Evidence-first notes on bioscience and deep tech, at the edge of the lab and the market. Information only — not investment advice.
The 30-second version
- What. In the 2025–26 GLP-1 commercial map, US prescription share has flipped: Lilly (tirzepatide) overtook Novo (semaglutide) — roughly 57%+ vs 43% of US scripts as of Q3 2025 (press / IQVIA-type data). But this figure is US prescription (script) share, not global revenue or profit share.
- So what. The real variable is not peak efficacy but access — supply (capacity), net price, and payer coverage. With a list price above $1,000/month and net prices that fan out from roughly $149 to $1,086 across channels, “the price” cannot be stated as a single number.
- Now what. Most new entrants (Amgen aside) are still unapproved Phase 2/3 — not approvals or revenue. The IRA taking effect in 2027 (negotiated semaglutide price), a Medicare bridge model (2026-07) and the oral-drug manufacturing ramp will set the pace of future diffusion. This post is for information only and is not investment advice.
Demonstration-gap note. The CAPEX figures in this part (Novo DKK 55–60B/year; Lilly $2–9B per site) are announced or in-progress commitments, and many sites come online only in 2028–29 — these are future options, not current operating capacity. Many pipeline “competitors” are in unapproved clinical stages, before any commercial entry.
The five-minute read
The headline is “reversal”; the bottleneck is “access”
Novo’s semaglutide opened the market, but 2025 US prescription share has flipped to Lilly’s tirzepatide (53% early 2025 → 57% in Q2 → 57%+ vs 43% in Q3 2025; CNBC, PharmaVoice, press / IQVIA-type data). This is US prescription-volume share only, not global revenue or profit share. Novo remains broad across the EU, emerging markets and its insulin franchise. On an FY2025 segment basis, Novo’s Diabetes & Obesity segment is about DKK 289B (≈$44B, approximate FX, company data) and Lilly’s Cardiometabolic Health is about $40B (company data) — but note that neither is standalone GLP-1 revenue; both are segment/aggregate totals.
The core argument is simple. Even a −20% weight-loss drug produces zero outcome if the patient cannot get it. The real bottleneck of the 2025–26 map runs along three axes: (1) supply (a three-stage bottleneck of peptide API, sterile fill-finish and pen assembly), (2) net-price durability (IRA, MFN, cash-pay, PBM rebate) and (3) payer coverage (Medicare exclusions, employer opt-outs).
Price is not one number
Zepbound’s list price is about $1,086/month, but LillyDirect cash-pay fell to $299–$449 (by dose) after a December 2025 cut, and Wegovy’s NovoCare cash-pay is $149–$349. So the same drug ranges from $149 to $1,086 depending on channel. On top of this, IRA Medicare negotiation (in effect 2027; semaglutide is on the second list, with an Ozempic negotiated price of $274/month vs a $959 list = about 71% lower, the CMS headline “70% discount”) and MFN/TrumpRx cuts (Novo cutting Wegovy −50% and Ozempic −35%) pull net price down simultaneously. A list-price headline ($1,000+) should not be misread as realized revenue (the gross-to-net gap).
| Channel | Approx. monthly price |
|---|---|
| Cash-pay (NovoCare / LillyDirect) | ~$149 |
| Medicare bridge (net) | ~$245 |
| IRA negotiated price (Ozempic) | ~$274 |
| MFN wholesale (starting) | ~$675 |
| List price (Zepbound) | ~$1,086 |
Supply is both moat and bottleneck
The three-stage chain of peptide-API synthesis → sterile fill-finish → auto-injector assembly is both the barrier to entry and the duopoly’s real moat. Both companies resolved the 2024–25 shortage through large CAPEX (Novo DKK 60B in 2025 / DKK 55B in 2026; Lilly’s Lebanon API site $9B, among others; company data), but because many sites only come online in 2028–29, this is a future option rather than current capacity. Small-molecule oral drugs (orforglipron) have a scale advantage because the API is chemically synthesized, but their efficacy is reported below the injectables.
Deep dive
1. Background — from pharmacology to access
Parts 0–2 of this series covered pharmacology (efficacy and safety). In Part 1, semaglutide and tirzepatide improved hard CV and renal outcomes (SELECT, FLOW), and KDIGO 2026 elevated GLP-1 to a standard CKM pillar. But when payer coverage, net price and supply are the bottleneck, population-level outcomes are delayed — that is, the commercial structure is the rate-limiting step of CKM epidemiology. Part 3 looks at the structure of the access bottleneck rather than at “market size / share” numbers.
2. What this synthesis newly clarifies — three axes shifting at once
In 2025–26 all three axes shifted at once: (1) Lilly’s US prescription-share reversal, (2) the IRA 2027 semaglutide designation and MFN cuts, and (3) the end of compounding, the start of a Medicare bridge model, and Pfizer’s acquisition of Metsera. The point is that these are not separate events but a restructuring of access.
