Upside comes from AWS AI re-acceleration plus retail automation and new trust/checkout rails for agent-led commerce. The payoff is gated by delivered power timelines and the shape of US/EU platform remedies.

Into 2031, Amazon compounds by turning AI-era scarcity (compute, power, secure execution) and commerce intent (Prime + delivery promise) into higher-trust rails for enterprises and agents, while robotics and network optimization lift retail/seller unit economics enough to fund outsized capex without permanently diluting cash returns.

Owns two durable Last-Economy control points—hyperscale compute APIs and a delivery/returns promise—plus the capital scale to build through compute/power bottlenecks and monetize trust/security.

The non-linear opportunity is Amazon re-pricing from “mixed retail + cloud” to “AI infrastructure + trusted commerce rails.” If AWS re-accelerates as new capacity is energized and higher-layer services (security, data, orchestration) grow faster than raw compute price compression, AWS becomes a larger share of gross profit. In commerce, Amazon’s edge is not just traffic; it is the verifiable delivery promise, returns resolution, and fraud/authenticity tooling that agents and brands will route through because it is reliable. Layering embeddable checkout/delivery promises off-Amazon, plus fulfillment robotics outcomes sold as services, creates incremental high-quality revenue while automation lifts cost-to-serve.

The binding risk stack is externally gated throughput plus permissioning: (1) delivered power and build timelines can cap AWS growth even with demand; (2) antitrust/platform remedies can limit marketplace and Prime monetization; (3) the 2026–2028 capex/depreciation wave can suppress FCF if utilization ramps slower than expected; (4) agents can shift shopping discovery and ad attribution off Amazon unless Amazon becomes the default trusted checkout + fulfillment endpoint.

$282.17