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

Cathie Wood trimmed stakes in more than three dozen Ark ETFs on Thursday, notably selling Netflix, Broadcom and AMD while only adding to Tempus AI. Netflix is up ~10% since walking away from the Warner Bros. Discovery deal and raised U.S. prices (standard ad‑free $17.99 → $19.99; ad tier +$1), after reporting Q4 revenue +18% and EPS +30% but guiding revenue growth to 12–14% and an operating margin of 31.5%. Broadcom is +73% over the past year with FY25 revenue growth slowing to 24% (from 44%) and trading near 17x next‑fiscal profit, while AMD posted revenue +34% for FY2025 with data center revenue +39% and trades ~31x trailing vs <19x next year.

Analysis

Netflix’s recent pricing action and margin leverage create a narrow path to upside: sustained ARPU growth funds content and dilutes fixed marketing spend, but the bulk of downside comes from higher-than-expected churn concentrated in the first 60–90 days after hikes. Monitor cohort-level retention and paid net adds two quarters out as the clean signal — if churn reverts to pre-hike rates, operating margins can sustainably expand 300–500bps over 12–18 months without incremental content investment.

For Broadcom and AMD the key second-order dynamics are different: hyperscaler AI spend dynamics are lumpy and front-loaded into GPU/accelerator refresh cycles, creating boom/bust inventory effects downstream (used GPU markets, channel destocking) that can swing near-term revenue by ±20% across a quarter. Broadcom’s software/recurring revenue mix can blunt this cyclicality, whereas AMD’s exposure to discrete GPU/CPU ASPs and foundry capacity tightness compounds upside but also amplifies downside if hyperscalers pause orders.

Consensus risk: markets have priced a smooth secular AI ramp; they underweight inventory and ad-revenue elasticity risks. Short-term catalysts that could flip sentiment are: quarterly guidance beats/misses from hyperscaler customers, public hyperscaler commentary on AI capex cadence, and 2–3 quarter changes in Netflix cohort churn. These are actionable within a 1–12 month window and justify asymmetric option structures rather than naked directional positions.

Key Decisions for Investors

• NFLX — Tactical long call-spread (6–9 month): buy 25% OTM calls and sell 55% OTM calls, size 1.5% NAV. Rationale: asymmetric payoff to capture margin re-rating if ARPU sticks; target 2.5x premium (~+150%), stop at 40% premium loss or if two consecutive months of paid net adds miss consensus by >20%.
• AMD vs INTC pair (12 month): long AMD equity (or 9–12 month LEAP calls) and short INTC equity equal notional, size net 2–3% NAV. Rationale: play secular share gains in data-center compute while hedging macro-capex risk; take profits when relative outperformance hits +25%, stop if AMD underperforms by 15% on fundamental misses.
• AVGO — Put-spread hedge (3–6 month): buy 10–15% OTM puts and sell deeper OTM puts to fund, size 1% NAV. Rationale: protects against an AI capex pause triggering rapid multiple compression in a richly priced name; target asymmetric downside capture if stock drops 20–30%, cost limited to paid premium.
• Portfolio hedge — long NVDA 6–9 month verticals (modest size 1% NAV) to protect AMD exposure. Rationale: NVDA moves are still the primary demand signal for the AI cycle; a continued outperformance would correlate with AMD upside, but rapid derates often start with NVDA guidance downshifts.