AWS is expected to generate $1 trillion in revenue over an 8-10 year horizon and Amazon stock could reach $500 by the end of 2027?
Morgan Stanley came up with this claim. More precisely, it's more of a bull scenario that, according to analyst Brian Nowak, is becoming increasingly likely if AWS continues at its current pace.
How far is AWS from $1 trillion?
At first glance, extremely far. AWS generated $42.2 billion in revenue in Q2, up 37% year-over-year, which was the fastest growth rate in the last 18 quarters. The current annual revenue run-rate is approximately $169 billion. To reach $1 trillion, AWS must grow approximately 5.9x more.
Profitability is also interesting. Operating margin was approximately 39%, while AWS accounted for only about 21% of Amazon's revenue but as much as 60% of its operating profit. That's also why AWS is much more important for Amazon's valuation than its share of total revenue suggests.
What is the Morgan Stanley model based on?
Two variables: how much compute capacity Amazon can build and how much revenue it can extract from each watt of that capacity.
Nowak expects approximately 6 GW of new capacity in 2026, 8 GW in 2027, and then approximately 8 GW annually. The second part of the equation is monetization. Morgan Stanley estimates that AWS currently generates approximately $8 in revenue for every new watt of capacity.
If that moved to $12 per watt thanks to more expensive AI workloads, better chips, higher data center utilization, and pricing, AWS could surpass $1 trillion in revenue around 2035 according to their model.
At a long-term EBIT margin of around 30%, AWS alone could generate approximately $300 billion in EBIT. After adding retail, advertising, and other parts of Amazon, Morgan Stanley sees potential for approximately $500 billion in EBIT for all of Amazon between 2034 and 2036.
What could go wrong?
The biggest risk today is capacity rather than demand. AWS needs enormous amounts of data centers, energy, and chips, so if construction slows or AI infrastructure is not sufficiently profitable, this whole scenario could be significantly delayed. Massive investments have already pushed Amazon's FCF into negative territory. Another risk I see is rising memory prices, which they also admitted on the last earnings call.
So are $1 trillion and $500 realistic?
Although the number sounds extreme, from today's approximately $169 billion, AWS would need to grow about 25% annually for eight years. For a company of this size, that's a huge pace, but today AWS is growing 37% and Amazon still says demand exceeds the capacity it can build.
What Amazon gets out of today's $220 billion in CapEx will be decisive. During the first big AWS buildout, it invested years ahead, free cash flow was under pressure, and the economics of those investments only showed up later. Today it's making a similar bet on AI infrastructure, just on a much larger scale.
For me, this scenario makes sense if new data centers fill up quickly, AWS maintains high margins, and AI continues to push cloud spending higher. Then today's extreme CapEx could turn into significantly higher operating profit and free cash flow in the coming years, and $500 per share seems quite realistic.
How do you see it? 👇
