Fantasy M&A: the recipe for a pharma champion

A note on what this is. Fantasy league, pharma edition. Every pairing below is invented. None is a reported transaction, none is a rumour we have heard, and none is investment advice or a view on what any board should do. Each entry gives the case for the deal, the reason it will not happen, and a note on which existing royalties, profit-shares and licensed IP it would reprice. The facts are sourced. The mergers are fiction. If one turns out to be right, the drink is on you.

The rules of the league

Fantasy football works because nobody has to pay the transfer fees. This is the same exercise for pharma: pick the pairings that would make sense on a whiteboard, then explain why the whiteboard is as far as they get.

Pharma had a live demonstration in August of why the whiteboard is where they stay. On 2 August press reports said AstraZeneca and Bristol Myers Squibb had discussed a combination worth close to $400bn. AstraZeneca's shares fell as much as 7% on the Monday. By Wednesday a senior source was telling Reuters there was no deal, there never was a deal to be done, and there were no discussions. Jefferies observed that AstraZeneca did not need financial engineering. Citi called the report a surprise given the pipeline.

Four days from leak to denial, and the acquirer's shares carried the cost of the rumour throughout.

That signal sets the rules. Regulators block therapeutic overlap. Shareholders sell the acquirer on announcement. Family and foundation control structures take several of the most obvious pairings off the board before anyone opens a spreadsheet. What follows is a tour of those obstacles, dressed as deals.

The pairings

1. Merck & Co and Bristol Myers Squibb

Two PD-1 cliffs, one balance sheet.

Why it would make a difference. This is the deal the AstraZeneca story was standing in for. Merck expects US Keytruda sales to decline from January 2028 under IRA pricing and to fall further after the compound patent expires in December 2028. Bristol Myers lists Opdivo's minimum exclusivity at 2028 in the US. One company could strip a shared cost base ahead of a shared cliff.

Why it might not happen. You cannot merge the world's two largest PD-1 franchises. A remedy would mean divesting one, and the PD-1 is the rationale. Bristol Myers also carries what analysts call the largest growth gap among large-cap peers, roughly $38bn of at-risk revenue. Combining the two moves the at-risk revenue onto one balance sheet without removing any of it.

Royalty read-through. Merck's lifecycle defence runs through licensed enzyme IP. Keytruda Qlex, approved in September 2025, uses Alteogen's berahyaluronidase alfa. Halozyme alleges infringement of 15 MDASE patents and won a preliminary injunction in Munich in December 2025. Its CEO has said a licence would carry royalties of 3% to 7%. The dispute attaches to the product and transfers with it.

2. Sanofi and Regeneron

Buying the other half of your own biggest drug.

Why it would make a difference. Sanofi's largest asset is half-owned. Dupixent's economics are shared with Regeneron, key patents expire in the early 2030s, and Paul Hudson's attempt to build a successor internally is the reason the board removed him in February 2026. Belén Garijo took over on 1 May with a mandate framed around rigour. Buying Regeneron internalises the full stream and the antibody engine behind it.

Why it might not happen. Regeneron's founders control it. The price would dwarf the $9.5bn Sanofi paid for Blueprint. A board that just fired a CEO for pipeline ambition is unlikely to fund a transformational one.

Royalty read-through. The collaboration is the clearest case in the industry of a profit-share behaving like a royalty. Consolidating it takes one of the largest shared-economics arrangements in the industry out of circulation.

3. Roche and Novartis

The Basel merger, floated every decade since the two started sharing a tram line.

Why it would make a difference. Combined, they would top the industry on R&D spend with breadth across oncology, ophthalmology, neuroscience, cardio and diagnostics.

Why it might not happen. The block is ownership. In November 2021 Roche repurchased Novartis's 53.3m bearer shares for $20.7bn and cancelled them, lifting the founding families' pool to around 67.5% of votes. Jefferies said at the time the deal put to rest speculation that Roche might pursue larger acquisitions. The pool spent $20.7bn to take Novartis off its register in 2021.

Royalty read-through. Novartis already licenses inclisiran from Alnylam. Its Avidity acquisition carved out Atrium Therapeutics, which kept the Bristol Myers and Lilly cardio collaborations and licenses the platform back. Any Basel combination layers Roche's partnered book on top.

4. Pfizer and Novo Nordisk

The cleanest fit on the list, blocked by a charity.

Why it would make a difference. Pfizer re-entered obesity by buying Metsera at $4.9bn plus a CVR of up to $22.50 per share, after discontinuing danuglipron on liver enzyme elevations. Novo has the manufacturing, the oral franchise and a collapsed share price: CagriSema's REDEFINE 4 missed its primary endpoint, and Novo guided to 2026 sales and profit falling as much as 13% while Lilly guided to 25% growth.

