CEOInterviews.AI
Start App
Dan Ives
Managing Director & Global Head of Technology Research, Wedbush Securities

Dan Ives Explains The Neocloud Debt Problem

📅 Oct 23, 2026 Einstein & Ives 17 MIN 19 VIEWS 141 SEGMENTS · 2 SPEAKERS
Dan Ives and I break down the neocloud layer of the AI buildout, how data center debt and the 10-year rattled names like CoreWeave, and why he keeps pointing to fundamentals. #einsteinandives #danives #ai Filmed October 21, 2026 📈Learn How to Trade the Stock Market from Peter Tuchman AKA the Einstein Of Wall Street: https://wall-street-global-trading-ac... 📲Follow Peter on Socials:   / einsteinofwallst     / einsteinofwallst   https://x.com/EinsteinoWallSt Check out my other Youtube Channel: @EinsteinofWallStreet DISCLOSURE: This interview is provided for informational purposes only a...

What Dan Ives said

Written from the verified transcript and checked against it. Every figure links to the moment it was said.

Dan Ives explained the recent selloff in neocloud stocks like CoreWeave and Nebius, attributing it to bond and credit investors reacting to rising debt costs rather than shareholders. He said these companies take on significant debt to build data centers, and as yields rise, debt becomes more expensive, with the percentage of debt issues under attack rising from roughly 1% historically to 10% now. Ives argued the nervousness was overblown, noting that demand for GPUs and data center capacity is accelerating, citing his recent trip to Korea. He compared the AI buildout to building the Las Vegas Strip in 1955, saying investors underestimate the scale of the AI revolution. He emphasized that a 75 basis point increase in rates does not significantly change the demand story, and that upcoming tech earnings will validate the fundamentals. He advised investors to focus on fundamentals and ignore narratives, acknowledging there will be some losers in the buildout.

Key takeaways

  1. Neocloud selloff driven by bond and credit investors, not shareholders, due to rising debt costs.
  2. Percentage of debt issues under attack rose from roughly 1% historically to 10% now.
  3. Demand for GPUs and data center capacity is accelerating, based on recent trip to Korea.
  4. AI buildout compared to building the Las Vegas Strip in 1955, with investors underestimating its scale.
  5. 75 basis point rate increase does not significantly change the demand story.

Numbers and commitments

FigureWhat it refers toTypeAt
1% Historical percentage of debt issues under attack metric 8:24
10% Current percentage of debt issues under attack metric 8:24
75 basis points Rate increase that does not change demand story metric 13:34
7,700 Current S&P level versus expected 6,500-6,700 metric 12:56

Chapters

  1. 0:00Neocloud debt problem
  2. 3:00Hyperscalers and neoclouds explained
  3. 4:40Bond market pressure on CoreWeave
  4. 5:52Capex versus monetization debate
  5. 11:05Demand acceleration in Korea
  6. 11:52AI buildout as Vegas Strip 1955
  7. 13:34Focus on fundamentals over narratives

