KROS

The headline numbers scream 'cheap' — 4.9x P/E, 0.9x P/S, 35.7% net margin, 28.7% ROE — but they are an accounting mirage. The MD&A states plainly that Keros 'has not generated any revenue from product sales as none of our product candidates have been approved for commercialization.' The entire FY2025 revenue jump ($3.55M in FY2024 to $244M, +6775%) and the swing to $87M net income come from the exclusive elritercept license signed with Takeda in December 2024 (effective Jan 16, 2025). That is a lumpy, largely non-recurring upfront event, not a durable earnings stream. Prior years — FY2024 -$187M, FY2023 -$153M, FY2022 -$105M — show the real run-rate of a company still spending $130M/year on R&D. Trailing multiples on one-time licensing income are meaningless for forward value.

What is real and supportive is the balance sheet. With $287M cash, only $34.9M total liabilities, and a 0.12x liabilities/equity ratio, Keros carries essentially no financial risk today. Strikingly, market cap ($222M) sits below net cash (~$14.57/share vs. the $11.26 price), so at the current quote the market assigns negative value to the pipeline — the Takeda-partnered elritercept (KER-050, MDS/myelofibrosis) plus the wholly-owned rinvatercept (KER-065, Duchenne muscular dystrophy and ALS). That cash cushion is why this is a hold, not an avoid.