Tech Stocks and AI Bonds Are Tumbling. Is the AI Trade in Trouble?

0:00Asa Fitch: The alternatives universe is vast. Knowing where to focus matters. From private equity and private credit to hedge funds and real assets, Morgan Stanley Investment Management helps advisors see the solutions clearly. With more than 40 years of alternatives experience and specialized global expertise, we address liquidity, growth, and diversification with conviction backed by capability. Explore Morgan Stanley Investment Management's approach to alternatives at morganstanley.com slash im slash alternatives.

0:33Telis Demos: Hi, Miriam. Hi, Talos. So it's been a tough couple of weeks, I'd say, in the markets for AI. And I think that the focus has really turned to what's going on on the credit side of AI. We've been talking about AI stocks, obviously, on this show and everywhere for quite a long time. But the focus has really shifted, I think, to the money being raised in the bond markets and the loan markets, the borrowing that needs to happen for this AI buildup. And you started to see it a couple of weeks ago when a bunch of AI bonds in particular, bonds of a company called CoreWeave, sold off pretty sharply. And you've continued to see it.

1:11Telis Demos: I mean, SpaceX reported earnings the previous week. They showed a profit. That was good. Their stock went up a little bit. But then they started talking about how they need to raise even more money to pursue a business that is sort of like that of CoreWeave, which we'll explain what that all is in a moment.

1:25Asa Fitch: And then investors started to get nervous again.

1:27Telis Demos: And then investors got nervous again and sold off the stock. So, I don't know. It seems like as all these AI companies are now reporting earnings, I don't think the market's in a very receptive mood.

1:38Asa Fitch: No. And then we also have, of course, Anthropic having filed its S1, its prospectus to go public. And, you know, all eyes are on that. But kind of lurking in the background are these fears around what is it going to take to finance this infrastructure that still has yet to be built. And I have to admit that when I heard that CoreWeave has kind of become the poster child for this, and I even wrote a story about CoreWeave, about Blackstone financing CoreWeave a couple of years ago. But I was like, wait, what does CoreWeave do again? What is CoreWeave? I need help to understand what's going on with this company.

2:16Telis Demos: Yeah, I think the question that I was having was, OK, I can see this stuff happening in the credit markets. I follow bonds and bond prices and what investors are talking about. But, you know, I'm not the expert on what's going on in the underlying market for AI and compute. Like, has something changed?

2:36Asa Fitch: Don't say compute. Compute is not a noun. Let's just say computing power, OK, just to please me. I'm a grammarian.

2:42Telis Demos: All right. All right. Point well taken, Miriam. I appreciate you standing up for the English language. But the question remains of whether or not there is something changed in the demand for computing power and chips. Has something happened that is fundamentally altering what investors' expectations are of the demand and the need for AI that then is being reflected in these bond and stock prices over these last couple of weeks? You know, I'd say that we follow the markets pretty closely, but we're not the experts in the world of computing power and GPUs and CPUs and TPUs and chips and all those sorts of things.

3:22Telis Demos: So we thought that we needed some outside perspective on what, if anything, has happened in the market for AI and computing power that we need to be talking about. And so to do that, we have phoned a friend, a good friend, our friend Asa Fitch, who is my colleague at Heart on the Street. And he writes about AI. He writes a great AI newsletter that comes out each week. And he has forgotten more about this stuff than I've ever known. So Asa, welcome to the show. Thanks for having me.

3:47Asa Fitch: Okay, Asa, let's start out with a really basic question. What the heck is CoreWeave? What does that company do?

3:53Miriam Gottfried: So CoreWeave was a company founded quite a while ago by some Wall Street guys to do Ethereum mining using GPUs. and back then in like 2019 2018 that was a big deal and you can make a lot of money just throwing computing power at the gpus these guys were wall street guys they were like we've got money we're going to do a side project in our little office here in menhav we're going to set up a little cluster of these these gpus and do this eventually of course ai came along and they went into ai computing and that's when they linked up with a company called nvidia which basically uh made CoreWeave what it is today, which is possibly the largest of these so-called neoclouds, companies that rent out computing power for people who want to do AI.

4:42Asa Fitch: So CoreWeave owns the chips that NVIDIA sells them. Is that how it works?

