[0:00]Hello fellow investors and welcome back to Borex Investing. This channel is all about the high conviction AI and tech stocks I personally invest in. So if you follow names like Palantir, Nebius, or Nvidia, you are in the right place. All right, so let's start this video off with a question: Can Nebius realistically hit $200 this year? Because after their latest earnings report, the narrative is getting louder. Institutions are building positions, new price targets are coming out, momentum is shifting. But price targets are easy to tweak. What's not easy is doing the math to explain the numbers. I already broke down the Q4 earnings in my last video. It was a very strong report. They executed, they reiterated the 7 to 9 billion in ARR framework for 2026, and they showed they are scaling aggressively. But today I want to go a little bit deeper. If you, like me and many others, see $200 as a realistic price target, well then what actually has to happen for that to make sense? How much revenue, what multiple, how much debt, what about dilution? Because the market doesn't pay for hype, it pays for forward earnings power. So in this video, I'm walking you through my model, the revenue scenarios, the enterprise value multiples, capital structure assumptions, and I'm going to show you exactly what combination of execution gets us there. Let's get to it.
[1:41]Okay, so something important is happening around Nebius right now. Institutional ownership is increasing meaningfully. BlackRock just disclosed a position of roughly 9.5 million shares. Their prior filing showed just under 24,000 shares, so this is a massive increase. When a firm of that size moves from a token position to a multi-million share stake, it suggests Nebius is starting to be evaluated differently at the institutional level. On top of that, Compass Point initiated coverage with a buy rating and a $150 price target, implying roughly 54% upside at the report date. So the conversation around Nebius is clearly shifting.
[2:27]But instead of reacting to headlines or price targets, I wanted to step back and ask more important question. Where can the stock realistically go based on the numbers? To answer that properly, we can't focus on 2026. The market is forward-looking. 2026 is a scaling year. There's heavy CapEx, there's infrastructure build out, there's execution. The real earnings power begins to show when you look at 2027 numbers. And this is the framework I'm using to evaluate the opportunity. If Nebius delivers 7 to 9 billion in ARR by the end of 2026, what really matters isn't the full 2026 number. It's the exit run rate. What does December 2026 revenue annualize to? Because that's what the market will start pricing. And that's why I'm using run rate revenue here. High growth infrastructure businesses are typically valued on forward earnings power, not the transitional year while they're scaling. Now, we also need to understand our Kadivalage and the management team here. Nebius isn't Silicon Valley hype culture. The management team here is extremely measured, they're very practical, they don't want to overpromise. So based on that, I built five revenue scenarios that I think are realistic. Starting at 7 billion and going up to 11 billion in 2027 annualize revenue. Just in case they materially over deliver. So from there, I had to address debt. We know CapEx guidance for 2026 is up to 20 billion. About 60% is being funded internally, which means roughly 40% or around 8 billion still needs financing. Now, I'm also assuming that if they hit the higher revenue scenarios, that likely required more infrastructure, which probably means more debt. So in these different scenarios in the model, I'm assuming around 8 to 10 billion in lingering total debt by 2027. Then we get to multiples, and this is where forward-looking variables drive the valuation. I used Enterprise Value to Sales multiples of 4x, 6x, and 8x. The reason I framed it this way is because the multiple is really a reflection of execution risk. If the market still doubts them, maybe it's 4x. If execution continues and risk compresses, maybe it's 6x. If they're viewed as a premium AI infrastructure platform, maybe it's 8x. Now, for context, companies like Snowflake have traded at significantly higher forward sales multiples during growth cycles. They are currently at 14x. But hyperscaler style businesses like Amazon's cloud segment have historically supported the mid-single digit multiple depending on growth and margins. So for me, I think 4x to 8x feels like a rational band. So you know, personally, I'm leaning towards the 6x band, the medium blue row, because I think that's a fair and realistic multiple if they continue executing. Then we have dilution. On the latest earnings call, management did not indicate any immediate plans for equity issuance. But we have to leave the door open for it in the future. So I modeled three cases: no dilution, 5% dilution, and 10% dilution. Under my realistic base case, around 9 billion in 2027 revenue, a 6x multiple, 8 billion in debt, and no dilution, I get very close to that $200 range price target with the math showing about $182. Now, of course, this model does not include potential monetization of assets like ClickHouse or Avride. I'm treating those as strategic holdings rather than liquidity events. You could argue they add additional value, maybe 8 billion combined depending on how you value them, but I didn't really bake that into the price target. I wanted to stay conservative here. And by the way, I've seen more aggressive price targets floating around, 300, 400, and I'm not saying these are impossible long-term by late 2027 or beyond, maybe. But based on assumptions I'm comfortable making today, the math supports something closer to the $200 range. Now there are plenty of variables that could change this. I mean, a new hyperscaler deal could materially increase revenue. But it also increases debt in the short term because more capacity would need to be built, so there's always a balance. But really what this model gives us with scenario 3 at a 6x multiple and no dilution, that seems like a realistic goal to strive for. So that's how I personally got my $200 price target. And remember, yes, you do have to price in a bit of premium as this is a high beta name that gets investors very excited when it takes off. It's not really a Palantir-like premium or multiple, but who knows, maybe we'll get there eventually.
