Amazon runs the world's biggest online store and the cloud (AWS) that powers much of the internet. AWS just posted its fastest growth in 18 quarters, yet the stock sits 14% below its highs because Amazon is spending $200 billion this year building AI infrastructure. Strong business, worried price.
Scale is 0–50%/yr. The shape of the verdict: real acceleration the market will not pay for yet. The base case clears neither hurdle; only the bull case clears the 12% growth bar.
AWS grew nearly 37% last quarter, its fastest clip in 18 quarters, on a $42 billion quarterly base with margins near 39%. The backlog stands at $496 billion. After two years of deceleration fears, the cloud engine is humming again. That is the single most important fact about Amazon right now.
The old story was "retail loses money, AWS pays for it." That era is over. Advertising and logistics efficiency have made the commerce business genuinely profitable, with operating income up 43% last quarter. Amazon is now two strong businesses, not one strong business carrying a weak one.
This is why the stock lags. Amazon plans to spend roughly $200 billion in 2026, mostly on AWS and AI infrastructure, and free cash flow has turned negative. The market sees the bill today and the payoff tomorrow. Every dollar of capex is a bet that AI demand fills the capacity.
Amazon's big investment in Anthropic ties it to one of the leading AI labs, as both a cloud customer magnet and a strategic partner. It also adds noise: a large mark-to-market gain flattered recent earnings, so the headline numbers overstate the operating trend. Read past the one-offs.
Fundamentals versus price is the whole trade. The business is accelerating; the stock is 14% below its high and trails the other giants this year. The base case (~10.1% a year) is solid compounding that clears neither hurdle. Weakness toward $230 would change the math.
$200 billion a year of datacenter spending is a staggering bet. If AI demand disappoints, Amazon owns the world's most expensive empty buildings. Negative free cash flow is fine during a buildout and damning if the buildout was wrong.
The reacceleration is one to two quarters old. Cloud growth has disappointed before. If enterprise AI budgets pause, AWS is the first line item that gets cut.
Advertising and logistics gains drove the retail turnaround. A consumer slowdown hits both ad budgets and order volumes at once. Retail profits are higher quality than they were, not invincible.
Amazon faces antitrust pressure in the US and Europe over marketplace practices and its cloud bundling. Structural remedies are the tail risk nobody can price.
It has lagged the other mega-caps for months despite strong numbers. Cheap-for-a-reason is a real outcome: if the market decides hyperscaler capex is value-destructive, the multiple stays compressed no matter what AWS prints.