Amazon’s Andy Jassy Just Made a Startling Prediction.

AWS has reached a $169 billion annual revenue run rate, while its AI revenue run rate has surpassed $25 billion roughly three years into the generative-AI boom. Amazon increased planned 2026 capital spending to $220 billion from $200 billion to expand AI and cloud infrastructure, yet says it cannot meet all demand this year and next. CEO Andy Jassy said AWS, which Amazon once expected to become a few-hundred-billion-dollar business, could ultimately generate at least double that and potentially $1 trillion in annual revenue.

Analysis

The investable issue is not AWS demand visibility but conversion of that demand into incremental return on capital. AMZN is choosing to fund capacity ahead of revenue recognition, which can depress near-term FCF and keep the stock’s valuation anchored to retail/advertising earnings rather than a clean AI-infrastructure multiple. The key 1-3 month catalyst is whether management quantifies backlog, utilization, and AWS operating-margin resilience; without those, a larger capex plan can be interpreted as competitive necessity rather than proof of pricing power.

AWS’s broad compute stack is strategically more important than a pure GPU reseller model: custom silicon and managed AI services can raise switching costs and protect gross margin if Nvidia supply loosens. Conversely, the same vertical integration makes AMZN a more direct competitive threat to NVDA at the inference layer and to MSFT/GOOGL in enterprise AI workloads. The second-order beneficiary is likely networking and power infrastructure rather than application software: sustained cloud build-outs support AVGO, ANET and data-center electrical suppliers, while enterprise software will not receive a durable benefit until AI workloads translate into customer productivity budgets.

Consensus appears too willing to capitalize distant cloud revenue while discounting the financing and depreciation burden required to earn it. A trillion-dollar AWS outcome is not a near-cycle valuation input: it requires sustained hyperscaler share, enterprise migration, and acceptable returns through at least one hardware-refresh cycle. The near-term asymmetry is therefore modestly favorable only if AWS growth reaccelerates without an accompanying step-down in consolidated FCF or operating margin; a capex increase alone is not a bullish catalyst.

Key Decisions for Investors

• Maintain or initiate a measured AMZN long on post-results weakness rather than chase promotional AI headlines; use the next earnings release as the decision point. Add only if AWS growth accelerates and management demonstrates stable-to-improving AWS margin despite capacity additions; reduce if consolidated FCF conversion weakens materially for two consecutive quarters.
• Express AI infrastructure demand through long AMZN / short MSFT in a 3-6 month relative-value sleeve only if AWS backlog or capacity constraints are disclosed as worsening. AMZN offers greater upside to incremental capacity monetization; the pair is invalidated if Azure growth materially outpaces AWS or AMZN’s cloud margin compresses.
• Use AVGO or ANET as higher-purity 6-12 month beneficiaries of continued hyperscaler build-out, but do not treat them as substitutes for AMZN: their upside depends on orders and component content, not cloud-service pricing. Trim if hyperscaler capex guidance shifts from expansion to optimization or lead times normalize sharply.
• Avoid adding to NVDA solely on this signal. AWS’s increasing use of internally designed chips is a longer-duration inference-margin risk; reassess the AMZN/NVDA relationship after disclosed custom-silicon adoption, GPU mix, or evidence that AWS can meet incremental demand without proportional Nvidia procurement.