowa Kirin. Kyowa Kirin later halted all trials on safety grounds in March 2026, which in hindsight makes Amgen's exit look prudent rather than a lost opportunity. Amgen also dropped bemarituzumab after two Phase 3 gastric-cancer trials (February 2026) and discontinued the follow-on obesity molecule AMG 513 (Q2 2026). Olpasiran's readout has slipped because cardiovascular events are accumulating more slowly than predicted (Q4 2025 call). Section 9 covers the history. The mechanism is concentration: with these exits, the future depends more heavily on two assets in two crowded classes. Calibration: moderate probability that at least one more key programme slips or disappoints. In combination with 8.3, high impact.

8.7 Tavneos and the integrity of acquired data. In April 2026 the FDA's drug centre formally proposed withdrawing Tavneos's approval. It stated that the drug has not been shown to be effective and that the application "contained untrue statements of material fact". The approval was originally won by ChemoCentryx before Amgen acquired it. Amgen refused a voluntary withdrawal, requested a hearing (June 2026), and submitted real-world evidence and a blinded re-adjudication of trial data (July 2026). The FDA had also issued a drug-safety communication on serious liver injury, including fatal cases (March 2026). Tavneos is a small product, so revenue impact is minor. The risk is reputational and legal: product-liability exposure from the liver-injury reports, and a precedent that acquired-asset due diligence can miss data problems. Amgen's growth strategy leans on acquisitions. Calibration: high probability of continued regulatory friction, low impact on the overall business.

9. Walk the Talk

The six calls used are 1 May 2025, 5 August 2025, 4 November 2025, 3 February 2026, 30 April 2026 and 4 August 2026. The most recent is 89 days before today.

The olpasiran slippage is the most telling miss, because it shows how management handles uncertainty it cannot control. Asked about timing on the Q3 2025 call, commercial head Murdo Gordon declined to commit:

"We won't guide today on that particular date, but we'll keep all posted as it comes into focus."

That restraint was justified. By the Q4 2025 call, management said events were accumulating more slowly than predicted. By April 2026 the readout that outside observers had pencilled in for late 2026 or early 2027 had become a promise of a timing update in early 2027. R&D head Jay Bradner told CNBC it was impossible to say why events were slow without seeing the data. The same thing has happened to Novartis's pelacarsen, so this is an industry effect of better background therapy (Section 6), not an execution failure. But it means olpasiran, a 2026-27 catalyst a year ago, is now realistically a 2027-28 one.

The dropped programmes were treated quietly rather than defended. Rocatinlimab had run a large Phase 3 programme with positive primary endpoints. Amgen handed it back to Kyowa Kirin, citing "portfolio prioritisation". Kyowa Kirin then halted the programme over safety in March 2026, so the call looks better in hindsight than it did at the time. Bemarituzumab was dropped after its Phase 3s, without the fanfare that accompanied its start. The pattern is that Amgen cuts programmes decisively once data disappoint, which is good discipline. But the pipeline that was presented in 2025 was more fragile than its length suggested.

The capital-structure promise was the most accurate call of the period. After Horizon, CFO Peter Griffith said repeatedly that deleveraging came first. On the Q3 2025 call he could say:

"We have returned to our pre-Horizon capital structure ahead of plan."

The cost is visible in Section 10: buybacks were deferred for three years to get there.

Assessment. On things management controls (commercial execution, launches, approvals, the balance sheet), this team is consistently conservative. It has raised guidance repeatedly and beaten the top of its original ranges for two years running. On things it does not control (event-driven trial timing and early-stage biology), it has been neither dishonest nor especially accurate. It declined to give dates it could not stand behind, but the pipeline delivered less than the 2025 presentations implied. The CFO transition is a continuity risk worth noting: Griffith is retiring and Thomas Dittrich takes over from 1 September 2026. Per Amgen's May 2026 announcement, Dittrich previously held senior finance roles at Amgen and was most recently CFO of Galderma, and before that of Shire and Sulzer; the announcement gives no dates.

10. Ownership, Governance and Shareholder Friendliness

Ownership and control. Amgen is widely held with no controlling shareholder, family, parent or state holder, and has a single class of common stock (one share, one vote). There were about 540.6 million shares outstanding as of 30 July 2026 (Q2 2026 10-Q). Institutions own about 76-81% of the shares, per 13F data aggregated by Fintel and Yahoo Finance as of mid-2026. Index managers top the list: Vanguard and BlackRock each hold roughly 9-10%, followed by State Street and Capital Group entities at about 5-6% each. Those aggregator figures are derived from 13F filings and should be read as approximate. For a minority shareholder, this means full proportional voting power, no related-party extraction through a controlling holder, and no free-float constraint. It also means no anchor owner with a long-term stake in board decisions.

