Today is the day I join the Substack family, and I have two main reasons: 1) Substack offers better options for sharing longer and more complex work. I want to share, for example, my Excel files, where I have all my calculations, and that works better on Substack than through X articles 2) Feedback from my followers - I was surprised, but many people already asked me if I wanted to create an account on Substack. I assume it is easier for people to read and share than using X. Please understand that all my articles, deep dives, valuation models, and further analysis will be free for everyone. I am new to Substack, and I must say the fact that the first thing it helps you with is setting up subscriptions is comical. I have some preset settings for subscriptions, but you don’t need to subscribe, as all posts will be open to everyone! The only limitation I actually set up myself is that commenting is open only to subscribers (the free tier is sufficient) to avoid annoying people. Feel free to share this so people can follow. Already uploaded my updated Nebius Thesis, and tomorrow I will release the margin impact analysis regarding Token Factory’s new joiner, Inferize. Thanks for all the support! funmentalist.substack.com/

We might have an ugly week again, you never know in this market. So let me give you some advice as someone who had everything going for him and suddenly lost most of it. I had to rebuild everything from scratch while suffering from terrible feelings of never being able to be happy again. Don’t be afraid to lose. Everyone who is truly successful, once lost something that meant a world to him… If you don’t lose money that you need to survive or someone you love, there is always a way to get everything back and not only that, even more. You have to lose in life and remember: The biggest risk is not taking any risks! See you tomorrow

$NBIS I’ll be honest, but after this week I am really upset with what is happening on X. Nothing against @brent_e_trader, first post that I see from him so he might be different, but I see so many negative people on X right now and it must feel terrible for people that are actually trying to get valuable insights from here. So many posts, that Nebius and other companies are done and that they will go way lower. Do you know what’s worst? That not a single argument/number or god forbid valuation explanation is provided by these people who claim that Nebius is done. All I read now is: “Nebius is dead” “You will lose all your money” “Look at the debt!!!” Not a single FCF model that would show why might the business not work, why the debt is a problem, nothing… I have done all the calculations myself, so I am not bothered by any of this, but I feel in the comments under my posts that some investors are genuinely worried. I don’t want to defend them as they should be responsible for their investments and they should have their own conviction, but man… so many posts also feel like the people are actually happy that others are losing money. Genuinely sad…

$IREN Disrespectful to shareholders to be fair… I hold $IREN for the immense potential upside and I was never hiding the fact that I see many red flags in the management’s ability to achieve their goals. I apologize to @danroberts0101 and his team, but there is no way these deals make any sense from the ROI perspective… Don’t use any “visibility” argument on me please. If you are in the compute sector right now when demand is through the roof and people don’t know your company, you are just a poor company and your product must be terrible. Come at me $IREN bulls, but I am really disappointed shareholder by these things and I think that any lack of execution in the actual compute side of the business will send $IREN back to single digits… Looking forward to Q3 earnings

This is the mistake that $CRWV and others did… Few understood ($NBIS, $IREN)

