Arcus Biosciences, Inc. (RCUS) stock intrinsic value fair value estimate & owner earnings analysis

Arcus Biosciences is a clinical-stage cancer drug developer with no approved products. Its revenue ($33 million in 2025) comes from collaboration payments and cost-sharing with Gilead Sciences and Taiho Pharmaceutical. The business consumes cash: on our owner-earnings basis (operating cash flow, less working-capital swings from collaboration timing, less stock compensation, less maintenance capital spending) it lost about $271 million in 2024 and $350 million in 2025. The only year in the last eight with a positive figure was 2021 ($0.88 per share), and that came from collaboration revenue recognised that year from cash Gilead had paid earlier, not from a self-funding business. Because there is no earnings stream to capitalise, we value what shareholders can demonstrably keep. At June 30, 2026 Arcus held $775 million of cash and investments against $100 million drawn on a Hercules term loan, or about $675 million net, roughly $5.30 per share on 127.3 million shares. Cash fell from $1,010 million at the end of 2025 as operating cash outflow ran at $257 million in the first half of 2026. The company says its cash funds planned operations until at least the second half of 2028, but every dollar spent before then must earn its keep. We credit no value to any drug that has not yet succeeded. The pipeline has already had a hard year: the Phase 3 STAR-221 trial of domvanalimab was stopped in December 2025 after an interim analysis showed no benefit over standard care, and the company has since discontinued the STAR-121 lung cancer trial and announced it will discontinue the PACIFIC-8 trial run with AstraZeneca. The two remaining Phase 3 efforts are PRISM-1 (quemliclustat in pancreatic cancer, fully enrolled in October 2025, results expected in the first half of 2027) and PEAK-1 (casdatifan). Taiho could owe up to $172 million of clinical and regulatory milestones on casdatifan, but these depend on trial outcomes, so we leave them out. Gilead, which held about 25% of the shares at the end of 2025, gave up its three board seats effective August 5, 2026. Each case is a story about how much cash is spent before management stops funding programs that have not delivered. Arcus's outside trial and service contracts are generally cancellable on 30 days notice, so spending is a choice. - Bear ($0.90 per share): PRISM-1 and PEAK-1 do not deliver, and spending continues at roughly $400 million a year through 2027 before a wind-down. About $114 million of the $675 million net cash is left, before any wind-down costs, which we could not quantify. - Base ($2.29 per share): spending runs at about $200 million per half-year through the PRISM-1 readout in the first half of 2027, then stops absent evidence of value. About $292 million remains. - Bull ($2.56 per share): the wind-down of the TIGIT programs lowers spending to about $175 million in the second half of 2026 and $190 million in the first half of 2027, leaving about $326 million. All cases discount at 6% and credit no growth. Stock compensation (about $60 million in 2025) is treated as a real cost in the historical figures, and the cases assume no new share issuance. What would make this wrong: a positive PRISM-1 or PEAK-1 result would create a drug worth far more than anything we credit, and this valuation would then be too low. We deliberately do not price that possibility until the evidence exists. The reverse risk is that cash is consumed faster than assumed, or that the Hercules loan end-of-term charge and the liability tied to previously sold future royalties reduce what shareholders keep; we have not deducted these. The 2026 figures come from the June 2026 quarterly report and the company's August 2026 results announcement; everything else comes from the annual reports.