Palantir Billionaire Peter Thiel Shifts His AI Bet. He Recently Dumped Nvidia Stock and Bought Shares of These 2 Tech Industry Stalwarts.

Thiel Macro, which reports more than $74 million in 13F securities, sold its entire Nvidia position of 537,742 shares (40% of the portfolio) in Q3 2025 after initially buying in Q4 2024, and redeployed capital into 49,000 shares of Microsoft (now 34% of the portfolio) and 79,181 shares of Apple (27% of the portfolio). The move reflects a rotation out of a high-valuation, AI-centric play—Nvidia posted >$130 billion in revenue last fiscal year and is seen as a primary AI chip provider—into more diversified tech incumbents with broader revenue bases, signaling a cautious repositioning in response to AI-driven valuations and risk.

Analysis

Market structure: Thiel’s rotation from NVDA (40% of his 13F) into MSFT (34%) and AAPL (27%) signals a flow preference for diversified, cash-generative mega-caps over pure-play AI semiconductors. Direct beneficiaries are Azure/Cloud exposure (MSFT), device and services cashflow (AAPL) and index/ETF products tracking mega-caps; losers include high-beta semis and smaller AI hardware suppliers whose valuations assume uninterrupted GPU demand. Supply/demand: investor demand may shift away from forward-looking GPU beta to cash yield, which could ease some near-term pressure on GPU lead times and compress NVDA’s premium pricing power for used/spot markets. Cross-asset: expect downward pressure on NVDA implied vol and option skew, modest tightening of tech credit spreads, and transient USD strength if safe-haven flows favor large-cap liquidity.

Risk assessment: Key tail risks are AI regulatory actions (US/EU export controls within 3–12 months), a sudden datacenter capex pullback (>20% QoQ), or geopolitically-driven supply shocks (Taiwan-strait). Immediate (days) reaction is flow-driven repricing; short-term (weeks–months) depends on earnings and inventory prints; long-term (quarters–years) hinges on durable moat: NVDA’s moat vs Microsoft’s platform and Apple’s hardware+services cycle. Hidden dependencies include hyperscaler procurement timing, cloud pass-through pricing, and Apple’s A-series roadmap/upgrade cadence. Catalysts to watch: next 2 quarters’ revenue mix, NVDA channel inventory reports, Azure AI contract disclosures, and export-control announcements.

Trade implications: Favor size into MSFT and AAPL as de-risked AI exposure and buy optionality via long-dated calls/LEAPS only if valuation dips ~10% post-earnings; protect NVDA exposure with short-dated put spreads to cap downside. Implement pair trades: long MSFT vs short NVDA to express rotation from concentrated AI beta to diversified cloud exposure for 6–12 months. Tactical options: buy 3-month NVDA 10% OTM puts and sell 25% OTM puts (cost-limited hedge) sized to cover 0.5–1% portfolio risk while deploying 2–3% into MSFT/AAPL.

Contrarian angles: The consensus ignores tax/liquidity or portfolio-construction drivers behind Thiel’s trades — his sale of NVDA may be portfolio rebalancing, not a negative fundamental call. NVDA’s AI revenue still has convex upside; a full exit by smart money can create contrarian buying opportunities if NVDA corrects >20% without an earnings miss. Historical parallels: semis overconcentration (2017–18 crypto/supply swings) show fast mean reversion; unintended consequence: momentum-driven outflows could temporarily overshoot intrinsic value, creating 3–9 month alpha windows for disciplined buyers.

Key Decisions for Investors

• Establish a 2–3% long position in MSFT (ticker: MSFT) with a 12–18 month horizon; target 15–25% upside and place a stop-loss at -10% to protect against cloud/AI execution risk.