## Detailed walkthrough
### [00:00](https://www.bilibili.com/video/BV1wuJ26pE9d?t=0s) Market Bleed and CoreWeave Bond Yields Spiking
The conversation opens on a day of intense portfolio carnage across the AI infrastructure sector, prompting the hosts to assess the broader structural health of the ecosystem. The focus turns to Neil, an investor holding shares in Nebius and January 2028 $100 call options on CoreWeave, to weigh in on whether CoreWeave's surging bond yields point to an underlying failure of the business or just another cycle of debt-fueled anxiety similar to past panics surrounding Oracle. Neil emphasizes that while today hurts, investors must keep context: “Second of all, for all of us that are bleeding today, we are probably bleeding today because we're still crushing it year to date.” He argues that if CoreWeave executes on its plan, its revenue will sufficiently service its mounting debt obligations, even though interest payments have rapidly escalated from $640 million last quarter to over $900 million in the current quarter.
```mermaid
flowchart TD
accTitle: CoreWeave Debt and Execution Risk Flow
accDescr: Visual diagram showing the progression from backlog to debt service risk.
subgraph Commitments ["Contract Pipeline"]
A["Contract Backlog ($104B-$130B+)"] --> B["Long-term Deals (Microsoft, OpenAI, Meta)"]
end
subgraph Execution ["Infrastructure Buildout"]
B --> C["Data Center Deployment & Power Access"]
end
subgraph Financials ["Balance Sheet Pressures"]
C --> D["Quarterly Interest Surges ($640M -> $900M+)"]
D --> E{"Project Delays or Non-payment?"}
E -->|Yes| F["Yield Spikes & Credit Panic"]
E -->|No| G["Revenue Recognized & Debt Serviced"]
end
```
The speakers debate the true scale and validity of CoreWeave's backlog, which was reported at $104 billion and grew to over $130 billion in the most recent quarter. The central concern remains execution: contracts stretched out to 2028 and 2029 generate revenue only when facilities go live and customers pay on time. Yet Neil maintains strong conviction in CoreWeave's medium-term survival, stating, “I just don't see how CoreWeave doesn't exist a year from now and isn't much, much bigger.” Even though bond yields have blasted through 15%, Neil does not believe it automatically implies the core business is collapsing, nor does he see a reason for Nvidia to orchestrate an emergency bailout right now. The other hosts push back, questioning whether the company's 100% to 200% year-over-year expansion is fundamentally unsustainable and built entirely on reckless balance sheet leverage.
- CoreWeave's interest expense escalated from roughly $640 million in the prior quarter to over $900 million in the current quarter.
- Contract backlog was previously reported at $104 billion and exceeded $130 billion as of last quarter.
- CoreWeave takes on debt specifically to backstop contracted capacity extending into 2028 and 2029 for hyper-scalers like Microsoft, OpenAI, and Meta, rather than purely speculative buildouts.
---
[05:00](https://www.bilibili.com/video/BV1wuJ26pE9d?t=300s) Credit Market Disconnect, Elon's Warning, and Neocloud Positioning
The panel shifts to examining why the credit markets are selling off CoreWeave bonds so aggressively, pushing borrowing yields into distressed territory between 13% and 15%. A key distinction is drawn between the equity market, which can swing wildly on irrational panic, and the credit market, which typically consists of deep-pocketed, conservative institutional capital like pension and insurance funds. One co-host recalls how an equity sell-off in Nebius and CoreWeave was triggered by an institutional player getting liquidated before being bailed out by Ken Griffin, after which stocks rebounded. However, credit distress is harder to dismiss: “Usually the people that are selling in the credit markets have significantly more information than equity investors because equity investors can sell or buy for any reason.” The co-hosts point out that while only about $1 billion of CoreWeave's debt carries an immediate 15% rate, with the remainder predominantly locked in at fixed, lower rates, taking on over $50 billion in total liabilities introduces extreme sensitivity to delays and cost inflation.
