Here's My Top Reason to Buy Amazon Stock Right Now

Amazon's AWS generated 60% of company operating income in Q2 despite representing about 21% of revenue, highlighting the cloud segment's materially higher profitability versus commerce. AWS revenue grew 37%, compared with 16% growth in North American commerce and 15% internationally, while its backlog reached $496B. Management indicated available computing capacity may remain insufficient to meet AI-driven demand through 2026 and potentially 2027, supporting a bullish outlook for AWS-led earnings growth.

Analysis

The investable issue is not AWS growth alone, but whether incremental AI capacity converts into operating-income mix expansion faster than it absorbs cash through capex and leases. AMZN can re-rate if AWS growth remains above 30% while segment margins hold despite accelerated depreciation, power costs, and customer incentives; the opposite outcome is a familiar hyperscaler pattern of strong revenue with weaker near-term FCF. The next 1-3 month catalyst is earnings evidence on capex intensity, committed-capacity conversion, and whether management raises 2027 infrastructure spending faster than AWS revenue expectations.

Competitive risk is more acute than the bullish framing implies: MSFT and GOOGL can subsidize cloud AI workloads through broader enterprise software ecosystems, while ORCL is targeting capacity-constrained customers with aggressive contract economics. AWS's push toward proprietary silicon improves long-run gross-margin control but may reduce the marginal revenue capture for NVDA if workloads migrate from GPU training toward lower-cost inference. Second-order beneficiaries are power and thermal bottleneck suppliers—VRT, ETN and GEV—because cloud providers cannot monetize booked compute demand without grid, cooling, and electrical equipment deployment.

Consensus likely underestimates the duration of physical infrastructure constraints but overestimates the certainty that backlog converts at current pricing. A sustained AWS margin decline, a meaningful increase in cancellation/renegotiation commentary, or capex growth materially exceeding cloud revenue growth would falsify the AMZN margin-expansion thesis over 6-18 months.

Key Decisions for Investors

• Maintain/add AMZN on 5-10% pullbacks rather than chase AI-news strength; target a 6-12 month position sized to a scenario where AWS growth stays above 30% and consolidated operating-margin expansion offsets higher depreciation. Reduce if AWS growth falls below 25% for two quarters or management guides capex materially above consensus without corresponding revenue acceleration.
• Pair trade for the next 3-6 months: long AMZN / short ORCL in equal beta-adjusted dollars. AMZN has greater margin diversification through advertising and retail efficiency, while ORCL is more exposed to execution risk from concentrated cloud capacity commitments and capital requirements. Exit if ORCL demonstrates sustained cloud growth acceleration without incremental margin or financing pressure.
• Own VRT or ETN as a higher-purity 6-18 month infrastructure expression of hyperscaler capacity scarcity; use a basket if possible. The key risk is project-delay rather than demand destruction, so trim if order growth decelerates sharply or backlog conversion weakens for two consecutive quarters.
• Do not add directional NVDA exposure solely from this development. Monitor AWS disclosures on Trainium/Inferentia adoption and GPU availability: accelerating proprietary-chip mix is a medium-term headwind to NVDA unit economics even if total AI infrastructure spend remains robust.