$NBIS Taking a step back after the drawdown What I always do for my investments when they experience such a dramatic fall every time is take a step back and look at everything again over the weekend with a cool head, so as not to mix emotions and facts. 1) Thesis First of all, I attach the updated thesis that I published at the end of September, and I think this is what you should go back to first, as it makes you remember all the verticals that the company is currently progressing on, and it shows you why I believed in the company in the first place (and still believe obviously)! I believe that nothing has changed since September 2026, and many would argue that things even improved. 2) CapEx/Debt debate A lot of things were said during this week regarding CoreWeave financials and their implications for the whole neocloud sector. I think it is fundamentally wrong to compare CoreWeave with Nebius and IREN, for example, because each company has a different strategy. In my opinion, the biggest problem for CoreWeave compared to these two is the fact that they sold out their whole 2027 capacity, and overall their deal profile is long-term contracts with terms that do not reflect the current prices. Combine this with their annual interest and potential difficulties to raise additional capital, and you have a problem. I absolutely disagree with statements that if Nebius were to succeed, CoreWeave must succeed as well. However, that’s not what I really care about today, as I want to discuss Nebius as my biggest position. (Good luck to CoreWeave, as it would definitely be better if they managed to get out of this negative perception!) So what is the current situation for Nebius? If we assume the high end of the 2026 CapEx guidance - $25B and look at the gap between what Nebius already spent on CapEx and what is left to spend (we don’t really know what happened since June 30, so we have to work with that): - CapEx spent in H1 2026 - $8.13B - CapEx required to meet guidance - $25B - 2026 Remaining CapEx (as of June 30) - $16.87B What is the cash situation? - Unrestricted cash on June 30 - $8.04B - July secured debt facility - $0.775B - August convertible financing - $5.75B This gives us approximately $14.5B in cash and financing resources, which creates roughly a $2.3B gap to cover the remaining CapEx. We also have to assume that Nebius wants to still finish the year with some unrestricted cash, so let’s assume the $8B similar to its June 30 position. This means that Nebius needs ~$10.3B of additional net cash inflow during H2, and let’s discuss where they could get this money before people start with the FUD that we will get diluted or Nebius will take on an additional $10B of debt (trust me, they will eventually, you have to account for that). 2.1) Customer prepayments Management expects more than $9B of upfront customer funding in 2026, but don’t forget that is for the entire year. If we look at the H1 cash-flow statement, we can see a $4.4B positive change in deferred revenue, so we don’t know for sure, but we can expect several billion dollars of additional customer funding arriving during H2, which helps significantly with this gap. 2.2) Operating cash generation H1 operating cash flow was $109M, and I believe that we can expect at least $400M of operating cash flow in the rest of 2026. Nothing significant, but still almost half a billion dollars. 2.3) Additional asset-backed financing People completely forget that this is actually the most important and plausible source of capital. Management said that more than $40B of additional customer commitments could support a similar financing structure to the $775M secured facility in July, where Nebius borrowed against deployed GPU infrastructure and contracted cash flows from investment-grade customers. Overall, to maintain that $8B cash buffer, I believe Nebius still needs to find approximately $5-$6B of further financing excluding operating cash flow and customer prepayments, which we will know more after Q3 earnings (second week of November most likely). Remember, $5-$6B of additional cash if they want to keep that financing buffer, but if they think that now is not the right time to seek capital, they are covered for 2026 despite what many tried to tell you during this week. So again, everyone panicking before double-checking the facts... 3) Mismatch between my expectations and the street’s consensus Probably something that makes me most bullish still, and I keep going back to this number is the extremely slow progression curve that the street is expecting. To be fair, I am not surprised that the consensus right now is that Nebius is not a great investment if they expect them to announce $16B in ARR by the end of 2027. I don’t know what the exact consensus on the active capacity at that time is, but with that estimate, I would be bearish as well, as the economics wouldn’t make much sense at that point, and this is where I believe the big re-rating will happen - Q3 earnings. I believe many saw the post from @SMASIMHO, where he showed the least bullish base case you can ever see (on purpose), but I will repeat it, so you don’t have to switch between posts: - 2026 ARR guidance is currently $7-$9B, and I will confidently say that we can assume the midpoint of $8B - Already confirmed capacity for 2027 is 800MW, and that will most certainly be around 1GW, and in my valuation model, I currently assume 1,700MW of active power by the end of 2027 You can probably understand by now that $16B in ARR is complete nonsense because that would imply -> $9.4M/MW, which is below the bare-metal deals around $11-$12M/MW I’ll just use my estimates of 1,700MW of active power and blended revenue per MW of $21.1M. First of all, I am already quite bullish in my base case, but we only see GPU prices increasing, and until I see actual cracks in the data, I won’t be lowering my expectations. For those who read my valuation model, I am way above $30B in exit 2027 ARR, but even if we assume my bear case of $25.8B, it would still be 61% higher than what the street is expecting - 61% My next point would be whether the deals are actually profitable and how much FCF is actually generated, but I am preparing a longer article on this topic because many people apparently can’t do math on X (spoiler: a lot of FCF for Nebius). So yeah, I am still calm and collected going into the next week and I would summarize it like this: - No signs of demand slowing from Nebius - Deals make sense economically - Sentiment shifted, which is normal - Thesis is still intact - Pumped for Q3 results Thanks for your attention!