Questions asked in this interview

12
  1. 0:42What are you seeing in the market?
  2. 2:07what do they actually do? What do those companies...?
  3. 2:55Or what role do they play?
  4. 3:04Those are tech names, but what do they do?
  5. 3:18What is a... what does hyperscale...?
  6. 4:18For companies that need to get space in the data center, right?
  7. 5:06Because they came out with bad earnings?
  8. 6:12why is it the shareholders of the company are being so impatient about that capex?
  9. 7:00Massive numbers in terms of GPU, gigawatt, in terms of all what they said that they're seeing from a demand and what they're going to build out, right?
  10. 7:46What was the... what was the initial catalyst?
  11. 12:03No one believed in it, right?
  12. 15:32Are there going to be bumps in the road and pebbles in our shoe?
Host 0:03 ↗
Einstein and Ives. We're here with Ives. Ladies and gentlemen, this man has got his whole world customized. Like we can all... We are just... This is Dan's world. We're just living here.
Dan Ives 0:13 ↗
No, it's your world.
Host 0:15 ↗
We're just renting. We're just renting space. We all... We all... We all just living. You're just paying the rent to be in the Einstein and Wall Street life. Anyway, this is this man. He has customized his life.
Dan Ives 0:28 ↗
Well, look, we're here to break all the barriers, to break down the market to you in a way that completely... we're going to unpack it and serve it to you on a silver platter like they used to do back at Peter Luger's back in the day. Remember that? Pouring a little butter on top of that steak.
Host 0:42 ↗
It's a beautiful thing. Creamed spinach for everybody out on Wall Street. Dan, welcome home. You are looking so F1. I don't even know what to say about it. You look amazing. Anyway, let's talk about it. Market is up today. We did have a little bit of a scare late afternoon, market sold off. It was definitely in the Middle East, right? I mean, the United States government put out a report saying everybody who's in the Middle East should come on home. Whatever that means, you know, we don't know what that's about, but that was definitely the market. However, you know, we know when that happens, that sort of... for the sort of the... I don't know. You know what? That's why we have humans trading this market because you know the... you know the bots will react to stuff like that, sell the market off when at the end of the day we know it's nothing that's really going to happen. Anyway, give me a breakdown. What are you seeing in the market? You just came back from Korea.
Dan Ives 1:33 ↗
Yeah.
Host 1:33 ↗
Okay. We're talking about neo... neo... what's the word? Neo...
Dan Ives 1:36 ↗
Neocloud.
Host 1:37 ↗
Neocloud. Please explain to me. These are... look, we have Mag 7, we've got tech, and then we've got the new neote. We've got neocloud, we've got hyperscalers, we've got software, we've got chips, we've got memory, we've got so many different things that fall under this new breakdown of what it's like. It's a new... it's a new sector. It's a sector inside of a sector.
Dan Ives 2:02 ↗
No doubt. And remember, Neoclouds...
Host 2:04 ↗
What are the Neo...?
Dan Ives 2:05 ↗
CoreWeave, Nebius. Okay.
Host 2:07 ↗
They... what do they actually do? What do those companies...?
Dan Ives 2:10 ↗
What they basically do is think about them renting, leasing space from a data center perspective. So...
Host 2:16 ↗
Okay.
Dan Ives 2:16 ↗
So those companies that ultimately need to build out data centers, you go to hyperscalers, and hyperscalers could also go to neoclouds. They're essentially no different than... they rent...
Host 2:31 ↗
Data warehouse space in these data centers.
Okay, so let's break this down. Then we've got tech, then we've got GPUs. GPUs is the meat and potatoes of what's going on. Okay. In order to have ChatGPT and in order to have AI, you need to build data centers. Data centers are the ones that produce the GPU.
Okay. And then so then you've got hyperscalers. Or what role do they play?
Dan Ives 3:00 ↗
Yeah. Think about hyperscalers: Microsoft, Amazon, Google.
Host 3:04 ↗
Those are tech names, but what do they do?
Dan Ives 3:08 ↗
Those hyperscalers are basically the foundation for all companies' infrastructure. So when they're moving to the cloud, when they're going down the AI path, they do it one way or another with a hyperscaler.
Host 3:18 ↗
With a hyper... So in order for them to... I need you to break it a little more down to me. So in order for... explain that part of it because I'm not clear. What is a... what does hyperscale...?
