Palantir or AMD: Cathie Wood Is Leaning Into One AI Bet

Cathie Wood added 85,485 shares of Palantir worth about $11.15 million while selling 44,446 AMD shares valued at roughly $10.52 million, signaling a rotation from AI chips toward software. Palantir reported revenue up 70% year over year to $1.407 billion and raised 2026 revenue guidance to $7.182-$7.198 billion, while AMD posted 34.1% revenue growth to $10.27 billion but saw ARK trim exposure ahead of earnings. The article is supportive for PLTR sentiment and more neutral-to-cautious on AMD, with analyst upside favoring Palantir at about 52% versus 16% for AMD.

Analysis

The signal here is not just a style rotation from semis to software; it is a bet that AI monetization is shifting from infrastructure scarcity to application-level pricing power. That usually favors names with direct workflow embed and recurring usage, while pressuring chip leaders that are still priced off a longer-duration capex cycle. If that rotation broadens, expect the market to reward companies that can show net retention and margin leverage rather than raw GPU demand growth.

PLTR is the cleaner expression of that trade because its upside is increasingly driven by proof of durable government and enterprise adoption, not just narrative multiple expansion. The risk is that the stock’s implied growth path leaves very little room for even a modest deceleration; a single guide-down or billings miss could compress the multiple sharply over 1-2 quarters. Still, Wood buying the dip after a sentiment shock can attract momentum capital if the next print confirms that demand is not being pulled forward.

AMD looks more like a “good business, slower re-rate” situation: strong fundamentals, but increasingly viewed as a second-order beneficiary behind the hyperscaler complex. Near term, the stock is vulnerable to expectations being too high into earnings, especially if investors start to focus on mix, gross margin, or the slope of AI accelerator contribution rather than total revenue. A soft guide would likely hit harder than the headline print because the market is paying for accelerating AI share gains, not just solid execution.

The contrarian takeaway is that the crowded long may actually be the chip exposure, not the software one. If AI capex growth normalizes over the next 6-12 months, the winners should become the platforms that sit on top of that installed base and monetize usage, which supports PLTR relative to AMD. However, if enterprise AI adoption remains slower than expected, PLTR’s valuation is the first to air-pocket.

Key Decisions for Investors

• Long PLTR on post-earnings weakness over the next 2-6 weeks; use a stop if the company fails to raise forward demand indicators, because the setup is asymmetric only while momentum remains intact.
• Short AMD into earnings as a tactical hedge over the next 1-3 weeks if positioning is extended; best case is a modest pullback from expectations reset, worst case is limited upside because much of the AI narrative is already in the price.
• Pair trade: long PLTR / short AMD for 1-3 months to express software-vs-semiconductor AI rotation; this is a relative-value trade with better payoff if AI spending becomes less capex-intensive.
• For options, consider PLTR call spreads instead of outright calls to cap valuation risk; the trade works if the stock reclaims trend but reduces damage if the market re-prices the multiple.
• Do not chase AMD strength ahead of the print; wait for either a guide-up or a post-earnings reset before adding, since the near-term risk/reward is skewed to disappointment rather than surprise.