Loading summary

Transcript52 lines

[00:00]
Capital One Advertiser
Think of a toilet as just an everyday object. Your entire perspective shifts the moment you experience a Kohler Smart toilet. Design changes everything, transforming a basic routine into something extraordinary with standout aesthetics and intuitive controls. These are functional works of art. Stunning design that never sacrifices performance. Enjoy customizable features for elevated comfort and convenience. It's more than a fixture. It's a cleaner routine and a more refined space. Experience the difference of Kohler Smart Toilets. Find more@kohler.com this episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love. It helps schedule a test drive, get pre approved for financing and estimate trade in value add. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.
[01:11]
Rob Guenther
It has been a wild week for the stock markets. The S&P 500 reached record highs, fueled by what sounds like good news coming out of the Strait of Hormuz. Although it does feel like we've been here before, but okay. And tech companies posted some strong earnings, so. So that's reassuring for any investor worried that the AI boom is going to stop booming. But if you go back not even that far, like a week ago, the picture wasn't so rosy. South Korea's stock market tanked, which pushed the NASDAQ way down toward corrections territory. This pattern has been on repeat all summer. Up, down, up and down. What's going on?
[01:48]
Ed Elson
This feels to me like the most volatile market of all time.
[01:53]
Rob Guenther
Ed Elson Co host, the Profit Markets
[01:55]
Ed Elson
Podcast I think what's happening here is that because the prize of AI is so gigantic, but also because the possibility of of the destruction if the AI thing doesn't work out, because that possibility is so gigantic as well, you have huge amounts of volatility. Well, these are massively money losing businesses. OpenAI is one of the most unprofitable companies in the history of companies. This is a company that lost more than $20 billion last year. That is based on their leaked financials. This is a company that is spending more than the US Government spends on the FBI and NASA combined. So you have to start asking the question, what happens if the money runs out?
[02:41]
Rob Guenther
Today on the show, did somebody say AI bubble? Don't tell the markets. I'm Rob Guenther filling in for Lizzie o'. Leary. And you're listening to what Next? Cbd, a show about technology, power and how the future will be determined. Stick around. So, Ed, we're talking on Wednesday morning this week. We've seen the S and P reach new highs. What's driving this growth?
[03:08]
Ed Elson
It's hard to fully understand at this point because we're kind of reacting in real time here. I think this is a lot of optimism coming off of a lot of pessimism that investors were feeling in the last couple of weeks. Basically the past month, a lot of concerns related to AI, related to the amount of leverage that was being poured into the AI trade. And we saw the ramifications of that in what was a pretty brutal market for the Nasdaq and also for the South Korean stock market, which is heavily leveraged on the AI trade. So we can maybe get to that in a moment. But I think what we're seeing now is investors are looking for any reason that they can to feel optimistic again and to believe that this AI thing is going to work out. Because, yeah, there are a lot of warning signs and there's a lot of instability right now. But the other thing that investors know is that if it does work out, if things happen the way they'd like, well, this is a multi trillion dollar opportunity. So I think investors are looking for reasons to feel optimistic again. They're looking for reasons to believe that this will work out over the long run. And I think that's what we're seeing this week.
[04:19]
Rob Guenther
You're talking about the NASDAQ going down recently. To me, that screams volatility. Do you see this market as somewhat volatile? And if so, do you think investors are starting to price volatility into their business?
