We're raising our price target on Amazon after its all-around killer quarter

Amazon delivered a strong Q1 beat, with revenue up 17% to $181.52 billion versus $177.3 billion expected and GAAP EPS up 75% to $2.78 versus $1.64 consensus. AWS growth accelerated to 28.4% with revenue of $37.59 billion, while operating income rose 30% to $23.85 billion and the company reported its highest operating margin quarter in history. Management held 2025 capex guidance at $200 billion and guided Q2 revenue to $194 billion-$199 billion, above consensus at the midpoint, prompting a price target raise to $300 from $250.

Analysis

The key read-through is that AMZN is no longer just a “cloud beta” name; it’s becoming a capital-allocation compounder with multiple self-reinforcing loops. The backlog surge and faster AWS growth imply management can keep leaning into capex without the usual fear of overbuilding, which should pressure competitors that lack AWS’s mix of custom silicon, enterprise lock-in, and demand visibility. That also matters for NVDA: the market narrative is shifting from “Nvidia wins every AI dollar” to a more fragmented spend stack where hyperscalers increasingly capture more economics in-house.

The second-order winner is the entire AI application ecosystem, but the second-order loser may be the margin structure of smaller cloud and infrastructure providers that cannot fund comparable buildouts. Amazon’s custom chip progress increases pricing discipline across the cloud market and could make MSFT/GOOGL defend share with more incentives and capacity commitments, which is mildly negative for near-term cloud margin expansion elsewhere. For retail peers, strong operating leverage in North America and International suggests Amazon is taking more share while still lowering unit costs, a combination that raises the bar for WMT and TGT to compete on convenience without sacrificing margin.

The most important risk is not demand—it is execution on capex intensity. Over the next 2-3 quarters, the stock can keep rerating if AWS backlog converts cleanly, but any sign of delayed monetization, power bottlenecks, or rising depreciation drag would hit sentiment fast because expectations are now stretched. The market is likely underestimating how sensitive AMZN becomes if AI infrastructure spend slows even modestly; after a strong run, the setup is more vulnerable to a “good but not accelerating” print than to a true miss.

Contrarian view: the move may be only partly about the quarter and more about a regime change in how investors value Amazon’s optionality. The risk is that consensus is extrapolating the current AWS acceleration and assuming capex automatically compounds into higher long-term margins. If AI spending normalizes or the Anthropic-related revenue contribution proves lumpy, the stock can still work, but the multiple expansion probably pauses rather than compounds at the current pace.

Key Decisions for Investors

• Stay long AMZN on 1-3 month horizon; buy on any post-earnings consolidation rather than chasing strength. Risk/reward favors add-ons on dips if the stock holds above prior breakout levels, with upside tied to backlog conversion and capex credibility.
• Pair long AMZN / short a basket of slower-growing cloud beneficiaries over 2-6 months. The trade expresses AWS scale and custom-chip advantage while fading firms most exposed to hyperscaler price pressure.
• Reduce tactical NVDA overweight on AMZN strength, not because Nvidia loses structurally, but because hyperscalers are increasingly internalizing more AI economics. Use a 1-2 quarter horizon; risk is that NVDA demand remains broad-based and offsets this rotation.
• For retail exposure, prefer AMZN over WMT/TGT for the next 6-12 months. Amazon’s logistics and subscription flywheel are compounding faster; if consumer spend weakens, its efficiency gains are better insulated than box retailers' traffic-dependent models.
• Consider selling 30-60 day out-of-the-money AMZN calls against a core long if implied volatility remains elevated. This harvests upside premium after a strong rerating while keeping exposure to the longer-dated AI/cloud thesis.