AbCellera Buries Its OX40 Bet and Prices a Hot-Flash Franchise Instead
A 50-point placebo-adjusted win turns a platform company into a late-stage biotech — and quietly kills the program management spent a year defending.
ABCL · Earnings Call · 2026-11-09
The quarter where the story changed shape
For eighteen months AbCellera's pitch was a two-handed one: a platform that makes antibodies, and two clinical shots on goal. On November 9 the company effectively told investors to stop looking at the second hand. Q3 delivered the Phase II readout for ABCL635 in vasomotor symptoms (VMS) — hot flashes — and the numbers were not "comparable to the small molecules," they were multiples better. The ABCL635 group showed an 83% mean reduction in moderate-to-severe hot flash frequency at four weeks versus 33% for placebo, a placebo-adjusted treatment difference of 50% against the roughly 20%–25% registrational benchmark for fezolinetant and elinzanetant. Severity split was wider still: 58% versus 12%, a 46-point delta where the small molecules delivered a fraction of that. And 37% of treated patients had complete resolution of moderate-to-severe events versus 2.2% on placebo.
That is the number that matters, because it is the number management had explicitly refused to promise. In August, Carl Hansen framed success as merely matching the incumbents: “on the frequency side, we are obviously looking for a response relative to placebo that's on the order of 20% at least, which is about what you see with the small molecules” — Carl L. Hansen, President and CEO · 2026-08-05. And back in May he set the celebration bar even more plainly, calling a product “with efficacy that is comparable or equal to the small molecules” the target product profile. Delivering double the bar is why the stock has run +62% in ninety trading days and why late stage development now leads the company's fresh keyword set with real force — up from a background theme to the central plan, with Phase III initiation guided to 2027.
The program that stopped being talked about
Here is the tension. On the same call, AbCellera buried ABCL575, its OX40 Ligand antibody. “Based on these external results, we will not pursue development of ABCL575 in atopic dermatitis” — Carl L. Hansen, President and CEO · 2026-11-09 — and, more damningly, no plans beyond Phase I. It still reported the Phase I cleanly (a ~65-day mean half-life, no serious AEs), which rather underlines that the science was fine and the commercial read was not.
Rewind to February and the same management was answering an analyst question about the OX40 class's Kaposi sarcoma signal with conviction. Sarah Noonberg: “I agree with Carl, our overall investment thesis in 575 hasn't changed. We believe that OX40 ligand is a pretty compelling target.” — Sarah Noonberg, Unknown · 2026-02-24 In November 2025 Carl was still arguing the molecule could enable dosing “perhaps 6 months.” The class has now been read out for them: amlitelimab discontinued in atopic dermatitis, rocatinlimab too. The OX40 ligand class has fallen out of the story as quietly as an investor relations team can manage — a keyword that topped the company's own thematic rankings a few quarters ago, now mentioned only as a post-mortem.
The real tell is where the money and language moved instead. The new program is ABCL992, a half-life-extended analog of 635 whose whole purpose is to protect the franchise: “we've had this molecule in the wings ready to go, and we're able to bring it into IND-enabling activities immediately following the data.” — Carl L. Hansen, President and CEO · 2026-11-09 That is a defensive move made in anticipation of fast followers. A dosing interval that stretches from monthly toward quarterly, built on a half life extension, is how you defend a mechanism rather than invent a new one. Alongside it, two undisclosed programs, one of which — ABCL646, an antagonist against an Ion Channel target — is the long-standing platform bet finally surfacing as a clinical asset.
A timeline that slipped, gently
The one soft spot in the Q&A. The 12-week washout and open-label extension data had been expected by end-2026; now it is early 2027. Kripa Devarakonda of Truist pressed directly, and Sarah Noonberg's answer was operational throat-clearing: “There is nothing unusual from an operational standpoint... it's the preferred timing of release of this data between late 2026 and early 2027. So there is nothing that fundamentally has changed.” — Sarah Noonberg, Chief Medical Officer · 2026-11-09 Possibly true; single-dose washout curves are only interesting insofar as they build the exposure response model for dose selection, and the company is mid-way through the FDA's model-informed drug development dialogue. But note what was declined: the head-to-head question against an active comparator, floated in August, is now "we will not be sharing any more information about our pivotal study." The next disclosure risk is real.
What the numbers underneath say
The financials confirm the pivot rather than merely accompany it. Revenue has collapsed as the partnership model wound down — total revenue of $4M in the June quarter, down 76% year over year — while spending stays pointed at the clinic. R&D held near $46M a quarter and the loss ran to $55M in the June quarter, narrowing to roughly $41M in Q3 per the call. SG&A actually fell from $25M to $14M as the Bruker litigation concluded — a real cost line removed.
The balance sheet is the whole argument. Over $790M in cash and securities plus $100M of committed Canadian government funding gives liabilities at just 31.5% of assets and, in Andrew Booth's phrasing, “we continue to believe that we have sufficient liquidity to fund at least the next 3 years of pipeline investments” — Andrew Booth, Chief Financial Officer · 2026-11-09 — topped up by the oversubscribed $200M raise that followed the Phase II data. Valuation now has nothing to do with earnings: 36.4x trailing revenue is a bet on 635 reaching a market, not on any current business.
The tape's verdict
The stock is up 62% in ninety days but sits 15.9% below its September 29 peak of $14.82 — the market celebrated the data in August, then marked time waiting for the design and the 12-week cut. Against a since-IPO return of -78.8%, this is a company trying to re-rate on a single asset in a category where nonhormonal treatment for hot flashes suddenly has a genuine efficacy leader. The story that matters: AbCellera was a platform company that found one drug better than it promised, and has now bet the narrative — and the OX40 program — on that being repeatable. Whether it is a “category-defining women's health product” or one very good asset and five programs still in the dark is the entire question, and it will not be answered until 2027.