[04:33]
Ed Elson
This feels to me like the most volatile market of all time. But I have to be cautious when I say that because most markets feel like the most volatile market of all time. But this market especially feels that way. I mean, if we just look at big tech earnings, which we saw just last week and the week before, and we look at the way that some of the largest names, some of the largest, most valuable companies in the world, the way they whip soared up and down. We looked at Amazon, for example, which literally rose 15% in a single day. It's kind of unheard of for a company that's worth literally trillions of dollars. We saw similar things with Meta. We saw similar things with, with Google and Microsoft. I think what's happening here is that because the prize of AI is so gigantic, but also because the, the possibility of, of the destruction if the AI thing doesn't work out, because that possibility is so gigantic as well, you have huge amounts of volatility. You do not have a real sense of consensus in the markets right now. Investors say, I think AI is going to be worth X trillions of dollars. I think it's going to be worth nothing. And no one really knows right now because we're all just making bets on the future. We're all speculating at this point. We haven't seen a genuine ROI on AI investment so far and people will debate me on that. But if we just look at some, some of the end products, if we look at OpenAI, if we look at anthropic, if those are sort of ground zero for what the AI future might look like, well, these are massively money losing businesses. Open is one of the most unprofitable companies in the history of companies. This is a company that lost more than $20 billion last year. That is based on their leaked financials. This is a company that is spending more than the US Government spends on the FBI and NASA combined. They spent $34 billion last year. And even that isn't enough. And we're seeing this now. We saw just this week that Anthropic is looking to receive a $36 billion loan from Blackstone because the nearly $170 billion that they already raised in the private markets, that's not enough money. That's not enough money to cover the cost of simply selling their product. So what we're starting to see here is, yeah, on the front end of AI, this business has not been proven whatsoever. Sure, you can, you can build data centers and you can sell the compute to those companies, but that business only works so long as an OpenAI and an Anthropic still exist. And a lot of these big tech companies are becoming increasingly reliant on those two AI startups for their growth. So you have to start asking the question, what happens if the money runs out for OpenAI? What happens if they're no longer able to go to the debt markets and raise tens of billions of dollars at a time? That's a real concern.
[07:35]
Rob Guenther
Well, I guess my question based on what you said is what is the timeline there? Because from my perspective right now, these companies are, they're chugging along on the faith that something good will happen in the future. Do we have a sense of how long that Runway is? How long can these companies maintain the status quo just by promising something shiny in the future?
[07:54]
Ed Elson
That is the multitrillion dollar question that Literally no one knows the answer to. But what I can tell you is that there is a timeline and there's gonna be a deadline on this. The optimism cannot exist forever. We've seen this with asset bubbles throughout the history of markets. There are a lot of people, what I can tell you, there are a lot of people who say, don't worry about it, they'll get there eventually. Just give them some time, it's gonna happen. Don't worry about the timeline. That's not a problem. Those people are for sure, certainly 100% wrong. There is a deadline on this. They cannot continue to raise hundreds of billions of dollars ad infinitum. And by the way, their next, their next place to go in order to get the money to continue building this dream of an AI future is the public markets. But what we are starting to see now is that they're actually getting anxious about that themselves. OpenAI had plans to go to the public markets, to go public, to issue an IPO and raise money there. But then they table the. Those plans because I think they probably realized, actually, we're not sure that we're going to be able to convince enough investors in the public markets that this is worth investing in. I think they probably saw some of the reaction to their leaked financials. They probably saw the way people were talking about OpenAI and the anxiety about it. And they said, you know what, let's, let's hold off for now. Let's continue searching for money in the private markets. Let's maybe go and figure out some debt financing as Anthropic is doing with Blackstone and Apollo. And they said, let's table this for now. And I think the real sort of canary in the coal mine here when it comes to AI existing in a fully disclosed world of public markets. If you want to get a sense of how an AI company like an OpenAI might trade on the public markets, how the public markets would receive that company, just look at SpaceX, which was cut in half in less than two months. And we're recording this on Wednesday, August 5th. SpaceX has lockup periods that expire tomorrow, which means that a lot of the insiders are going to be allowed to sell legally for the first time ever since the company went public. Which means we're about to see even more selling pressure on this stock. Billions of dollars of SpaceX shares that will likely be sold by investors who invested a long time ago and want to cash in because they want, they want their paycheck.
[10:25]
Rob Guenther
Yeah, I want to drill down on that because you had a guest on recently who said that even if everything goes right for SpaceX, which includes, you know, building these AI compute centers in space, which seems like an unrealistic proposition so far, even if that comes to light, you say he, your guest told you that the valuation should be like a quarter of what it is right now. It's just so overvalued.