4:47Miriam Gottfried: Yeah. NVIDIA has the chips manufactured. Those chips are put into servers. Those servers are sold to CoreWeave. CoreWeave rents out that compute or computing power, as those who have discerning grammatical taste might say, to AI labs like Anthropic, which just filed for its IPO. So that's essentially the business model. In a way, it's very simple. But when you start digging into the financial dynamics of it, it gets pretty complicated. And the folks at CoreWeave, having come from Wall Street, really understand that this is a financial business and that these GPUs are financial assets. And that's kind of one of the more interesting things about CoreWeave.

5:29Telis Demos: So what has happened in the last few weeks that has changed the market's view on CoreWeave and I would say other kind of neocloud businesses, these people selling computing power to AI? Was there something that one of them said about demand? Was there some other piece of data that came out? Like what has happened that suddenly made them this pinata being knocked around by the street?

5:52Miriam Gottfried: Well, I think the first thing to say really is that there hasn't really been a clear ebbing of demand for computing power for AI. Like, let's just state that from the start. Like, Meta Platforms has just reported its results, raised the floor of its CapEx guidance for 2026. So it's not like people are saying, oh, goodness, we don't think AI is a big thing anymore or as big of a thing. we're going to like take our foot off the gas pedal with, with AI. But, you know, CoreWeave is an interesting business because it is so connected with, um, sort of AI computing and the risky side of AI computing, like the leveraged play in AI computing, you know, it's sort of like, um, on paper, at least it looks heavily indebted.

6:45Miriam Gottfried: It looks like, you know, it's heavily dependent on NVIDIA, but it doesn't really have a huge track record as a business. It hasn't been around doing anything really before AI of any magnitude. So it's getting punished, I think, in a way, because it's the clock that ticks when sentiment around AI gets a little bit iffy.

7:08Asa Fitch: And is that reasonable? I mean, all of those things sound like something that would provoke anxiety in me if I owned that stock or the debt of that company?

7:19Miriam Gottfried: I mean, there, there, okay, look, there is some, there's some, in some sense, justification to it. But I think if you really look under the hood, it gets a lot more complicated and nuanced in that, you know, it's, Corweave is a company that's executing a financing model. And you have to think about how its financing is structured. There are different ways that Corweave raises money for itself to do what it wants to do. Okay, one of those buckets is it does these things called delayed draw term loans. That's what it did with Blackstone. Those deals are very like sort of closed off deals, because basically everything is done through a special purpose vehicle that that handles everything, even the contracts that CoreWeave has with its end customers are put into this SPV.

8:08Miriam Gottfried: and then that SPV makes money by renting out that computing power. And then some of that money gets kicked back up to Coral Weave, but the SPV retains what it needs to retain to pay off its debts and things like that to pay off these loans. So those loans are like really, if you really stand back and look at it, they're quite secure.

8:26Telis Demos: And by the way, I just want to chime in to say that the market pricing on those loans has not changed nearly as much as the pricing on stocks and other types of debt. So the market understands that those loans are more secure. I mean, it seems like it.

8:39Asa Fitch: Presumably. Yeah.

8:41Miriam Gottfried: So it's an interesting business model. It's kind of a mix of things that look pretty bulletproof from a financial perspective. And then there are other areas where it looks a little more vulnerable. I mean, bottom line, this company is heavily indebted. Like if you look at the debt to equity ratio, it's like, I don't know, it's like depending on which data source you look at is like eight or ten times.

9:04Telis Demos: What would a typical like big established hyperscaler company like Microsoft be?

9:08Miriam Gottfried: I mean, let's say Meta. For example, people think of Meta as like this, like they're kind of they've been borrowing too much and they're kind of in somewhat some kind of trouble. They're at like, I don't know, 36 percent or, you know, it's like it's less than one times on that ratio.

9:24Asa Fitch: Let's take a moment to pause right here. When we come back, we'll have more with Asa Fitch.

9:33Asa Fitch: In 1851, a group of neighbors came together to pool resources and strengthen their community. Guided by the belief that progress meant putting people first, they became mass mutual. 175 years later, we're still committed to the interests of our clients, members, and policy owners, and to supporting the financial professionals who serve them. Because when you're built to last, you don't think in quarters, you think in generations. Learn more at MassMutual.com.

