Cathie Wood’s ARK buys OpenAI Group stock, sells Veracyte shares

ARK's largest trade was the sale of 115,718 shares of Veracyte (VCYT) across ARKK/ARKG for $3,522,455; other notable sales included 612,483 shares of Strata Critical Medical (SRTA) for $2,480,556 and 6,204 shares of Teradyne (TER) for $1,714,475. On the buy side ARK purchased 348,995 shares of OpenAI Group PBC Series C across ARKK/ARKW/ARKF, added 26,515 shares of CoreWeave (CRWV) for $1,833,512 and 14,729 shares of DoorDash (DASH) for $2,180,039; smaller buys included 90,514 Kodiak AI (KDK) for $585,625 and 4,559 GeneDx (WGS) for $271,761. The trades signal a continued tilt into AI/tech and private-venture exposure; these moves may shift small-cap/security-level prices (~1–3%) but are unlikely to drive broader market moves.

Analysis

ETF re-allocation into AI/compute themes creates a two-speed market: small-cap diagnostics and niche biotechs are liquidity-susceptible and will likely trade on fund flows more than fundamentals in the next 1–3 months, while AI infrastructure and private-to-public bridge names can see multi-week squeezes as capacity constraints surface. Expect order book dislocations: widened spreads and lower depth in < $1bn market-cap diagnostics names, and transient price spikes in GPU-access plays when large passive/active reallocations require quick exposure.

Second-order supply effects matter: sustained incremental demand for hosted GPU capacity or custom silicon raises lead times for data-center buildouts and pushes cloud providers to favor elastic third-party partners, improving revenue visibility for dedicated infra providers but pressuring margins for generalist cloud operators over 6–18 months. For diagnostics, reimbursement cadence and a single favorable clinical readout can trigger outsized mean reversion rallies because float is thin and investor base is narrow.

Key tail risks and catalysts to watch are funding/flow reversals from major active managers, clinical/regulatory updates in diagnostics, and a softening in AI capex driven by macro (rates or enterprise demand). A flip in ETF flows could reverse moves inside days; fundamental catalysts (earnings, FDA/reimbursement) operate on 3–12 month horizons and will determine whether repricing is permanent or a transient liquidity event.

The pragmatic trade framework is to own optionality into AI infra capacity tightness while hedging or shorting small-cap diagnostics exposure that lacks clear short-term revenue catalysts. Position sizing should assume 30–40% intraday volatility for small caps and a 10–20% event risk gap for mid-cap AI names on M&A signals or supply surprises.

Key Decisions for Investors

• Long CRWV (CoreWeave) via 3–6 month call spread (buy ATM, sell 1.5x OTM) — objective: capture 40–80% upside if GPU-hosting demand accelerates; max loss = premium paid (~100% of premium). Stop-loss: close if spread cheapens by 50% in first 30 days.
• Pair trade: long NVDA or a broad AI infra ETF (3–12 months) / short VCYT (VCYT) — target relative outperformance of 20–30% in 3 months. Size short at 25–50% of long notional to limit tail risk from binary clinical outcomes; tighten if VCYT shows sustained buy-side depth >$500k/day.
• Directional short VCYT via 3–6 month puts (buy puts at ~10–15% OTM) — thesis: liquidity-driven downside; target 30–50% price fall if flows persist. Risk: binary clinical/reimbursement catalyst can gap price +40–80%; cap position size accordingly (max 1–2% portfolio).
• Volatility sell on small-cap biotech through covered-call overlays or short-dated iron condors (30–60 days) — collect elevated IV premia from flow-driven events; keep max loss defined and avoid around known clinical/regulatory dates.