[10:45]
Ed Elson
I mean, this is the guess you were referring to as Nicholas Owens. He's an equity research analyst at Morningstar and his fair value estimate. And this is not an ideological and biased person, this is someone who has simply run the numbers, who actually has a pretty middle of the road view on what AI could do. I mean, I'm someone who thinks that OpenAI and anthropic are running into trouble. His belief is that they're going to be okay. And yet he believes that the fair value estimate for SpaceX is $62 per share.
[11:23]
Rob Guenther
And it was up as high as
[11:24]
Ed Elson
it went to 225 was its high. And when that company, when that stock went out and when I saw the IPO prospectus, the original IPO price was $135 per share. It went out immediately at around 150. It went up on the first day, hitting 225. My prediction before the company went public was that the stock was going to rise 25%. That happened. My second prediction before the company went public was that then it would get cut in half. That did happen. It went down to around 107. We've risen back up a little bit more because they reported earnings recently which were decent ish. I mean, behind some of those decent ish numbers are some scary numbers related to the amount of money that they are spending to build the data centers to build the AI. But that stock has gotten hammered and a lot of retail investors have lost a lot of money on that thing. And it's largely because they were sold a dream. A dream that didn't really have any basis in reality.
[12:34]
HeyGen Advertiser
This episode is sponsored by HeyGen. HeyGen turns a script, photo or presentation into a polished video of you in minutes. No camera or crew required. It's how real estate agents, financial advisors, attorneys and creators show up consistently on every feed and build a brand that brings in clients without spending their whole week making content. Rated the number one AI video platform for small businesses on G2, with the most realistic AI avatars in the industry. Over 30 million people use Heygen, including 85% of the Fortune 100. It's built to work in over 175 languages so you can reach your audience wherever they are. Here's the wild Record yourself for just 15 seconds and Heygen builds an AI avatar that makes professional videos for you on demand so you can post everywhere your audience is as yourself without filming every time. Your first three videos are free@heygen.com pod that's H E Y G E N.com pod starting your own business is never easy. Starting your own podcast? That seems easy, but actually there are a ton of landmines to step on along the way. Finding producers, selling ads, and connecting to WI fi. Oh, does that sound straightforward? It's not. I'm talking about sitting in coffee houses for hours after buying one scone. I'm talking about sitting in hotel lobbies and pretending your backpack is luggage. It's torture. I spent so much time making my home office look professional, but my connection didn't get the memo. The last thing you want during a major interview is for your guest's voice to turn into a stutter. When your bandwidth can't keep up with your ambition, your home office starts feeling like an amateur operation pretty fast. And for a podcast, the Internet is key because the Internet is how we talk to almost everyone. And no matter the guest, a laggy connection can ruin an exclusive interview. Great connectivity isn't a bonus, it's the whole game. And ATT Business is here to help. They've got the tools, team and expertise you need for a stable network you can rely on. And when you can rely on the network, you can get back to thinking about the more important stuff, like nabbing that great guest and getting back to work at AND T Business. Built to Work get at AND t business@business.att.com. This episode is brought to you by Bill, the intelligent finance platform that helps businesses and accounting firms scale with proven results here at what Next TBD we know business. And in businesses across America, smart people are stuck doing the grunt work. You know the drill. Those hours when you could be brainstorming big ideas, you're instead filling in spreadsheets, filling out invoices, or hunting down somebody else's signature. Bill wants to change that. With AI powered automation, Bill removes the busywork from your account's payables workflow. They handle capturing invoices or routing approvals and syncing with your accounting software so that your team can focus on growth instead of paperwork. Bill is so reliable, 98 of the top 100 accounting firms in the US trust it to simplify and secure their bill payment processes. Bill's handled over a trillion dollars in secure payments and is ranked number one overall on G2's 2025 list of best accounting and finance products. So stop the guesswork and start scaling with the proven choice. Go with a company whose financial infrastructure is trusted by nearly half customers. Ready to talk with an expert? Visit bill.comproven and get a $150 gift card as a thank you. That's bill.comproven. terms and conditions apply. See Offer page for details.
