Merck & Co. is the one company in this oncology story for which we hold audited, fidelity-checked numbers, and they show a large swing. Its cash and equivalents stood at $18.169 billion at the end of Q3 2025.1 By Q1 2026 they were $5.327 billion, and in Q2 2026 they were $6.849 billion.1 That is a fall of roughly two-thirds in two quarters. The dossier does not say why the balance fell, so we will not guess. What it does show is that the company heads into a cluster of cancer-drug catalysts with far less cash on hand than it had last autumn.
For scale, $6.8 billion is about 37 times the $184 million in quarterly revenue that Novocure just reported.8 The sector is not uniformly rich or uniformly short of money. Capital is unevenly spread, and this piece looks at where it sits.
The Merck balance sheet: what the filings say
Merck's year-end cash was $6.841 billion for FY 2023, $13.242 billion for FY 2024 and $14.565 billion for FY 2025.1 The company is clearly not cash-starved: its year-end cash more than doubled over two years. Quarter-end figures swing a lot, though. Q1 cash was $5.579 billion in 2024, $8.629 billion in 2025 and $5.327 billion in 2026.1 Single quarters are therefore a poor guide to the underlying position, and the Q3 2025 peak of $18.169 billion looks like an outlier in a series that mostly sits between $5 billion and $15 billion.1
Capital spending on plant and equipment rose in FY 2025. Capex was $3.863 billion in FY 2023, $3.372 billion in FY 2024 and $4.112 billion in FY 2025, an increase of about 22% over the prior year by our arithmetic.2 The first quarter of 2026 saw $991 million of capex, down from $1.328 billion in Q1 2025.2 A single quarter's $991 million is more than five times Novocure's entire quarterly revenue.8
The cost line is rising faster
Cost of revenue, the direct cost of making and delivering what Merck sells, was $16.126 billion in FY 2023, $15.193 billion in FY 2024 and $16.382 billion in FY 2025.3 The 2026 quarters stand out. Q1 2026 was $4.195 billion against $3.419 billion a year earlier, and Q2 2026 was $4.395 billion against $3.557 billion.3 By our calculation that is an increase of about 23% in Q1 and about 24% in Q2. The filings do not say what drove it, and the dossier carries no revenue figures for Merck. We therefore cannot say whether sales grew as fast, which is the figure that decides whether this squeezes profit. Treat it as a flag to check when Merck's next filing arrives, not as a conclusion.
The catalyst calendar
Three events fall into the next several weeks, and one has already landed.
The first is intismeran, the personalized mRNA cancer vaccine Merck is developing with Moderna. Moderna and Merck plan to present detailed results at ESMO on October 24, 2026 and to engage with regulators about potential filings.4 That is a presentation and a conversation, not an approval. A filing, if one follows, would come later.
The second is a regulatory date. Summit Therapeutics' Biologics License Application for ivonescimab, based on the HARMONi trial, has an FDA action date of November 14, 2026.5 Summit has also announced a clinical trial collaboration with Daiichi Sankyo and AstraZeneca to test ivonescimab alongside Datroway across several solid tumor settings, intended to include breast and lung cancer.6 The dossier gives no deal terms, so we cannot tell readers who pays for what.
The third has reported. The EXCALIBER-RRMM Phase 3 trial showed a median progression-free survival of 42 months for the ZDd regimen against 20 months for DVd, a hazard ratio of 0.49 and a 51% reduction in the risk of progression or death, with median follow-up of 23 months.7 This was a dual-primary endpoint and the reported p-value was below 0.000001.7 Follow-up of 23 months is shorter than the median PFS in the winning arm, so the long-run picture is still being filled in.
Who is spending and who is cutting
The commercial companies look strong. Novocure reported second-quarter revenue of $184 million, up 16% year over year, driven by 18% growth in active patients.8 Its CEO Frank Leonard said: "This was our strongest quarter to date, with record net revenues and active patients on therapy."8 Novocure also received a CE Mark for Optune Pax in locally advanced pancreatic cancer, with Germany the first EU market to launch.8
Smaller developers are managing their cash. ADC Therapeutics announced a workforce reduction "projected to deliver $10 million in annualised cost savings", and said it "maintains a cash runway into 2028, supporting ongoing regulatory and clinical plans for ZYNLONTA combinations."9 The savings are small next to Merck's numbers, but for a company of that size they are the difference between a 2028 runway and an earlier funding round. The same RTTNews roundup reports a week of FDA and EU approvals, layoffs, acquisitions and late-stage readouts, which is the sector's usual mix of progress and pressure.9
New programs are still being funded. Telix dosed the first patient in its Phase 3 LUTEON trial of TLX250-Tx in kidney cancer, which it describes as the first radiopharmaceutical therapy to enter Phase 3 in that disease.12 In Europe, the CHMP recommended Gilead's Trodelvy plus Keytruda for first-line PD-(L)1-positive triple-negative breast cancer.11
The AI pitch, and how much weight it can bear
AstraZeneca's Puja Sapra argues that AI can shorten discovery. She describes a "lab of the future" in which "this system uses AI to make predictions, robotic systems to execute experiments, and instruments to generate data."10 She says that "Data is our differentiator", meaning AstraZeneca's proprietary datasets are the edge for training its models.10 She is also candid about the limits: "One of the hardest problems in de novo design is predicting whether a computationally generated molecule will be safe in the human body."10 And she stresses that "human oversight remains at the heart of this approach."10
Readers should know how this source performs in our checks. The article, published by MIT Technology Review, has a measured reliability of 0% across the 11 claims we checked against it.10 That does not mean the quotes are misattributed, since we quote them as the speaker's words. It means that claims drawn from this piece should be read as one executive's pitch, not as established results. The dossier contains no figure for how much AI has actually shortened any timeline or cost, and we do not supply one.
The press-release sources are also imperfect. The NewsEOD documents behind the Novocure, Gilead, Telix and ADC Therapeutics items held up on 56%, 56%, 31% and 45% of the claims we checked across those feeds.8,11,12,9 Company-issued numbers deserve cross-checking, which is why we lean on SEC-filed figures where we have them.
What to watch
- October 24: the ESMO presentation of intismeran data from Moderna and Merck.4
- November 14: the FDA action date for ivonescimab.5
- Merck's next filing: whether cash stays near the $5-7 billion of the last two quarters and whether cost of revenue keeps rising about a quarter faster than a year ago.1,3
- ADC Therapeutics: whether its 2028 runway holds as the $10 million in savings takes effect.9
If you hold a broad index fund you may own Merck indirectly, but the dossier contains no index weightings, so we cannot say how much. Nothing in these sources is investment advice, and the dated catalysts above are the points at which the picture will change.
