Amazon Stock Fell 7.6% This Week. The Real Reason Is More Interesting Than It Looks.

Amazon (AMZN) fell 7.6% this week. That made it the worst performer in the Magnificent 7. It also made it one of the most interesting stories, because by almost every business metric, Amazon had a good week.

AWS grew 17% year-over-year last quarter, the fastest pace in two years. Truist Financial raised its price target on the stock. Cathie Wood's Ark Innovation ETF (ARKK) bought $6.5 million in Amazon shares on the dip. The company's advertising business keeps taking share from traditional media. Alexa+ expanded internationally.

And the stock still fell nearly 8%.

The Real Story: $200 Billion in Annual Capex

Amazon disclosed it will spend approximately $200 billion on data centers and AI infrastructure in 2026. That number is the problem. It is bigger than most companies' entire annual revenue. For a stock trading at 35x earnings, every dollar of capex that doesn't immediately translate to revenue is a dollar of future earnings deferred.

The market is running a simple calculation: if AWS growth requires $200 billion per year to maintain, what are the actual economics of that business long-term? The answer is still favorable, but it's a more complicated math than it looks from the outside.

The Bull Case

AWS's 17% growth rate on an already massive base is impressive. The advertising business is structurally taking money from TV and social. Amazon's logistics network is a moat that took 20 years to build. Cathie Wood doesn't buy stocks because she's being polite.

The structural case for Amazon is intact. The free cash flow pressure is a timing problem, not a business quality problem.

Bottom Line

AMZN sits at $250, down 7.7% year-to-date after entering 2026 near $270. The analyst consensus remains Buy with an average target above $310. The capex number is real but it is building assets that will generate returns for a decade. At $250, investors are pricing in a lot of bad news on a company that is still executing.

P.S. AWS outages make headlines. AWS growth doesn't. The Virginia facility thermal event in May got more coverage than the 17% growth number. That's the information asymmetry investors can exploit.