10:07Asa Fitch: Talk to us a little bit more about the relationship between NVIDIA and CoreWeave, because I think that's actually a really important point. Why does NVIDIA want CoreWeave to succeed?

10:19Miriam Gottfried: I wrote a piece not too long ago saying essentially that NVIDIA needs CoreWeave, and they do. I mean, you know, NVIDIA is selling chips to a lot of customers that want to compete with it. You know, its main customers are Google, Microsoft, of course, Meta. But all of these companies have programs where they're making their own AI chips. You know, NVIDIA has a gross margin of greater than 70%. if you're nvidia's customer and you're paying this huge margin and you're a huge tech company with a lot of resources you'd be insane not to try to develop your own chips and save some money especially when people are deploying ai training and inference at this scale they're doing it at so so nvidia kind of sees a future i think where it could be a little bit in trouble if everybody kind of does their own stuff.

11:17Miriam Gottfried: So they need an outlet where, you know, they have companies to sell to that aren't trying to replace them. And CoreWeave for NVIDIA is really the primary outlet for exactly that.

11:30Telis Demos: So it needs to see companies, it needs to see new entrants in the market for computing power in order to have healthy customers, right? It needs to see the success of these guys.

11:40Miriam Gottfried: Oh, absolutely. And just think about it even broader terms. It's a power thing, right? I mean, And NVIDIA has enormous power over CoreWeave. These big tech companies don't need NVIDIA quite as much. And NVIDIA can't really call the shots with these guys quite as much. So if you're in NVIDIA's shoes, I can certainly see why you'd want CoreWeave to succeed versus these other guys.

12:01Telis Demos: But this, though, gets me back to like, so what exactly is the market's concern here? Right. Why, given what we've just talked about here, what does it mean when we see a lot of stocks? Like, and I was just, I just did a little screen of, you know, kind of all the different buckets of AI stocks. And like, everyone's down this month. Some of them more significantly than others, but, you know, the chip makers, the neoclouds, the trad clouds, you know, everyone in that business is affected. So is it a situation where people are worried about something in the future?

12:40Miriam Gottfried: Yeah, I think that's part of it. I mean, if you look at CoreWeave, for example, in its model, it's all about from the start, it's always been about, you know, how do we maximize our return by driving down the cost of capital? Driving down the cost of capital is basically the lever for CoreWeave to get bigger, to get more wealthy, et cetera, et cetera. We're in an environment where the cost of capital is not really going down. I mean, the Federal Reserve is trying very hard to raise the cost of capital right now. Yeah, exactly. Exactly. And that trajectory doesn't seem to be poised to reverse anytime soon, barring some sort of recession.

13:14So let's see where that goes.

13:18Miriam Gottfried: But in this environment we're in now, it certainly seems likely that the future is going to look more difficult for the core weeds of the world. So if these guys are looking at paying extremely high interest rates on their debt or not being able to borrow at all, what does their future look like? you know how are they going to expand you know these guys have like of course they have these contracts with these big hyperscalers that are triple a rated many there are some of them at least microsoft um just microsoft you know there's only two or three it's just microsoft yeah okay very highly rated double a it's highly rated you know well into investment grade etc but there's only so much of this activity that's going to be there for them and they need to expand their own facilities you know they need to borrow money to uh get ready for the next stage of demand if they can't do that, or if it costs too much to do that, you know, it's really, it's kind of a death knell for the business.

14:12Miriam Gottfried: I don't want to overstate it, but it's the future looks bleak. And if you're an equity investor, it's all about the future. You know, I mean, if the business just stops, you're in trouble.

14:23Asa Fitch: Is there also some sense that everything has become too concentrated as a bet on AI? I mean, I wrote in a story last month that 49 % of new investment grade bond issuance so far this year has been AI linked. So if, you know, I cover wealth management, if your wealth advisors out there telling you, oh, you need to buy more bonds and then you also own the S &P 500, you're basically your entire portfolio is like levered to AI. And so if something happens, if there's a slowdown, if there's if the bubble bursts like they're like, that's all you have. I mean, is there some sense from investors that like they need to be putting their money somewhere else right now?