Why it might not happen. At end-2025 Novo Holdings held about 28.1% of the share capital and about 77.3% of the votes through unlisted A shares. The Foundation's articles oblige it to keep a voting majority and forbid selling the A shares. Doustdar has said Novo is not in a position to pursue large-scale M&A as a buyer. The charter closes the other direction.

Royalty read-through. Semaglutide sits in the IRA's second negotiation cycle. CMS announced the negotiated prices on 25 November 2025, effective 1 January 2027, at $274 for a 30-day supply against a $959 list price. Any acquirer would be buying a franchise whose largest US payer price is already fixed by statute.

5. Bayer Pharma, carved out

The break-up Anderson said was "not now", which he asked not to be read as "never".

Why it would make a difference. On 25 June 2026 the Supreme Court held in Monsanto v. Durnell that FIFRA pre-empts state failure-to-warn claims over glyphosate. The litigation had wiped more than $60bn from Bayer's market value since 2018. That reopens the structural question Anderson deferred in 2023. The fantasy: separate Crop Science as a Corteva rival, sell or list Pharma.

Why it might not happen. Sequencing. Anderson pushed back on break-up talk in August while reporting 5.5% Crop Science growth. Barclays flagged possible dividend restoration in FY2027 but noted generalist investors still treat Bayer as uninvestable while any overhang remains. Until the liability is quantified, neither half can be priced at full value.

Royalty read-through. Bayer Pharma's growth products are partnered. Nubeqa is licensed from Orion, which manufactures the product, co-promotes it in Europe and receives tiered royalties averaging about 20% of global sales, rising above 25% if annual sales reach EUR 3bn. A carve-out is a change of control for every one of those agreements.

6. GSK and AstraZeneca

The UK champion, with a lawsuit in the family.

Why it would make a difference. AstraZeneca's Nasdaq ADS listing ceased on 30 January 2026 and its ordinary shares began trading on the NYSE on 2 February, with the company keeping its London primary listing, Cambridge headquarters and UK tax residence. GSK's new CEO Luke Miels started on 1 January and has been buying: Nuvalent at $10.6bn, RAPT at $2.2bn (tracker). GSK's dolutegravir franchise loses protection between 2028 and 2030. Combined scale would smooth two cliffs and give Britain a company to point at.

Why it might not happen. Pfizer tried in 2014 and AstraZeneca declined twice; the politics run the same way in reverse. AstraZeneca sued Miels in 2017 when GSK poached him. The respiratory and oncology overlap would keep three regulators busy for years.

Royalty read-through. ViiV is a GSK-majority venture with Pfizer and Shionogi as minorities. AstraZeneca's Enhertu and Datroway alliance with Daiichi Sankyo is a profit-share with its own governance. In July 2026 Royalty Pharma bought a 3% to 4% royalty on cliramitug, AstraZeneca's Phase 3 ATTR-CM antibody, from originator Neurimmune for up to $425m, $125m of it upfront. Each of those agreements has consent and assignment terms that an offer would trigger.

7. Eli Lilly and Vertex

The only buyer on the list that could pay in paper for anything it wanted.

Why it would make a difference. Lilly became the first drugmaker to reach a $1trn market value on 21 November 2025. Q2 2026 revenue rose 48% to $23.0bn, with Mounjaro at $9.9bn and US Zepbound at $4.9bn, and full-year guidance was raised to $85bn to $87bn. Vertex would add a cystic fibrosis franchise with no near cliff, Journavx in pain and Casgevy in gene editing, none of which Lilly has.

Why it might not happen. Lilly does not need it, and the market is pricing organic incretin growth a dilutive integration would distract from. Vertex prefers to be the buyer: Alpine for about $4.9bn in 2024, and on 6 July 2026 Crinetics at $85 per share, about $10bn in equity value, expected to close in Q3. A company that has just spent $10bn on an acquisition has signalled which side of the table it sits on.

Royalty read-through. Royalty Pharma holds tiered royalties on Vertex's CF franchise from Kalydeco through Alyftrek, its largest single marketed royalty, and is in a dispute-resolution process with Vertex over whether deuterated ivacaftor in Alyftrek is royalty-bearing, which would lift the blended rate from about 4% to about 8%. Casgevy is co-developed with CRISPR Therapeutics, with Vertex leading and profits and costs split 60/40. A Lilly takeover changes the obligor on the first and forces CRISPR to face a far larger partner on the second. Economics travel with the product; the disclosure and audit provisions get renegotiated.