[16:20]
Rob Guenther
You had mentioned asset bubbles before. I didn't want to say the bubble word. It sounds a little, maybe a little cliche. Only because I feel like we've been hearing about an AI bubble for over a year now. Right. And I'm just thinking of. Paul Krugman recently talked about the end of this optimism that you're kind of describing right now. He said that this is like not the beginning of the end, it's the end of the beginning because these are such slow moving phenomenon. What are your thoughts on that?
[16:45]
Ed Elson
I think that's probably right. I mean, I can understand why someone would feel frustrated by the idea that we keep on hearing this word bubble. And you're right, we have been hearing it for, I would say it started to be thrown around in September of last year or in August of 2026. So it's been almost a year of people saying, is there a bubble? Is there not a bubble? Are we starting to see signs of a bubble? And so I'm sure that people look at the markets and they see the markets going up and they go, you guys are wrong. There is no bubble because the markets are up today and the markets are up from where they were a year ago. That doesn't negate the possibility of there being a bubble. Bubbles can inflate over long periods of time. They can also deflate over long periods of time. Just because the markets have increased since we started hearing murmurs of there being a bubble doesn't mean that there isn't a bubble. It doesn't mean that asset prices have become inflated to a, to a level that is not based on the reality on the ground.
[17:57]
Rob Guenther
So if we try to define like this bubble, right? The idea, I think we've already said it, but the idea is that we just have billions and billions of dollars going into this idea that AI is going to become this product that just generates way more money than it than it's showing the potential to right now. Let's say I'm not a retail investor. Let's say I feel like I don't have any skin in this game, should I be freaked out? Anyway,
[18:21]
Ed Elson
it's tough because I think the bubbles are located in different pockets of the market and the market is grappling with this question in real time. I mean, the nice thing about this bubble, and I believe that it is a bubble, the nice thing about this one is the markets aren't totally clueless here. If you go to 2008 and you look at the housing bubble that developed, there was such an amount, such a, a large amount of misinformation and obscurity that led the entire market in one fell swoop to believe something that was totally wrong. I don't think that's exactly what's happening here. I think that people are. I think investors in some corners of the market are somewhat aware of what is happening here. And then there are other corners of the market where they just have their blinders totally on and they don't really care about the fundamentals. They don't really care about what might be going wrong over at the Frontier Labs. And I think that you start to see that in some of the more obscure AI names that have become more popular over the past year or so. Some of the, The Neo Cloud names like Core, Weave and Nebius, some of the chip names like sandisk. There are a lot of these stocks which have just become synonymous with AI is going to the moon. And you have a lot of investors who are just covering their eyes and just saying, buy, buy, buy. I don't really care. We did see the ramifications of that, by the way, last week in South Korea because of course, the top two stocks there are.
[19:47]
Rob Guenther
Yeah, let's dig, let's dig into that. What happened in South Korea, I know that so much of the value of their stock market was wiped out. It sent the NASDAQ briefly toward correction territory.
[19:57]
Ed Elson
First off, the South Korean stock market got eviscerated. It fell about 44% in 40 days. It wiped out $2 trillion in market value. There were two main reasons for that. And the reasons are named Samsung and SK Hynix. And these are these two chip stocks which became extremely valuable in a very short amount of time because of their role in AI. Because everyone's trying to build a data center because there is a shortage of, of chips to go into those data centers. Samsung and SK Hynix make those chips so they are able to ratchet up their prices well while selling a lot of. A lot of chips. So their businesses are flat out crushing right now. The trouble is that a lot of investors are now extrapolating that into the future and going, oh my God, these businesses are going to crash for the next 2, 3, 4, 5, 10 years. They believe that this is perpetual, this is never going to end. Which is a bold prediction to say the least. So those two stocks absolutely skyrocketed and eventually they breached roughly half of the value of the entire South Korean stock market. They're worth around 55% of the entire stock market. So this is the most over concentrated stock market in the world. Suddenly we started to see some murmurs of how the AI trade might not be working out. We saw some sketchy stuff with some circular financing over in video. We saw some kind of shady levels of debt that were being raised. Investors got a little bit of the jitters and that sent the AI stocks back down again. The thing that was really the problem was leverage, and that is that the South Korean retail investor community had gotten so excited by this idea of leverage, which is where you basically just borrow money to amplify your returns. You, you can do 2x leverage, you can do 3x leverage, you can do 5x leverage. They employed this strategy of leverage. What they forgot about is that it also amplifies your losses. Yes, it can. To your return. But also.
