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/

$NBIS I agree with @InvestNorthwise, as we were among the most bullish ones regarding the Nebius stock price and, more importantly, ARR/revenue estimates. The analysts are not paying enough attention to what Nebius says, and they are for some reason, modelling everything with numbers that are not correct in the current supply-constrained environment. While street consensus is $16B in ARR, my base case is $35B. While many argued about this number being way too high, it is a simple combination of two metrics that Nebius needs to achieve: 1) 1700MW of active power by the end of 2027 To be honest, I don’t see any issues with that, as currently confirmed is roughly 1550MW and with the speed of AI adoption in Europe, we will most likely see additions to this number and we already have clues for several sites across Europe and also Asia. 2) $21M/MW blended revenue This sounded unrealistic to many when I first published my valuation model, but with Nebius confirming 1GW of Vera Rubin to be deployed during 2027 and many of these MW already active by the end of the year, it will not be that difficult for Nebius to achieve that. Short term deals will be signed for $50M per MW in 2027, while medium-term deals will be around $25M per MW. More than 500MW of active power will be signed for either one of these types, and if you do the math, we are around $20M/MW already. For clarity, I said this multiple times: Nebius won’t guide for $35B, but I believe they will achieve it in 2027. $641 is the price target for EOY27

$NBIS 1GW of Vera Rubin in 2027 Huh? We knew that Nebius wants to add >1 GW of connected power each year starting in 2027, but we definitely didn’t know that there would be at least 1 GW of VR GPUs going online... Just so you understand the implications (as of today): Pricing: at least $40-$50M/MW CapEx: roughly $50M/MW (including all costs of the buildout) 1GW of VRs will add roughly $50B of revenue for Nebius...

$NBIS Just one hyperscaler asking another hyperscaler Nothing unusual

$HIMS Both injectable and oral testosterone are now available through the HIMS plaftorm!

$NBIS $IREN Agents used 34x more tokens today than at the start of Feb You can see that in August, we started going vertical and there seems to be no slowing down Tokens being cheaper causes them to be also more valuable, which means that people might suddenly see reasonable ROIC. This is extremely healthy for the long-term AI adoption

$NBIS Startup? Hyperscaler, you mean?

$NBIS $IREN It wasn’t that hard to predict, and it shows how stupid and inefficient the market currently is... What did everyone expect? That Elon will say: "Yeah, we don’t see massive demand anymore"?

$NBIS $IREN There is a lot of quotes floating around on X right now regarding what was said during $SPCX’s earnings call. I don’t really care about Elon saying that ARR will be next year $500B - $1T and that they plan to have 13GW of power by the end of 2027. What is important is the fact that there is still significant demand and nothing changed since Q2 earnings, which means that we can expect strong Q3 earnings from other AI companies as well. We are still far from a bursting bubble…

Huh? $SPCX “13GW of compute by the end of next year” But they told me we are overbuilding Elon!!! $NBIS $IREN

Investing in $NBIS finally feels like going against the masses, which is an important indicator that there is still room to go You can have your price targets, your earnings estimates, anything you want, but as long as the sentiment is extremely bullish, there is usually a limited short-term upside. Tables have turned, and I still stand by my price targets: $300 EOY26 $641 EOY27 People just don’t get it... that $16B ARR estimate was a confirmation for me

$NBIS $IREN $SPCX already has positive income from operations 5.8% when in full build mode... Just so we adress the discussion that this business can never be profitable... sure, sure

$SCPX Massive beats! All we need is Elon to confirm that compute is still in high-demand territory, and we can start the reversal on $NBIS and $IREN

$SPCX We could theoretically see the narrative shift in the neocloud sector ($NBIS, $IREN, $CRWV) today already. Elon knows how to pump his stock, and he will definitely talk about the development of AI compute demand. Remember what he dropped last earnings call? He was the first one to hint that VR could be priced around $40-$50M/MW. We will see in 15 minutes!

$NBIS $IREN This is what happens when someone who has no clue about the topic, starts talking… I know that @edzitron would see something bearish in literally any piece of information just so he can promote his AI doomerism, but these takes just make him look dumb. These GPUs are in warehouses and can’t be installed because other components are not available… Every single GPU that can be installed is installed and monetized immediately (confirmed by Amazon)

$NBIS $IREN Absolutely no impact for Nebius or IREN I know that Amit is trying to find some reasons why we falling like this, but this is definitely not it. If something, it would be the whole KOSPI going down

$HIMS It is just beginning… So excited for HIMS, everything is just starting but in 5-10 years we could be looking at $1T company

$HIMS Official meeting with White House’s Office of Information and Regulatory Affairs Peptides soon…

$NBIS $IREN If Europe wakes up and finally starts investing into the infrastructure, it should be priority for Nebius and IREN to capture the market. There is very little competition compared to the TAM. Germany was always a very strong economy, but the country’s development over last 10 years is truly sad. Hopefully, they will wake up and AI would be first logical step…

$NBIS Literally impossible Current cash/debt position is roughly net equal And the annual interest is nothing compared to the value of owned assets

$NBIS Spot pricing - initial reaction Just for people who tend to draw conclusions early... do not panic or celebrate this until we get at least 2-3 months of data. Prices look extremely volatile in the first two weeks, which means that it needs to run for a little longer before we can say something definite From the fluctuations, you can tell that spotting the correct PVM price by setting a fixed price is nearly impossible; we can see that immediately