Adicet Bio, Inc. (ACET) stock intrinsic value fair value estimate & owner earnings analysis

Adicet Bio is a clinical-stage biotechnology company in Boston, Massachusetts and Redwood City, California. Its lead candidate, prula-cel (formerly ADI-001), is an allogeneic (donor-derived) gamma delta T cell therapy that targets CD20. The U.S. FDA cleared its application to test prula-cel in lupus nephritis in December 2023, and in August 2024 the company added systemic lupus erythematosus, systemic sclerosis and ANCA-associated vasculitis to the program. The autoimmune work is early-stage (Phase 1). The company has no approved product and no product sales. Its only income has been payments from a collaboration with Regeneron Pharmaceuticals, which had paid a $25.0 million upfront fee and $20.0 million of research funding as of December 31, 2025. How we value it. We measure earning power as the cash a business generates for its owners after stock compensation, working-capital swings and equipment spending. Adicet has not produced positive owner earnings in any of the last five years. On that basis its result per share, restated for the 1-for-16 reverse split of December 30, 2025, was about -$36.76 in 2021, -$32.59 in 2022, -$44.16 in 2023, -$20.74 in 2024 and -$16.86 in 2025. The middle of those five years is about -$32.59 per share. Cash used in operations was $95.2 million in 2025. Stock compensation of $14.3 million is treated as a real cost. We did not credit any earning power from prula-cel or the Regeneron collaboration, because there is no approval, no signed commercial partnership and no product revenue behind them. With no earnings to discount, the question becomes what shareholders could keep: the cash left after another year of spending, less what the company owes. The starting point is this. The company held $158.5 million of cash and short-term investments at December 31, 2025. It used $41.3 million in operations in the first half of 2026, which leaves roughly $117 million. That first-half spending equals about $83 million a year. In the first half of 2025 the company used $52.0 million, so spending is about 21% lower this year, after the July 2025 workforce reduction. Total liabilities were $28.0 million at June 30, 2026. There were about 9.59 million common shares at December 31, 2025, plus 1.15 million pre-funded warrants, or roughly 10.74 million shares in all. The October 2025 offering raised about $74.8 million net, selling 4,375,062 shares plus 625,000 pre-funded warrants. Bear case: $0 per share. Spending continues at the 2025 pace of about $95 million for another year, leaving roughly $22 million, which is less than the $28 million of liabilities. The company would owe more than its remaining cash, so nothing is left for shareholders, and any new financing would dilute current holders further. Neither this case nor the others counts any value for prula-cel, because nothing signed or approved supports it. Base case: $0.60 per share. Spending continues at the first-half 2026 pace of about $83 million a year for one more year, leaving roughly $35 million of cash. After $28 million of liabilities, about $6.6 million remains, or about $0.60 per share on 10.74 million shares. The one-year horizon is our modelling choice. We credit no revenue because there is nothing yet to credit. Bull case: $4.46 per share. The company stops spending after only six more months at the first-half 2026 pace, a modelling choice. That leaves about $76 million of cash and about $48 million after liabilities, or about $4.46 per share. The cost cuts behind this pace are already visible in the reported cash flow. This case still gives no credit to future drug sales. What would make this wrong. A partner paying real cash for prula-cel, or clear efficacy data in lupus nephritis, would give the company earning power we have not credited. Balance-sheet cash is not owner earnings, and the cases above are cash-based, not profit-based. The company states that its existing cash is enough to fund operations for at least the next twelve months. We could not tie the one-year horizon to a dated trial result, so it is a modelling choice and not a forecast of an event. Further share sales would dilute the per-share figures above. The December 2025 reverse split combined 16 old shares into one, so all per-share figures here use the new share count.