Is Amazon Stock at $255 a Share an Obvious Buy Right Now?

Amazon is projected to spend $220 billion on capex in 2026, up from $132 billion in 2025, largely to expand AI capacity; the investment has contributed to trailing-12-month free cash flow turning negative by $7.6 billion and tens of billions of dollars in borrowing. Growth has accelerated, with Q2 2026 net sales rising 20% year over year versus 13% a year earlier, while AWS growth increased to 37% from 17.5%. At roughly $255 per share and 21x earnings, the article views Amazon as a long-term buy despite risks that AI spending could prove difficult to justify.

Analysis

AMZN’s valuation discount is only investable if AWS incremental returns remain visible despite the capex step-up. The key mechanism is depreciation and financing: even if revenue growth holds, the larger asset base can pressure EBIT/FCF conversion for several quarters, limiting multiple expansion relative to capital-light AI beneficiaries. This favors firms monetizing Amazon’s buildout with faster revenue recognition—NVDA near term, but also networking and power/cooling suppliers not represented in the ticker set—over AMZN until utilization data confirms demand is absorbing capacity.

Near-term, the market will focus less on reported sales growth than on AWS backlog, capacity availability, operating-margin trajectory, and whether management can bound 2027 capex. A 1–3 month upside catalyst is evidence that enterprise AI workloads are moving from experimentation to sustained inference consumption, which would improve AWS utilization and reduce the “overbuild” narrative. The adverse case is a second consecutive quarter of negative free-cash-flow conversion alongside rising debt, likely producing a rerating toward infrastructure-heavy cloud peers regardless of topline growth.

Consensus may be over-indexing to headline capex and underweighting Amazon’s ability to cross-subsidize AI infrastructure through advertising, third-party marketplace economics, and retail logistics density. But the low P/E is not automatically cheap: earnings can be flattered while cash economics deteriorate through depreciation timing. The 6–18 month payoff is attractive only if AWS growth remains above 30% while segment margin stabilizes; otherwise the investment case shifts from AI platform leader to a lower-multiple utility-like infrastructure owner.

Key Decisions for Investors

• Maintain a neutral-to-modest long AMZN rather than chase: add only following an earnings print showing AWS growth above 30%, stable/improving AWS operating margin, and no upward revision to next-year capex. Target a 6–12 month rerating if FCF turns positive; exit/reduce if AWS growth falls below 25% or debt-funded capex accelerates.
• For a cleaner 1–3 month AI-spending expression, pair long NVDA against AMZN in equal beta-adjusted dollars. NVDA captures accelerated infrastructure orders earlier, while AMZN bears utilization and depreciation risk; reassess if hyperscaler order commentary weakens or NVDA data-center guidance decelerates materially.
• Use AMZN downside protection rather than outright short exposure: buy 6–9 month put spreads around the next two earnings dates if implied volatility is below its post-results range. The asymmetric risk is a capex/FCF disappointment, while outright short risk is strong AWS demand or an AI monetization announcement.
• Set an earnings watch item—not a trade trigger—for disclosure of AWS backlog, customer commitments, server useful-life assumptions, and leasing obligations. These items determine whether current spending is demand-backed capacity or speculative supply, and are more informative than consolidated P/E.