CoreWeave Credit Risk Hits Record, Pricing 60% Odds of Default

0:02 Bloomberg Audio Studios.

0:04 Podcasts.

0:05 Radio.

0:06 News.

0:09 You're listening to the Bloomberg Intelligence Podcast. Catch us live weekdays at 10 a.m. Eastern on Apple CarPlay and Android Auto with the Bloomberg Business App. Listen on demand wherever you get your podcasts or watch us live on YouTube.

0:24 So one of the big stories you just mentioned, CoreWeave, its credit risk hitting an all-time high, pricing at 60% odds of a default. Here to dig into that with us is Vasu Katsibhotla. Vasu, he's Bloomberg Intelligence Senior Industry Analyst. You have a piece out today speaking about how CoreWeave sell-off may reflect funding but not compute demand. Walk us through that.

0:48 Yeah, so clearly, you know, Corby stock and the, you know, some of the parent unsecured debt has come under pressure in the last few days. Our opinion is that this has been, you know, building up. You know, the key catalyst actually has been the rising interest rates, close to 6% now. So I think that's going to put pressure. across the AI infrastructure ecosystem. So CoreWeave as one of the marginal players in this build out, I think is coming under pressure and particularly within CoreWeave. You know, CoreWeave has multiple bonds and loans outstanding and the one which is unsecured at the parent level That was issued at 9.6%. Now it's trading close to 15%. So I think, but there are other core loans and bonds which are actually trading

1:35 much better. It all depends on what sort of collateral, what sort of covenants are there and what particularly is it at a project level debt or is it a parent debt? A lot of things matter there.

1:48 A lot of things matter. I'm curious what you think about the stock, because there are some analysts who now say that there's a buying opportunity here. Deutsche Bank, in particular, says because the sell-off has reached more than 26%, this is now a rare opportunity to get in. How do you think about that?

2:05 Well, as you know, we don't particularly issue ratings or price targets, but Corby, at this point, compared to even its new cloud peers, obviously, as a discount currently at a relatively low valuation compared to its peers. So that's one thing you can clearly observe. And the valuation compared to its own history is also under tremendous pressure. Whether it's a buying opportunity depends on what happens in the coming two, three years. CoreWeave, I also hear CoreWeave may go to zero. There are some extremist views like that.

2:39 I think that's pretty unlikely. The refinancing aspect of CoreWeave, I think, is pretty under control. They have very strong EBITDA growth in the next two, three years. That's more than sufficient to pay off the principal. The key risk is really Is the growth profile going to hold up? Because to continue to grow at the numbers that consensus has for them, or even the buy side who's expecting from them, they need to continue to build out this AI infra. The CapEx this year itself is guided to somewhere in the high 30s billion. And I think in the next two years, it's expected to grow even more.

3:16 We have calculators which show that even a single gigawatt data center capacity with Vera Rubin couldn't be as much as $ 60 billion. So that doesn't come cheap. So CoreWeave has aspirations of building close to 9 gigawatts capacity in a few years. I think so it's... So all of that requires a lot of capex. And so the key question is really, can they actually fund it and fund it at a relatively attractive yield? Because currently, if some of the parent debt is trading at 15%, the real question is not really what happens to the existing debt that they already have, which I think they will pay, but really more about can they finance the next wave of capex.

3:56 And as we look at whether or not they can finance the next wave of capex, as there do seem to be some fundamental issues here, But CoreWeave says the market is acting irrationally.

4:05 Is it?

4:08 Well, you know, again, you know, so CoreWeave is one of the players, but everybody is exposed to the same theme. CoreWeave is coming under more extreme pressure compared to some other stocks. But, you know, on the other hand, a lot of other, you know, semiconductors and other stocks in the AI complex have also come under a lot of pressure.

4:25 So really, I think if you look at anthropic revenue last night for the, you know, they used to disclose the error, but now they disclose that, you know, it seems like they have 16 billion dollars of 16 plus billion dollars of revenue in the quarter. And so. It's a tremendous growth, and the numbers for Anthropik are fantastic.

4:42 OpenAI is likely to, I think they are coming at 70 billion by the end of this year. Next year, some analysts have that growing multiple times again. So as long as the demand exists, and if you see the numbers from NVIDIA and other companies, the optical companies, HBM companies, Memory, Micron, SK Hennix, and so on, they all sold out capacity in 27 and probably a lot of 28 as well. So, I think if you look at it, compute continues to be under tremendous demand still at this point. There has not been any cancellation. Is the sell-off extreme? Is this a buying opportunity? That all depends on what investors are thinking.