15:05Telis Demos: I want to step in for a second, Miriam. I think that's a great point. And I did quote, you know, one of the top folks at PIMCO in one of my stories recently, Lodfi Karawi, talking about there being crowding out kind of within the AI trade, basically that investors now have a lot of different choices for types of credit, types of business models they can back, et cetera. And so from a financing point of view, he did think the market was going to be more discerning and start to really pick out things that, okay, we think this should have, you know, be kind of a low cost, low risk form of financing, but we think this is higher risk.

15:37Telis Demos: So I think you are seeing that. And so then I'll turn it back to Asa and I'll say, what are we seeing then? Are we seeing a preference for different business models? Do people start to think that the hyperscalers are going to be the ones that will win and that there won't be room for upstarts in computing power or something like that?

15:56Miriam Gottfried: Well, think about what are the purest, pure play AI companies on the market. Like there's the obvious one is NVIDIA because they sell chips that power AI. And that's pretty much all they do these days. They have a gaming business that's relatively small at this point. But the others really are folks like CoreWeave and Nebius. These sort of neoclots have come along. I mean, the hyperscalers of the world are like Google and Amazon and, you know, all those folks and Microsoft have much more diversified kind of businesses that have exposure to all kinds of different things. Now, we are going to see when Anthropic lists, that's another very, very pure play sort of AI name.

16:38Miriam Gottfried: But, you know, I think there is a discernment around, you know, people looking at the most sort of deeply exposed to the AI boom kind of names like CoreWeave versus places where you might have a little bit more margin of safety. Like, you know, Google, if the AI bets go wrong, they can fall back on, you know, regular cloud computing. They've got a productivity suite. They've got their huge search business that drives most of their revenue. You know, Microsoft has its installed base of corporate software. You know, all these guys have other bets going on. CoreWeave doesn't really, basically. And Nebius is the same way.

17:20Miriam Gottfried: It's another company that's kind of like CoreWeave in some ways, although less heavily indebted. These neoclouds are, you know, it's not hard. It wouldn't be hard to see a future where a lot of these guys went out of business or were just consolidated or rolled into other businesses because they have, you know, they've gotten the computing power set up. And so they have an asset that they can then use for something. But as businesses, they might not survive. I don't think that's unreasonable to think. I'm not sure CoreWeave would be the one to fail.

17:52Asa Fitch: For NVIDIA then? I mean, given what you just said about how NVIDIA really needs CoreWeave to exist, and given what you just said about how NVIDIA is also kind of a pure play AI bet at this point, like, I mean, I know it's like the biggest company in our stock market driving our stock market, but like, isn't what you just said kind of bearish for, you know, how the market might treat NVIDIA itself?

18:17Miriam Gottfried: yeah it is i mean to be quite frank it's you know nvidia's involved also in a lot of these deals that people are calling circular where they're they're backstopping guaranteeing uh you know helping people get financing and other other tools to buy their own chips uh you know that's great for nvidia it's driving it's helping boost their revenue but it's also driving you know the production and sale of chips that eventually might not be necessary or or that might have to be, you know, people might need to find new uses for them. I mean, one thing you could see is that a scenario where, you know, NVIDIA doesn't kind of lose a ton in this scenario, but some of the neoclods, their business models just don't really work out, but their assets are taken over by others who can manage them more effectively.

19:08Telis Demos: I want to pull out something you just said, which is that, you know, these assets will have value to someone, right? And that's actually what I have trouble wrapping my head around when I see people selling off, you know, kind of selectively or, you know, that, oh, this bond we've repriced, but we haven't repriced other people's bonds, is that I understand that like maybe the relative success of these folks, you know, is up for debate. But if we're living in a world where we still think that there is going to be like a major industrial revolution and, you know, transition to a lot of things that are done today by humans or algorithms are going to be done by AI in the future, that someone will have a use for these things, right?

19:50Telis Demos: That this computing power, these data centers, these chips are useful to someone. And so I would argue that that's maybe a reason why a lot of the credits here are protected by this circularity, right?

20:03Asa Fitch: If you're a bondholder, you will get paid back, most likely.

