Mind the middle
The likeliest place for a European biotech to die in 2026 is not the laboratory. It is the gap between one funding round and the next.
A good result is no longer enough. The money that carries a biotech from promise to proof has thinned in the middle, and that, not the science, is where companies are now most likely to fail.
The fix is not better fundraising. It is better design.
Ask a biotech founder what keeps them awake, and most will describe an experiment that might not work: the result that will not replicate, the endpoint that refuses to move. It is a reasonable fear. Fewer than one in twelve drugs that enter human trials ever reach approval (BIO, Informa Pharma Intelligence and QLS Advisors, 2021). And yet, through 2025, it was the wrong fear. The thing now most likely to kill a European biotech is not its science. It is the money running out in the gap between one round and the next.
The missing middle
The money has changed shape. Venture firms raise their funds from pension funds, endowments and wealthy families, and those backers apply a simple test: did the last fund hand cash back? Through 2025 the answer was mostly no, because almost nothing went public and few companies were bought. So the backers tightened the taps. American healthcare funds raised less in 2025 than in any year for a decade, money going into new drug companies fell by nearly a fifth, and nearly half of what remained chased artificial intelligence (Silicon Valley Bank, H1 2026).
Scarcer money behaves predictably. It gathers at the two ends of a company's life and avoids the middle. Investors will still back a clean idea at the start, when the cheque is small and the dream is large, and they will pay well at the finish, for a test or a therapy whose data already works. What they are reluctant to fund is the long, costly stretch in between: the years when the first money is spent, the results are promising but unproven, and more cash is needed to find out whether the promise holds. Biologists already have a name for that stretch. It is the valley of death, and in 2025 the bridge across it was quietly pulled up.
The risk that moved
This is why the founder's instinct misleads. A drug takes about a decade to reach approval and burns through more than two billion dollars along the way (BIO, Informa and QLS, 2021; DiMasi, Grabowski and Hansen, 2016). No founder carries that sum. The journey is made in stages, each paid for by a separate raise. Remove the financing in the middle and the company dies with its science perfectly intact. Diagnostics is no gentler: investment in tests and tools fell by 15% in 2025, and the usual obstacle is not the assay but the slow road to reimbursement (Silicon Valley Bank, H1 2026).
The company dies with its science perfectly intact. It is the cruellest way for a good idea to go.
A colder question
Investors have drawn the obvious conclusion, and now ask a colder question. Not merely does the science work, but can this company raise again? In 2025 they favoured biology already proven somewhere over anything novel, and clubbed together so that no company they backed would be left short of cash (Silicon Valley Bank, H1 2026). The person across the table is no longer only judging the experiment. They are judging whether the next round exists. A founder who has not thought about that has already lost half the meeting.
Mike Carusi of Lightstone Ventures put the mood plainly: there is no "rising tide lifts all boats" environment yet. The companies that raised easily in 2025 had real clinical data, a large market, and some distance already travelled. The rest fought over what was left.
The buyer pays late
The reward at the end has moved in the same direction. When Pfizer agreed to buy the obesity company Metsera, it handed over only about three-quarters of the headline price at signing; the rest, more than two billion dollars, arrives only if the drug clears the hurdles still ahead (Silicon Valley Bank, H1 2026). This is now the habit of large buyers. They increasingly license rather than purchase outright, and release most of the money only as each result lands. The buyer, like the investor, has decided to pay for evidence, and to pay late.
An architecture problem
Set the two against each other and the lesson for a founder is not financial but architectural. Each raise has to buy a result clear enough to make the next raise possible, in a market that no longer pays for effort or promise. The reward at the end will arrive in instalments, each released against a result a partner has named in advance. Both ends of the company now follow one rule: money follows evidence, and it comes in stages.
A broker finds the money for the round in front of you. The harder and more useful task is to design the company so that the round after that, and the buyer after that, are within reach: to assemble backers who can keep funding you, to spend at a pace that buys time to reach a result rather than merely to stay busy, and to choose which result to chase first because it unlocks the next cheque as surely as the next paper. In this market, surviving the middle is not luck. It is something you design in from the first round.
If your next round depends on a result you have not yet planned to produce, that is the conversation to have before you raise, not after.
Book a 30-minute call →Like what you're reading?
Subscribe for more strategic notes on biotech and venture design.
Frequently asked questions
Sources: Silicon Valley Bank, Healthcare Investments and Exits, H1 2026 (January 2026); BIO, Informa Pharma Intelligence and QLS Advisors, Clinical Development Success Rates 2011 to 2020 (2021); DiMasi, Grabowski and Hansen, Journal of Health Economics 47 (2016).