Amazon Just Proved It's No Longer an AI Underdog

Amazon said AWS generated $14.6 billion in operating income, or 59% of total company operating income, while AWS revenue grew 28% year over year in Q1. Total Amazon revenue rose 17% year over year, marking its best quarter since 2021, with management highlighting surging demand for custom AI chips and Trainium adoption by OpenAI and Anthropic. The article argues AWS is becoming Amazon's key AI growth engine and supports a buy-and-hold view on the stock.

Analysis

AWS is becoming the earnings flywheel that matters for AMZN valuation, and the market is still under-assigning value to the quality of that mix shift. The second-order effect is that higher-margin cloud growth can mask slower commerce unit economics, allowing Amazon to compound operating income faster than headline revenue suggests; that tends to support multiple expansion because the market pays for durability and optionality in AI infrastructure, not just retail scale.

The real strategic edge is less “AI demand” in the abstract and more Amazon’s ability to monetize pick-and-shovel infrastructure while also using chip design to reduce dependence on the current GPU bottleneck. If Trainium adoption broadens, the implication is not just incremental AWS share gain; it can pressure the effective price/performance benchmark across cloud providers, forcing competitors to either cut margins or accept slower AI workload migration. That dynamic favors AMZN and is mildly negative for hyperscaler rivals and for NVDA at the margin if custom silicon captures a larger share of inference workloads over the next 12-24 months.

The key risk is execution latency: capex can outrun near-term monetization if customers remain in pilot mode or if AI workloads shift toward lower-margin inference faster than expected. In that case, AWS growth can stay strong while incremental returns on invested capital disappoint, and the stock could de-rate on “good revenue, mediocre FCF” concerns over the next 1-2 quarters. The contrarian point is that the consensus may be too focused on AI upside and not enough on how much of the current enthusiasm is already embedded in the stock; upside likely needs another earnings revision cycle, not just narrative improvement.

Key Decisions for Investors

• Maintain/accumulate a 3-6 month long AMZN position on weakness; use pullbacks tied to broader mega-cap rotation as better entries, with upside tied to continued AWS margin leverage and AI capex monetization.
• Pair trade: long AMZN / short a basket of lower-quality cloud or software names that are more vulnerable to AI spend normalization and margin compression over the next 6-12 months.
• For tactically bullish exposure, buy AMZN call spreads 3-6 months out rather than outright calls; the stock likely needs another earnings beat/revision cycle, so defined-risk convexity is preferable.
• Trim or hedge some NVDA exposure against AMZN strength if you think custom silicon adoption accelerates; the trade is not bearish NVDA outright, but a relative-share shift in AI workloads could cap upside over 12-24 months.
• Set a risk trigger to reassess AMZN if AWS growth remains strong but operating income guidance or capex-to-revenue conversion deteriorates over the next two quarters; that would signal the market is paying up for growth that is not translating into cash flow.