20:07Telis Demos: Someone will want this stuff, right? Or this business has value to be acquired by somebody. And I understand some of the bonds are directly secured by assets. Others are unsecured bonds. But it seems like someone will see value in this somewhere along the way, and especially from a credit point of view, which is more about the things that can go wrong versus stocks, which are about the things that can go right. I mean, I would say that that's an argument that wherever you pick to play in this, you're still going to benefit from the overall kind of success and growth of this market. And so why not buy the bonds that are -

20:42Asa Fitch: So you might as well click this high coupon that you're getting paid from a junk bond like Corweave's debt, right? Yeah.

20:48Telis Demos: Maybe that's actually, you're kind of taking advantage of an inefficiency there to some extent. Asa, does that sound crazy to you?

20:56Miriam Gottfried: No, it doesn't sound crazy. I mean, it could be a good bet. a lot of these companies are doing pretty well in the end and will continue to have assets that have value, and even if those assets have to be sold off. Now, of course, one thing that's kind of a question mark over all of this is this notion that NVIDIA wants to promote, that the GPUs, the chips are effectively commodities. I mean, Jensen Huang, the CEO of NVIDIA, has repeatedly talked about how these chips are quote unquote fungible you know it almost as if they're a currency that you can just exchange from one thing to the next that is to some degree true um you know especially you're talking about just nvidia's chips moving from one place to another one owner to another it's not always the case though because you have uh setups like you know google's tpus for example

21:51Asa Fitch: um sorry what's a tpu i'm like out of the way sorry that's it's it's an abbreviation for tensor

21:57Miriam Gottfried: processing unit, which it's basically just a, it's a special AI chip that Google has developed over the years.

22:02Asa Fitch: It was in fact, a TPU only comes is only made by Google. Is that?

22:06Miriam Gottfried: Yeah, exactly. Yeah. It's sort of Google's in-house AI chips, but they're, they're very, you know, widely used Anthropic uses a ton of them has whole data centers that are sort of, you know, dedicated to them. But, you know, those, those chips may or may not be as fungible if you will, as, as other ones because, you know, client A might not be able to go take over client B's thing that was made with TPUs, et cetera. So it's a little bit, it kind of makes sense, you know, that there is this asset that has value and will continue to have value to the future. There are some wrinkles there, you know, you can't completely say it's true as though it's just like money that you can exchange for other money.

22:47Miriam Gottfried: You know what I mean? It's not quite that clean.

22:49Telis Demos: Let's take a quick break here when we come back more with Hurt on the Streets' Asa Fitch.

22:58Asa Fitch: In 1851, a group of neighbors came together to pool resources and strengthen their community. Guided by the belief that progress meant putting people first, they became MassMutual. 175 years later, we're still committed to the interests of our clients, members, and policy owners, and to supporting the financial professionals who serve them. Because when you're built to last, you don't think in quarters, you think in generations. Mark Benthien, Ph.D.: My other question is, in general, is computing power fungible?

23:39Telis Demos: Or are there differentiations or are there other layers to what, you know, you get at if you're a customer of AWS or Azure at Microsoft or of CoreWeave or Nebius? Like maybe investors would be well served to understand some of the nuances between them and then kind of say, actually, I think what these folks do is something that is, you know, potentially has a broader application than what those folks do. Is that a fruitful thing to pursue?

24:10Miriam Gottfried: Yeah, I mean that's an enormously complex question because like in theory, like philosophically, let's say, the fungibility of everything just depends on the level of abstraction. You could have a software layer that abstracts away everything let's say. And so then everything is fungible because you tell the software to do something and it knows that, you know, AWS does this differently than this other guy. And it accounts for that. Now in reality, in the real world, it just doesn't work that way because you can't plug and play everything and expect it to work. So yeah, in reality, there is a, um, there's a competition between these companies on the software level.

24:45Miriam Gottfried: Um, and it takes many dimensions. I mean, one attraction of like an Amazon, for example, or AWS, which is its cloud computing arm, is that it has a lot of other stuff. It's not just the AI stuff. It's also linked to, you know, it has your regular, you know, non-AI computing in there. And you can access a lot of that very easily on demand. So there's a reason for people like stick with Amazon. It's kind of like your easy, you know, choice. It's like you have Amazon Prime or whatever. You easily get your paper towels from there. You don't go to like the paper towel store, you know. Some of these guys, though, say that they have software that's better than their competitors for managing workflows.