8. Everyone and Gilead

The HIV annuity that funds whatever you want to buy next.

Why it would make a difference. Gilead's Q2 2026 revenue was $7.8bn, up 8%, with HIV sales of $5.7bn up 12% and operating margin around 49% before deal charges. Yeztugo is guided to about $1bn for the year, and a once-weekly oral islatravir plus lenacapavir regimen is targeted for 2027. For AbbVie it would be post-Skyrizi diversification. For Amgen it would be cash to fund MariTide against Lilly and Novo.

Why it might not happen. Gilead has positioned itself as an acquirer. In Q2 alone it booked $11.2bn of acquired IPR&D: $7.0bn for Arcellx, $3.1bn for Tubulis, $1.0bn for Ouro. A company spending that on cell therapy and ADCs has a plan of its own.

Royalty read-through. Gilead pays Royalty Pharma on Trodelvy, which grew 26% in Q2. Its Q2 release also records an IPR&D impairment on assets acquired from Immunomedics and a collaboration with Lakefront Biotherapeutics. Any acquirer consolidates those lines plus the Arcellx, Tubulis and Ouro earn-outs.

9. Johnson & Johnson and Alnylam

The most royalty-entangled target on the list.

Why it would make a difference. Post-Kenvue, J&J has a clean balance sheet and a taste for size: Intra-Cellular at $14.6bn closed in April 2025. Stelara biosimilars are eroding Innovative Medicine. Alnylam brings an RNAi platform and a growing TTR franchise.

Why it might not happen. Alnylam trades at a premium and its best cardiovascular asset is already licensed out. J&J prefers a bolt-on cadence. Talc litigation constrains large capital commitments.

Royalty read-through. Alnylam licenses inclisiran to Novartis, and in November 2025 Royalty Pharma paid Blackstone Life Sciences $310m for a 1% royalty on Amvuttra running to March 2035, a stream that originated in Blackstone's 2020 financing of the HELIOS-B trial. A takeover tests the change-of-control provisions in the Novartis licence and puts the Amvuttra stream under a new obligor.

10. Zoetis and Elanco

The two companies that were once Pfizer Animal Health and Lilly Animal Health.

Why it would make a difference. Zoetis is the largest animal health company, at $9.2bn of 2024 revenue. Elanco became the second-largest by buying Bayer Animal Health for $7.6bn in 2019, and is now shutting its German R&D site and consolidating R&D and manufacturing in Indiana. Zoetis trades at a forward P/E below the S&P 500, with a Librela safety review as a 2026 catalyst. Number one buying number two would be the Bayer/Merial/Novartis roll-up completed.

Why it might not happen. The sector has already consolidated to five: Zoetis, Elanco, Boehringer's Merial, Merck Animal Health and the former Bayer unit, now inside Elanco (Fenwick). Number one and number two overlap in companion-animal parasiticides and vaccines, which is the overlap regulators examine first. Elanco's leverage from the Bayer deal is the other constraint.

Royalty read-through. Animal health runs on licensed biologics and monoclonal antibodies. Elanco has flagged five to six differentiated approvals between 2026 and 2031, much of it immunotherapeutics on in-licensed platforms. A combination reprices every one of those licences and every human-health cross-licence behind them.

11. Lonza and Samsung Biologics

A Swiss CDMO champion for the capacity glut.

Why it would make a difference. Lonza is the largest CDMO at around $7.5bn of revenue, has completed its transformation to a pure-play CDMO and guides to 11% to 12% growth in 2026. Samsung Biologics grew 30% in Q2 2026 with Plants 1 to 4 fully utilised. That is the top tier. Underneath it, the industry is pouring concrete: PharmaSource counted 732 CDMO announcements in 2025 with $24.86bn of disclosed capacity investment, Samsung's fifth plant adds 600,000 litres from H2 2026, Fujifilm is building $3.2bn in North Carolina, and Lonza bought Genentech's Vacaville plant for $1.2bn. WuXi Biologics carries a $23.7bn backlog that BIOSECURE may or may not move West. The case for a merger is the case every capital-intensive industry makes at the top of a build cycle: consolidate before the utilisation curves roll over, and own the customer relationships when the price war starts.

Why it might not happen. The build-out carries multi-year ramp and utilisation risk, and a buyer of Samsung would be paying peak multiples for capacity that is about to compete with itself. The largest players have already absorbed the obvious bolt-ons. Samsung Biologics is a Korean strategic asset. Sponsors dislike supplier concentration, and a merged entity would serve competing customers under one roof.

