𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬

Institutional-grade research for serious investors. Covering $IREN since $4.80, and hunting the next under-the-radar multi-bagger.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

$NBIS: First Ever Deep Dive

It’s no secret that I’ve shared a fair bit of criticism toward $NBIS over the years. Yet until now, I’ve never actually published a full deep dive on the stock.

Well, that changes today.

We’ve just released our brand new report to X and Substack subscribers, covering $NBIS's credit situation, funding strategy and broader fundamental setup.

The central question behind the deep dive is whether the company has what it takes to weather the current financing storm, and at what cost, if any.

I’ll be the first to admit that the operational growth and sheer revenue ramp at $NBIS have been extremely impressive, and I’m not trying to take anything away from that.

That said, there are some cracks beginning to appear, particularly around the company’s funding strategy, which I went to great lengths dissecting throughout the report.

A core principle behind our research is making even finance-heavy topics easy to understand without dumbing them down or stripping away the intricacies where much of the alpha often lies.

For this deep dive, we went above and beyond on that front.

Across 6 in-depth chapters, we created more than 20 custom graphics to help readers follow the analysis all the way through, from Nebius’s convertible structure and funding costs to its lease exposure, capital requirements and broader credit setup.

I believe any investor in this space, whether you own $NBIS, $IREN or another AI infrastructure name, will get a lot of value from this report.

And to be clear, this is NOT a bearish report by any means.

I’m giving $NBIS a fair shot here and analysing the company from as objective a lens as possible.

After previously breaking down $CRWV's debt spiral and contrasting its model with $IREN's approach, I think this new $NBIS report may surprise some people.

I’m fully aware that most of you are probably $IREN bulls, but understanding the competitive landscape is imperative for anyone invested in this space.

With financing conditions and credit markets under increasing stress, I believe this sector is ripe for major disruption and potentially significant consolidation.

Understanding how today’s largest players are positioned for that environment is tremendously valuable for any serious investor.

With $NBIS earnings coming up in just a few days, I’ve also laid out some of the key areas investors should be watching, especially given the current macro and credit backdrop.

To the subscribers who have already read the new piece, I’d love to hear your take below.

I genuinely believe this is one of our most insightful releases to date

If you’re interested, feel free to check out the new deep dive in the pinned comments.

PS: We’re also currently in the middle of putting together our $IREN earnings breakdown, which I’m personally very excited about given all the new developments from the quarter.

That report will be out to all our subscribers in a few days. Cheers! ✌️

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@FransBakker9812 @unprofitablepnl I believe Vineland delays are until Q2 2027. As for macro / financing environment, I agree, they are in a great spot.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@cvel317 I'm simply clowning wallstreet analysts who had an awful 2027 capex estimate going into these earnings...

There are many reasons why to be upset about $NBIS's results, but I'm sorry, if you expected 2027 capex to come in at $29b, you were sleeping at the wheel.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

Anybody selling $NBIS today because of its 2027 capex guidance is an idiot...

Management guided for $40B-$45B in capex, while the Street was expecting around $29B.

But those analyst estimates were terrible to begin with.

In the $NBIS deep dive we published earlier this week, I estimated capex at roughly $50.5B in the “central” case.

And to be fair, the actual guidance came in about 15% below that, likely because of the roadblocks the company is starting to encounter with third-party developers, or because some of its self-built capacity is progressing slower than expected.

In any case, I hope this teaches people to stop blindly relying on analysts’ expectations.

If you actually read some of the research notes they put out, you’ll quickly realize that many of them have no clue...

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@cvel317 @ScalpinRoc What does any of this have to do with $NBIS 😭

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@rafal_robert Dude... All of this may be true, so what? What's your point?

I was simply highlighting the fact that the street's capex estimates where awful to begin with and that the $40-$45b guidance shouldn't surprise anyone.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@Eltotti1984 All in amigo

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

Is $NVDA doomed without Jensen?

With Jensen potentially losing his CEO position at $NVDA, the main question for investors is what happens to the company and its stock going forward.

I think $AAPL is an interesting case study for both sides of the bull / bear argument.

When Steve Jobs was originally ousted as CEO, $AAPL didn’t do too well.

Yet when he was succeeded by Tim Cook in 2011, $AAPL continued to thrive all the way through today.

And yes, if Steve Jobs were still alive and leading $AAPL, the company might be an entirely different beast. But I think we can all agree that the stock itself has done fantastically over the past 15 years.

It’s also worth pointing out that even the first time around, when Steve got kicked out of his own company, $AAPL stock continued to perform extremely well for the first 6 years.

In fact, it went up by roughly ~7x during those initial years.

There is no doubt that Jensen is a massive asset to $NVDA, and what the company has been able to pull off under his leadership is nothing short of amazing.

But I believe the key-man risk to the company isn’t as high as it was a few years ago.

$NVDA is in the most dominant position it has ever been, and I think it would be difficult for Jensen’s potential successor to mess that up.

Assuming $NVDA will go to ruins without Jensen as CEO takes away from the amazing business culture he created.

