Morgan Stanley: AWS to Reach Trillion-Dollar Revenue in a Decade, Amazon Stock Could Double to $500 by End of 2027
Nashnova编辑部
Morgan Stanley projects AWS can reach $1 trillion in annual revenue within eight to ten years, setting a $500 year-end 2027 price target for Amazon — implying 93% upside from the current ~$258.68, but Wall Street is sharply divided on whether the path holds.
Where does the trillion-dollar number come from?
Analyst Brian Nowak starts from Amazon CEO Andy Jassy's own words: AWS will "very likely" become a $1 trillion-a-year business.
AWS currently runs at roughly $170 billion in annualized revenue. This means → reaching a trillion requires nearly 5× growth over eight to ten years.
Nowak's math from there: if AWS hits $1 trillion, Amazon's overall EBIT could reach $500 billion, supporting a year-end 2027 target of $500 per share.
What drives the growth?
The core variable is how fast compute capacity can expand — specifically, how quickly data centers get built and how much power comes online.
Nowak forecasts Amazon adding 6 GW of data-center capacity in 2026, 8 GW in 2027, and 8 GW annually after that. Jassy has confirmed the company expects to double its power capacity by end of 2027 versus 2025 levels.
The other key metric is "revenue per watt" — how much income each watt of compute generates. In plain terms = how much money a single unit of electricity earns from customers sets the ceiling on growth.
Nowak estimates AWS currently earns roughly $8 per watt. If that rises to $12 per watt, AWS could hit $1 trillion as early as 2035. He argues AI innovation will keep pushing this efficiency higher.
What do the skeptics say?
D.A. Davidson analyst Gil Luria calls the trillion-dollar estimate "a bold extrapolation."
He sees AWS growing at a healthy 40%–50% this year, but says "extrapolation beyond that range is already quite aggressive."
Luria's core challenge: "Mr. Jassy — or anyone else — has no hard data to quantify a market that didn't exist three years ago." This reflects a fundamental uncertainty over whether AI compute demand can sustain at this pace.
What unknowns lie beyond 2028?
Nowak himself acknowledges significant unknown variables in any forecast past 2028.
These include: server and rack procurement scale, pace of energy-efficiency gains, regulatory hurdles, and the actual speed of data-center construction.
In plain terms = even if the direction is right, any single bottleneck — slower power approvals, a server supply-chain break — could push the trillion-dollar timeline out significantly. Whether AWS can keep converting demand into effective compute supply is the central test of this long-term thesis.
AWS at $1 trillion — achievable path or over-extrapolation?
BULL
CEO endorsement
Jassy publicly stated AWS will 'very likely' hit a trillion — management is betting on this path.
Capacity ramp is concrete
2026–2027 data-center expansion plans are specific; power capacity doubling in two years.
Revenue per watt can rise
From $8/watt to $12/watt — AI innovation is the driver.
BEAR
Extrapolation stretches too far
From $170B to $1T requires ~5× growth — beyond reasonable extrapolation range.
No historical data for this market
The AI compute market didn't exist three years ago — no hard data to quantify it.
Execution variables are dense
Regulation, supply chain, efficiency — any bottleneck breaks the timeline.
Put simply = the bulls are betting on 'right direction + fast execution'; the bears say 'direction maybe right, but the timeline is a guess' — the disagreement isn't about the destination, it's about the path.
Content is for reference only, not financial advice.