Wood and Druckenmiller Quietly Bet Big on Amazon’s Hidden Compounder
Second-quarter 13F filings just revealed an odd bit of convergence: Cathie Wood’s ARK Investment Management and Stanley Druckenmiller’s Duquesne Family Office — two investors who almost never agree on anything — both piled into Amazon and Alphabet last quarter. Druckenmiller, a macro skeptic who has been warning about AI overreach, raised his Amazon stake by a startling 1,083% to 541,600 shares worth $129 million, and opened a brand-new $120 million position in Alphabet. Wood, the growth evangelist, added 18% to ARK’s existing Amazon holding (now $379 million, 2.46% of the portfolio) and boosted Alphabet by 45% to $369 million. When investors this different reach the same conclusion, it’s worth asking what they’re actually seeing in the numbers.
The answer looks less like AI hype and more like a genuine shift in Amazon’s underlying economics. AWS revenue growth has now accelerated for five straight quarters, hitting 37% year-over-year, up from 28% the quarter before — and it’s happening alongside margin expansion, not at the expense of it. AWS operating margin climbed to 39.4%, up 6.5 percentage points year-over-year, driven by efficiency gains and tighter capacity management. The backlog jumped to $496 billion, up $130 billion in a single quarter. Amazon’s push into its own silicon — Trainium and Graviton chips — is a quiet moat-widener here: Graviton already powers 98% of the top 1,000 EC2 customers, and Anthropic alone has committed to spending more than $10 billion a year on Trainium capacity Amazon doesn’t have to buy from Nvidia.
Valuation tells a more nuanced story than the headline multiple suggests. Amazon trades at a forward P/E near 21, about 32% above the sector median of 16 — expensive at first glance. But its forward PEG ratio of 1.03 is actually in line with the sector’s 1.40, meaning the premium is backed by real earnings growth, not sentiment. Against its own history, the stock looks downright cheap: that forward P/E is 87% below Amazon’s five-year average of roughly 160. The risk side of the ledger is real too — trailing free cash flow turned negative $7.6 billion as capex guidance jumped to $220 billion for 2026, and AWS’s growth is increasingly concentrated in a handful of AI labs whose staying power is unproven.
So what for long-term investors: it’s rarely meaningful when one high-conviction manager buys a stock, but it’s a signal worth studying when a hyper-growth investor and a hardened macro skeptic land on the same name for overlapping reasons — durable infrastructure economics, not momentum. This isn’t a call to chase the stock; it’s a reminder that Amazon’s cloud business has quietly evolved from a growth story into a margin-expanding compounder, even while heavy AI capex keeps near-term free cash flow under pressure. Investors who already hold it have more reason to watch capex normalization and free-cash-flow trends over the next few quarters than to fixate on daily share-price swings.