I also don’t think Jensen leaving $NVDA is comparable to $TSLA losing Elon. Not because of what each person brings to the table, but because of valuation.

$TSLA trades at a forward P/E of ~193, while $NVDA's is just ~11…

$TSLA trades at an “Elon multiple,” based on hopes and dreams that Elon might be able to fulfil one day.

$NVDA trades as if earnings will fall off a cliff tomorrow, while its projected earnings growth for next year is actually around 70%.

So at this valuation, and given the state of $NVDA today, I’d say this is an incredible buy-the-dip opportunity.

[Serenity @aleabitoreddit]

I disagree, Jensen Huang is practically irreplaceable from $NVDA.

And it's like if you took out Elon from $TSLA, the valuation and confidence in the company would collapse.

The foresight to bet on CUDA before AI took off, build the Neocloud ecosystem...

The move to secure critical supply chain capacity years ahead of demand, and sell rack-scale systems (with CPUs, networking, etc).

Turn the ASIC threat into another opportunity by investing in $MRVL, MediaTek, to compete against $AVGO. Bringing them all into Nvidia's NVLink ecosystem...

Is absolutely insane foresight. Jensen practically architected both the financial and industrial ecosystem surrounding the current AI trade.

Jensen took Nvidia from a startup to a $5T+ company, and the fact he managed to keep growing that with projected FY2028 70% constrained forward growth... at these levels.

Is equally as absurd.

I normally don't post negative things about companies... but even if you don't see material revenue or roadmap changes in the next 2 years.

Ousting the face of the AI trade and a technology visionary might spell disaster for Nvidia long term.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@bullabulla1004 CFO temporarily, but I think one of the front runners of a more permanent replacement could be Ian Buck, who helped develop Cuda & Nvidia’a transition towards AI

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@zoezkzk Yup, indeed

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@jiahanjimliu Exactly 💯 Without knowing TCV or even $/MW rate for that matter, it’s difficult to compare the two references in an apples to apples manner

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@jiahanjimliu Wow Jim, thank you so much! Really means a lot to me & my business partner who creates all the graphics used throughout our deep dives!

Very cool for you to post this 🙏🤝

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@_Sgr_A_Star @BitWhitman @bitcoinbutcher1 @PhadsEth @SCurveCapital Excellent post GP!

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@si_yokohama0101 Exactly, 10/10 comment

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

$CRWV's Q3 results are VERY bullish on the sector

Demand for AI compute is truly through the roof:

- $20M-$25M / MW pricing for 4-5 year contracts
- $40M / MW for short-term contracts
- Hopper & Ampere pricing rose 30-50% QoQ
- Very strong prepayment environment

Management also made clear that the GPU / asset-backed financing market remains robust.

My issue with $CRWV has never really been demand. It’s how the company funds the enormous amount of capacity it has already committed to deliver.

$CRWV signed a >$100B backlog BEFORE its cost of capital exploded at the corporate level.

During the Q4 2025 call in February, management signalled it was de-prioritizing customer prepayments in favor of better headline economics.

That means a large portion of the existing backlog was signed under a very different financing backdrop and likely without the kind of upfront customer funding CoreWeave is aggressively leaning on today.

So how does $CRWV fund all of that existing backlog?

A meaningful part of the answer now appears to be taking large prepayments on newer contracts, and using that liquidity to support the broader buildout.

In a perfect world, that’s simply smart capital management.

The problem is that $CRWV is also one of the least vertically integrated neoscalers, relying heavily on third-party data center operators to actually deliver much of its capacity.

That introduces another layer of execution risk.

If a major site gets delayed badly enough to breach CoreWeave’s delivery obligations, customers would be able to walk away from the affected capacity and demand their prepayments back.

If that upfront cash has already been deployed elsewhere in the buildout, things can get dicey very quickly.

Besides that, $CRWV has already ‘largely sold out’ its 2027 capacity. You probably don’t need me to tell you why I think that’s a major strategic blunder.

I'm really not a fan of the way $CRWV does business...

I much prefer $IREN's fully vertically integrated, bottom-up approach or $NBIS's more hybrid strategy.

That said, $CRWV's call made it very clear that the AI compute market remains in very strong shape.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@typical90sguy @ilzmcfly And Anthony highlighting non equity sources specifically as it relates to the $6b in unfunded capex:

“The additional $6 billion of planned expansion CapEx sits against $12 billion to $15 billion of sources before converts or equity.”

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

"My personal view is that this dislocation between the fundamentals in the market sentiment at the moment is unlikely to persist very long. We are prepared if it does. But there are a number of options that we can continue monetizing that 5 gigawatts, even in the absence of equity market recovery."

The “even in absence of equity market recovery” is what I was referring to

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@RichardNem1338 @ilzmcfly Well said

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

Yeah, and that funds roughly half the backlog (50% of backlog is billing by year end). For the remainder they'll have to issue new DDTL's and possibly more corporate-level debt down the line. = more interest to come for today's backlog.

Remember, $CRWV pre-contracts YEARS in advance. These fools are already contracting 2028 capacity... no way they have drawn the debt for those contracts yet.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

@MikeKuczborski @ZeekTyt exactly