Royalty read-through. CDMO deals rarely carry drug royalties, but supply agreements carry change-of-control and assignment clauses, and some licences disable a buy-out option if the acquirer is a pure CDMO. Novo Holdings owning Catalent while Novo Nordisk is a customer shows the conflict any such contract now has to address. In a glut, the take-or-pay terms in those supply agreements carry most of the value.

12. An Indian buyer of a Western specialty pharma

The precedent is five months old.

Why it would make a difference. On 27 April 2026 Sun Pharma agreed to buy Organon at $14.00 per share, an enterprise value of $11.75bn including roughly $8.6bn of net debt, the largest overseas acquisition by an Indian pharma company. Indian generics majors generate cash and were exempt from the Section 232 tariffs. The fantasy is the next rung: a specialty company with patented growth.

Why it might not happen. Organon's sales had been flat around $6.2bn to $6.4bn for four years. Indian buyers have so far paid for cash flow at low multiples. Anything much larger would stretch even Sun.

Royalty read-through. Organon carries Merck-era brand licences and a biosimilar partnership with Samsung Bioepis. The merger agreement carries a $120m termination fee. Deal protection is a contingent payment like any other.

13. A Chinese buyer of a Western pharma

The licensing boom, turned into ownership.

Why it would make a difference. China's out-licensing went from 94 deals worth $51.9bn in 2024 to 157 deals worth $135.7bn in 2025. Hengrui overtook AstraZeneca as the world's largest trial sponsor in 2024, and BeOne runs a global commercial model from Switzerland. A Western platform would let a Chinese innovator keep economics it currently licenses away.

Why it might not happen. The BIOSECURE Act and CFIUS make a Chinese acquisition of a US pharma unworkable in 2026. Chinese companies chose licensing over ownership for a reason.

Royalty read-through. A Chinese acquirer inherits every out-licence its target signed with Bristol Myers, Pfizer, GSK and others, each with territory and change-of-control terms. A CFIUS-forced divestiture would reprice the licensed asset, with the Western licensee in the stronger position.

14. A Gulf sovereign buys a listed European mid-cap

The least fantastical entry, because the first rungs are already climbed.

Why it would make a difference. ADQ bought Egypt's Amoun from Bausch Health in 2021, then Swiss-based Acino, Türkiye's Birgi Mefar, UAE-based Pharmax and a minority stake in Biocon Biologics. In April 2024 it consolidated them into Arcera, with 6,500 staff, 2,000 branded medicines and seven sites. Mubadala agreed to buy Kelix Bio in March 2024 and has since added Julphar's DiabTec biologics plant. MGX co-invested in Isomorphic's 2026 round. The next step would be a listed European specialty company at Ipsen or Recordati scale.

Why it might not happen. Everything bought so far is private, emerging-market or generic, and priced accordingly. A listed European target brings a public premium, a foundation or family on the register, and a government that notices. Arcera's stated emphasis is affordable medicines for underserved markets, which differs from running a rare-disease franchise in Paris or Milan.

Royalty read-through. Acino's model already runs on in-licensing, so a sovereign platform is by construction a book of royalty obligations to Western licensors. A step up to patented products adds the reverse: royalties owed to the target, now flowing to a state-owned obligee with no cost of capital.

15. A frontier AI lab buys a pharma company

The first rung has been climbed here too.

Why it would make a difference. In April 2026 Anthropic acquired Coefficient Bio in a stock deal of about $400m. Nature Biotechnology has since written up "frontier AI companies as biotech acquirers" as a category. Claude for Life Sciences counts AstraZeneca, Sanofi, Novo Nordisk and Genmab as clients, and OpenAI released GPT-Rosalind in April (PharmaVoice). Isomorphic raised $2.1bn in May and expects its first drugs in the clinic by year-end. Lilly opened a $1bn AI lab with Nvidia; Roche built an AI factory with the same supplier.

The scale on the AI side is a different order. Anthropic raised $30bn at a $380bn valuation in February. Nvidia put $10bn into Anthropic and $30bn into OpenAI in one quarter, and Jensen Huang said both were likely to list in 2026. A listed lab with a $400bn equity currency could buy a mid-cap pharma in paper the way Lilly can.

Why it might not happen. Everything bought so far is a platform. Coefficient Bio had no drug in the clinic and no revenue. Observers quoted by PharmaVoice do not expect the labs to run drug programmes of their own, and point to Verily and IBM Watson Health. The path already visible is that labs buy biology talent and pharma buys compute.

