Back to News
Market Impact: 0.32

Billionaire Stanley Druckenmiller Sold Nvidia and Palantir and Piled Into One of Wall Street's Hottest Drug Stocks Ahead of 2026

NVDAPLTRTEVANFLXNDAQ
Artificial IntelligenceTechnology & InnovationInvestor Sentiment & PositioningCompany FundamentalsHealthcare & BiotechLegal & LitigationManagement & GovernanceMarket Technicals & Flows
Billionaire Stanley Druckenmiller Sold Nvidia and Palantir and Piled Into One of Wall Street's Hottest Drug Stocks Ahead of 2026

Duquesne Family Office's Stanley Druckenmiller materially rotated away from AI names, selling the fund's remaining 214,060 Nvidia shares in the September 2024 quarter and exiting 769,965 Palantir shares between July 1, 2024 and March 31, 2025 amid valuation and bubble-risk concerns (Palantir P/S ~127 as of Dec. 19; Nvidia briefly topped P/S 30). At the same time he aggressively accumulated Teva Pharmaceutical — purchasing 1,427,950 (Q3 2024), 7,569,450 (Q4 2024), 5,882,350 (Q1 2025), 1,089,189 (Q2 2025) and 625,000 (Q3 2025) shares for a total of 16,593,935 shares between July 1, 2024 and Sept. 30, 2025 — making Teva the third-largest holding in his ~$4.1bn fund as Teva rallies ~191% since the start of 2024, trims net debt to $14.6bn (Q3 2025) and trades at a historically attractive forward P/E (as low as ~5, ~11x at writing).

Analysis

Market structure: Druckenmiller's exit from NVDA and PLTR and simultaneous accumulation of TEVA signals a rotation from hyper-growth AI names into beaten-up healthcare value. Direct winners: Teva, large-cap drug developers with brand assets (pricing power); losers (near-term): highly valued AI/compute plays with elevated P/S (PLTR P/S 127, NVDA briefly >30). Supply/demand: GPU demand remains structurally tight supporting NVIDIA pricing, but investor willingness to de-risk increases selling pressure and raises near-term liquidity-driven volatility. Risk assessment: Tail risks include an AI re-rating/bubble (20–60% implied drawdowns in high-multiple names), regulatory actions targeting government data contractors (PLTR) or pricing scrutiny on drug launches (TEVA), and drug-trial or patent setbacks. Time horizons: immediate (days) = elevated IV and tradeable spikes; short-term (weeks–months) = re-rating and earnings guidance; long-term (quarters–years) = NVDA structural moat vs TEVA execution on Austedo and balance-sheet repair. Hidden dependencies: TEVA’s upside relies on Austedo sustaining >$2B sales and stable FX; PLTR/NVDA downside amplifies if corporates pause AI capex. Trade implications: Rotate exposure toward selective healthcare value while hedging tech concentration. Favor limited, sized entries (2–3% positions) into TEVA; trim or hedge NVDA/PLTR exposure via puts or call overwriting rather than outright shorts on NVDA given supply-led demand. Watch catalysts: NVDA guidance (next 30–90 days), PLTR government contract disclosures, and TEVA pipeline/earnings over next two quarters. Contrarian angles: Consensus may underprice NVDA’s durable pricing power from constrained supply — a full exit is premature for long-term allocators. Conversely, PLTR’s valuation appears disconnected from revenue reality and is vulnerable to multiple compression if ROI on enterprise AI stalls. TEVA’s +191% run creates short-term mean-reversion risk; a disciplined entry with debt/earnings thresholds (net debt <~$14B, EPS growth >10% next 12 months) is prudent.