25:24Miriam Gottfried: So for maximizing your spending on AI, you have a certain amount of money you're going to spend on AI computing. And so if you can make the best use of that time and money with CoreWeave, because they have a great bunch of software that helps make it more efficient, then you're more likely to go with CoreWeave in the end. Having said that, a lot of customers of these neoclouds use their own software because they have such massive contracts and they just have massive computing setups. They need that software that they're running on CoreWeave infrastructure to work with their other stuff. So, you know, CoreWeave and other NeoClouds have their own software plays, which are important for them.

26:01But those are really more important, I think, for their like medium to smaller size customers

26:07Miriam Gottfried: because they do less of their own software development. I think a lot of them would like to kind of diversify their customer base because they've been too exposed to these big tech companies. kind of a double-edged sword because those big tech companies are so creditworthy and you can borrow more money if you have contracts with them. But yeah, it's definitely a point of differentiation that investors should think about, but it's complicated. You can really get in the weeds in it. But that's a great point because it's a question of in the years to come, will

26:39Telis Demos: every business need to develop its own kind of AI strategy and things like that. And as the ability to use it kind of filters down from like the very largest companies, right? Like, you know, obviously, whatever, JPMorgan Chase has, you know, like an IT budget the size of a small country. And so like, of course, it will be using AI and it has the ability to like shop around and say, oh, let's try this for that and that for this. But as you get into, you know, mid-cap companies and then, of course, like smaller businesses. Like maybe there is just a long enough tail of these things that there's a place for all of these different AI players to, you know, to serve a certain segment of customers or something like that.

27:22Telis Demos: So I guess it all, again, it comes back to like what do you think the future of the AI business is. But I think the bigger point that you're making and Asa's making is that this this sell off in CoreWeave debt is not the investors saying we don't believe in AI and the growth of the

27:43Asa Fitch: AI industry and the need for financing. We are just kind of maybe like a little bit nervous on the edges about what the future could hold for certain AI business models.

27:56Miriam Gottfried: Yeah, I think that's exactly right. I mean, that's sort of where the, you could see it as an indicator of the broader market, for the broader market, sentiment in the market, if you will. I think CoreWeave and these other neoclouds are sort of barometers of that sentiment in some ways. So there is maybe a little bit of a broader bleed over to other things. But I think what you're saying is exactly right. There's nothing fundamentally here that says the AI boom is turning to a bust today. You know, even though the bond prices are going down and the share price is going down, you know, we've had these moments in the boom.

28:33Miriam Gottfried: I think we've had them. I can't count the number of times where we've had these like, you know, sort of market swoons where people get all up in arms about, you know, oh, goodness. Everybody's trying to figure out where the top of this thing is and nobody seems to be able to call it right. So, you know, I think you'd be hard pressed to find somebody who'd say definitively like this is going to be the one. But clearly there's some there's definitely some fear out there. And, you know, it's it's always been somewhat justified because of how much money is pouring into this thing. There's no real commensurate amount of returns coming in.

29:08Miriam Gottfried: So it's like, you know, the writing is sort of on the wall, but it's been on the wall for a long time. And maybe it will be on the wall for some time to come.

29:15Telis Demos: I guess what we're saying is that if I can summarize. Feel free to disagree.

29:20Asa Fitch: Tie it all up in a neat bow for us. Feel free to disagree.

29:22Telis Demos: But it sounds like what we're saying is that what we've seen over the last few weeks is a lot more about financing and the markets and interest rates than it is about, like you said, Asa, this isn't the moment of the bubble burst and that nobody wants AI and that the whole thing is kind of been overstated from the beginning. that this is a bump along the way more so than it is a real kind of fundamental turning point.

29:49Asa Fitch: It's also, I mean, in line with interest rates, it means everything's going to be more expensive. Like we know there's these huge costs. And the fact that the market now wants more, you know, more yield to cover this means that it's probably going to be expensive. It's going to be even more expensive for these companies to finance their AI build out than it was before. The fact that we've seen longer term bond yields like 10-year treasuries and stuff start to decline, like they've had some big down to yields have dropped sharply in some recent sessions.