Royalty read-through. Platform collaborations generate milestone-and-royalty income that now sits inside an AI company's revenue line. If a lab bought a wholly-owned pipeline, the royalties owed and earned would flow through a non-pharma obligor with a share price set by compute demand. Existing royalty agreements were drafted without that counterparty in mind.

16. Private equity takes a large pharma private

The Recordati precedent, scaled up.

Why it would make a difference. CVC and GBL's take-private of Recordati, at around $12.4bn (EUR 10.7bn), is the largest 2026 transaction on most trackers. A sponsor that can underwrite rare-disease cash flows at that size can look at Ipsen, Jazz, Perrigo, Elanco or Grifols.

Why it might not happen. R&D intensity and binary pipeline risk are the wrong inputs for an LBO model. Brookfield's approach to Grifols ended in November 2024 when the board declined to recommend an indicative offer of about $6.8bn (EUR 6.45bn) as an undervaluation, and Brookfield walked. MFN pricing and tariffs compress the cash-flow certainty sponsors need.

Royalty read-through. Sponsors increasingly fund take-privates by selling the target's royalty book to an aggregator. A take-private is also a change of control for every in-licence the target holds. The diligence question for the buyer and the royalty investor is the same: which clauses fire.

17. The royalty aggregators merge

Where the royalty read-through is the whole deal.

Why it would make a difference. Ligand closed its acquisition of XOMA Royalty on 14 July 2026 at $39.00 per share, about $739m in equity, plus a CVR on pending litigation, taking its portfolio past 200 assets. Royalty Pharma internalised its manager for about $1.1bn in 2025, raised 2026 Portfolio Receipts guidance to $3.4bn to $3.5bn in August, and has around $22bn of invested capital at work. A merged aggregator, or one that bought an operating company for its royalty book, would change the pricing of every stream in the market.

Why it might not happen. The model is buying royalties one at a time. Aggregators differ in structure and tax. Royalty Pharma has shown no intent to become an operator.

Royalty read-through. Aggregator M&A transfers portfolios wholesale, and the XOMA CVR shows that even a royalty book can carry contingent litigation value that has to be priced separately.

18. Merck & Co and Merck KGaA

A century-old family separation, reversed.

Why it would make a difference. The two Mercks share a founding in Darmstadt in 1668 and have been separate since the US seized the American subsidiary in 1917 (pharmaphorum). A re-merger would end a century of trademark litigation over who gets to be called Merck where. It would give the German company a growth engine and the American one a tools business.

Why it might not happen. E. Merck KG, the family partnership, controls around 70% of Merck KGaA and prizes independence. Merck KGaA just spent $11.3bn on Bio-Techne to become more of a tools company. And a tools business does not replace Keytruda revenue.

Royalty read-through. Merck KGaA's Erbitux rights outside North America trace to the ImClone licence now owned by Lilly, with Japanese rights shared with Lilly and Bristol Myers. The Erbitux chain is one of the oldest licence structures still paying.

19. Roche buys the rest of Chugai

The Genentech playbook, run in Tokyo, on a company that now owns the obesity pill's royalty.

Why it would make a difference. Roche holds 59.89% of Chugai under a 2002 alliance that left Chugai listed, autonomous and Roche's exclusive Japanese channel. Chugai's market value was about $74bn in July 2026 on $8.4bn of trailing revenue, which prices the minority at roughly $30bn. Chugai invented Hemlibra, Actemra and orforglipron, which it licensed to Lilly in 2018, and it began recognising royalty income on the product, sold as Foundayo, in Q2 2026. Roche paid $47bn for the 44% of Genentech it did not own in 2009 and has been asked about repeating it ever since; Bloomberg reported a $10bn approach in August 2014, which Chugai denied.

Why it might not happen. Severin Schwan has said the Chugai situation is very different from Genentech, with a different contract structure and Japanese duplication already removed. The alliance agreement commits both parties to keeping Chugai listed on the Prime Market. And the minority has tripled in price since the last rumour.

Royalty read-through. Chugai is a royalty company wearing a Japanese pharma's clothes. It collects on Hemlibra from Roche, on Nemluvio from Galderma, and on Foundayo from Lilly, and Chugai's Q2 slides describe growing royalty streams from out-licensed products as a pillar of the business. A buy-out would put the orforglipron royalty, the single largest new royalty stream in the industry, inside Roche, which competes in obesity. The Lilly licence's change-of-control terms are the document that matters.

20. The vaccines champion

CSL Seqirus, GSK Vaccines and Sanofi Pasteur, in the year the US market for their product shrank.

