Back to RXRX overview

Recursion Pharmaceuticals (RXRX): Partner Validation Arrives as Collaboration Revenue Fades

Published September 21, 202617 min read·TickerFile Research · Recursion Pharmaceuticals (RXRX)
ShareXLinkedIn

Recursion Pharmaceuticals is no longer asking investors to underwrite an abstract artificial-intelligence discovery factory. The second-quarter print is the first clean look at whether partner science, not partnership accounting, can carry the equity after three mid-stage programs were killed last year, Exscientia was absorbed, and founder Chris Gibson left the board. Genentech optioned the collaboration's first previously unexplored neuroscience target into a joint early discovery program, which is the first external scientific vote that the Recursion maps can produce biology a large partner wants to own. That proof arrives while collaboration revenue is falling because prior Roche project phases are already complete.

The tension is that the income statement is still recognizing leftover deferred work rather than new cash. Total revenue was $8 million in the quarter, down sharply from the year-ago period as Roche recognition rolled off completed map-build phases. Cash, cash equivalents, and restricted cash stood at $557 million at the June close. Management lowered full-year cash operating expense guidance to $375 million. That guide is enough on the current plan to fund operations into early 2028. Operating cash outflow still widened because last year's quarter collected a United Kingdom research-tax credit that did not repeat. The equity is therefore a duration trade on platform validation, not a revenue-growth story.

What the next several months resolve is whether the familial adenomatous polyposis MEK program can convert a median polyp-burden reduction into a defined Food and Drug Administration registrational path, and whether Sanofi takes differentiated oral molecules to development-candidate status. Additional TUPELO data is slated for a Presidential Plenary in November. If those events stay qualitative only, the market is left holding a large cash pile against a still-unremediated control weakness and an unused at-the-market facility. Does partner science start to pay, or does the platform remain an expensive option on future maps?