```mermaid
flowchart LR
accTitle: Neocloud Risk Profiles Comparison
accDescr: High debt versus cash-balanced model between CoreWeave and Nebius.
subgraph CoreWeave ["CoreWeave Profile"]
CW1["Over $50B Debt Exposure"]
CW2["Gigawatt Power Scale Leader"]
CW3["Vulnerable to High Rates"]
end
subgraph Nebius ["Nebius Profile"]
NB1["Shorter-term Contracts"]
NB2["Cash Matches Debt Total"]
NB3["Insulated from Credit Squeeze"]
end
```
The discussion grows heated over Elon Musk's recent remarks on ZeroHedge regarding data center debt, highlighting the immense difficulty of physical execution. While raising debt or announcing data center ambitions is one thing, procuring gigawatts of power, securing GPUs, and bringing complex facilities online on schedule is another matter entirely. The speaker highlights that bondholders may believe CoreWeave cannot successfully deliver the capacity it borrowed against: “They're sensing that these guys levered up their balance sheet and they can't actually execute on what they levered up for.” While CoreWeave remains the largest neocloud player by active gigawatt power capacity, the panelists conclude that buying CoreWeave equity or debt right now carries unnecessary risk. Instead, Nebius represents a far safer vehicle to play the neocloud boom: “'Cause Nebius is not going to have the same debt problems. They have essentially as much cash as they do debt. They could pay it all off.”
- CoreWeave carries over $50 billion in total debt obligations, with around $1 billion currently exposed to a punitive 15% rate.
- Elon Musk warned on social media that while pitching and raising money for data centers is straightforward, executing and building them is extraordinarily difficult.
- CoreWeave holds the market lead among neoclouds in terms of active power capacity measured in gigawatts.
- Nebius holds roughly equivalent cash reserves to its debt burden, insulating it from the solvency shocks rattling CoreWeave.
这一局面直接引向了对英伟达自身商业闭环的质问：英伟达未来的业绩指引中，究竟有多大比例是绑定在 CoreWeave 庞大的未交付积压订单（Backlog）之上的？如果债券市场的严峻形势迫使英伟达不得不亲自下场为 CoreWeave 提供信用兜底，整个科技板块的估值逻辑将遭受剧烈冲击。一位发言者不可思议地质问：“And the fact that we're even having the conversation that Nvidia needs to come in and backstop CoreWeave is crazy to me.”
分析师们对即将来临的财报季抱有极强信心。随着台积电（TSMC）与 Meta 财报拉开帷幕，紧接着 Alphabet、亚马逊、AMD 等巨头接棒披露，市场情绪极大概率迎来大反转。发言者断言：“There's no way this happens. And if it does, then I guess we're all fucked, but I don't see it happening.” 届时大型科技公司绝不可能突然改口宣称算力需求归零，相反，它们很可能再次确认订单已排满至 2027 或 2028 年。
subgraph PowerGrid["电力与审批瓶颈"]
A["ERCOT 电网总容量: ~91 GW"]
B["排队调研项目需求: ~450 GW"]
C["数据中心许可证审批推迟"]
A -.->|"严重超载差距"| B
B --> C
end
subgraph FinancialImpact["债务与商业履约危机"]
D["运营商举债建设 (如百亿级发债)"]
E["数据中心无法按时通电并网"]
F["下游客户行使拒绝支付权利"]
G["云厂商独自承担巨额利息敞口"]
C --> E
D --> E
E --> F
F --> G
end
```
> “The fact that there's theories that that might need to happen. A, what does that mean for the rest of NeoClouds? B, what does that mean for all the deals that Anthropic and OpenAI have with CoreWeave, which ultimately goes back to the hyperscalers?”
然而，对于英伟达可能下场施救的核心逻辑，讨论呈现出戏剧性的多空对立。部分观点认为，CoreWeave 最大的投资安全边际（Margin of safety），本质上就是黄仁勋绝不可能坐视这家核心标杆企业倒闭的“隐性看跌期权”（Jensen Put）。当救助传闻出现时，市场条件反射般地将其视作“世界末日”，但反向思考，这恰恰证明了英伟达正在积极行动以消除系统性崩溃。无论是通过直接注资、收购还是联合财团重组，兜底举措在短期利空消化后，反而确认了公司的生存底线。尽管如此，讨论最终以理性的审慎态度收尾——动用救助手段本身就坐实了 CoreWeave 商业模式的脆弱与失衡，且由于该传闻信源尚非顶级权威媒体，投资者必须等待官方正式公告与落地细节，切忌在迷雾中盲目下注。
> “The margin of safety on CoreWeave was the idea that Jensen would never let this company fail. And here they are saying, you know, NVIDIA might not need to let this company fail. And everyone's like, oh, my God, that it's the end of days. It's like it's the opposite of end of days. This is them coming in to make sure this company won't fail.”
