NBIS
Nebius Group N.V. Class A
stockNASDAQ

Hey guys, It has been a while since my last post! Last month, I ran a contest on my YouTube channel: 'Pick 7 stocks and win $2,500.' We ended up with 750 participants and a total of 5,250 stock picks. I broke down the results in a new video today, but I figured you would like to see where retail sentiment actually sits right now: |**Rank**|**Ticker**|**Points**| |:-|:-|:-| |1|SOFI|317| |2|APP|301| |3|AVGO|281| |4|AMZN|224| |5|UBER|222| |6|NVDA|202| |7|CELH|185| |8|META|172| |9|MU|153| |10|NFLX|146| |11|MELI|128| |12|GOOGL|116| |13|AMD|115| |14|NBIS|106| |15|RKLB|97| |16|NU|94| |17|NOW|87| |18|DUOL|82| |19|ADBE|79| |19|GRAB|79| Honestly, the community picks do not look terrible to me. * Do you think people are right this time? * Which one looks like the best value bet on the list? At this point, I really do not care about chasing views or subscriber counts. My channel is doing just fine, and I would honestly rather keep the real discussion right here. But if you are curious to see the breakdown with my italian accent, you can check out the video here: [I Asked 750 Investors for 7 Stocks (Here’s What they Picked)](https://youtu.be/eh61xUzx6mw) Peace ✌️

NBIS !!!!!! Just bought 150 shares on Friday

except in 3years from now NBIS may go bankrupt lol

NBIS can and will come back 20% in no time though. Goog will take 3 years to move up 20%. You're essentially just capitulating at or near the bottom

Down 20% on NBIS. Thinking about realising the loss and rotating into GOOG lmao. If the bubble pops NBIS will get annihilated but GOOG ain’t going anywhere

NBIS should be fine, couple good earnings and decent macro and this shit is back to 200, ASTS and RKLB you will baghold for some time but again should be fine

The IV on those NBIS calls lol

I don’t think our timeframes differ particularly. Where I disagree is the implication that “Microsoft needs compute today” somehow gives NBIS/IREN a comfortable 3–5 year runway. You’re analysing the demand side and barely considering the supply response. We’re not talking about Nebius and IREN operating in a vacuum. There are shedloads of neoclouds, the existing hyperscalers are spending absurd amounts themselves, infrastructure capital is piling in, and you’ve omitted probably the most awkward entrant of all: SpaceX/xAI. SpaceX already has Colossus at 220,000+ Nvidia GPUs, is supplying compute to Anthropic, explicitly says it intends to serve external cloud customers, and has just secured roughly $40bn of financing for next-generation Nvidia GPUs. And unlike a pure-play neocloud, SpaceX has Starlink and launch revenues underneath the financing. The COMPANY is already profitable. That’s the competitive problem. I wouldn’t say with certainty that SpaceX “cripples” NBIS or IREN on price, because nobody knows future pricing. But it plainly has the scale, vertically integrated infrastructure, diversified cash generation and financing capacity to price extraordinarily aggressively if it wants market share. And that’s before Oracle, CoreWeave, the hyperscalers themselves, Nscale, the miners converting power positions, infrastructure funds etc. So yes, I completely understand why Microsoft rents incremental capacity TODAY. There’s a shortage TODAY. That’s literally my thesis. What I don’t understand is why people take: **“**Microsoft cannot source enough compute today” and turn it into: **“**therefore Nebius/IREN have attractive economics for the next five years**.”** The former causes the latter only for as long as supply remains constrained. And your own 3–5 year site-build point cuts both ways. Five years is a very long time for tens/hundreds of billions of dollars of competing capacity to come online. The whole industry is currently responding to exactly the price signal you’re capitalising. This is basic commodity-cycle stuff: shortage goes to extraordinary returns goes to capital floods in goes to supply expands goes to bargaining power shifts goes to returns compress. The fact that new capacity takes years to build doesn’t invalidate that cycle. It’s WHY the shortage can produce exceptional rents before the supply response arrives. NBIS may absolutely make a fortune during that window. IREN may too. I’ve never disputed that. I’m disputing the leap from “these businesses are monetising a historic bottleneck very effectively” to “therefore they are structurally attractive businesses.” And this is where “NBIS has better financing than CoreWeave” still doesn’t solve anything for me. If SpaceX can finance $40bn of GPUs against Starlink/launch cashflows, hyperscalers can fund their own buildouts from enormous existing businesses, and infrastructure capital keeps adding competing supply, then the relevant question isn’t: **“**Can Nebius get financing?**”** Clearly it can atm. It’s: “What sustainable economic rent does Nebius retain once all this competing capital actually turns into compute?” That is still unanswered. And if the answer to the 5-year thesis is essentially “supply won’t catch demand quickly enough”, then call it what it is: *a very attractive scarcity trade***.** That’s completely different from a durable compounder. My objection has always been to people confusing the first with the second.