[22:04]
Rob Guenther
Yeah, let's like. I would love to like translate this for like a non business audience. My understanding is that in May, South Korea was given the ability to do what, what you're talking about. These leveraged ETFs, they're allowed to borrow a lot of money to make a bet, and then if they win, which you know, the whole market had been winning on AI, they'll get a bigger, they'll get a bigger piece of the pie. Right? And what you're describing here is when you have too much leverage though, like when things turn on a dime, you could just like get wiped out, which is what happened.
[22:35]
Ed Elson
360,000 accounts in South Korea went to zero because they became so hooked on the leverage, they became so obsessed with it that as soon as the stock started to enter a downturn, suddenly they were getting margin calls from the bank and they had to give up all the money back that they borrowed because they had levered up so high. There were 1.2 million South Korean investors that received margin calls. It's more than 3% of the entire adult population. This is a nation that became addicted to a financial instrument that is really akin to gambling.
[23:09]
Rob Guenther
So South Korea, everybody's pissed off. They've, they've halted trading. My understanding from watching your content is that the government has banned these leveraged ETFs. Right. But you've called this crash a warning for U.S. investors. And I want to, I want you to articulate how that is so.
[23:25]
Ed Elson
Well, if we look at what brought down the stock market in South Korea, it was an over concentration into a handful of AI stocks. It was a prolifera proliferation of leverage because they got obsessed with the gambling. And there is one other aspect to this which I think is relevant, but it's more of a cultural sociological point, which is that South Korea has a population of young adults who are aimless and lonely. And we are seeing this in the fertility rates in South Korea, which are, I believe they're 0.7 right now. It's the lowest fertility rate of any OECD D nation. Their population is entering into a state of structural decline. Their marriage rates have fallen 40% over the past year. There is a very serious loneliness crisis that is developing in South Korea. And it's also related to the fact that housing prices in South Korea have risen. Inflation is also rising. Young people are generally struggling economically. And so there is a large swath of the population, the young population specifically, that is sitting at home, especially if you're a young man, because young men are more interested in these and more have more of a penchant towards these gambling tendencies. They're sitting at home and they're gambling on these 2x and 3x and 5x leveraged ETFs. And when you look at the, the demographic makeup of the accounts that were wiped out, a significant portion of those accounts were young people. It's people under the age of 30. So I believe that that is a big contributor to the problem, which is that you have a lot of young people. They don't feel that they have economic prospects. They don't feel that they have much of a purpose, much of a mission in life. And so they see these opportunities for 200% returns, 2000% returns. This is how I'm going to do it. This is my lottery ticket out to make my life better. And they get it hooked on these strategies. And it's not entirely their fault, by the way, because also they're being sold these strategies. They're being, it's predatory dollars that are going in to putting these leveraged ETFs out there and getting people to believe that this is your ticket out. This is how you're going to make money. Now, why is that a warning to the U.S. well, I look at the U.S. and I see all of the same trends. I see a population of young people that is unprecedentedly lonely. One in five of us young people, I'm Gen Z. One in five of us say we have zero close friends whatsoever. I look at a market that is unprecedentedly over concentrated into a handful of names. The top 10 stocks in America now make up 40% of the entire stock market. 30 years ago that number was 20% given. Less concentrated than South Korea, but still very very concentrated. And then we're also seeing a rise in leverage, specifically leveraged ETFs, which have jumped from $120 billion in AUM in April to more than $200 billion today. It's up nearly 70% in just a few months.
