Schrodinger, Inc. (SDGR) stock intrinsic value fair value estimate & owner earnings analysis

Schrodinger sells physics-based computational software used to design molecules, mainly for drug discovery and also for materials science. It also runs a drug discovery business that earns fees and potential milestones and royalties from partners (Bristol Myers Squibb, Lilly, Novartis) and develops a few of its own drug candidates. In 2025 total revenue was $255.9M: software $199.5M (on-premise licences $101.4M, hosted subscriptions $45.1M, maintenance $27.3M, services $9.6M, and $16.0M of Gates-related contributions) and drug discovery $56.4M. Software gross margin was 74%, and the company reports 96% gross dollar retention among commercial software customers. How we value it. We value shares on owner earnings: cash from operations, minus the cash effect of working-capital swings, minus stock-based compensation (a real cost), minus the part of capital spending needed just to stay in business. On that basis the company has not earned anything for its owners in any year we could measure. Per diluted share it was roughly -$0.77 (2020), -$1.47 (2021), -$1.87 (2022), -$2.31 (2023), -$2.60 (2024) and -$2.06 (2025). Reported cash flow can look better than this because of timing. In 2024 Novartis was billed a $150M upfront fee that was collected in January 2025, so 2025 operating cash flow was a positive $13.9M even though excluding that collection and other working-capital movements the business consumed about $151M after stock compensation. In 2023 the company reported net income of $40.7M, but that came from $147.2M of gains on equity stakes plus $53.5M of fair-value gains, not from operations. The $48.2M gain on the Structure Therapeutics stake in 2025 is likewise not an operating earning. Operating cash flow for the first half of 2026 was again negative (about -$49M) in the company's June 2026 quarterly report, though the first half is seasonally lighter for on-premise software renewals. What drives value. The software franchise is real: segment gross profit was $148.5M in 2025 versus $143.5M in 2024 and $129.6M in 2023. But the company spent $309.5M on research, sales and administration in 2025, of which the filing does not let us separate platform costs from drug-program costs. Drug discovery itself had negative gross profit of $5.9M. Management has been cutting: a May 2025 restructuring removed about 60 employees and is expected to save roughly $30M a year, one clinical candidate (SGR-2921) was discontinued in August 2025, and the company states it does not intend to start further clinical trials beyond the ongoing Phase 1 studies of SGR-1505 and SGR-3515. Even so, the 2025 operating loss was $166.9M; if every dollar of external program spending ($49.3M) disappeared, an operating loss of roughly $118M would remain before adding back non-cash stock compensation of about $44M, so on our measure earnings would still be negative. We found no signed contract, backlog or track record that turns this into positive owner earnings. The three cases. Bear: cash burn continues near recent levels, the company depends on its own cash and capital raises, and the on-premise to hosted software transition lowers reported revenue near term, as management itself says it expects. Value to shareholders on our measure is $0. Base: software keeps growing modestly as the record suggests, costs come down as announced, but no evidenced path takes owner earnings above zero. Value is $0. Bull: the announced cost cuts and the end of new clinical spending are fully realised and software continues to grow; even then, with the cost structure disclosed, we cannot build positive owner earnings from named, evidenced drivers, so this case is also $0. We did not credit BMS, Lilly or Novartis milestones (for example up to $2.272B potential milestones under the Novartis agreement, of which none had been recognised as of December 31, 2025), because they depend on partners' programs succeeding and have not yet happened; we did not credit software share gains or a hoped-for margin recovery for the same reason. An important limit of this approach: we do not add the company's balance-sheet cash and securities ($402.3M at December 31, 2025) on top of any case, and we do not subtract debt. Our method values only the earnings stream. The company's losses are large relative to that cash, and management states the cash should cover operating needs for at least 24 months, so the $0 values reflect that the business does not currently earn a return for owners, not that the shares have no claim on that cash. The company does not appear to owe more than it is worth; the cases simply floor at zero. What would make this wrong: a sharp cut in spending that leaves the software business (about $148M of gross profit) carrying only modest overhead, or a large milestone or royalty stream from partners. Neither is yet in the filings as a signed or delivered fact. The valuation would be revisited if owner earnings turn sustainably positive.