NBIS has been whipping around because the tape is emotional and the headlines are loud. But Nebius is not a mystery box. It’s a capacity business in the most capacity-starved corner of tech: AI compute. If you can get GPUs, power, and data center throughput online, you can grow. If you can’t, you stall. That’s the whole game.
So here’s the trade idea: ignore the volatility and follow the capacity. The stock is back near a widely watched area around $94-$95 after a sharp drop, with heavy volume and an RSI that’s cooled off to 44.7. That combination often creates a tradable floor if the fundamental narrative stays intact. I’m looking for a defined-risk long, not a marriage.
There’s also a positioning angle that matters. Short interest climbed to 37.8M shares as of 12/31/2025 with 3.89 days to cover. In a stock that can move 5% in a day on a headline, that’s enough dry tinder to turn a normal rebound into a faster squeeze if buyers show up.
Thesis: Nebius is priced like a high-expectation AI infrastructure story (because it is), but the recent pullback has brought the stock back toward a technically defensible level. If NBIS holds support and recaptures key moving averages, the next leg higher can be a “capacity bid” rather than a hype bid. The trade is simple: buy near support, stop out if it breaks, target a move back toward the low $110s.
What Nebius actually does - and why the market cares
Nebius Group N.V. provides infrastructure and services to AI builders. The core is Nebius AI, an AI-centric cloud platform offering large-scale GPU clusters, cloud services, and developer tools. On top of that, it operates specialized brands: Toloka AI (data partnerships for generative AI development), TripleTen (tech reskilling/edtech), and Avride (autonomous driving tech).
From a market perspective, the reason NBIS matters is that it’s positioned as a “pick-and-shovel” play in AI. While model winners rotate (and LLM narratives age fast), compute demand has been sticky. The practical bottleneck has been supply: advanced chips, power availability, and the ability to deploy clusters quickly. That’s why “neocloud” names can trade with such torque: when capacity is scarce, customers pay and contracts get big.
Recent coverage has pointed to Nebius benefiting from explosive growth tied to AI compute capacity demand, with mentions of large customer relationships and very large deal sizes. The market is essentially trying to handicap one question: can Nebius keep scaling capacity fast enough to justify the valuation? If yes, the stock can look cheap in hindsight. If no, the multiple compresses brutally.
The numbers that matter right now
| Metric | Value | Why it matters |
|---|---|---|
| Current price | $94.46 | Back near a key support zone after a sharp selloff |
| Market cap | $23.03B | Large-cap expectations - the market wants proof of scalable economics |
| 52-week range | $18.31 to $141.10 | Shows how violently sentiment swings in this name |
| Volume (most recent session) | 14.68M | Active participation - the move isn’t happening on air |
| Average volume (30 days) | 11.23M | Liquidity is strong enough for defined swing execution |
| RSI | 44.7 | No longer overbought - momentum cooled, room to rebound |
| 10-day SMA | $100.83 | First upside “regain” level if a bounce sticks |
| 20-day SMA | $96.26 | Nearby friction/resistance - reclaim helps confirm trend repair |
| 50-day SMA | $93.00 | Below current price - a useful reference for support structure |
| MACD state | Bearish momentum | Trend is damaged - we want confirmation, not blind catching |
| P/E ratio | 99.77 | Market is pricing strong future earnings power - little patience for stumbles |
| P/B ratio | 4.79 | Rich vs traditional infrastructure - justified only if growth stays exceptional |
Two things jump out.
First, the stock is sitting near a psychological and technical battleground. A recent news-driven drop referenced $94.50 as key support. We closed at $94.50 and are now around $94.46. That’s not a coincidence level. It’s a “market is voting right here” level.
Second, momentum is weak but not broken beyond repair. MACD is bearish (line 1.10 vs signal 1.81), but the RSI at 44.7 suggests selling pressure has cooled from the frothier levels you’d see near blow-off tops.