- Duopoly structure. The US prescription-share reversal is confirmed, but with a definitional caveat — US script share ≠ global revenue/profit share. A GxP News report (2026-05) that “Lilly is overtaking Novo outside the US too” is a separate item with a different timeframe and definition from the US Rx share (Q3 2025).
- Triple price pressure. IRA (in effect 2027, Part D only; Ozempic $274, high-dose Wegovy $385), MFN (Novo US wholesale Wegovy −50%, Ozempic −35%, starting around $675/month), cash-pay (from $149) and PBM rebates pressure net price simultaneously. Novo’s 2026 guidance of −5% to −13% adjusted operating profit (CER) reflects the net decline (company data).
- End of compounding. After the FDA declared the tirzepatide shortage resolved (2024-12) and semaglutide (2025-02), it proposed on 2026-04-30 removing sema/tirze/liraglutide from the 503B bulks list. An OFA lawsuit’s injunction request was denied. Closing the legal path for low-cost compounded supply supported branded net-price defense, but from an access standpoint (loss of low-cost supply) it may be a headwind to CKM diffusion — a two-sided effect.
3. Strengths and limits of the method — rigor on attribution and timing
The strength of the source asset is that each quantitative claim is attributed by source type (company data vs analyst estimate vs press), timepoint and definition (net vs list, US vs global, segment vs total). As both a limit and a required caveat, three cross-source variances are flagged.
- Variance ①: Novo segment $44B vs separately reported net sales $43.27B — a difference of definition, FX and aggregation basis (a segment cannot exceed total company revenue, so one is likely a segment and the other a company total). Neither figure is standalone GLP-1 revenue.
- Variance ②: market-size estimates — Goldman ~$95B in 2030 vs Morgan Stanley ~$77B (obesity-only) vs $190B by 2035 (combined). Definitions (obesity-only vs all GLP-1) and years differ, so no single number should be cited. These are analyst estimates, not company data.
- Variance ③: the “share reversal outside the US” report (GxP, 2026-05) differs in timeframe and definition from the US Rx share (Q3 2025).
4. Neighbouring domains
- Materials / process engineering. The peptide-API and auto-injector manufacturing bottleneck connects to materials and process engineering. The scale advantage of oral small molecules provides a commercial basis for Part 2’s “oral-access thesis” (though efficacy is below the injectables).
- Economics / policy modeling. IRA and MFN price negotiation is a policy-simulation domain. The structure whereby Medicare excludes weight loss alone but covers it indirectly via CV (Wegovy) or OSA (Zepbound) labels shows that “indication = access right” — payer policy, not pharmacology, sets the pace of CKM diffusion.
5. Commercialization and investment view (TRL, companies)
Quantitative claims are attributed by source and timepoint. Net vs list and US vs global are distinguished. Unapproved pipelines are marked by phase. The table below is a neutral compilation of public information, not a ranking or superiority judgment.
| Company | Commercial position (timepoint) | Key assets | Open questions |
|---|---|---|---|
| Eli Lilly (LLY) | US prescription share 57%+ (Q3 2025, press / IQVIA-type); Cardiometabolic ~$40B (FY25, company) | tirzepatide (Mounjaro/Zepbound), retatrutide, orforglipron (P3) | net-price defense, oral manufacturing ramp, IRA (future rounds) |
| Novo Nordisk (NVO) | US prescription share 43% (Q3 2025); D&O segment ~DKK 289B (FY25, company) | semaglutide (Ozempic/Wegovy), CagriSema (FDA filing) | IRA 2027 semaglutide negotiated price, MFN cuts, share defense |
| Pfizer (PFE) | new entrant (M&A) | Metsera ($10B, 2025-11) — unapproved | integration/development timeline, market several years out |
| Amgen (AMGN) | differentiated entrant (once-monthly) | MariTide (P3 MARITIME, ~2027) | GIP-antagonism reproducibility, dosing-differentiation evidence |
| Roche | M&A + partnership entrant | CT-388 (Carmot), petrelintide (Zealand, P3 planned) | Phase 3 efficacy, late-entrant penetration |
| Boehringer / Zealand | MASH-axis entrant | survodutide (P3, 2026 data) | obesity vs MASH positioning |
| Viking (VKTX) | small challenger | VK2735 (SC + oral) | manufacturing, funding, partnership/M&A |
| Structure (GPCR) | oral small molecule | aleniglipron, ACCG-2671 | efficacy gap, capital |
M&A landscape. Pfizer–Metsera $10B (2025-11; $86.25/share = $65.60 cash + up to $20.65 CVR) was the outcome of a bidding contest with Novo (final $7.6B), where Metsera’s board judged the Novo offer to carry “unacceptable regulatory risk” over FTC antitrust concerns (StatNews, BioPharma Dive). This suggests an antitrust wall against horizontal consolidation of the duopoly, meaning organic entry is the main path. Earlier waves include Roche–Carmot $2.7B (2023-12, CT-388) and Lilly–Versanis up to $1.93B (2023-07, bimagrumab). Most pipeline entrants are unapproved Phase 2/3, before commercial entry — counted as “competitors” but not approvals or revenue.