Governance. Robert A. Bradway is Chairman, CEO and President. Per the 2026 proxy, he has been a director since 2011, chairman since 2013, president since 2010 and CEO since 2012. He was earlier COO (2010-2012) and CFO (2007-2010) after joining in 2006 as VP of Operations Strategy, and is 63. The combined chair and CEO role is the main structural flag:

• Counterweight. Robert A. Eckert has been lead independent director since 2016 (2026 proxy).
• Board independence. Eleven of twelve 2026 nominees were independent. The board then appointed David W. Gibbs on 26 October 2026, bringing it to 13 directors, 12 of them independent. Per the 8-K of that date, Gibbs was CEO of Yum! Brands from 2020 to September 2025.
• Shareholder pressure. A shareholder proposal for an independent chair was on the 2026 ballot; the board opposed it (2026 proxy).

Beyond that, governance is unremarkable:

• Ernst & Young is the auditor.
• Pay is largely tied to measurable performance. About 80% of long-term equity is performance-based, and the 2025 annual incentive paid out at 136% of target, weighted 60% on financial metrics (2026 proxy).
• The CEO must hold six times his base salary in stock.
• The October 2026 8-K reports no related-person transactions with the new director.
• No insider share pledges were identified.

Dividends. Amgen has raised its dividend every year:

Source: FY2025 10-K. The dividend has grown steadily at about 6% a year with no cuts or specials.

Buybacks and dilution

• Last ~90 days (since 3 August 2026): MoatMap's database records no buybacks. Amgen discloses repurchases in its quarterly 10-Q rather than as separate exchange notices, and the Q3 2026 filing is due after 3 November.
• Older programmes: Per the FY2025 10-K, Amgen bought back nothing in 2023 (the Horizon year), 0.7 million shares for US$200 million in 2024, and nothing in 2025. It made no repurchases in Q1 or Q2 2026 (Q2 2026 10-Q), even though 2026 guidance allows up to US$3 billion and US$6.8 billion of authorisation remained at the end of 2025.
• Share count: rose from 533,976,238 shares (6 February 2023, FY2022 10-K) to 539,067,675 (10 February 2026, FY2025 10-K), and about 540.6 million in July 2026. That is a net increase of about 1% over three years from stock-based compensation.

Cash that would normally have gone to buybacks went to repaying Horizon debt (Section 9).

Verdict: Returns Capital. The most important reason is a dividend raised every year at about 6%. Buybacks were suspended for three years to deleverage after Horizon, so the share count has drifted slightly upward rather than shrinking.

11. Insider Activities

Source: MoatMap's US insider database (SEC Form 4), current to 31 October 2026, which recorded 57 transactions by 21 insiders over 12 months. The latest Form 4s were cross-checked on EDGAR. No newer filings were found for the past two weeks, which is consistent with the pre-results blackout ahead of the 3 November Q3 release.

Buys. There were no open-market purchases by any director or officer in the 12-month window, and nothing qualifies as a conviction signal.

Sells. The real market selling is small: about US$4.5 million across three officers.

• The August exercise-and-sell transactions by Derek Miller and Nancy Grygiel are classic option management. They exercised options granted in 2019-21, some within two years of expiry, and sold exactly the shares acquired, leaving their underlying holdings unchanged. The sales came after a strong post-earnings share-price move. The Form 4s say this was not under a 10b5-1 plan, so the timing was discretionary, but the economics are housekeeping.
• Rachna Khosla's 11 August sale of 2,000 already-owned shares is the one genuine reduction of a holding. It cut her direct stake by about a quarter. The reason was not disclosed.
• Nancy Grygiel's May sale of 1,237 shares likewise has no disclosed reason.
• The large-looking "Other" lines for Bradway and Gordon in March are tax withholding on vesting awards, not sales. Bradway did not sell shares on the open market in the window.

Net assessment. Insiders are small net sellers in market terms, concentrated in three non-executive-committee officers. The CEO, CFO and every independent director made no discretionary market trades. Nothing has changed: Amgen's senior team has a long pattern of holding vested equity and not buying in the open market. The read is neutral. The selling is routine option monetisation and modest diversification, and there is no buying to signal conviction either way.