Remember me when NBIS reaches $700+ by 2031

NBIS

Small nitpick, but I think it changes which of these are actually interesting: most of these numbers are distance from the 52-week high, which tells you how much air came out, not whether it's cheap now. A test I like better: is the stock trading below the average price people paid *after* the last earnings report came out? That's the market's own estimate once all the new info was public. Using Friday's close: - KLAC: ~48% off its high, but it reported Oct 28 and is only ~2% under the average price since then (just two sessions, so early). The big drop happened before the report; on the new numbers the market isn't marking it down much. - AMAT: ~47% off, ~15% under its average since its Aug filing. - MELI: ~33% off, ~9% under its average since its last major filing in September. - UBER: ~35% off, ~9% under. - CAT: reported Thursday, so there isn't a settled post-earnings price yet. - NBIS: the drawdown is real, but trailing free cash flow is about -5.9B, so "discount" depends entirely on what you think the buildout earns later. Not advice, just a different way to sort the list.

I don’t actually disagree with the first part. NBIS could do extremely well over the next few years. That’s not my objection. The current environment is basically perfect for a competent neocloud: compute remains supply constrained, demand is enormous, customers are willing to prepay substantial amounts to secure capacity, utilisation is high and Nebius is currently able to finance expansion on relatively attractive terms. Nebius is also clearly better positioned financially than CoreWeave. I’ve never argued otherwise. But “it’ll do well as long as funding and execution stay on track” is almost the point I’m making. The long-term question is what happens when the economics are no longer being supported simultaneously by: scarce compute; excellent utilisation; customers willing to pre-fund capex; cheap/available external financing; and rapidly increasing demand absorbing every new GPU that comes online. Those are favourable conditions. They are not a moat. Take NBIS specifically. Q2 revenue was $582m and adjusted EBITDA was excellent. But it spent roughly $5.7bn of capex in the same quarter, mainly on GPUs and data-centre expansion. D&A was $260m — about 45% of revenue. And Nebius itself says the revenue growth was driven primarily by capacity scaling. That’s why “multi-year billion-dollar contracts” don’t answer my objection either. They de-risk near-term utilisation and make the underlying infrastructure easier to finance. Fantastic. But I’m asking what return on invested capital Nebius earns across the full hardware/financing cycle once capacity becomes less scarce and customers regain negotiating leverage. Those hyperscaler contracts are especially interesting because the customers underwriting the current growth are themselves some of the businesses most capable of eventually routing around the neocloud layer. And yes, NBIS has much better financing access than CRWV. It literally secured another $3.5bn senior secured facility last week at SOFR + 275bps initially. But look at what the money is for: GPU procurement and data-centre buildout. Again. So “NBIS has better financing than CoreWeave” answers: Which neocloud is better positioned to fund the treadmill? It does not answer: Why is the treadmill itself an attractive long-term place for shareholder capital? That distinction is basically my entire thesis. And the customer-prepayment economics reinforce it. Nebius says it expects more than $9bn of prepayments this year and that some Q2 deals covered 50–60% of associated capex upfront. That is outstanding financing in a supply-constrained market. But I’m not willing to assume customers will permanently finance half of Nebius’s capex once they have abundant alternative capacity. Maybe they will. Prove it through a cycle. The thing that COULD change my view is Nebius’s asset-light/platform model — if third parties increasingly fund the GPUs/data centres while Nebius retains the customer, software and high-margin economics. But that remains precisely the thing that hasn’t been demonstrated financially yet. And extraordinary AI demand can make a structurally mediocre layer look economically fantastic during a shortage. The investment question is who still owns the economics when the shortage ends. Nvidia owns the scarce silicon. The hyperscalers own enormous customer distribution, internally generated cash flows and increasingly their own silicon. Nebius currently sits between them, buying/financing the scarce asset and competing to sell access to it. Maybe its software/platform eventually changes that. Until it does, “funding and execution stay on track” is not the answer to my bear case on NBIS and the neoclouds more generally.