5:25 This can be funded in 27, 28, and 29. I think 27 is more or less likely to be funded, but I think 28 and 29 are still up there. And at what rate? Again, one thing is it can be funded, but if the yields continue to be high and if the 10-year and 30-year rates continue to go 6% and above. I think the cost of financing, cost of capital is coming under a lot of pressure for everybody.

5:48 So I think, you know, and so Corbio, again, Corbio as one of the marginal players is coming under more pressure than other companies.

5:57 Stay with us.

5:58 More from Bloomberg Intelligence coming up after this. You're listening to the Bloomberg Intelligence Podcast. Catch us live weekdays at 10 a.m. Eastern on Apple CarPlay and Android Auto with the Bloomberg Business App. Listen on demand wherever you get your podcasts or watch us live on YouTube.

6:19 Joining us now, we have Charlie Pellett. Charlie, thanks for being here.

6:21 I am always delighted to be with you, Emma. Let us continue the conversation on Bloomberg Intelligence this morning now with a focus on technology. Anthropic, that is the company behind Claude, formally begins marketing its IPO as soon as the week of November 9th. Got a possible debut by the end of the year. Some reports say perhaps by Thanksgiving.

6:42 Joining us now, Bloomberg's Bailey Lipschultz and Bailey, thanks for joining us. Now, Anthropic has been meeting with investors. What is it telling investors about key metrics such as revenue and profitability?

6:53 Yeah, Charlie, they held their big investor day yesterday at their headquarters in New York, disclosing at the end of what really was almost a five-hour kind of ordeal explaining the company's ambitions, where they see AI going, disclosing to prospective investors that in the third quarter, revenue hit $ 16. 8 billion this up about 45% quarter over quarter. Also disclosing that they are generating positive free cash flow. They had a gross margin of 85% and a greater than 11% positive adjusted operating margin as of the third quarter. So really disclosing some of those latest metrics and what the three months ending at the end of September look like.

7:32 Just trying to get investors really up to speed answering questions around, again, where they see this business firing, where they see it going. And this has been a tape where there's been so much of a focus on, can they justify a potentially up to $ 2 trillion valuation? Can they justify... a record-setting IPO in terms of proceeds, the folks I talked to who were in attendance came away pretty bullish, but there was still a little bit of uncertainty around what this ultimately means for fourth quarter in 2027.

8:03 And so these results might potentially help the IPO, but what does it mean for the broader AI trade as this is such a bellwether for what might come?

8:11 If we just look at strictly market reaction, we saw futures fall after these headlines were set late yesterday evening. We saw weakness overnight in Asia. Do you want to call out? Yeah, the NASDAQ 100 is still down about six-tenths of 1%, but it has kind of recouped those losses since about 10 o'clock. So I think the big debate that we're hearing from investors is, How do you position in some of these AI trades that have been such big winners over the last year and change? And we've seen them really falling pretty flat. And how does that ultimately impact kind of expectations, again, going into a critical earnings season? So you have kind of like this tale of two cities, if you will, where these are

8:49 pretty solid numbers for Anthropic. It's super impressive what they've been able to do over the last few years. But if you've been bulled up in the Corweaves, the Cerebrases of the world, you're kind of taking your medicine and we're seeing that fall out today.

9:00 Speaking of being bulled up, gross margin of 85%. I heard that correctly from you.

9:06 You did.

9:07 How does that compare to rivals such as OpenAI?

9:09 Oh, man, Charlie, putting me on the spot. I think... It's tough when we don't have real numbers to compare apples to apples. So when you're looking at these companies, you have to think through some of the metrics they point to. You can point to GAAP gross margins, which we reported for Bloomberg. When you think about what that ultimately means for a company like OpenAI, we've been focusing so much on ARR or the prospect of something resembling ARR.

9:36 So I don't know that we have a precise comp.

9:39 Right but one cannot deny 85 that is a that is an unbelievably quote healthy number yeah no i mean that's great any industry for any industry they also rolled out that their net revenue retention again measuring their ability to keep customers increased spending was 900 like these are huge numbers again when you're talking about a single quarter that they'd revenue of almost $ 17 billion, like these are big numbers.

