Cathie Wood‘s ARK Invest made its most consequential single-day portfolio realignment in months on Friday April 24, simultaneously placing a $92.5 million bet on nuclear energy startup X-Energy (NASDAQ: XE) on the day of its market debut, adding $71.5 million in Amazon (NASDAQ: AMZN) across multiple funds ahead of the company’s April 29 earnings release, while divesting $65.8 million worth of Advanced Micro Devices (NASDAQ: AMD) and trimming its Rocket Lab (NASDAQ: RKLB) position by $9.6 million.
X-Energy’s IPO priced above its preliminary range, with the small modular reactor company backed by Amazon entering the market at $23 per share on strong investor demand, and Wood’s entry as a key early investor purchasing more than 4 million shares positions ARK as one of the most prominent public institutional backers of the company’s entry into the nuclear energy space, reinforcing a broader ARK thesis that emerging clean energy technologies represent one of the defining multi-year investment opportunities.
The Amazon position, built on top of a smaller 3,492-share purchase earlier in the week, continues a pre-earnings accumulation strategy that Wood has deployed before several of Amazon’s strongest quarterly results, with AWS’s trajectory as the company’s dominant growth engine the primary driver of ARK’s conviction, and the April 29 earnings report expected to show continued double-digit cloud growth.
The AMD sale tells a different story: ARK sold into a 14 percent single-session gain for the semiconductor company, which itself had risen approximately 70 percent over the prior month on the back of Intel’s strong earnings triggering optimism about data centre CPU demand across the sector, and AMD’s positive ripple effect generated exactly the kind of price appreciation at which Wood’s approach involves taking profits from positions that have moved significantly toward or beyond her team’s assessed fair value.
AMD closed at approximately $347.81 on the day of the sale, against GuruFocus’s GF Value estimate of $212.86, making the stock 63.4 percent overvalued by that metric and giving Wood a concrete framework for the decision to reduce exposure at the current price even though AMD remains a top-three holding in the ARKK fund at a market cap within the portfolio of approximately $416 million.
The Rocket Lab trimming runs against the broader direction of market enthusiasm for the stock, which has been one of the best performers in the Nasdaq over the past year, but ARK had been steadily reducing RKLB across multiple funds throughout 2026, with the firm having sold a total of 235,705 shares since January through its ARKQ and ARKX ETFs, suggesting Wood’s team believes the stock’s extraordinary run has moved its price ahead of the incremental fundamental developments that remain ahead.
Rocket Lab closed down approximately 5 to 6 percent on Friday, with the sell side maintaining a Buy consensus and a median price target well above current trading levels, but KeyBanc analyst Michael Leshock’s assessment that “the equity’s risk/reward profile appears balanced in the near to medium-term” with major catalysts already widely recognised provides context for why even a bullish institutional owner might choose to reduce at current levels.
The broader ARK performance context matters for understanding these moves: ARK Innovation ETF (ARKK) is down 1.76 percent year-to-date in 2026 while the S&P 500 has gained 4.67 percent, and ARK’s five-year annualised return has compressed to negative 9.01 percent against the S&P 500’s 13.01 percent annualised return over the same period, creating a fundraising and redemption pressure environment that arguably constrains the firm’s ability to hold positions through extended drawdowns and incentivises taking gains when they are available.
Wood told a Bloomberg podcast in March that she sees the global economy entering what she calls a “great acceleration” rather than a recession, arguing that AI, robotics, energy storage, blockchain, and genomics are all entering commercial prime time simultaneously in a way that has historical parallels with the periods of explosive productivity growth driven by the Industrial Revolution, a framing that underpins her continued willingness to rotate aggressively between positions rather than adopting the kind of passive index-tracking approach her long-term performance figures suggest would have served her investors better.
The X-Energy bet in particular reflects Wood’s conviction that the AI data centre power demand crisis creates a once-in-a-generation commercial opportunity for small modular reactor companies that can provide reliable, carbon-free baseload power at the scale hyperscalers require, a thesis that aligns with the broader nuclear energy re-evaluation happening across institutional investment circles as the gap between renewable intermittency and data centre power needs becomes impossible to ignore.