30:21Telis Demos: And I think what that tells you is that maybe what we're learning here is that as the Fed ratchets interest rates a little higher, people were at the beginning of that saying, well, that won't do a thing because AI, there's so much demand for AI and AI financing. people will pay any price like the Fed's going to have to hike interest rates to 10 percent to have any chance of slowing AI. Well, maybe that's not quite the case. And so what the longer term bond market is telling you is like, OK, maybe actually the Fed doesn't have to do that much more tightening before we do see things start to slow down.

30:51Telis Demos: And so we don't need to buy 10 year bonds like these huge yields because, you know, the Fed's going to stop maybe at an earlier point.

30:59Asa Fitch: I like that idea.

31:01Miriam Gottfried: Well, I think that one of the takeaways here is that is that capital is king, right i mean the capital window is is wide open people can raise money i mean anthropic can at least they hope you know they're they're hoping to raise up to 100 billion dollars in this in this ipo so what matters now for these companies like corley like anthropic like others is capital and if you look at other players like meta platforms wants to you know they're raising 35 billion in equity at the moment. And they plan to raise another$60 billion through at-the-market transactions, meaning basically they just issue stock at different times strategically.

31:39Miriam Gottfried: You know, this is a lot of money. It's a lot of money to ask investors to give them, essentially, to pay for AI. So there's this wave of people begging for money, if you will, and investors, whether you're a credit investor or on the equity side, having to weigh whether to give them that money or name your price for giving you that money. So I think that's the thing I'm watching the most here looking forward, especially as we get into Anthropics IPO, which probably doesn't happen the next couple of weeks, maybe two or three weeks or something like that. That's really the next big question, I think.

32:18Miriam Gottfried: How much capital can you raise? What price?

32:21Telis Demos: I would say if they can raise equity, like if Meta can really raise all that equity, if Anthropics can go public successfully, Like if the equity markets are there for everyone, that's a pretty good sign for the bond market because that's just more cushion for bondholders who are in a capital stack above the shareholders. And raising equity, I mean, just to put that plainly, that means people are giving them tons of money, right? The worry about, oh, they have negative cash flow. They don't. Well, one way to get cash flow is to sell stock. And people give you a ton of money, right? And so if they can successfully do that, I don't know, maybe that's a sign that this thing can go on further.

33:00Asa Fitch: And we'll have to see. As I've said, people I talk to are still super excited about the Anthropic IPO. There's a lot of buzz around it in the retail investing community. So I don't know how institutional investors are thinking about it, but I think it could go well.

33:17Miriam Gottfried: Yeah, it seems ludicrous. I mean, think about when was SpaceX's IPO? That was only a number of months ago. You know, it's like it's the summer. So it's like it just seems ludicrous that people can just keep raising money, but they can't. So I wouldn't say that it's over and this IPO is not going to be met with sufficient supply of capital. But, you know, it seems like there's a lot of there's a lot of money that needs to be raised. All right.

33:47Telis Demos: Well, we'll have plenty to talk about in the years and months to come on this. Asa, thanks so much for joining us. Thank you.

33:58Asa Fitch: And that's everything you need to know to take on your week. The show is produced by Alexis Moore and Michael LaValle. Michael LaValle is our sound designer. He also wrote our theme music. Aisha Al-Muslim is our development producer. Chris Sinsley is our deputy editor. Lee Tal Malad is our senior director of shows. and Samantha Hennig is the Wall Street Journal's head of multimedia. For even more, head to WSJ.com. I'm Miriam Gottfried.

34:23Telis Demos: And I'm Telestivos. Until next time. You know, I think we probably... Sorry, I'm getting lost in my... These are tougher to do at 4 p.m. than they are at 10 a.m. We need your coffee. Woo!

34:44Asa Fitch: Harvard Business School Executive Education delivers a world-class learning experience that energizes aspiring and established changemakers. Prepare for the next elevation for your organization and for yourself. Learn more at hbs.me slash breakthrough. That's hbs.me slash breakthrough.