저자: 미상 · 외국인
요약 보기
캐시 우드(ARK)가 최근 주가 하락 시 AI 관련 3종목(TSMC, Pony.ai, X-Energy)을 매수했다고 소개. TSMC는 첨단 반도체 파운드리 점유율과 높은 성장률, 애리조나 투자 확대를 근거로 장기 보유 적합 종목으로 평가. Pony.ai는 로보택시 매출 급성장에도 적자가 지속되어 투기적 소액 베팅으로 제시. X-Energy는 AI 데이터센터용 SMR 원자로 개발사로 2030년 이후 상용화 전망, 장기 고위험 베팅으로 소개. 전체적으로 투자 권유가 아닌 개인 판단 필요 메시지로 마무리.
다듬은 글
Cathie Wood just dropped more than $20 million into three falling stocks this past week alone. While most investors are running from volatility, she's buying the dip hard. And every single one of these names is a pure play on the future of artificial intelligence. The AI boom is no longer a story about software alone.
It's about power, chips, autonomy, and the physical infrastructure that has to scale with demand. Data centers are hungry. Robots and robo taxis are moving from prototypes into real fleets. And the companies that sit at the bottleneck of that growth are the ones that can either print money or burn it for years before the payoff arrives.
Cathie Wood has always been willing to buy when others are fearful. She left Tesla and SpaceX off this particular shopping list because everyone already knows she owns those. She also skipped Meta after her recent activity there. Instead, she focused on three names that give a cleaner window into how ARK is positioning for the next phase of AI.
Let's walk through them one by one with the latest numbers and the real risks. First is the giant, Taiwan Semiconductor, ticker TSM. Over the past week, Cathie poured more than $15 million into TSM. The stock isn't crashing the way the smaller names are, but it did pull back more than 15% over the past month, and at one point was down around 22% from recent highs.
For a company of this quality, that kind of dip is rare and Cathie treated it like an opportunity. TSMC is the world's largest contract chip manufacturer. It produces roughly 70% of the global advanced semiconductor supply. Every time Nvidia, Apple, AMD, or any other AI leader designs a new chip, those designs land on TSMC's factory floor.
The company collects high margin manufacturing fees while the design houses fight for capacity. Demand remains relentless. Spending on AI infrastructure is projected to reach several trillion dollars by 2030. And TSMC sits at the absolute center of that spend. Last year, sales and profits jumped more than 30 and 40% respectively.
Management recently raised full-year revenue guidance to over 40% growth driven by stronger than expected demand for its newer, more energy-efficient 2 nanometer chips. The company is also doing something strategically important. It is pouring another $100 billion into expanding manufacturing in Arizona, bringing its total commitment there to more than $265 billion.
That move helps reduce the long-standing geographic risk of having almost all advanced capacity concentrated in Taiwan. TSMC chips already power a huge percentage of US precision-guided weapons, and the company is even launching a dedicated space chip division for satellites. Valuation still looks reasonable.
The PEG ratio sits under 0.7, roughly half the sector median. Yes, competitors like Samsung and Intel are trying to steal share, and Apple has explored alternatives. But the overall market is large enough that TSMC's leadership position looks durable. This is the highest-quality name on the list, and the one that fits cleanly inside a long-term AI portfolio.
From the absolute giant, we move to a much smaller, much riskier name, Pony.ai, ticker Pony. Cathie bought roughly $600,000 worth this past week. Since its IPO about a year ago, the stock has lost more than 70% of its value from the peak. That kind of drawdown is exactly the environment where Arc tends to nibble.
Pony is a Chinese autonomous driving company with Silicon Valley roots. It does not build its own vehicles. Instead, it runs an asset-light model focused on the AI software and platform for robotaxis and robotrucks. The actual cars and trucks are manufactured in partnership with major automakers, including Toyota.
The business itself is growing fast. Last quarter overall sales jumped triple digits. The robotaxi segment alone climbed nearly 400% to $8.6 million. Fair charging fees rose more than 450%. The RoboTruck division grew 31% to over $10 million, and the intelligent solution segment soared 246% to more than $15 million.
Management is targeting a fleet of roughly 3,500 RoboTaxis across 20 cities by year-end. In some major Chinese cities, the company already claims unit economics break-even, meaning the fare charged covers the operating cost of that specific vehicle. The problem is the other side of the income statement.
Pony is still losing more money than it generates in revenue. Analysts expect negative cash flows for years. The primary market is China, which carries clear geopolitical and regulatory risk. Even so, the current share price sits below even the lowest analyst price targets, and the average target implies triple-digit upside over the next 12 months.
This is a pure speculative bet on the RoboTaxi future. Cathie is willing to take a small position. Most conservative investors will want to size it carefully or skip it entirely if they already own broader autonomy exposure through Tesla, Google, or others. The third name is the one that ties AI directly to the power problem.
X-Energy, ticker X. Cathie added more than $5 million this past week. The stock has been a rough ride, losing well over half its value since its April 2026 IPO. Yet, the underlying thesis is straightforward. AI data centers need enormous amounts of reliable, carbon-free electricity, and traditional grids cannot keep up.
Small modular reactors are one of the few technologies that can scale on the timeline required. X-Energy develops advanced SMRs that run on its proprietary TRISO-X fuel ceramic coated particles that can withstand much higher temperatures without melting. That design is considered safer than conventional nuclear fuel.
Amazon owns roughly 20% of the company, giving X Energy a direct relationship with the world's largest cloud and data center operator. The US Department of Energy projects hundreds of gigawatts of new nuclear capacity will be needed by 2050 just to meet data center demand. X Energy was also selected for a $200 million federal program alongside Microsoft and Nvidia aimed at accelerating reactor deployment for AI facilities, including help cutting through regulatory red tape.
The financial reality is still early stage. In its first quarter as a public company, the firm reported revenues and grant income of $43 million versus $21 million a year earlier. But losses remain wide. The first commercial reactors are not expected until at least 2030. That long timeline is the main reason the stock has been crushed.
Liquidity, however, is strong, roughly $2 billion after the IPO, and the company carries no debt. For Cathie, this fits the disruptive energy and AI thesis. For most individual investors, it remains a high-risk, long-duration bet that should only be a small sleeve of a portfolio. So, those are the three AI-linked names Cathie Wood has been buying on the dip.
Taiwan Semiconductor, the indispensable chip foundry. Pony.ai, the speculative Chinese robotaxi pure play. And X-Energy, the Amazon-backed advanced nuclear bet that wants to power the next wave of data centers. One is a high-quality compounder trading at a temporary discount. The other two are early-stage, high-volatility stories that could either multiply or continue to struggle for years.
Cathie's style has always been to own the future before the consensus fully prices it in. That approach creates both the biggest winners and the biggest drawdowns in her funds. None of this is financial advice. Do your own work, understand the risks, and size positions according to your own timeline and risk tolerance.
The AI transformation is real, but the path will be messy. The investors who succeed will be the ones who separate temporary fear from permanent impairment of capital. If this breakdown helped you see the opportunity more clearly, hit like, subscribe, and turn on notifications. Drop a comment and tell me which of these three you're watching most closely or which one you're avoiding.
I'll keep tracking Arc's moves and the latest numbers so you don't have to dig through the filings alone. The next decade belongs to the companies solving the hardest problems in AI infrastructure, autonomy, and energy. Make sure you know which ones Cathie is actually buying when the market is scared.
Comments 0