Morgan Stanley Initiates Nebius at "Equal-Weight" with $144 Price Target
N.R. Finch
Morgan Stanley initiated Nebius with an Equal-Weight rating and a $144 price target — a roughly 47% discount to the $211.69 close — citing an attractive AI-cloud sector but a stock that has priced in too much optimism while compute supply bottlenecks cap near-term growth.
Why is Morgan Stanley cautious on Nebius despite liking the sector?
Morgan Stanley calls the AI-cloud industry "attractive" overall, yet rates Nebius only Equal-Weight with a $144 target.
This means → the bank sees the right sector story, but believes the current share price already bakes in upside that hasn't been earned yet.
In plain terms = good track, expensive ticket — the stock needs results to catch up with the price.
What did customers reveal at the summit?
Shopify said legacy hyperscalers "either lack capacity or charge too much," pushing it toward Nebius.
Higgsfield noted its demand growth has outpaced available compute; Rhoda praised platform stability but said it still cannot secure enough capacity.
This means → strong demand is good news in theory, but when supply can't keep up it becomes a ceiling — customers want to spend, and Nebius can't serve them all.
How is Nebius trying to move from selling raw compute to selling a platform?
The company has assembled a full AI-cloud stack: GPU servers, AI-optimized storage, serverless AI, an MLOps platform (tools that manage AI models from training to production), inference services, and an AI-agent tool called NebiusEcho.
Management explicitly prefers selling directly to end-customers rather than relying on large wholesale deals with hyperscalers.
This reflects a key bet — direct-to-enterprise AI-cloud services carry far higher margins than wholesale bare compute.
How steep is the projected revenue ramp?
Morgan Stanley forecasts Nebius revenue rising from roughly $530 million in 2025 to $17.93 billion in 2028 — approximately a 34× increase in four years.
Operating margins and adjusted EBITDA margins are expected to improve each year, with losses narrowing toward eventual profitability.
In plain terms = the forecast draws an extraordinarily steep growth curve; if delivered, the valuation makes sense — but "if" is doing a lot of work.
How wide is the gap between bull and bear scenarios?
Bull case: target price $400, assuming compute capacity scales up and software monetization beats expectations.
Bear case: target price $70, assuming demand fades, competition intensifies, and execution disappoints.
This means → the spread from $70 to $400 is nearly 6× — reflecting deep disagreement on Nebius's outlook. Currently 65% of covering analysts rate it a buy, 35% neutral, and no one rates it a sell — but no one is going all-in either.
What metrics should investors watch next?
Compute delivery: whether Nebius can close the supply gap over the next few quarters directly determines if revenue forecasts are achievable.
Software monetization pace: how quickly high-margin services like inference, intelligent routing, and AI agents scale is the key test of the shift from "selling compute" to "selling a platform."
In plain terms = compute is a "can they do it" question; software is a "how much can they earn" question — both need real numbers.
Content is for reference only, not financial advice.