> “It's not good that they have to bail them out. By the way, I don't even know what that source was. Like, we'll have to see. I'm going to wait for a press release to see what actually happens.”
## Notable quotes
> [05:00](https://www.bilibili.com/video/BV1wuJ26pE9d?t=300s) Usually the people that are selling in the credit markets have significantly more information than equity investors because equity investors can sell or buy for any reason.
> [10:00](https://www.bilibili.com/video/BV1wuJ26pE9d?t=600s) I don't know if any neocloud is safe if CoreWeave goes under. I think that's wishful thinking.
> [12:15](https://www.bilibili.com/video/BV1wuJ26pE9d?t=735s) The fact that we're even having the conversation that Nvidia needs to come in and backstop CoreWeave is crazy to me.
> [00:00](https://www.bilibili.com/video/BV1wuJ26pE9d?t=0s) I just don't see how CoreWeave doesn't exist a year from now and isn't much, much bigger. — Neil
[5:00] Usually the people that are selling in the credit markets have significantly more information than equity investors because equity investors can sell or buy for any reason.
[10:00] I don't know if any neocloud is safe if CoreWeave goes under. I think that's wishful thinking.
[12:15] The fact that we're even having the conversation that Nvidia needs to come in and backstop CoreWeave is crazy to me.
[0:00] Neil: I just don't see how CoreWeave doesn't exist a year from now and isn't much, much bigger.
Original transcript
[0:00] Yeah, looking at portfolios definitely makes people sick today. Let's get started with you, Neil, because Tanner and I have been streaming all day, so I think people have heard from us. But Um, Thank you. I don't know if you own CoreWeave. I definitely know you own Nebius, so it's obviously in that same world. I'm curious on your thoughts on CoreWeave today, curious on your thoughts about their bonds exploding. and curious if you actually think there are some structural concerns that maybe obviously don't end the AI revolution, but definitely are making people think twice about how capital is being deployed. in this AI ecosystem right now. Okay. So, I do own some call with. I have some call options for 2028 January $100 ones. Second of all, for all of us that are bleeding today, we are probably bleeding today because we're still crushing it year to date. Let's not forget about that. Then with regards to the concerns for CoreWave, which are the same concerns that we've talked about for Oracle. And the answer is always the same. if they continue to execute, if they can continue to grow and generate the amount of revenue that they're expected to generate, They have over 100 I've seen this $104 billion backlog, but they have way more than that, right? Because last quarter... $130 billion. Yeah, so that number gets bigger. Of course, that number doesn't mean that it gets recognized on time and the revenue gets translated into earnings and et cetera, et cetera. So I get it. But if the backlog does change, get recognized on time based on their own assumptions, then yes, I do think that they will be able to pay down the debt, the interest that they need to pay every single quarter, which, yes, last quarter has already doubled to 640 million. This quarter is already... above 900 million etc etc But the fear is always the same. It's not new. We know that if there are major delays, at CoreWeave or one of their customers cannot pay CoreWeave or a data center cannot come online on time. We know those fears already from a long time ago. It's not just CoreWeave, it's every other player out there. But if you're expected to spend an insane amount of money and suddenly a project is delayed and suddenly you cannot recognize your revenue on time, yes, it's an issue, but... I just don't see how they don't exist anymore. I don't see how CoreRift doesn't exist a year from now and isn't much, much bigger. unless something major, major happens. Yes, there still is Nvidia that could save them, but I just don't see why Nvidia would step in right now. Because again, just because we're seeing 15% on their bonds doesn't mean that the business is failing. But Neil, but what is the market? There's a reason those yields are spiking today. And maybe it has nothing to do with the business. I was just about to say, maybe there is a reason that we don't know yet, that the bond market knows. But as of right now, until today, CoreWeave has delivered. Well, I guess that's the question then. Like, do you think they've only delivered because of the insane amounts of debt that they put on the balance sheet? to be able to put up this 100, 200% year of year growth. But in reality, it is like quite literally not sustainable and someone has figure that out at this point. which is why the yields are up. meaning not sustainable, meaning they cannot recognize the revenue that's supposed to recognize or they will be... Meaning they can't make the interest payments on the debt even if they keep growing at 2x, 3x every single year. Why not? Because the numbers right now are not that scary for them not to be the numbers. Yeah, so... Unless suddenly the number blows up and then the business crumbles. then okay, but... There's no indication of that happening at the moment. And why would there be? Because if that would be the case, it would be an issue for everyone, not just for CoreWeave. Right, right. Yeah, like this is the weird thing that I'm – thinking about, which is like So Corweave takes out debt whenever they... are signing new contracts. They're not doing it ahead of these contracts. They're doing it for those contracts into the future. The deals that they're signing are way out into 2028 and 2029 because The current data centers that they're setting up are for already sold capacity to companies like Microsoft or OpenAI or, you know, Meta and existing long term contracts, which is a little bit different than how Nebius does it, which is. you know, more shorter term contracts. CoreWeave, for the most part, has already been sold out to really, really large companies. So, The debt that they're signing... is this, like what are the signs that would have changed the metrics? GPU pricing is going up. What is it, the cost of power that's now coming in? Like I don't know the new metric that could be coming in that would change their financials other than potentially this debt situation, which seems like a lot of money.