10:03 These are strong numbers. The big question does go back to, what actually is happening under the surface. Some numbers that were also disclosed was compute margin, something I admittedly had no idea what that meant, Hema. But according to our colleagues, it's the revenue left over when you strip out their computing costs and other costs. That was more than 40%, which I'm told is still solid.

10:25 All of this in the midst of what we're seeing with CoreWeave struggling pretty significantly. Jensen Huang had said at an event in Sweden that the cost of capital of CoreWeave is actually decreasing, that NVIDIA won't need to backstop CoreWeave. How do we think about Anthropic in the context of kind of a pretty significant blow up on the other side, literally within days and weeks of what we're expecting in IPO.

10:50 I know there are different companies and different exposures, but it does speak to the underlying sentiment of quite a great deal of concern.

10:59 No, definitely.

11:00 When you look at CoreWeave stock, again, it's down more than 25% or close to 25% this week alone. We did see in Australia, I believe last week, Firmis, which was an AI data center, play. Their IPO was ultimately pulled. They wanted to raise like $ 5 billion. So sentiment as it relates to data centers. data center read sentiment as it relates to these Neo clouds is pretty darn sour.

11:23 When I talk to people about AI infrastructure, they're not optimistic at all right now. Again, if you look at a CoreWeave chart over the last week, but if you look at it really sensitive with public, it's insanely volatile. The counterpoint that people make when we talk about an anthropic or when we talk about expectations for Meta's Muse and what they need to do is that these are companies that are fulfilling a need and they're still seeing strong growth. There are still questions around what does 27, what does 2030 look like? But the fundamental argument for an anthropic, the fundamental argument for real rival peers, meta, Google, you name it, is still solid. They still are struggling to get as much compute as they would need, but

12:04 the fundamentals are great from all intents and purposes.

12:06 Bailey, two other big questions. Unfortunately, only a minute to go here. First of all, we've got New York City's controller saying that this IPO ought to be paused because of safety issues. That's concern number one. Concern number two, the broader market backdrop. Is this a time to be going public?

12:22 I think people think that Anthropa could go public at any point. We could be in the midst of a terrible bear market and there still would be buyers for this IPO. Do they meet investors on valuation? Does that look closer to $ 1. 8 trillion than $ 2 trillion? That's the big question. But for all intents and purposes, I have not received any kind of pushback on expectations that this is going to be a public company before Thanksgiving.

12:45 And very briefly, I can't help but bring up SpaceX. Any comparisons here just in terms of the size, the scope? They want to be.

12:52 We reported this earlier in the summer. They want to raise more money than SpaceX, which was 75 billion before the green shoe. They want to be comparable in size. So, again, different but similar companies. But the backdrop in terms of size and scope should be pretty darn similar.

13:07 Stay with us.

13:07 More from Bloomberg Intelligence coming up after this. You're listening to the Bloomberg Intelligence Podcast. Catch us live weekdays at 10 a.m. Eastern on Apple CarPlay and Android Auto with the Bloomberg Business App. Listen on demand wherever you get your podcasts or watch us live on YouTube.

13:28 Let's talk a little bit about our Bloomberg Big Take. These are the stories that Bloomberg News has deemed to be what we want to feature based on our reporting, based on the information that we've gathered. And today's Big Take is about Volkswagen, the German automaker, and how it is investing heavily in the U.S. market because it sees a huge growth opportunity in the U.S. Gabriella Coppola is our Bloomberg auto reporter. Coppola, excuse me. And she joins us now from Detroit. Hi, Gabrielle. How are you?

13:56 Good.

13:56 How are you guys doing?

13:57 Great, great. So tell us a little bit about VW's plans and how it's intent on taking on Wrangler and Bronco on U.S. home turf.

14:06 Yeah.

14:06 So we had a great exclusive interview with the CEO, Oliver Bloom, at the Paris Auto Show earlier this week, and he kind of confirmed for us that Volkswagen is looking at getting into light trucks, body on frame. So think of like a Toyota Tacoma, like a midsize truck, but also things like the Ford Bronco or the Jeep Wrangler, which is really just like the most American and kind of really well-defended niche in America because this is what Detroit's bread and butter is. But they say that they are going to try to enter these segments. And I think it's a response to a lot of different things.

