Amazon's post-results research notes were notable for how similar the conclusions looked and how different the underlying valuation logic has become. Citi, Telsey, RBC, Mizuho and TD Cowen were among firms raising targets, with several clustering in the low-to-mid $300s. The ratings were mostly unchanged. The important shift was what analysts were willing to pay for AWS and the advertising business.
Amazon used to be difficult to value because retail profits were thin and cloud profits were growing quickly. It is now difficult to value because several high-margin businesses sit inside one company while management is also committing enormous sums to AI infrastructure. The model has become more profitable and more capital intensive at the same time.
AWS is doing more of the valuation work
Cloud growth carries disproportionate weight because incremental AWS revenue arrives at much higher margins than the core retail business. AI workloads add another layer: they can support demand for compute, storage and networking, but they also require Amazon to build capacity well before customers use it fully.
That timing mismatch is central to the analyst debate. A high target assumes the investment cycle translates into durable cloud growth and strong utilisation. A lower target can accept the same strategic logic while demanding more evidence that capital expenditure will earn an adequate return.
Advertising quietly improves the downside case
Amazon's advertising operation is often treated as a secondary business because AWS gets more attention. It matters to valuation because advertising monetises traffic the company already owns. That means the incremental economics can be attractive even when merchandise volume slows.
This makes Amazon less dependent on one growth engine than it was several years ago. It also means a simple ecommerce multiple is increasingly unhelpful. Analysts are effectively valuing a retailer, a cloud platform and a large advertising network under one ticker.
The next downgrade trigger is likely to be returns, not revenue
Revenue misses would matter, but the more interesting risk is a deterioration in returns on invested capital as AI spending rises. If AWS demand remains strong while cash generation weakens because capacity arrives too early, target multiples could compress even without an earnings recession.
For investors, that makes free-cash-flow conversion and AWS margin progression more useful than another round of target raises. The positive ratings tell you the Street likes the direction. The capital intensity tells you why the valuation debate is not finished.