Amazon CEO sees AI spending 'meaningfully' higher in 2027

Amazon reiterated plans for $220 billion in 2026 capital expenditure and said 2027 spending will be meaningfully higher; trailing-12-month free cash flow was negative $42.4 billion. Third-quarter revenue and earnings missed Wall Street expectations, and shares fell roughly 5% in after-hours trading, despite AWS revenue rising 39.3% year over year to $46 billion. CEO Andy Jassy projected AWS revenue growth above 40% in both 2027 and 2028 and said Amazon can fund the AI infrastructure buildout.

Analysis

The key risk is execution timing, not simply the size of the investment: servers delivered ahead of powered, commissioned sites tie up cash before they can generate cloud revenue. That can widen the gap between booked demand and realized returns, while depreciation and operating costs may begin pressuring margins once capacity is activated. Large-customer reservations improve demand visibility, but do not by themselves establish utilization, pricing, or attractive returns across the full buildout.

Over the next 1–3 months, the market is likely to focus on whether AWS growth and operating-margin progress can absorb the cash-flow deterioration; a spending plan is credible only if capacity converts into billable workloads on schedule. Over 6–18 months, sustained demand could make today’s spend a durable competitive advantage, while power and construction bottlenecks could delay payback and leave Amazon carrying underutilized equipment. Nvidia and data-center suppliers may receive orders before deployment translates into customer usage, so shipment momentum is not equivalent to end-demand monetization. Microsoft and Alphabet face similar capital-intensity trade-offs; Amazon’s relative performance will hinge on returns per unit of deployed capacity, not capex alone.

Contrarian point: negative free cash flow is not inherently bearish if capacity is rapidly monetized, but management’s long-range growth confidence is not a substitute for evidence of utilization and cash returns. The thesis improves if AWS growth remains strong, margins hold, and capex converts to operating cash flow; it weakens if commissioning delays persist or growth slows while spending rises.

Key Decisions for Investors

• Avoid adding to AMZN solely on management’s long-term demand outlook. Reassess after the next earnings release using AWS growth, operating margin, cash capital expenditure, and free-cash-flow trajectory together.
• For a tactical bearish expression, consider a small, defined-risk AMZN put spread only after post-earnings volatility settles; the setup is a further deterioration in cash conversion without corresponding AWS growth or margin support. Risk is capped at premium paid, and the trade is falsified by improving free cash flow alongside sustained AWS growth.
• Set an execution alert for evidence that delivered servers are becoming billable capacity: data-center power/commissioning updates, utilization, and customer deployment cadence. Continued delays alongside rising spending would strengthen the downside case; timely deployment and margin resilience would invalidate it.
• Do not treat supplier order strength as proof of end-user monetization. Monitor whether AI infrastructure shipments convert into deployed capacity and customer workloads before extrapolating supplier demand or Amazon’s return profile.