Amazon shares fell 11% in after-hours trading February 5 after revealing 2026 capital expenditure plans of $200 billion, roughly $50 billion above analyst expectations. The selloff came despite Q4 2025 results that were objectively solid: revenue of $213.4 billion (up 14% year-over-year, beating estimates), and AWS revenue growth accelerating to 24% from the prior quarter's 20%.
The real story: AWS added $7 billion in revenue year-over-year and maintained 35% operating margins while scaling infrastructure. The cloud segment's backlog reached $244 billion, up 40% annually and ahead of Google Cloud's $240 billion. CEO Andy Jassy said the company is "monetizing capacity as fast as we can install it."
Why the market rebelled: Amazon's massive capex increase wasn't matched by materially higher Q1 2026 guidance. Meta and Alphabet both announced aggressive AI spending this earnings season and got passes. Amazon didn't, largely because its guidance suggested returns won't flow to profit quickly enough.
The enterprise angle: Amazon's spending signals the hyperscaler infrastructure arms race is intensifying. For CTOs evaluating cloud strategies, three things matter here. First, AWS capacity constraints are easing, which should improve availability for enterprise workloads. Second, this level of capex ($200 billion represents roughly 40% more than 2025's spend) suggests Amazon sees sustained AI infrastructure demand through 2027. Third, the market's rotation from tech to value sectors means CFOs will face harder questions about cloud spend, even as providers scale capacity.
Jassy's assertion that "customers really want AWS for core and AI workloads" is backed by the backlog numbers. The question isn't whether demand exists, it's whether Amazon can convert $200 billion in infrastructure into returns fast enough to satisfy investors pricing in near-term earnings growth.
For enterprise buyers, this likely means AWS will remain aggressive on enterprise deals to fill that capacity. History suggests the best pricing comes when hyperscalers are racing to monetize fresh infrastructure. The trade-off: longer commitments in exchange for better rates.
We'll see if Amazon's bet pays off. The company has deep experience reading cloud demand signals. But in a market that's lost patience with "trust us" narratives, execution matters more than promises.