[5:00] which by the way, you know, I think it's like $1 billion that they have at a 15% interest rate. The rest is like mostly fixed at much, much lower rates. interest rates, what part of this would would Corweave not know before signing up this debt, that they would also have Jensen as an advisor and many of these other people who are part of this build out seeing that. Hey, maybe... don't go to 50 plus billion dollars of debt if you can't afford it is okay so is this not like a single player or like you know we ended up seeing nebius and core weave and all these companies sell off and we were like what the heck are people seeing We come to find out it was Leopold. And he was getting sort of liquidated. And Ken Griffin came in and saved him. and then the market ended up shooting up the day after. Is there a possibility that there's a big company that's getting out for their own reasons? Maybe, but the difference here is that was in the stock market. This is on the credit markets. But why can't there be an investment? I'm not saying this is Leopold. I'm saying... I don't know. I used to be a... Financial institution that is trying to wrap up some of their debts. So you're saying a financial institution that... bought CoreWeave's bonds at 9%. Yeah, they want to de-risk a bit. To the point where they're now at 15%, 13%, 15%. It I just I don't know. Maybe. But usually these are not funds that are super volatile. They're not levered up on micron options. They're pension funds, insurance funds like. They tend to be a bit more stable and can handle the volatility, which is why this one feels a little different than the traditional headlines that cause semi-stocks to go down. - But someone's selling into this increased yield. you know yeah and and we don't know who but Usually the people that are selling in the credit markets have significantly more information than equity investors because equity investors can sell or buy for any reason. Which is why... So I agree with both of you. Both of you are saying... And basically you're saying what the CEO said today, our commitments, our backlog, everything is good to go. We're strong. We have our pipeline. We're going to make our debt payments. But the market is not believing it. And so the question that becomes like during deep seek, the market didn't believe when Jensen put out that statement that, you know, the demand for compute will only go higher because the market's irrational fear that Chinese GPUs would take away from the efficiency of American GPUs and that caused that sell-off, which was a buying opportunity. But again, that was in the stock market. It's not like NVIDIA's bonds exploded that day. This time it feels like, and no one has an answer for it, but it feels like there's something weird happening behind the scenes. even though The fundamentals of the business seem to be fine outside of the massive debt that they have. Yeah. And then we also got clarity from Elon saying that there's no lack of demand for... data centers Like, I, I, I'm just, but the people with the answers are also kind of going, Hey, what's, like, Elon also said, Elon also said core weave is full of shit yesterday. Where? You saw that tweet, right? No. so under i think neil saw it under zero hedges posts around core weaves debt blowing up elon was like it's really easy to raise money to build a data center it's much harder to do it spacex is having troubles executing even though we're like world class at this What are you talking about? That's the opposite of what the Post said. He said it's extremely hard to raise capital. Read the tweet. Elon said... You know, we are... Uh, Our team is cracked, and it's really, really hard to... raise capital right now. It's easy to say you're going to do it. It's hard to actually do it. Well, that's my point. He's saying he's implying Corweave has raised a ton of money. They've raised a ton of debt, but actually executing on building the data center is hard. And if there's someone in the bond market that's saying Corweave is not executing on their data center ambitions, which is the part all three of us are confused about because everything we're seeing says that they are. then maybe that's why they're selling off the bonds, causing the yields to explode. Because, you know, They're sensing that these guys levered up their balance sheet and they can't actually execute on what they levered up for. Okay, so let's say, worst case scenario, a couple of weeks from now we're covering CoreWeave's earnings report and they don't meet the numbers. for XYZ reasons. Maybe there is a delay, maybe something happened, I don't know. Since when did CoreWeave become the center of all of these AI build-outs? I mean, they're the biggest neocloud. Thank you. Are they still? Ah, not by company. by marketing, but by... By active power. Gigawatts, yeah, yeah, yeah. But honestly, the truth is, I'm not even buying Core Reap here, because I do think that there should be a level of... Um, A level of worry. you know because i i do think that the bond market has way more information than us i don't know what they see And I think that the clearest way for me to benefit off this is Nebius, to be honest. 'Cause Nebius is not going to have the same debt problems. They have essentially as much cash as they do debt. They could pay it all off.
