Cathie Wood Goes On a Selling Spree: 3 Stocks She Just Sold

Cathie Wood's Ark sold stakes in more than three dozen names, including Netflix, Broadcom and AMD, while adding only Tempus AI. Netflix is up ~10% since terminating the Warner Bros. Discovery deal and raised US prices (standard ad-free from $17.99 to $19.99; ad tier $7.99 to $8.99); Q4 revenue +18% and EPS +30% but guidance implies revenue growth slowing to 12–14% and operating margin ~31.5%. Broadcom is up ~73% Y/Y with revenue growth decelerating to 24% in FY2025 and trading near 17x next fiscal profit, while AMD reported revenue +34% (data center +39%) and trades ~31x this year vs <19x next year. Overall the piece signals continued AI-driven demand for chip names but mixed fundamentals/guidance and active ETF rebalancing that could pressure individual stocks.

Analysis

ARK’s concentrated, high-turnover selling is a liquidity event more than a new information set — it creates transient price dislocations in names with high retail/ETF ownership and thin intraday liquidity. That amplifies downside moves but also seeds short-lived mean-reversion opportunities as programmatic desks and index rebalancers mechanically buy the dip once realized flows subside.

For streaming/media, the durable source of upside is not a single product event but the spread between ARPU growth and content-amortization cadence; the crucial monitors are short-term churn inflection (days–weeks signal) and the 6–12 month backlog of content cash vs GAAP amortization (earnings/cash conversion signal). A small move in churn (low single-digit percentage points) can pivot consensus margin forecasts by hundreds of basis points, changing the investment case quickly.

In semiconductors, structural AI demand masks two divergent business models: vertically integrated, software‑anchored incumbents with recurring revenue capture more of system-level margin versus pure-play silicon vendors exposed to inventory cycles and competitive ASP pressure. That asymmetry creates a convexity trade — you get defensive cashflow exposure in AI through companies that monetize both hardware and firmware/software, while GPU/CPU specialists face sharper peaks and troughs driven by OEM build cycles.

Timing matters: expect microstructure-led opportunities in the next 2–6 weeks (flow-driven) and fundamental re-pricings around quarterly guidance and large OEM design-win announcements over 1–6 months. Position sizing should reflect whether you are trading a flow squeeze (small, tactical) or a fundamental view (larger, multi-month) with explicit stop rules tied to churn/GTM metrics or inventory disclosures.

Key Decisions for Investors

• Pair trade (3–6 month): Long Broadcom (AVGO) 2% NAV / Short AMD (AMD) 1% NAV. Rationale: overweight recurring software/firmware margin capture vs cyclic GPU/CPU exposure. Target +18% gross on AVGO and -12% on the pair if AVGO outperforms; hard stop if AVGO falls 12% or AMD rises 20% from entry.
• Netflix hedge (0–3 month): Buy a 3-month put spread sized 1–2% NAV (buy 7–15% OTM put, sell lower strike to finance). Rationale: asymmetric protection vs a guidance/churn miss that would compress multiples; max loss = premium (~limited), payoff if shares drop >7–15% in next quarter.
• AI convexity trade (1–3 month event play): Buy a small NVDA 1–3 month long strangle (10% OTM call + 10% OTM put) sized 0.5% NAV. Rationale: capture elevated event/earnings-driven volatility without directional bias; close into realized vol or after earnings. Risk = premium paid, reward = unlimited upside on call or deep put payout if drawdown.
• Flow-fade tactic (days–weeks): On low-volume sell days where ARK/ETF outflows dominate, opportunistically buy into depth-probed selling in high-ARK-ownership names with a 2–4 week horizon and tighten stop to 6–8% to capture mean-reversion. Focus execution size to avoid catching a structural reversal.