Back to News
Market Impact: 0.25

Cathie Wood’s ARK sells AMD stock, buys CoreWeave and Kratos

AMDCRWVKTOSKDKNETNTLARKLBTERTWST
Market Technicals & FlowsInvestor Sentiment & PositioningTechnology & InnovationInfrastructure & DefenseHealthcare & BiotechEnergy Markets & Prices
Cathie Wood’s ARK sells AMD stock, buys CoreWeave and Kratos

ARK ETF disclosed a mixed daily trade list for Friday, May 8, 2026, led by the sale of 46,034 AMD shares for about $18.8 million and the purchase of 113,076 CoreWeave shares worth roughly $14.6 million. Other notable buys included Kratos Defense ($9.6 million), Kodiak AI ($3.6 million), Cloudflare ($3.5 million), Intellia Therapeutics, and X-Energy, while sales included Rocket Lab ($14.7 million), Teradyne ($3.1 million), and Twist Bioscience ($675,141). The filing mainly reflects portfolio rebalancing and sentiment shifts rather than a company-specific catalyst.

Analysis

The portfolio shift reads less like generic risk-taking and more like a rotation toward “enablers” of compute, autonomy, and national-security capex. The beneficiaries are the names that monetize picks-and-shovels demand: cloud/network infrastructure, defense electronics, and AI-dependent infrastructure builders. The losers are the legacy semis and space hardware names that are being treated as funding sources rather than core conviction holdings, which can matter because repeated fund-flow selling often compresses multiples before fundamentals visibly deteriorate. The second-order effect is that this mix is a barometer for where growth capital still has a bid: mission-critical software, secure connectivity, defense autonomy, and nuclear/energy optionality. That matters for supply chains because these subsectors tend to have longer procurement cycles and stickier backlog than consumer tech; once allocations shift there, the marginal dollar is less likely to reverse quickly on a single quarter of misses. The weakest signals are in the capital-intensive hardware names where valuation support depends on a clean execution path and continued narrative momentum. Near term, the main risk is that these are crowded “good story” trades without immediate fundamental re-acceleration, so any tape weakness can hit them harder than the averages. Over the next 1-3 months, the market will likely distinguish between names with visible contract conversion and those relying on sentiment alone. The best contrarian setup is that the sell-side may still be anchored to AI/defense enthusiasm, underestimating how quickly capital can rotate away from high-beta hardware if funding conditions tighten or guidance disappoints. For the oil/shipping backdrop, the embedded geopolitical premium is a potential cross-asset tailwind for defense and cybersecurity while remaining a tax on the most cyclically exposed industrial and transport names. If the conflict premium persists for weeks rather than days, defense/autonomy baskets should outperform on relative revisions, not just headlines. If the Strait risk de-escalates quickly, these trades can give back a meaningful chunk of gains because the premium is being paid on narrative, not current earnings.