Amazon / Assumptions, evidence and sensitivity
A strong thesis should survive a change in assumptions.
Separate what Amazon reported, what the original post assumes, and what the arithmetic actually establishes. All scenario values below use the same end-2026 reference date.
The reported starting point
Amazon reported 2025 AWS revenue of $128.7 billion and AWS operating income of $45.6 billion. Group revenue was $716.9 billion. Subtracting AWS gives about $588.2 billion of non-AWS revenue for that year, not the post's undated $630 billion. These are historical figures from the February 5, 2026 release, not current run rates. AWS operating income is not standalone net income.
Where the $530 billion comes from
The source post attributes this 2030 forecast to Morgan Stanley. The post establishes that attribution was made, not that the research estimate has been independently confirmed. No verified Morgan Stanley report is supplied here. We retain $530 billion as an explicit demo assumption. Moving from $128.7 billion in 2025 to $530 billion in 2030 requires about 32.7% compound annual growth across five years.
The valuation bridge
$530B × 25% net margin × 25 = $3,312.5B in 2030. Divide by 1.10 to the fourth power to get $2,262.5B at end-2026. Against the fixed $2,800B Amazon reference, the residual is $537.5B. Rounded card values are $3.31T, $2.26T and $538B. The original post rounds to $2.2T and $600B without specifying an exact discount period. Five years of discounting would instead give about $2.06T.
Three scenarios, one method
Thesis: $530B revenue, 25% net margin, 25x earnings gives $2.26T. Margin sensitivity: $530B, 20%, 25x gives $1.81T. Slower case: $350B, 20%, 20x gives $0.96T. Each discounts four years at 10%. These are illustrations, not probability-weighted forecasts or Amazon share-price targets. The slower case changes three assumptions; the margin sensitivity isolates one.
Why cash matters
Data centers, chips and power connections require investment before they produce revenue. Cash leaves upfront, while depreciation recognizes costs over time. Strong accounting earnings can therefore coexist with heavy reinvestment. Amazon reported $139.5 billion of operating cash flow and $11.2 billion of free cash flow in 2025. Those group figures should not be treated as AWS-only cash flows.
What this model leaves out
This discounts a terminal earnings valuation; it is not a full discounted-cash-flow model. It does not explicitly include interim distributions or allocate corporate costs, debt and cash to a hypothetical standalone AWS. The residual cannot be treated as a separate quoted valuation of retail and advertising. Their own economics still need to be modeled.
What would strengthen or weaken the case
Look for paying demand keeping pace with installed capacity, healthy unit economics and cash generation catching up with investment. Sustained pricing pressure, idle capacity, rising power costs or repeated construction delays would make the path harder. A lower market earnings multiple can reduce value even if operations hit their targets.
Second-order effects
If cloud demand expands, power, cooling and equipment suppliers could gain orders. Power bottlenecks could then delay cloud revenue. Cheaper compute could stimulate usage while reducing revenue per unit. Enterprise migration could create implementation work, followed by demand for cloud cost-control tools. These are conditional economic mechanisms, not reported outcomes or predictions that particular stocks will rise.
Reading the stock card
The AMZN card uses the app's market-chart provider. Its displayed date and price may differ from this fixed demo scenario. A share price is not market capitalization.