On 26 August 2026, SK Biopharmaceuticals agreed to license an epilepsy platform from Biohaven. Sixteen days earlier, Jazz Pharmaceuticals agreed to buy an epilepsy biotech outright. Same disease area, same month, two opposite structures. The licensing versus acquisition decision now shapes almost every pipeline strategy in large pharma. In 2026 the split between the two routes has grown sharper, and the sums involved have grown considerably larger.
The numbers behind the licensing versus acquisition split
J.P. Morgan’s half-year biopharma deal report tracked both routes through the first half of 2026. Announced biopharma licensing carried $166.7 billion in research and development value. Upfront cash accounted for only 6% of that figure. Mergers and acquisitions reached $96 billion across 80 deals over the same period. Of that, $55.1 billion came from 48 transactions in the second quarter alone.
The headline gap flatters licensing. Licensing totals are quoted as potential value, including milestones that may never fall due. Acquisition totals reflect committed capital. Read properly, licensing buys broad optionality cheaply. Acquisition buys certainty expensively. Both routes are expanding at once, and neither is displacing the other.
EY’s 2026 Firepower report recorded the same pattern a year earlier. Life sciences companies signed $240 billion in acquisitions during 2025, up 81% year on year. Deal volume actually fell 12%. Mean deal size reached $2.1 billion, more than double the 2024 average. Alliance biobucks broke every previous record in the same year. Average alliance size rose 46%, yet upfront payments represented roughly 7% of potential value.
Why the growth gap forces both routes at once
EY projects a $370 billion growth gap by 2032 for the 25 leading biopharma companies. The shortfall reaches roughly $100 billion as early as 2028. PwC puts more than $300 billion of branded revenue at risk from loss of exclusivity this decade.
Capital is not the constraint. EY put available industry firepower at a record $2.1 trillion in December 2025. Scarcity sits on the supply side instead. PwC’s US deals midyear outlook put first-quarter 2026 pharmaceutical and life sciences deal value above $65 billion. That figure was close to double the same quarter of 2025. Sixteen billion-dollar biopharma deals were announced in those three months.
Investors remain unconvinced. EY reviewed analyst commentary on the top 25 biopharma companies. More than half of it was neutral or negative on pipeline quality. More than 40% took the same view of their dealmaking strategies. That scepticism raises the cost of getting the licensing versus acquisition call wrong. The billion-dollar deals of the year’s opening months show the same pressures at work.
Policy turbulence has not slowed either route. Tariff threats and most favoured nation pricing proposals rattled valuations through 2025. EY concluded that such disruption had become the industry’s new normal. Acquirers kept signing regardless.
What actually decides the licensing versus acquisition call
Deal structure follows value creation rather than fashion. Dr Caroline Austin, vice president and head of transactions, business development at AstraZeneca, set out the test plainly. Speaking to EY, she said the company weighs where post-deal value will be created and who can best create it. In practice, she said, “We typically consider strategic fit, respective capabilities, speed and risk/valuation.”
Nauman Shah, who leads Janssen business development at Johnson & Johnson, described a deliberately agnostic toolkit. It spans internal research, acquisitions, licences, collaborations and alliances. He told EY that deal activity was picking up clearly, with most recent transactions falling below $5 billion.
Historic returns explain the caution. EY analysed 225 transactions completed between 2010 and 2024. Only 32% delivered the revenue forecast in the deal model. Among the successful minority, 86% involved therapeutic areas where the buyer already competed. Buying into unfamiliar territory rarely worked, even for approved products. Licensing offers a cheaper way to learn an unfamiliar field first.
Two epilepsy deals, two structures
SK Biopharmaceuticals took the licensing route. It secured worldwide rights to Biohaven’s Kv7 ion channel platform, led by opakalim. That candidate is a selective Kv7.2/7.3 potassium channel activator in Phase 2/3 development for focal epilepsy. Total payments run to $795 million, with $400 million in near-term cash. Of that, $350 million falls due at closing and $50 million in 2027. Royalties on US net sales range from the mid-teens to the low twenties. Closing depends on antitrust clearance.
Biohaven keeps running the ongoing trials, with SK covering costs from here. Topline data from the RISE3 study is expected in the second half of 2026. In the deal announcement, chief executive Vlad Coric called opakalim “one of the most exciting assets we have advanced”. In SK Biopharmaceuticals’ own statement, chief executive Donghoon Lee said the agreement “carries immense strategic significance to our company”.
