TD Cowen reiterated a Buy rating and $300 price target on Amazon after Amazon and Anthropic expanded their partnership, including Anthropic’s commitment to use up to 5GW of Trainium capacity and spend more than $100 billion over 10 years on AWS. Amazon is also investing $5 billion in Anthropic, with the option to invest another $20 billion, reinforcing the company’s AI infrastructure monetization story. Separate bullish analyst updates from Oppenheimer, UBS, and BMO lifted price targets to $275, $304, and $315, respectively.
The market is starting to treat frontier-model capacity as a long-duration utility contract rather than a cyclical cloud win, and that shifts the investment case for AMZN from ‘AI optionality’ to ‘AI annuity.’ If Anthropic really anchors multi-year training and inference demand on AWS, the second-order benefit is not just revenue growth but higher switching costs across the developer stack, which should support mix and pricing power in adjacent AWS services over the next 6-18 months. The bigger implication is competitive: if one hyperscaler can bind a top-tier model lab this tightly, rivals will be forced to spend harder on capacity, incentives, and custom silicon to avoid being left with lower-quality workloads.
The market is likely underestimating how much of this value accrues through Trainium utilization rather than headline model revenue. That matters because accelerator fill rates and cluster longevity can improve return on invested capital even before revenue catches up, so the near-term upside is more about margin durability than pure top-line acceleration. A key second-order beneficiary is any supply-chain name tied to networking, power, and data-center buildout; if this scales, the bottleneck moves from model demand to physical deployment capacity, which tends to support the broader AI infrastructure trade for several quarters.
The main risk is timing mismatch: capital intensity and capacity commitments are visible now, while monetization may take multiple quarters to prove out, creating a window where sentiment outruns earnings revisions. Another risk is that large model customers may diversify across clouds to preserve bargaining power, which would cap AWS’s share of wallet even if the partnership deepens. If AI spend broadens but enterprise adoption slows, the narrative can revert to ‘too much capex, not enough ROI’ quickly.
The contrarian view is that consensus may be too focused on the headline partnership and not enough on the economics of marginal growth. If AWS’s incremental AI revenue is being sourced at lower contribution margins due to silicon subsidies, customer incentives, or expensive energy/interconnect buildout, the earnings benefit may lag the stock multiple expansion. That creates a setup where the equity can remain supported near term, but the next decisive move depends on evidence that AI workloads are lifting operating leverage rather than simply inflating capex guidance.
AI-powered research, real-time alerts, and portfolio analytics for institutional investors.
Request TrialOverall Sentiment
moderately positive
Sentiment Score
0.62
Ticker Sentiment