12. Scenarios

Bull case. MARITIME-1 and -2 read out in 2027 with weight loss competitive with tirzepatide and nausea rates that the titration scheme has clearly fixed. The maintenance and switch studies show patients holding their weight on a handful of doses a year. Payers, tired of weekly-drug adherence and cost, give MariTide a maintenance niche that grows into a broad position. The capacity built in North Carolina, Ohio and Puerto Rico turns out to be an advantage rather than a stranded cost, because Amgen launches with supply in hand. Repatha's VESALIUS-CV label widens its patient pool faster than Lipfendra can generate outcomes data. Olpasiran reads out positively in 2027-28 on the back of a positive Novartis readout that validates Lp(a) as a target. Imdelltra becomes first-line therapy in small-cell lung cancer. The Tax Court rules largely for Amgen. In this world the 2028-29 patent expiries are absorbed by a new cardiometabolic franchise, and Amgen enters the 2030s with obesity, cardiovascular prevention and bispecific oncology as three distinct growth legs.

Base case. Management keeps doing what Section 9 shows it does well. The in-market growth brands keep growing, guidance is met or modestly beaten, and denosumab erosion runs its predictable course. MariTide succeeds in Phase 3 with a profile that is good but not class-leading. It becomes a meaningful product as a monthly or maintenance option, behind Lilly and Novo rather than in head-to-head competition with them. Olpasiran's readout lands in 2027-28 with the outcome still open. Repatha holds share against the oral entrant on evidence. Tepezza loses some new-patient share to Lumvoa. The dividend keeps rising at its usual pace, and buybacks resume once the tax ruling is known. The 2028-29 patent expiries produce a flat-to-slower period, and the business that emerges is more cardiometabolic and less inflammation-weighted than today.

Bear case. 8.1 and 8.6 compound. MariTide's Phase 3 shows either weaker efficacy or tolerability problems that the titration did not solve, while olpasiran slips again or reads out flat. That leaves the 2028-29 patent expiries in 8.3 arriving with nothing ready to replace them. At the same time 8.5 bites harder than expected: Lipfendra gains outcomes data and payer preference, and Lumvoa takes a large share of new thyroid-eye-disease patients. 8.2 then makes it worse, as Medicare price setting reaches further into the portfolio and MFN commitments widen. An adverse ruling in 8.4 forces a large back-tax payment, so the balance sheet must again be repaired rather than used for acquisitions. Amgen would then have to buy growth from a position of weakness, paying full price for late-stage assets, just as it did with Horizon.

Sources: Amgen Q2 2026 results release · Q2 2026 8-K product table · Q2 2026 10-Q · Q3 2026 webcast notice · Q2 2026 call transcript · Q2 2026 coverage, Yahoo Finance · Q1 2026 call transcript · Q1 2026 transcript, Investing.com · Q4 2025 call transcript · Q4 2025 coverage, Yahoo Finance · Q3 2025 call transcript · Q2 2025 call coverage · Q1 2025 call coverage · FY2025 10-K · FY2022 10-K · 2026 proxy · 8-K, director appointment, 26 Oct 2026 · CFO transition 8-K · Form 4: Khosla · Form 4: Miller · Form 4: Grygiel · Form 4: Busch · FDA: CDER proposes Tavneos withdrawal · Amgen Tavneos update · Fierce Biotech: Amgen exits Kyowa pact · Healio: Kyowa Kirin halts rocatinlimab · Fierce Biotech: MARITIME launch · Fierce Biotech: MariTide Phase 3 titration · MariTide 52-week Phase 2 · Clinical Trials Arena: MariTide at JPM26 · CNBC: Lp(a) drugs · OCEAN(a)-Outcomes on ClinicalTrials.gov · Imdelltra full approval · Uplizna gMG approval · Lipfendra approval · Lumvoa approval · White House MFN fact sheet, Dec 2025 · Pharmacy Times: MFN deals · Health Affairs: MFN deals · Amgen: US$650m Puerto Rico expansion · Amgen: additional US$300m US manufacturing · Xtalks: Ohio investment · BeOne Imdelltra royalty deal · J.P. Morgan: obesity drugs · Goldman Sachs: obesity market · ZeroHedge: Goldman obesity forecast · CNBC: obesity pills 2026 · Motley Fool: largest healthcare companies, Oct 2026 · Sandoz market value · Viridian quote data · Fintel: AMGN ownership · Yahoo Finance: AMGN institutional ownership · Amgen history, Wikipedia