Why it would make a difference. CSL announced in August 2025 that it would demerge Seqirus, the world's second-largest influenza vaccine maker, into an ASX-listed company by June 2026. In October it shelved the timetable, citing US flu vaccination rates expected to fall 12% overall and 14% among over-65s, and guided Seqirus revenue down by a mid-teens percentage. The shares fell as much as 17% to their lowest since 2018. HHS had earlier cancelled $500m of mRNA contracts, one of them Seqirus's. A business that cannot be floated can still be merged. Three vaccine franchises with sub-scale US demand and shared manufacturing economics would have the balance sheet to wait out a policy cycle.

Why it might not happen. GSK's vaccines are a core segment, Sanofi Pasteur is the reason Sanofi exists in Lyon, and a three-way combination of the largest flu, RSV and paediatric franchises is a monopoly by construction. CSL has said it will revisit timing when conditions support maximisation of shareholder value, which is a statement about waiting.

Royalty read-through. Vaccine franchises run on adjuvant and platform licences, government advance-purchase contracts and pandemic-preparedness agreements; Seqirus claimed 90% of global avian flu contracts in FY2025. Those contracts have assignment terms, and the counterparties are governments.

21. Bayer Consumer Health as the next Kenvue

The consumer-health separation that has already happened once, with a buyer waiting at the end.

Why it would make a difference. Kimberly-Clark agreed in November 2025 to buy Kenvue for $48.7bn, 14.3 times trailing EBITDA, in the largest consumer-health transaction on record; shareholders approved it in January 2026, the HSR waiting period expired on 4 February, and closing is expected in the second half of 2026. A household-staples group paying that multiple for Tylenol and Band-Aid sets a price for every other consumer-health division. Bayer's is the largest one still inside a pharma conglomerate. Combined with the Supreme Court outcome in entry 5, the sequence writes itself: settle the liability, separate Consumer Health, sell it at a Kenvue multiple, and fund whatever Pharma needs.

Why it might not happen. Anderson has declined to entertain break-up talk. Kimberly-Clark's deal carries a $1.1bn termination fee either way, which is the market's estimate of how much a consumer-health buyer can be made to pay for certainty. The next buyer would have to be Unilever, P&G or a sponsor, and each of them watched Kenvue trade below its spin price for two years before the bid arrived.

Royalty read-through. Consumer-health portfolios carry brand licences, trademark royalties and legacy Rx-to-OTC switch agreements. The Kenvue separation left J&J with continuing obligations and Kenvue with licensed brands; a Bayer Consumer separation would do the same for Aspirin, Claritin and Bepanthen.

22. Pfizer, in pieces

The demerger an activist wanted and could not get.

Why it would make a difference. Starboard Value built a $1bn stake in Pfizer in October 2024, arguing that $70bn of post-COVID acquisitions, Seagen at $43bn among them, had not restored growth. Its presentation cited a 132-point lag in shareholder returns against the NYSE Arca Pharma Index. Pfizer has an oncology business built on Seagen, a vaccines business built on Comirnaty and Prevnar, a primary-care business with Eliquis and Vyndaqel, and now an obesity programme built on Metsera (entry 4). Three or four listed companies would each have a coherent investor base. The combined one has an activist-shaped hole in its share register.

Why it might not happen. Starboard's two former-Pfizer allies withdrew their support within days, and the fund sold its entire position by Q3 2025. Pfizer's answer was Metsera, which is a bet on concentration, and its cost programme, which is a bet on integration. Neither is a demerger.

Royalty read-through. Eliquis is a Bristol Myers collaboration; Comirnaty is a BioNTech profit-share; Vyndaqel and the Seagen ADCs carry their own licences. A demerger allocates each of those agreements to a successor entity, and every one of them has an assignment clause that was written for a Pfizer that stayed whole.

23. Roche buys Zealand Pharma

Converting the largest obesity licence in the industry into ownership before Phase 3 reprices it.

Why it would make a difference. In March 2025 Roche paid Zealand $1.65bn upfront for petrelintide, with $1.2bn of development milestones, $2.4bn of sales milestones, a 50/50 profit share in the US and Europe and tiered royalties up to the high teens elsewhere, which William Blair called the highest total consideration in the obesity field. Petrelintide is now heading into Phase 3 with trials starting in H2 2026. Buying the rest of Zealand collapses the profit share and the ex-US royalty into one line and removes a partner from the fixed-dose combination with Roche's own CT-388.

Why it might not happen. Roche already controls the asset through the licence and would be paying a second time for economics it half owns. Zealand has been monetising rather than selling: on 12 August 2026 it sold rusfertide's economics to Royalty Pharma for $100m, keeping a 1% royalty. And the amylin re-rating since the Roche deal has raised the price of the half Roche does not have.