[10:00] I promise you they won't. But, Um, And plus they have things like their asset light model, And high prepayment, I don't know how they're getting this, better than Corweave. But Um, Nebius looks like a safer investment, even though they're at a much higher premium. So I'm buying Go. I don't know if any neocloud is safe if CoreWeave goes under. I think that's wishful thinking. You mean the stock? Not the business. The stocks, yeah. Not the business. I don't know why. Because if CoreWeave's trying to get debt today and they're getting a yield of 12% to 15%, When Nebius goes out either to raise convertible debt or they're just going to issue shares, which will tank the stock price as well. What? What, like... firms are giving nebius eight to nine percent yields which is what core we've got in june versus 12 to 15 percent Unless those issuers just think Nebius is so much better than Coral Weave, which doesn't make sense. Well, Let's talk through that a little bit. I think Corweave had... Um, essentially sold all of their capacity way too quickly back in a market where pricing was not as high as where it is today. So they're still setting up data centers to fulfill on contracts that were signed with Anthropic, OpenAI, Meta. from years ago. at way, way lower essentially dollars per megawatt that they're setting up with those companies. So still at a profitable level, but not nearly what is coming to market right now. And Nebius has essentially said, hey, we have customers that want to sign extremely long-term deals with us, but we're actually turning them down because we can make more on short-term deals. And so their average... essentially deal million dollars per megawatt I think for Nebius now is somewhere between 20 to 25 million where I think CoreWeave is anywhere between like 15 to 16. So they're actually making much more money per GPU. which means, you know, margins are higher, depreciation as a percentage of revenue is lower, like, They... I think that they're a better operator. Next, I don't think that these companies... are Um, that they need to raise debt. They want to because their competitors are raising debt and it's a prisoner's dilemma. And if they don't, someone else is going to take the market share. OK, but if there becomes a just complete standstill, no bank is giving out money, no one wants to take the risk. What happens to the data center space? We all don't want to use AI anymore. Like GPU prices go up, more auctions, people will go to the highest bidder, and actually the margins on the NeoClouds get even better. So yeah, maybe the stock price gets hurt, but I think the NeoCloud space is going to be still very much in demand. Look at Elon's tweet. He says there's there's There's zero lack of demand, essentially insatiable demand. Mm-hmm. I do think they all trade with each other, though. That's the issue. I think a lot of your points are right. The problem is if Corby actually gets down to the 30s, 40s, that beast is gonna get hit so will i run i mean yeah the problem is core we've The thing I can understand, and Neil, I'm curious about your perspective on this, like you and Tanner both were very early Nebius because I think you saw the fundamentals of the company and were like, this is way better than the fundamentals of CoreWeave. But the people that went on the CoreWeave bandwagon saw Jensen. as like the person who was anointing them as like the core cream of the crop Neocloud. I don't get how, Jensen's pick of the litter can't get access to cheap capital. And if they can't get it, Even if there's nuances between the rest of the neoclouds, first of all, it begs the question of how much NVIDIA's future revenue is tied to CoreWeave's backlog and selling all those gpus but it also means if the rest of the neo clouds blow up a little bit maybe not go out of business but definitely uh you know face some some pain in the bond market How much will that affect Nvidia's broader GPU sales, which brings up the question if Nvidia has to save CoreWeave. And the fact that we're even having the conversation that Nvidia needs to come in and backstop CoreWeave. is crazy to me. Like it makes no sense And that's why I think the rest of the neoclouds, you know, are going to be in a tough spot if that actually does have to happen. Thank you. We still need to see what happens with that $500 billion program thing that Jensen wanted to do with all the big finance projects. Folks, maybe that's one way to get cheaper capital, we'll see. Um, But on the first point, I think both Tanner and myself, I just speak for myself, but I chose Nebius because of the team. because the team had years and years of experience, building huge huge i mean yandex's base it was called the google of russia and when you have the exact same team trying to build something similar but new for the age of AI, I guess.