I think they'll be good for the next 5 ish years, the current AI infrastructure build out is going to generate a lot of income for some of these business. The financing across the sector is concerning, but NBIS and IREN are far more robust in that department than others like CoreWeave which has massive debt. Many smaller "neoclouds" will fall by the wayside, but I think those two will do quite well in the short to medium term so long as funding and execution stay on track. Both companies are expanding globally into new regions and already have multi year billion dollar contracts with major hyperscalers and AI labs. IREN earnings next week and NBIS the week after, it'll be interesting to see how they're both doing after spending the last two months getting absolutely hammered in the stock market and sentiment hitting rock bottom.

NBIS 😂

More than NBIS and RDDT? Now that’s comical.

Is the plan to sell the NBIS calls before earnings?

NBIS and the other neoclouds are not good long term investments imo.

Calls on this dip. $NBIS $DRAM $HOOD

NBIS at a 49% discount? You mean 49% off the ATH. Is it discounted from intrinsic value at mkt cap of 42 billion? Well it provides infrastructure and services to AI companies worldwide. Sounds promising. It is being forecast to have about 30 billion revenue by 2030. A CAGR of just over 32% from 2026 numbers. Tough ask but let's assume they can do it with great certainty. Some forecasts stretch to 90 billion. Let's call the first the base and bear case with 95% probability, and the latter the bull with 5%. Because to get to the latter in that time requires CAGR of revenue over 130%. Neoclouds like Nebius are predicted to have operating profit margins of 25% (base bear) to 35% (bull) in the long term, but let's say they get there by 2030. So base bear is 95% chance of 7.5 billion operating profit, versus bull 5% chance of 30 billion. Overall likely 10 billion let's say. Net profit margin is likely to be 15% of operating profits are so high so net profit would be 6 billion in 2030. And if they started returning capital straight away they could do say 50% return, so buy backs could be 3 billion. Or about 7% yield. So, you can buy treasuries at just under 5% and start getting that yield today. Not have to wait 4 years for a concatenation of unlikely events (consistently grow revenue at never before seen rates, achieve long term profit margins in the short term, return capital aggressively immediately even as likely still growing, and no dilution from the expected 30 billion capex required to get there). So is there a relative undervaluation that causes one to fund a purchase of Nebius? No. Buying here likely only yields a little more than 10 year Treasuries without any of the certainty, and certainly without any cash back in the next 3 years. What about capital appreciation? Well maybe they double back to ATH. Is that the market feeling at the moment, no but it is a possibility. Maybe they halve again to early 2026 levels. That seems a possibility too. So that's a wash. So you're left to compare with Treasuries. Good luck with that first place.

Nebius? That garbage is a dead duck. Why buy some over-debt ridden flea-bag smaller-cap startup when the market and macro's are in the state they are in today. You don't happen to be waiting to offload NBIS, do you>

Classic WSB portfolio 😂 Relax, NBIS is a good company and should recover. ASTS and RKLB are memes though. I do think the space sector will continue to have violent booms and busts. I would not buy them again and just drop them as soon as they recover. The era of making easy money with meme stocks is pretty much over for now. Gotta wait for the next time it happens. (Could still come back soon) Unironically i like TLT here lol

UBER is now a big position in my pf @70 Consudering SOFi, RDDT, NBIS

**FIG** – 84% (added to ) (***second*** place) **LYFT** – 81% (New position) **NKE** – 80% **SEA**– 80% ( new position) **RDDT (Reddit)** – 66% ( trimmed significantly ) **TTD** \-66% (new position) • **NBIS** – 49% (added significantly, ***First place)*** • **RKLB** – 64% (considering) • **ORCL** – 60% • **INTC** – 54% • **NFLX** – 53% • **WDC** – 53% (new position) • **SNDK** – 51% • **CIEN** – 48% • **SOFI** – 48% • **KLAC** – 48% • **AMAT** – 46% • **LRCX** – 45% • **HOOD** – 45% • **ITR** \- 42% (new position) • **CAT** – 41% (new position) • **MELI** – 39% • **DASH** – 37% • **UBER** – 35% (considering)

AVGO, NVDA, ORCL, GOOG, META, NBIS, RKLB, AMZN.

Buy NBIS until it hits $500