Royalty read-through. The cleanest on this list. A takeover extinguishes the 50/50 US/EU profit share, the rest-of-world royalty and the $350m combination-product payment Zealand owes Roche, and transfers Zealand's rusfertide arrangement with Royalty Pharma and its legacy glucagon economics to a new obligor.

24. Someone buys Telix

The vertically integrated radiopharma platform, with a supply chain everyone else is trying to build.

Why it would make a difference. Telix guided FY2026 revenue and other income above $1bn, with Q2 revenue of $247m up 21%, a Regeneron collaboration that paid $40m upfront, and manufacturing capacity expanding in Australia, Belgium and Japan on top of the RLS radiopharmacy network. Novartis is the buyer with the most reason: Pluvicto sold $651m in Q2 2026, up 43%, Lantheus has tentative approval for a Lutathera generic, and Narasimhan told Fierce the commitment to radioligands remains completely unchanged. Lantheus is the other: Pylarify sold $240.9m in Q1 2026, down 6.5% on competition from Telix's Illuccix and Gozellix, and a merger would end the PSMA imaging price war by owning both sides of it.

Why it might not happen. Telix has been the acquirer, buying RLS and issuing $600m of 2031 convertibles to fund its own pipeline. A Lantheus/Telix combination puts the two largest PSMA PET agents in one company, which is the overlap regulators look at first. Novartis has said its own supply network is the moat, which is an argument for building rather than buying.

Royalty read-through. Telix's Regeneron collaboration and Lantheus's GE HealthCare licence on piflufolastat in Japan are the two live agreements; either transaction reprices them. Radiopharma also runs on isotope supply contracts, and those carry assignment terms that a change of control triggers.

25. Lilly buys the channel

The manufacturer that built its own pharmacy decides to own the rest of the pipe.

Why it would make a difference. On 6 November 2025 Lilly and Novo signed pricing deals with the White House that put Zepbound and Wegovy on TrumpRx at $350 a month or less, with the lowest Zepbound dose at $299 through LillyDirect. TrumpRx launched on 5 February 2026 as a direct-to-consumer channel designed around, and partly against, pharmacy benefit managers. A company already fulfilling a third of its new obesity starts outside the PBM system has the volume to justify owning a pharmacy, a fulfilment network or a small PBM outright.

Why it might not happen. The three PBMs that matter are inside UnitedHealth, CVS and Cigna, and no carve-out has been offered. Experts quoted at the TrumpRx launch said the direct channels mostly benefit the uninsured for now. Owning the channel puts Lilly across the table from the payers who reimburse the rest of its portfolio.

Royalty read-through. Thin on royalties, heavy on contracts. The telehealth and fulfilment partners behind LillyDirect, the rebate structures in existing PBM formularies, and the Chugai royalty on Foundayo (entry 19) all sit downstream of who owns the channel.

26. L'Oréal takes the aesthetics, Galderma keeps the drugs

A carve-out that separates a cosmetics franchise from the second-largest new dermatology royalty.

Why it would make a difference. L'Oréal raised its Galderma stake to 20% by buying a further 10% from the EQT-led consortium, completed in February 2026. Galderma's 2025 net sales passed $5.2bn and it doubled its Nemluvio peak sales target to above $4bn after the drug went from $23m to $452m in a year. Jefferies has said L'Oréal is unlikely to want the prescription business. The structure writes itself: L'Oréal absorbs Restylane, Sculptra and Cetaphil, and a Nemluvio-led dermatology pharma lists on its own.

Why it might not happen. L'Oréal has said it is not contemplating a further increase in its stake. Galderma's strategy is built on keeping aesthetics, consumer and prescription under one roof, and its CEO has defended that. A carve-out at the point where Nemluvio's economics are accelerating would be selling the growth to fund the mature business.

Royalty read-through. Nemluvio is licensed from Chugai, which books export sales and royalty income on it every quarter. A carve-out moves that licence to a new obligor and puts the Cetaphil and filler brand economics on the other side of the line. This is the same Chugai as entry 19, which means one Japanese company holds royalties on the industry's largest new obesity product and one of its largest new dermatology products, and is itself 60% owned by Roche.

Mooted, and outlandish

Deals that have been floated, tried, or that only look absurd until you check the register.

Roche buys Illumina, again. Roche launched a hostile tender for Illumina in January 2012 at $44.50 a share and raised it to $51, about $6.8bn, before walking away after Illumina's board called the offer grossly inadequate. Illumina's rationale then was a sequencing-to-diagnostics pipeline; Roche's diagnostics division still exists and Illumina's post-Grail valuation is a fraction of its peak. The absurd part is the memory, on both sides.