[15:00] You see the performance of Nebius, compare that with the performance of every other NeoCloud out there. it is quite clear to see which company has an insane amount of engineers with years and years of experience in the field. This is also why SpaceX can run much faster than everybody else. It's a game of can I get or can I raise a lot of capital faster than the others at better rates, But it's also a race of, do I have the best talent inside my company? There's a reason why Meta spent tens of billions of dollars acquiring business. great engineers, geniuses, etc., etc., you're now seeing the results with Muse. So yes, on the one hand, it's capital, but on the other hand, capital gets you part one. Part two is is building a product someone wants in an efficient way where your return on investment is much better than the rest, because then you can offer better prices and still be a good business, which Nebius fits all of the categories. well fits all of these points. Doesn't mean that Core Review is a bad business, because if it was a bad business, I highly doubt Jensen would give them. Uh, It's time of day. Thank you. It's a fair point. It's just... It's weird, man. Last October, we had soybeans and a crypto liquidation in this October, I mean, things were doing okay until today. I know memory has been stagnant and stuff like that, but... I mean, things were okay until literally today, losing almost a trillion dollars of market cap. I think we added a trillion like in the first couple days of October, And it seems like it's because of Corweave. I mean, luckily for us, we have TSMC tomorrow with Meta next week, and then the whole earnings season starts. So we'll have way more answers to the stupid questions that... some of the headlines are asking right now. Just have to wait a couple of... a couple of days and then maybe it might turn out to be a nothing burger. Maybe a week from now when we cover Meta, Meta knocks it out of the park. Mark talks about how great everything is, how they are seeing great return on investment. Then the week after that, we'll probably have Google and Amazon and Alphabet and all of the other companies and AMD and Intel talk about, yeah, there's just too much demand. Yeah. Supply cannot meet it, were sold out until 27/28 Whatever, and then suddenly, oh, sentiment changes. Because at the end of the day, results matter and I don't see a scenario where suddenly All of the companies that have gone to analyst conferences that had their own investor days suddenly flip the switch and say, "Oh, actually, there's no demand anymore." Actually, we're not sold out anymore for the next three years. Actually, prices are not increasing. They're all decreasing. There's no way this happens. And if it does, then I guess we'll We're all fucked, but... I don't see it happening. Well, I'm happy you said that, Neil, because... And it's kind of the same point Tanner made early on, which is like... This is a bond market issue, which is something we can't really explain, but nothing around backlog, earnings growth, margins, data center build. I mean, you know, if the if the if the administration came out today and said we're banning data centers, I could understand like, OK, the new clouds are like. That's not happening, right? So it's like you can't point to anything that we all see as retail investors and we're all covering the news pretty deeply. That is like, I mean, maybe CPI. But I don't know if CPI takes off a trillion dollars off the S&P when everyone already knew it was going to be a little. It feels like there's something deeper. And we're in a stage of uncertainty because none of us really have the answer to what that deeper reason is right now. Yeah, I think like Um, and What parts of the market can we pay attention to in this space? You know, we can look at demand, we can look at CapEx. This is still the exact same market that, you know, ASML just reported and said, you know, customers are now booking into 2029 and 2030. They're seeing insatiable demand. We're gonna see TSMC. We already know what the revenue is 'cause they report it on a monthly basis. They're growing at some of the fastest levels ever, 55% year over year. Like they're going to put up amazing numbers and hopefully commentary is good as well. Next comes big tech and like last quarter we heard Google say that they're growing so fast that they're going to look towards third-party demand, meaning they're going to tap neoclouds for potentially even more supply. I just don't know where you look to see, okay, well, this area over here is a little shaky. You know, it's like, where's the shakiness? We're seeing nebbiest do auction sales that are continuously are breaking records. Um, I don't know. You know where I'll say their shakiness is –