Dan Ives 3:29 ↗
Those are big cloud players. So think about them like they're the core foundation of...
Host 3:36 ↗
For every company in terms of where they build their tech infrastructure.
Okay. So those are companies like Micron, like Google, like Meta...
Dan Ives 3:45 ↗
Microsoft and Meta being really a consumer play that's really turning into a quasi-hyperscale.
Host 3:53 ↗
SpaceX being a space company that's more or less turning into a hyperscale...
Dan Ives 3:58 ↗
Hyperscaler also.
Host 4:00 ↗
So it's just... it's hard for me to understand. I don't know if the mainstream people are going to be understanding what does that actually mean. So a Neocloud is renting. Give it to me one more time.
Dan Ives 4:11 ↗
Yeah. Neoclouds are basically... they're all part of capacity. When you think about GPU capacity, data center...
Host 4:18 ↗
For companies that need to get space in the data center, right?
Dan Ives 4:22 ↗
That need to ultimately have their...
Host 4:25 ↗
Sort of workflows that they build out in AI.
Dan Ives 4:28 ↗
Okay, they could go to hyperscalers or neoclouds, names like...
Host 4:34 ↗
So they're serving a similar purpose.
Dan Ives 4:35 ↗
They're basically serving a similar purpose...
Host 4:38 ↗
About capacity.
Dan Ives 4:39 ↗
Okay.
Now the thing is, if you look at this past week, the big... sort of the big issue is that Core... because these companies also take on a good amount of debt...
Host 4:49 ↗
Okay.
Dan Ives 4:50 ↗
To build out the data centers.
Host 4:52 ↗
Okay. Oh, so Core is participating in building out data centers...
Dan Ives 4:56 ↗
And that's why, if you look at Core, to borrow money to build out the data centers, you're taking on debt right now. Debt is more expensive than we anticipated, and that's why it was under a lot of pressure.
Host 5:06 ↗
100%. And then... and then it's the whole view. But why did they get singled out? Because they came out with bad earnings?
Dan Ives 5:12 ↗
No, because CoreWeave and Nebius are... those are your two bigger... two of the bigger Neocloud plays.
Host 5:18 ↗
Okay. So there's a view where, okay, if they have a hundred billion of backlog...
Dan Ives 5:25 ↗
Right.
Host 5:26 ↗
And they take...
Dan Ives 5:27 ↗
Of demand...
Host 5:28 ↗
For GPUs.
Dan Ives 5:28 ↗
For GPUs, and if they take on 5 billion, 10 billion, 15 billion, it's investors basically being like, look, they're taking on the debt now...
Host 5:39 ↗
That revenue they're getting here...
Dan Ives 5:40 ↗
Right.
Host 5:41 ↗
Investors basically saying for the risk...
Dan Ives 5:46 ↗
Between this...
Host 5:46 ↗
Right, and there...
Dan Ives 5:47 ↗
And there. I want...
Host 5:50 ↗
I don't feel... I don't feel comfortable.
Dan Ives 5:52 ↗
So it's back to that capex versus monetization. It's a capex versus monetization story once again.
Host 5:57 ↗
No doubt.
Right. So, okay. So the investors are going, 'Wait a minute, guys. You haven't...'
Dan Ives 6:01 ↗
Those are bond... those are really bond and credit investors.
Host 6:04 ↗
Okay. Bond and credit investors...
Dan Ives 6:06 ↗
Because of the debt. Because of the debt, when they raise debt, they're not talking to stockholders.
Host 6:10 ↗
Are they issuing bonds...?
Dan Ives 6:12 ↗
Bonds. Debt.
Host 6:12 ↗
Okay. So those companies are issuing bonds to raise money, right? And they're... they're okay. But so... but why is it that the... why is it the shareholders of the company are being so impatient about that capex?
Dan Ives 6:25 ↗
It's actually less the... it's actually less the shareholders. It's more the bond market or the bond vigilantes or whatever you want to call it, almost making shareholders nervous, right? Because of what's happening with CDS spreads and yields on the bonds. That's... that's essentially what's happened with CoreWeave and Oracle.
Host 6:46 ↗
Wow.

66 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Ives, D. (2026, October 23). Dan Ives Explains The Neocloud Debt Problem [Interview transcript]. Einstein & Ives. CEOInterviews.AI. https://ceointerviews.ai/interview/3121207/

MLA

Dan Ives. "Dan Ives Explains The Neocloud Debt Problem." Einstein & Ives, 23 Oct. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/3121207/.

BibTeX
@misc{ives2026_3121207,
  author       = {Dan Ives},
  title        = {Dan Ives Explains The Neocloud Debt Problem},
  howpublished = {Interview transcript, Einstein \& Ives. CEOInterviews.AI},
  year         = {2026},
  month        = {oct},
  url          = {https://ceointerviews.ai/interview/3121207/},
  note         = {Speaker-attributed transcript with timestamps}
}