Berkshire Hathaway buys a pharma. Berkshire put $5.66bn into AbbVie, Bristol Myers, Merck and Pfizer in Q3 2020 and had sold most of it by the end of 2021. The one position it built while selling the others was 13.1m shares of Royalty Pharma for $475m in Q3 2021. Berkshire kept the royalty vehicle and sold the operating companies.

The Novo Nordisk Foundation buys Lilly. Absurd on scale: Lilly has been worth more than $1trn since November 2025. Less absurd than it sounds: the Foundation controls 77.3% of Novo Nordisk's votes, owns Catalent through Novo Holdings, and answers to no shareholder. It has the patience. It lacks about $900bn.

Regeneron buys Sanofi's half of Dupixent. A reversal of the power dynamic in entry 2, and structurally the simpler transaction: a buy-out of a partner's share of an existing collaboration is a contract amendment with a price attached, and Regeneron would be acquiring an asset it already co-owns and co-runs. The obstacle is that Sanofi would be selling its largest product at the moment its new CEO has been told to fix the pipeline around it.

Lilly buys Novo. The obesity duopoly collapsed into one company. Blocked by the same 77.3% as above, and by every competition authority on Earth.

Also on the list

Pairings the survey would have produced, each blocked for a reason already covered above.

• Takeda, Astellas and Daiichi Sankyo. The Japanese champion. Blocked by Daiichi's Enhertu economics already sitting with AstraZeneca.
• Teva, Sandoz and Viatris. The biosimilar champion that would collect the 2028 cliff. Blocked by molecule-by-molecule generics antitrust and three balance sheets.
• Moderna and BioNTech. Two mRNA platforms with shrinking vaccine revenue. Blocked by founder control and the fact that a merger validates no solid-tumour data.
• Novartis and Alnylam. Deepens the RNAi bet. Blocked by Novartis already holding the asset it wants under licence.
• UCB and Argenx. The Benelux immunology champion. Blocked by Argenx being coveted by everyone.
• Ipsen and Servier; Chiesi, Menarini and Angelini. French and Italian champions. Blocked by foundations and families that do not sell.
• AbbVie and Vertex. Post-Humira diversification. Blocked by AbbVie's leverage and Vertex's independence.
• Merck KGaA and Bayer. The German champion. Blocked by E. Merck KG and by Bayer's own unfinished sequencing.
• AstraZeneca or Amgen and Viking. The last large independent GLP-1/GIP asset. Blocked by two binary readouts and by both buyers already running their own Phase 3 obesity programmes.
• Novo and Structure. The oral GLP-1 catch-up after losing Metsera to Pfizer. Blocked by Novo's own oral franchise and a target funded into 2028.
• BioMarin and Ultragenyx. A rare-disease combination that Martin Shkreli publicly urged after Ultragenyx's Angelman failure. Blocked by BioMarin having just spent $4.8bn on Amicus.
• Amgen enters radiopharma. Every other large cap has a platform. Blocked by capital committed to MariTide and no manufacturing base.

Why none of it happens

The scepticism has numbers attached. H1 2026 biopharma M&A totalled $96bn across 80 deals on J.P. Morgan's count. The largest transactions of the year sit around $10bn to $12bn. Buyers favour bolt-ons in the $1bn to $10bn range.

Structure is doing the work equity used to do. Bristol Myers handed five immunology assets to a Bain-backed company, kept close to 20% of the equity, and took royalties and milestones on each asset. That is a large pharma converting pipeline it will not fund into a royalty position, at no cash cost.

Christensen and colleagues put the M&A failure rate at between 70% and 90% in 2011; Lev and Gu, on 40,000 deals over 40 years, put it at 70% to 75%. Pharma investors have read both. AstraZeneca fell on a rumour and rebounded on a denial.

For a royalty desk, the useful version of a fantasy list is the inverse one: which existing streams would be repriced if any of these happened. The answer lives in change-of-control, assignment and competing-product clauses that nobody models until the announcement. A Merck and Bristol Myers combination alone would put two PD-1 divestitures, an enzyme licence dispute and a set of alliance revenue lines into play on the same day. That mapping can be built from filings.

If any of the twenty-six does happen, the drink is on you.

All information in this report was accurate as of the research date and is derived from publicly available sources including court opinions, regulatory guidance, academic literature, SEC filings, and financial news reporting. Information may have changed since publication. This content is for informational purposes only and does not constitute investment, legal, or financial advice. The author is not a lawyer or financial adviser.