[7:40]And I have seen other models floating around online where people try to price Nebius based on capacity, meaning they look at how many data centers are going live, they estimate the megawatts deployed, they multiply that by assume GPU density, and then they project revenue from there. And honestly, that can be a very powerful way to model this business. In theory, pricing capacity is probably the most infrastructure native way to value a company like Nebius. But here's my issue with that. Right now we don't have enough visibility. We don't know the exact timeline for when all nine new data centers will go live. We don't know how much capacity each one will ultimately support. We also don't know if they'll phase in contracted power gradually or ramp up in larger blocks. And we definitely don't have clarity on GPU delivery timing, Nvidia supply constraints, or how aggressively customers will scale into that capacity. So while a capacity-based model might be more precise long term, today it just would require layering assumption on top of assumption. You'd be guessing on build timelines, GPU availability, utilization ramp, pricing per unit of compute, contract structures. There's a lot of moving parts. So for now, I prefer this for more pragmatic approach. You take the December 2026 run rate revenue, you annualize it into 2027, you assume no dramatic changes to deal solution or capital structure beyond what we already know, and then apply a rational multiple. It's not perfect, but it's grounded in numbers we actually have visibility on. As more data centers go live and we get clearer disclosures around capacity and utilization, we can absolutely evolve the model into something more granular, but today I think this is the cleaner and more defensible framework. Now, let me say something important here: this is not an easy company to value. Really, in fact, the entire Neo cloud space is extremely difficult to model. You're dealing with massive revenue growth with heavy infrastructure build out, with front-loaded CapEx, with leverage decisions, with GPU supply dynamics, it's very complex. And its complexity tends to scare off a lot of traditional analysts. But everyone can see the growth, the revenue ramp is obvious. It's just the CapEx, the debt, and the dilution questions make the valuation kind of messy. And I think when valuations get messy, the market tends to move the entire sector together and, you know, we've really seen this repeatedly. Negative headlines around Oracle or CoreWeave, well, that means the whole AI infrastructure space sells off. Bad news for a competitor becomes a red day for Nebius, even when Nebius hasn't delivered any negative updates itself. And if you think about that, it's a bit strange. I mean, imagine if Nike had a weak quarter and because of that, New Balance or A6 suddenly dropped 10% in sympathy. That wouldn't make much sense, right? Different companies, different execution. And in fact, I'm reading Nike is struggling now because of New Balance, A6, and Onrunning having record high quarters that are stealing their market share. But of course, in AI infrastructure right now, that's the complicated environment we have. The market is still treating this as a thematic trade rather than separating winners from laggards. And that creates volatility, but it also creates opportunity if you believe execution will eventually differentiate. And not only that, but there is also a lot of external pressure here. There's the obvious endless flow of geopolitical issues we have now. There is also a lot of AI FUD coming from the media and from haters like Michael Bury who try to create panic while they short. And we know when macro uncertainty combines with sector-wide skepticism, multiples compress even for companies that are executing well. In my view, that's also a part of why Nebius is trading at a discount right now to what the forward revenue trajectory suggests. But remember, those pressures are cyclical, they are not structural. At some point in 2026, geopolitical tensions will stabilize, monetary policy will be clearer, and the market will shift back towards fundamentals. When that happens, companies that are executing, that are delivering revenue, scaling capacity, maintaining discipline, tend to re-rate higher. And that's the environment I think Nebius eventually benefits from. So I want to ask you, what do you think of this model presented today? Am I being too conservative? Am I being too aggressive? What assumptions would you change? Because at the end of the day, valuation is about assumptions, and I want to pressure test mine. So please let me know in the comments. And please remember to like and subscribe if you found this video informative in any way. Thank you.
