Palantir Billionaire Peter Thiel Dumped Nvidia and Bought This Other Magnificent Stock Instead -- Even as Warren Buffett Was Selling It Before Retiring

Thiel Macro's latest 13F shows a full exit of Nvidia—selling 537,742 shares in Q3—and the initiation of a position in Apple, signaling a rotation from high-beta AI exposure into a more predictable, cash-generative large-cap. Nvidia has grown from a $345 billion market cap at ChatGPT's commercial launch to about $4.6 trillion as of Jan. 7, yet shares have only risen ~1.7% since its fiscal Q3 earnings (Nov. 19, 2025) amid concerns about competition from AMD and custom ASIC makers like Broadcom. The piece highlights valuation differentials (forward P/E ~24 for Nvidia vs ~32 for Apple) and frames Apple’s >2 billion active device base and steady cash flow as a defensive, lower-volatility alternative should momentum stocks correct.

Analysis

Market structure: Thiel's rotation from NVDA to AAPL signals a tactical shift from high-beta AI-capex exposure to cash-flow defensives. Direct beneficiaries in a rotation are AAPL (services monetization, recurring revenue), Broadcom/AMD (as alternative ASIC/GPU suppliers), and legacy semicap names if capex staggers; losers would be momentum-sensitive NVDA and levered AI play stocks if sentiment reverses. Expect short-term demand softness for premium GPUs (pressure on ASPs) while long-term data-center demand remains structurally strong.

Risk assessment: Tail risks include tighter US export controls to China, a large hyperscaler capex pause, or a NVDA earnings guide-down—each could move NVDA >30% in weeks; conversely, a stronger-than-expected Apple AI product rollout could re-rate AAPL by +20% over 12–18 months. Immediate (days) risks = volatility around earnings/announcements; short-term (weeks–months) = rotation and options-volatility repricing; long-term (years) = structural share gains for incumbents or ASICs. Hidden dependency: NVDA’s valuation is more sentiment-dependent than revenues; Apple’s upside depends on measured uplift in services ARPU, not just device sales.

Trade implications: Size trades defensively and time around catalysts. Tactical ideas: modest core long AAPL (2–4% portfolio), hedge existing NVDA exposure with 3–6 month put spreads (10/20% OTM), and run a dollar-neutral pair trade long AAPL vs short NVDA for 3–6 months to capture rotation. Add selective longs in AMD/AVGO (1–2% each) as asymmetric plays on competitive GPU/ASIC wins and semiconductor pricing power.

Contrarian angles: The market underprices Apple’s optionality to monetize AI via a 2B+ device base—if Apple lifts services ARPU by $2–5 annually, EPS could expand materially over 24 months. NVDA pullbacks of 20–30% historically present high-quality re-entry points given secular GPU demand; avoid one-way thinking that NVDA is “too expensive” without a 12–24 month operational bear case. Unintended consequence: crowded AAPL longs could steepen skew and raise hedge costs, compressing realized returns for momentum allocators.

Key Decisions for Investors

• Establish a core long in AAPL representing 2–4% of portfolio within 2 weeks; buy 12–18 month LEAPS ~5–10% OTM or purchase stock with a stop-loss at -18% and a 12–18 month target of +20% (driven by services AI monetization).
• If you hold NVDA, trim position size by ~30% immediately and buy 3–6 month put spreads sized to hedge 50% of remaining notional (e.g., buy 10% OTM put / sell 20% OTM put) to protect against a >15–25% downside in the next 3 months; redeploy to NVDA only on a drop >20% from current levels.
• Implement a dollar-neutral pair trade: long AAPL (equal $) and short NVDA via a 3–6 month bear put spread (short leg 10–15% OTM) to capture rotation risk over the next 3–6 months; size to 1–2% net market exposure.
• Add small asymmetric longs in AMD and AVGO (1–2% each) as share-shift/ASIC exposure; exit or re-weight if AMD/AVGO report sequential revenue misses >5% vs guide or if NVDA files patent/legal setbacks within 60 days.
• Monitor US export-control announcements and hyperscaler capex commentary over the next 60 days; if new controls explicitly restrict advanced-node shipments to China or hyperscalers cut GPU orders by >10% YoY guidance, reduce NVDA exposure by an additional 50% and increase cash allocation.