[20:00] a lack of of power, a lack of being able to even get permits to set up data centers. We're seeing that in Texas. I think Texas is total... ERCOT's Total power on the grid is like something like 91 gigawatts. And they have right now like 450 gigawatts worth of surveys, like future data centers that are wanting to be set up. How are we possibly going to be able to meet that demand? But that doesn't seem negative to me to the... to the actual... you know, data center plays other than to potentially slow down growth. Like, for example, If Oracle goes out and raises $155 billion worth of debt, and their New Mexico data center is not getting set up in time, the companies that have those backlogs will essentially... have the right of probably refusal to pay the interest and and Oracle's the one sitting with the debt, waiting for this data center to come online. They can't get the permits because that's being pushed by. So that is a definite concern. But I mean, maybe that's it. Maybe CoreWeave, we just found out they're about to not get a permit that's going to slow down data center build-outs. Maybe we're about to see... Um... Some sort of hell happened over on that side of the company. I don't know. I have two headlines. Bloomberg says just 10 minutes ago, Tesla SpaceX merger announcement possible before December. That'd be crazy. if that's actually announced this year. Number two, is QTR news. NVIDIA may be forced to bail out CoreWeave after $250 million initial investment and multi-billion dollar follow-on. The fact that... The news wires are reporting. Can you read that again? Yeah, NVIDIA may be forced to bail out Corweave after $250 million initial investment. And this is not Bloomberg, but this is a smaller newswire putting it out right there. Um, I mean, why are we talking about bailouts? Right? like like like why is that entering into the news flow and like and and does that speak to someone in the credit markets yesterday selling off their bonds because they're afraid of an event that's happening the reason it's so confusing for all of us is because we can't really point to anything because it's in the credit markets. But like, well, here's another question. How the hell will Anthropic go public in the midst of all of this? I mean, who's buying that IPO? You got headlines on NVIDIA potentially considering bailing out Corvive, which again, NVIDIA Corvive hasn't said anything about that, but the fact that there's... theories that that might need to happen. A, what does that mean for the rest of NeoClouds? B, what does that mean for all the deals that Anthropic and OpenAI have with CoreWeave, which ultimately goes back to the hyperscalers? Like, a lot of this stuff is still interconnected. And I think it's why the Anthropic IPO needs to be unbelievably successful to get the sentiment to kind of die down. But right now, no one believes in this anthropic IPO. It's funny, right? The investors of CoreWeave, which I'm a part of, And I'm down on the name now. It's just funny listening to Neil going like... Yeah, we're all up on the year and everything like this. I'm like, I'm down on Corby. This is not an exciting thing. Okay, stop complaining. You're up substantially on other names year to date. So you're back on the 2020. I'm not here to celebrate. I'm actually getting worried I'm not going to get that $250 deal. Me and Steve made a bet to see if NVIDIA was going to pass 250 by the end of the year. Put it in the chat. Let me definitely do a poll right now. I'm doing a poll in the chat. Nvidia 250. I remember being at CES with Tanner watching Jensen speak and, And Tanner was saying NVIDIA 2. That's when it was like a 170. Tanner was like, it's going to happen. So you've been on this 250 train for a while. I remember sitting there. We were listening. I think they were showing off the Verirubit. I was like grabbing you by the shoulders. I was like, we're going to be fucking rich. Here we are at the same price. God. Oh, five days ago. It's up from where it was back then. It was like 180, but it's not at 250. Oh, sorry. the margin of safety on core weave was... the idea that Jensen would never let this company fail. And here they are saying, you know, NVIDIA might not need to let this company fail. And everyone's like, oh, my God, that it's the end of days. It's like it's the opposite of end of days. This is them coming in to make sure this company won't fail. Yeah. Thank you. That's a great point. That's a legit great point because the entire, one of the entire quote-unquote puts you had on Corweeb is that Jensen's going to backstop. And now... that you've got headlines of Jensen potentially coming in to save them. Which I don't know if that means an acquisition. I don't know if that means giving them... $10 billion from the $500 billion he got from his financier buddies. Like that should be a bullish argument. Even if it feels bearish in the short term, it's a margin of safety
[25:00] It's not good that they have to bail them out. By the way, I don't even know what that source was. Like, we'll have to see. I'm going to wait for a press release to see what actually happens.