14:46 It's trying to find profits here because the China business is falling apart. It's a response to tariffs. It's a lot of different pressures here. on Volkswagen kind of coming together. And now they're going to try to do what basically everybody in the global auto industry is doing, which is try.

15:00 To make more money in the U.S.

15:01 Market because this is the big profit pool for the global auto industry. It always has been, but even more so now that people are losing share around the world because of competition with China.

15:11 Gabrielle, execs had urged the CEO to make this pivot years ago. Why didn't they then? And isn't the task just so much harder now? especially because of the competition that you just mentioned?

15:24 Yeah, it is.

15:25 And, you know, I think, you know, you can always look at a car company and say, well, how much things were in their control and how much of this was just, you know, crazy factors beyond their control, like governments changing policy. And in this case, I think with Volkswagen, a lot of it really was in their control. And I think it was just a lot of, I mean, they're kind of notorious for having a very slow moving kind of bureaucratic process of decision-making. And that part of that is because Volkswagen is a group, you know, it's not just a Volkswagen brand, it's Audi, it's Porsche, it's Skoda, it's all these different things. And all of those different brands have to compete. It's like, they're all kind of

16:03 competing for resources, you know, they're all at the same time. And so to get people to, um, kind of pull the trigger, make a decision, you know, and, and one of the examples we give in our story is, um, I spoke to a former head of the Volkswagen. He was like the COO of the Volkswagen brand in North America.

16:23 And he talked about how he was really pushing, uh, management in Wolfsburg and, you know, VW's headquarters saying, we got to do a hybrid version of the Atlas SUV.

16:33 We got to do it. This is, you know, where the market's going.

16:36 And they did, discussed and discussed and okay, we'll talk about it. And then it was like, well, now it's too late because now the, you know, the car program is being built and it's too late to do that.

16:45 Right.

16:45 Yeah.

16:46 So now hybrids are taking a huge chunk of the market.

16:49 They're growing like wildfire in the U S because gas prices are high and.

16:52 Whoever has a hybrid is winning share.

16:54 So hello, Toyota, Hyundai, you know, and that there's no reason why Volkswagen couldn't have done that too, if they could get out of their own way and make a decision.

17:03 Right.

17:03 And, um, that is, they're suffering for that. And yeah.

17:07 So inertia got in the way there. So it sounds like they are behind. How did they get up to speed quickly? Are they going to build it from scratch or are they going to partner up with someone?

17:18 You know, that is a great question. That is the question that we're all following very closely. So the most obvious easy way in would be to partner with Ford because outside of the United States, they already, Volkswagen makes a midsize truck called the Amarok that's built at a plant in South Africa. That's basically a Ford truck that, you know, done with a Volkswagen body on it, but a Ford chassis. And so the easiest thing would be to say, hey, Ford, can we borrow your truck chassis?

17:45 Yeah.

17:46 But the reality is I don't think with the U.S. becoming so competitive with, I mean, this, everyone's trying to do.

17:54 Off-road vehicles, light trucks. It's popular.

17:56 You know, the Koreans are starting to get into it.

18:00 Everyone's trying to be more rugged.

18:01 So why would Ford, which has a major winner, you know, they have a hybrid Maverick, like why would they give the crown jewels to VW in their home market, which is the source of most of their profits? So I think it's, I say they're looking for partnerships, they're in talks, but at the same time, they're saying, well, we could do this ourselves.

18:17 They certainly have room in their plants.

18:19 But my question is to build a body on frame truck, meaning you don't just put a chassis and then you stick a top hat on it. You got to build separate stuff because it has to be able to tow and do all this tough stuff.

18:30 So To do that, it.

18:32 Takes billions of dollars, separate engineering tooling. So they say they can do it. I have a lot of questions about how they're going.

18:39 To pull that off. Stay with us. More from Bloomberg Intelligence coming up after this.

18:47 You're listening to the Bloomberg Intelligence Podcast. Catch us live weekdays at 10 a.m. Eastern on Apple CarPlay and Android Auto with the Bloomberg Business App. Listen on demand wherever you get your podcasts or watch us live on YouTube.

19:02 We do have some breaking news. I have a headline here that says the European Union will begin assessing additional sectors for China trade quotas. The region is evaluating trade data to detect surges in imports from China and other countries, which could lead to quotas on additional sectors of goods, sectors including chemicals and machinery. And the European Commission may use a recent understanding with China on hybrid vehicle imports as a blueprint for other sectors. We know that Chinese EVs have been flooding into the Eurozone and perhaps posing a lot of competition to European-made electric vehicles. And now the EU is looking to take a closer look at this as it tries to address its trade deficit with China.