6. The skeptic’s bottom line
Inherits and quotes the §7 skeptic gate of the source asset.
- Guard against supply overstatement. Announced CAPEX figures are commitments or in progress, and many sites come online only in 2028–29 — a future option, not current capacity. The 2024–25 shortage resolution is confirmed, but next-gen (oral, once-monthly) volume is a separate ramp.
- Doubts on price durability. Net price is falling simultaneously through IRA, MFN, cash-pay and PBM rebates. The claim of “maintained revenue growth” (some press) is an assumption that volume growth offsets unit-price decline, and needs verification. Do not misread a list-price headline as realized revenue.
- Cross-source variance (3 items). Specified in §3: segment vs total, differing market-size definitions and years, and differing timeframes for in/out-of-US share reports — no single number should be cited.
- Competitive-entry realism. New entrants (Amgen aside) are mostly Phase 2/3, with approval and revenue years away. Metsera is a “market several years out” (StatNews). The end of compounding defends branded net but, in terms of loss of low-cost supply, may be a two-sided headwind to CKM diffusion.
7. What to watch (three falsifiable predictions)
- [Price] After the IRA takes effect in 2027, if semaglutide US net revenue is maintained or grows on volume, “elastic demand” is confirmed; if a net-unit decline more than offsets volume and segment revenue falls, “weakening price durability” is confirmed.
- [Supply] After Novo/Lilly API sites come online in 2028–29, if oral and once-monthly formulations do not re-enter shortage, a “capacity moat” is confirmed; if shortages recur on the new-formulation ramp, the bottleneck persists.
- [Competition] If 2027 Amgen MariTide Phase 3 and Roche petrelintide data demonstrate differentiation (once-monthly, muscle preservation, efficacy) vs tirzepatide/semaglutide, a “crack in the duopoly” is confirmed; if they fall short, the duopoly consolidates.
References
- CNBC. 2026. “Eli Lilly, Novo Nordisk earnings and the GLP-1 market.” 2026-02-04. link
- PharmaVoice. 2026. “GLP-1 market changes: Eli Lilly, Novo Nordisk earnings.” link
- Morgan Stanley. “GLP-1 weight-loss market may double to $190 billion by 2035.” link
- Goldman Sachs. “The anti-obesity drug market may prove smaller than expected.” link
- BioPharma Dive. 2025. “Medicare price negotiation: Wegovy, Ozempic, Trelegy (2027).” link
- Managed Healthcare Executive. “CMS negotiates a 70% discount for Ozempic and Wegovy.” link
- KFF. “What to know about the BALANCE model for GLP-1s in Medicare and Medicaid.” link
- FDA. “FDA clarifies policies for compounders as national GLP-1 supply begins to stabilize.” link
- Pharmacy Times. “FDA moves to permanently close the door on compounded GLP-1s.” link
- Pharma Manufacturing. 2025. “Novo Nordisk to spend about $9 billion in 2025 to create additional capacity.” link
- StatNews. 2025. “Pfizer beats Novo Nordisk to acquire Metsera obesity.” 2025-11-07. link
- BioPharma Dive. “Metsera, Pfizer accept offer; Novo, FTC, obesity drugs.” link
- BioPharma Dive. 2023. “Roche, Carmot acquire deal, obesity drug.” link
- Pharmaphorum. 2023. “Lilly bulks up in obesity with $1.93bn Versanis buy.” link
- Managed Healthcare Executive. “CVS Caremark to put Zepbound back on formulary and add Foundayo.” link
- CNBC. 2025. “Eli Lilly prices Zepbound weight-loss drug vials.” 2025-12-01. link
- Pharmaphorum. 2026. “Novo Nordisk cut GLP-1 prices, tough US market.” link
- BioSpace. 2026. “Lilly, Novo face off at ADA 2026 as others seek to compete in obesity.” link
Disclosure
This post is for information only and is not investment advice. Statements about market share, competitive superiority, M&A and pricing should be read as neutral framing in particular, and must not be interpreted as buy/sell implications for any security.
The author holds no position in the mentioned securities (Novo Nordisk NVO, Eli Lilly LLY, Pfizer PFE, Amgen AMGN, Roche, Viking VKTX, Structure GPCR, and others).
COI note (inherited from the source-asset frontmatter coi): this content describes listed pharmaceutical companies and their pipelines and deals factually and neutrally. Company IR/press figures are attributed as “company data,” analyst market-size estimates as “estimates,” and press reports with the outlet name. Every quantitative claim specifies net vs list, US vs global, segment vs total, and timepoint; items that could not be confirmed or that differ by definition are flagged as cross-source variance (no softening language).
Source knowledge-asset verification grade: VERIFIED (22 confirmed / 0 refuted / 3 cross-source variance). Inherited blog skeptic verdict: proceed-with-caveats (conditional) — the three cross-source variances and the net/list and US/global caveats govern the published narrative, so insertion of the required caveats is a condition.
Uthando Nethemba 
Leave a comment