AbCellera and Recursion Pharmaceuticals: two AI-drug stocks, two different stories

The article favors a long Recursion (RXRX)/short AbCellera (ABCL) pairs trade after ABCL rose 169.7% over one year while RXRX fell 28.2%. RXRX trades at $3.66 versus a $3.94 fair-value estimate and has 97.7% consensus target upside, while ABCL at $12.53 is estimated to be 44.4% above fair value and missed Q2 revenue expectations at $4.1M versus $8.1M consensus. Key risks are RXRX's accelerating EBITDA loss of $564M in 2025 and ABCL's potentially stock-moving Phase 2 ABCL635 data expected in Q3 2026.

Analysis

The proposed pair is directionally sensible but is not a clean factor hedge: ABCL is principally a pipeline-readout and platform-licensing exposure, while RXRX is valued on partnership monetization, compute-enabled discovery credibility, and financing runway. In a broad AI-biotech risk-on tape, ABCL can continue to outperform despite weak near-term operating leverage; conversely, a sector-wide funding scare would likely hurt RXRX disproportionately because its valuation depends more heavily on future capital availability. The market should discount generic “fair value” outputs in both names: neither has sufficiently stable cash flows for a terminal-value model to be a decision tool.

Near term, ABCL’s crowded momentum setup creates asymmetric downside on any delay, ambiguous efficacy signal, or lack of follow-on partner economics, but a naked short into a clinical catalyst has unfavorable gap risk. The more investable 1-3 month catalyst for RXRX is evidence that collaboration revenue converts into durable contracted economics rather than low-margin, milestone-dependent revenue; absent that proof, the revenue trajectory does not offset operating-loss expansion. Over 6-18 months, the differentiator will be whether either platform reduces development cost and time enough to create repeatable economics—not whether it can generate isolated AI-discovery announcements.

Contrarian point: analyst target dispersion is usually least informative in pre-profit biotech, where targets embed unobservable probability-of-success assumptions and update slowly after a repricing. ABCL may be overextended tactically, but the catalyst can invalidate a valuation short overnight; RXRX may appear inexpensive on book value yet still be expensive if its burn rate requires equity issuance before self-funded milestones. The thesis is falsified if ABCL delivers clearly differentiated clinical data with a credible commercialization path, or if RXRX’s next results show material acceleration in contracted backlog/revenue alongside narrowing cash burn.

Key Decisions for Investors

• Do not initiate an unhedged ABCL short ahead of the pending clinical readout. If the stock remains extended, use a defined-risk ABCL put spread dated at least 30-60 days beyond the disclosed data window; only enter after implied volatility versus realized volatility is reviewed, since unavailable option pricing prevents a recommendation on strike selection.
• For a market-neutral expression over the next 1-3 months, initiate a small long RXRX / short ABCL dollar-neutral pair only after confirming ABCL catalyst timing and borrow availability. Size ABCL at no more than 50-75% of the RXRX long notional until the binary event passes; review or close if ABCL reports compelling efficacy/safety data or if the spread widens another 20% without a fundamental change.
• Treat RXRX as a watch-list long rather than a standalone core position until the next earnings release discloses cash burn, cash runway, collaboration backlog, and revenue concentration. Add only if management demonstrates sequential burn improvement or funded milestones sufficient to reduce dilution risk; exit on a meaningful acceleration in operating cash use without offsetting contracted revenue.
• After ABCL’s clinical event, reassess the short leg: negative or equivocal data would create the cleaner opportunity to short rallies or add to the pair, while positive differentiated data should prompt immediate cover rather than reliance on technical mean reversion.