19:46 All right, let's move on to our next guest. And this is something that John Tucker was talking about earlier, which is that with Grand Theft Auto 6 coming out, people are straying away from video game consoles and moving more towards cloud gaming, because a lot of this can be done without a PlayStation or an Xbox. Are you a big video game person?

20:04 I am not.

20:06 I am probably not the demographic.

20:08 Nor am I.

20:08 Personally, I don't have the interest.

20:11 There you go.

20:12 I'm kind of in the same boat. So that means we need to speak with an expert. And our expert today is Nathan Nadeau. He is our Bloomberg Intelligence Technology Research Analyst. They published a deep dive into cloud gaming. And Nathan, how much is GTA 6, the long-awaited arrival of GTA 6, kind of pushing us in that direction of killing the console?

20:33 As a gamer myself, I think GTA is kind of unique on its own. That, you know, 70... Upward of 70% of gameplay is going to take place on console because that game is that high fidelity, that much graphic processing. But there are simpler games, simpler GTA games from the same franchise. We have seen how that has worked well on Netflix, which actually had GTA trilogy games for a short while. And actually, that was around the time when Netflix actually experienced a spike in its games activity. So we can see how that model would work for the future, but just not for, you know, just this GTA 6 that is highly anticipated coming.

21:17 What is the.

21:18 Benefit from streaming games from the cloud? If you already own a console and you have a really good experience from that, what is the value add and the growth opportunity from the cloud.

21:27 Side of it?

21:28 Yeah, for someone, you're right on the nail there. For someone who already owns a console, you know, cloud gaming is mainly about convenience, assessing the same games across TV, phones, PC, handhelds, without being tied to a home console, the box that's literally feeder to the TV. One example of that use case is as a gamer, hopefully at some point you get your hands on and try some games. The use case is that you could be completing the side task of a game while commuting to work on a train while leaving the main story, the juicy bits for when you're back and you're at your main gaming rig. To your questions, you know, we see the bigger opportunity for cloud gaming is actually reaching people who never own a console.

22:09 in the first place. So these are especially population in countries where most people access games through mobile devices, speak about India or Brazil. And the main gaming, cloud delivery can make high-end games in this case accessible without the need for an expensive local hardware or consoles. And we have seen recently how Sony and Microsoft have actually high prices of consoles and their consoles is more than midway through their life cycle. Traditionally speaking, consoles become cheaper as the components become cheaper, the components to make them. But in this latest generation, the Xbox Series X, S and the Sony PlayStation 5, we are seeing the reverse. So they're becoming less affordable. So that's really a use case

22:56 here for cloud streaming gaming to really pick up speed in the future.

23:00 You mentioned how GTA 6 is kind of an exception and outlier because of the graphics and, you know, the visuals that are involved there. What kinds of games will be more popular, will proliferate with the advent of cloud gaming? I mean, is it going to be, you know, really basic puzzle type games?

23:19 Definitely puzzle games and also games with a huge, with a big social graphs. You know, talking about Roblox where, you know, anywhere from kids to teenagers to people who play Roblox, sometimes will literally just access the environment, a gaming environment to just hang out, you know, and actually a lot of Roblox activity already come through mobile. And we see the same for other games with a huge social mode or social experience to it that will proliferate on cloud gaming. And also simpler games like you mentioned, whether it's puzzle games, farming games, you know. And for game makers, really just an additional distribution of their games to be able to spread wider beyond the console or PC.

24:06 Nathan, before we let you go, very quickly, you're a gamer, so what's your favorite game?

24:11 Well, I'm obliged to say GTA, but really, I'm deep down a Pokemon fan, so I'm not going to have anything to play this year. But I guess, finally, I have the GTA V, but I guess, for research purpose, I have to try the VI.

24:24 This is the Bloomberg Intelligence Podcast, available on Apple, Spotify, and anywhere else you get your podcasts. Listen live each weekday, 10 a.m. to noon Eastern on Bloomberg.com, the iHeartRadio app, TuneIn, and the Bloomberg Business app. You can also watch us live every weekday on YouTube and always on the Bloomberg Terminal.