Jazz Pharmaceuticals chose ownership instead. It agreed on 10 August to acquire Actio Biosciences for $820 million upfront. Up to $500 million more depends on approval and sales milestones. The prize is ABS-1230, an oral inhibitor for KCNT1 epilepsy. In the announcement, chief executive Renee Gala said the “emerging clinical profile of ABS-1230 is highly encouraging”. The transaction is expected to close by the fourth quarter of 2026.
The structural contrast is instructive. SK bought access to a platform it can build alongside its existing epilepsy franchise. Jazz is buying a single asset outright, in a field where it already sells Epidiolex. Both choices satisfy the therapeutic alignment test that EY identified.
GSK and AstraZeneca ran both playbooks in one year
Britain’s two largest drugmakers have not picked a side. GSK agreed in June to acquire Nuvalent outright for $10.6 billion, or £8.0 billion. Net of cash acquired, the investment falls to $9.4 billion. The purchase completed on 15 July. The transaction brought two late-stage lung cancer inhibitors already under FDA review. Zidesamtinib was approved as Jideytro on 22 July 2026, two months ahead of its target date. It is GSK’s first approved lung cancer medicine. Neladalkib’s decision is due on 27 November 2026.
GSK said the purchase fits a strategy of buying validated targets that improve on standard of care. When an asset is that close to launch, licensing rarely satisfies the buyer. Ownership captures the full commercial margin.
AstraZeneca has leaned the other way for cross-border assets. In January it agreed to pay CSPC Pharmaceutical $1.2 billion upfront. That secured rights outside Greater China to eight obesity and type 2 diabetes programmes. Milestones could lift the deal to $18.5 billion. In July it agreed a licence for Zegfrovy from Shanghai-based Dizal. AstraZeneca paid $600 million upfront, with up to $900 million in milestones. AstraZeneca confirmed completion on 1 September, ahead of a US launch planned for the fourth quarter. It has still acquired outright, buying the cell therapy developers EsoBiotec and Neogene Therapeutics, and the radioconjugate specialist Fusion.
The middle ground is expanding fast
Neat categories are dissolving. Johnson & Johnson announced agreements with Sail Biomedicines on 29 July 2026. Initial payments total $785 million, including a $465 million equity investment. A further $140 million rests on development milestones. The company also secured an exclusive option to acquire Sail for $2.58 billion.
That structure is neither a licence nor an acquisition. It is a priced call option on an in vivo CAR-T platform. Jazz used a related device. At closing, Actio will spin out a separate company holding its remaining assets, with Jazz retaining a minority stake.
Contingent value rights, milestone earnouts and option frameworks all serve one purpose. They bridge the gap between what a seller expects today and what a buyer will commit today.
China has widened the gap between the two routes
Geography now bends the licensing versus acquisition equation. EY put inbound China alliance value at $94.3 billion in 2025, across 48 deals. That represented 34% of all biobucks spent by US and European biopharma, up from 21% a year earlier. Five of the ten largest alliances of 2025 involved Chinese partners.
Outright acquisition of Chinese biotechs remains rare. Qiang Li of DLA Piper, interviewed for the EY report, points to unavoidable diligence and regulatory complexity. Continuing geopolitical friction compounds the problem. Licensing ex-China rights sidesteps much of that, and lets buyers test unfamiliar data packages before committing further capital. On that reading, the licensing versus acquisition boundary in China is shaped by politics as much as by science.
What to watch through the rest of 2026
Private company buyouts are accelerating. HSBC Innovation Banking’s mid-year healthcare venture report counted 19 venture-backed pre-clinical and Phase 1 biopharma acquisitions. Those came in the 18 months to mid-2026, at a median upfront of $1 billion. Early-stage assets are changing hands at prices once reserved for later-stage deals.
Three tests now loom. Whether Johnson & Johnson exercises its Sail option will show how hybrid structures resolve in practice. The RISE3 readout will test SK’s licensing bet on opakalim. Full-year figures will show whether licensing keeps outpacing acquisition on headline value while trailing badly on committed cash. The licensing versus acquisition question is not settled. It is simply being asked deal by deal.
