Amazon reported a strong fourth quarter, with revenue up 14% to $213.4 billion and net income climbing to $21.2 billion, yet the market reaction was swift and negative when the company outlined a radical increase in capital spending for 2026. Management told investors it expects to boost capital expenditure to roughly $200 billion next year, up sharply from about $131 billion in 2025 — a jump that markets interpreted as an aggressive, expensive commitment to AI infrastructure.
Behind the headline numbers there are mixed signals. Amazon Web Services (AWS) remained the company’s cash engine: fourth‑quarter sales at AWS rose 24%, the biggest jump in more than three years, and the division still provides more than 60% of Amazon’s operating profit despite representing only 15–20% of overall revenue. Advertising also continued to perform well, with sales up 22% to $21.3 billion, even as the company booked a $610 million impairment linked to its physical stores and continued to restructure its retail footprint.
The $200 billion capex plan reflects Amazon’s place in an escalating infrastructure race among hyperscalers. The company is expanding data centres, networking and custom silicon — including launching the “Rainier” AI infrastructure project and deploying nearly 500,000 in‑house Trainium2 chips, some of which power work for Anthropic. Management singled out investments in high‑speed consumer internet (project Leo) and “instant retail” services as additional near‑term cost drivers, including roughly $1 billion in new Leo spending baked into guidance.
Investors punished the stock. Amazon fell about 4.4% in regular trading and plunged further — roughly 9% in after‑hours trading — as markets focused less on the quarter’s top‑line strength and more on the cash flow implications of near‑term spending. Street analysts flagged that the company’s midpoint guidance for operating income in the first quarter of 2026 sits below consensus, and commentators warned the planned capex could exceed Amazon’s operating cash flow if sustained at the announced scale.
The reaction underscores a wider market dynamic: Wall Street will tolerate soaring AI budgets only if they translate into commensurate revenue and margin gains. Alphabet and Meta won investor approval for their large capex plans because each paired spending with visibly strong cloud or ad revenue growth, while Microsoft has seen its share price punished when cloud growth merely scraped expectations. Amazon’s risk is that a big infrastructure lead will be expensive and slow to show returns in per‑share cash flow terms.
The strategic calculus for Amazon is straightforward but high‑stakes. Owning and operating extensive AI infrastructure — custom chips, dense data centres and high‑bandwidth networks — could lock in long‑term advantages for hosting large models, attracting model developers and capturing higher‑margin cloud workloads. But the company must also contend with potential overcapacity, steep depreciation, and the need to maintain profitable growth in legacy retail and advertising businesses while financing this build‑out.
Beyond Amazon, the spending surge has supply‑chain consequences. Semiconductor suppliers, power equipment vendors and data‑centre construction firms stand to benefit from multi‑year orders; conversely, a miscalculation could leave the industry with excess capacity and downward pressure on utilisation and prices. Amazon’s pivot in physical retail — closing Fresh and Go locations, converting some to Whole Foods, and opening a large new warehouse-format store — shows the company attempting to rebalance capital allocation across a sprawling portfolio of retail, cloud and ad assets.
The near‑term story is therefore one of tension between ambition and patience. Amazon’s executives are buying an early seat at the AI table by investing heavily in the underlying plumbing; investors are asking for a clearer path from billions of dollars of capex to sustainable earnings and cash flow. How convincingly Amazon can translate raw infrastructure into monetisable services will determine whether the $200 billion wager becomes a defining strategic advantage or a source of prolonged valuation pressure.