Valuation framing: expensive on purpose
At roughly $23.0B in market cap, NBIS is not being treated like a niche hosting business. It’s being valued like a scaled AI infrastructure platform in the making. The headline multiples reinforce that: a ~99.8 P/E and ~4.8 P/B are not “cheap.” They’re the market pre-paying for growth, and paying up for the idea that Nebius becomes strategically important to customers who need capacity now.
The bull case is that AI compute is not a normal cyclical demand curve. When customers are racing to build, they don’t optimize for pennies per GPU-hour - they optimize for availability. If Nebius keeps selling out capacity and expanding, the multiple can stay elevated longer than skeptics expect.
The bear case is just as clean: if big customers decide to internalize more compute, or if the supply side loosens faster than expected, pricing power fades and “growth at any price” names get repriced quickly.
Trade plan (actionable)
This is a mid term (45 trading days) swing trade. Why that horizon? Because NBIS needs time for either (1) a technical repair back above the $96.26 20-day and toward the $100.83 10-day, or (2) a catalyst-driven repricing. A 2-3 day window is too tight for a stock with this kind of headline risk and intraday range.
- Direction: Long
- Entry: $94.46
- Stop loss: $90.90
- Target: $112.00
How I’d manage it: If NBIS can reclaim and hold above the $96.26 20-day average, that’s a basic “trend repair” signal. A push back toward the $100-$101 area is the next test (the 10-day SMA at $100.83). If the stock fails repeatedly under that zone, I’d tighten risk rather than “hope.” On the flip side, if it breaks $91.37 (the most recent session low) and can’t bounce quickly, the trade thesis is wrong and the stop should do its job.
What could push NBIS higher (catalysts)
- Capacity wins or expansion updates. Any credible signal that GPU clusters are scaling smoothly tends to matter more than general AI talk.
- Big-tech demand read-through. The market has been rewarding AI infrastructure when large platform companies signal ongoing enterprise AI spending.
- Short-covering fuel. With 37.8M shares short and 3.89 days to cover, a strong up week can force reluctant buying.
- Technical reclaim of moving averages. A sustained move back above $96-$101 can pull in systematic buyers that currently see “bearish momentum.”
Risks and counterarguments (don’t skip this)
The cleanest way to lose money in NBIS is to pretend it’s a stable compounder. It’s not. It’s a high-expectation infrastructure ramp story, and the stock trades like it.
- Headline and policy risk. NBIS has already shown sensitivity to macro-political headlines, including tariffs tied to European nations. More policy noise can hit multiples regardless of execution.
- Momentum is currently bearish. The MACD setup is negative, which means rallies can fail quickly and turn into lower highs. This is why the stop matters.
- Valuation gives little margin for error. At a ~99.8 P/E and ~4.8 P/B, the market is assuming meaningful future earnings power. If growth disappoints, the compression can be violent.
- Competition is structural. Nebius competes not just with other specialized AI cloud providers, but also with hyperscalers and large tech firms building their own infrastructure. That can cap pricing power over time.
- Crowded retail attention can amplify swings. NBIS appearing in popular retail stock lists is a double-edged sword: it can add demand, but it can also intensify air pockets when sentiment flips.
Counterargument to the long: The recent break from the $141.10 52-week high suggests the market may be de-rating the entire “AI capacity” complex. If the regime shifts from “growth at any cost” to “show me durable profitability,” NBIS could stay heavy for months even if the business keeps growing. In that world, the better trade is patience, not catching rebounds.
Conclusion: Follow the capacity, but trade the level
I like NBIS here as a defined-risk long because the stock is sitting on an obvious decision point around $94-$95 with strong liquidity, a cooled RSI, and a real possibility of short-covering if the tape improves. The market may debate narratives, but capacity businesses ultimately trade on one thing: can they deliver more supply into urgent demand.
What would change my mind: A clean break and failure to reclaim the low $91-$92 area would tell me support didn’t hold and sellers are still in control. Separately, if NBIS rallies but repeatedly fails under the $100-$101 zone (near the $100.83 10-day SMA), I’d treat that as a warning that the bounce is just a bounce, not a trend restart.
Until then, I’m comfortable taking the trade with the stop doing the heavy lifting.