$NBIS As long as people think about Nebius like this, I am happy to hold because the upside is still significant

I see a lot of people upset about me and other FinX influencers for saying that $NBIS was a great buy since $200 (or even higher) and now at $165. First of all, if you bought a stock simply because someone told you on X and you did 0 DD on the company. I hate to break it to you, but your chances of making serious money in the stock market are pretty low. Second, I am a long-term investor in $NBIS and I don’t expect to sell my position for several years, which means that if I buy Nebius at $200 and it goes lower, I’ll just simply buy more and won’t be upset about being down 20%. I can’t predict the market in the short-term and while there are indicators and TA works to a certain extent, it is still just psychology and if the majority of the market believes something structurally changed, they will sell/buy beyond what TA suggests. My valuation of Nebius suggests that we are far away from its current fair price and I’ll happily continue buying.

For complete transparency, which I promised to my followers, I ended the week down 15.3% You get used to it and I stand by my decision to be heavily allocated in $NBIS and $IREN I wouldn’t be able to look in the mirror if Nebius or IREN go way higher from these prices because all I see is, in my opinion, asymmetric opportunity

I had almost zero time today, so I had only time to buy more $NBIS Will write my thoughts tomorrow, but I really think that market is overreacting way too much Q3 earnings soon and I believe that management will AGAIN show everyone that we need compute and there is still extreme supply shortage This week everyone suddenly turned bearish AI trade, which is healthy long-term…

Lol I absolutely love people like this If the market goes up, this exact person will tell you that the market is irrational. Now, @daniel_koss provided a simple argument on his take and this person uses market as his argument because he actually has no idea what to answer… Comical

$IREN This is scary to hear as an IREN investor. Regardless of who’s fault it is, ANY issues with Microsoft as a client will punish the stock heavily. Not that my thesis is heavily connected to this bare-metal deal, but it is/was the only validation of IREN’s technology for the broader public. Will have to investigate this further!

$NBIS Beam Day 0 Should I even comment this anymore? Token Factory needs time and investors need to be patient, but it will be a huge business

$ZETA Upgraded to $45 by GS I studied a lot of material from @wealthmatica back in May/June and even asked him a few questions afterwards. The thesis was clear to me and it was one of those buys where you immediately get a big return Lot of people asking me if the stock makes sense after the latest run, but I believe it does (long-term). Back in May when price was around $20, I got $60 as a fair price given the growth and FCF $ZETA will soon generate. Management is fire, CEO loves the company, pretty simple long-term position if you want exposure outside the AI buildout and also have exposure to a company that benefits directly and significantly from AI itself.

$NBIS $IREN Ross apparently knows nothing about AI/GPU technology because he provides a one-sided argument that works only if the output price of tokens is linear, which it obviously isn’t, with Vera Rubin and others producing cheaper tokens than previous generations. However, he is right that at this point no one probably doubts the demand, and the main skepticism is around "who is going to finance all of this?" and "is there enough capital?" This is why the sell-off was so aggressive due to potential rate hikes (which I disagree with) and CoreWeave bond yields surging. If there are signs of capital not willing to finance the buildout anymore, we would have a structural problem. However, both Nebius and CoreWeave were able to raise a meaningful amount of money in the last 2 months, so I don’t think we are this far in the "bubble" cycle. It will definitely come one day because even among "neoclouds", I see companies that just want to take advantage of the current supply shortage, and their strategy/financing structure makes 0 sense. Only the strongest will prevail, but as long as we see Nebius/IREN increasing their prices and customers willing to pay a significant premium for short-term compute, there is no reason to panic

$NBIS "Demand is essentially modeled as infinite right now. Generally, for every contract that's about to expire, there's about 15-20 other customers waiting to bid for that." I think I don’t have to comment on this one

$NBIS Muse Spark Open on TF It was already leaked earlier today, but the best US open-weight model is on Token Factory Day 0 AGAIN

$HIMS I kept saying this since May Only a matter of time for HIMS Peptides are first vertical that will start spinning the wheel

$NBIS $IREN The market loves the B300

$NBIS And you are not adding in this pullback? Day 0 or 1 every single new model