[26:42]
Capital One Advertiser
This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One. This message is brought to you by Cologuard. If you recorded songs off the radio for a mixtape, it might be time to screen for colon cancer. The Cologuard test is non invasive, requires no special prep or time off, and you start screening right from your home. It's just three simple setup, sample and ship. If you're 45 or older and at average risk, ask your doctor about the Cologuard test, available by prescription only. Learn more or request a prescription@cologuard.com podcast.
[27:47]
Rob Guenther
All this behavior that you're identifying, all this predatory behavior and our sort of being habituated to participate in it, is this the type of activity that's sort of providing the scaffolding for the AI boom here in the US
[28:01]
Ed Elson
that's the scarier implication and that is the scarier implication of South Korea, because that is what created those phenomenal returns in South Korea. Yes, there was AI optimism, but at the same time there was huge amounts of leverage that was juicing the trade upward when things were going well. And it's, you know, if, if people are making 2000% on Dogecoin or Comerocket and then losing all of their money like it's not a good thing. But at least it's not really systemic to the wider stock market. At least that's just happening out in crypto land and people are kind of messing around with those nonsense coins. But in the case of, of the AI trade, which is not just juicing the stock market, but it's juicing the economy itself. I mean, this is not Samsung SK Hynix. These are not meme stocks per se. Yes, they have meme like qualities, but they also make up literally half of the entire Korean stock market. And they're also inking deals with some of the most important systemic companies in the world. They're inking deals with Nvidia, they're inking deals with the big tech companies. So these companies and these stocks actually matter. And so when we use financial products and we use leverage and we use all of these different derivatives and instruments and vehicles to make the line go up higher, we're starting to play a very dangerous game which could result in the evisceration of a lot of portfolios and retirement accounts and pension funds, all of the things that actually do matter in the real economy.
[29:36]
Rob Guenther
So you've identified leverage, you've identified this gambling on your programing. You've also identified a couple of other phenomena that seem to be playing a big role. One is through these SPV special purpose, special purpose vehicles. And then the other one is called circular financing. Maybe let's start with circular financing. What is that?
[29:54]
Ed Elson
So circular financing is a phenomenon that arose actually more than a year ago. And it was the reason that a lot of people started to get worried about there being a bubble in AI, and rightly so. Essentially what you do say, I'm, say I'm Google. We'll use a recent example. I'm Google and I want to increase my revenues this year. What I'll do is I will invest $10 billion into an AI startup called Anthropic, and then a few months later, Anthropic will turn around and they'll buy $10 billion worth of my AI chips. So what happened? There is one. Google looks good because they're investing in Frontier Technologies and they're participating in the upside of Anthropic. And they also look good because their revenue went up by $10 billion. They sold $10 billion more worth of chips. So on the surface, it looks really good. The trouble is the money's just moving in a circle. There was no actual new value that was created. I'm just sending you the money and then you're wiring the money back to me, which creates a false sense of growth, a false sense of business, because
[31:01]
Rob Guenther
the numbers do go up and people do get returns. Is that the idea?
[31:05]
Ed Elson
They don't look enough into the numbers, they don't register the fact that actually the money is moving in a circle. And so how sustainable is that revenue growth? Yeah, I can just send you money and then you can send it back to me and I'll say, look, my revenue went up, but if you're going to keep on doing that, then I mean, eventually the revenue is going to run out and eventually people are going to realize that actually we're not creating new economic value here. So this was the concern and it was a strategy that was employed by many big tech companies.
[31:33]
Rob Guenther
And it's one thing when we know what's happening, right, we can see these trades taking place. But there was another piece of news recently, and it was, it touches on the SPVs, which is the second thing that you, that I learned from watching your show. But it turns out that a lot of these companies had even more debt than we imagined. So it was debt that was hidden. It was a massive amount of debt. And they're all being done through these vehicles called SPVs. What's going on there?
[31:56]
Ed Elson
Yeah, so one of the reasonable bull cases in the markets right now is that you could say, okay, yeah, they're spending all of this money on AI and who knows if the AI is going to generate a return, but at least they're not taking on too much debt. And that was a fair argument that I subscribed to for a long time. Because the reality is these big tech companies, they are cash cows. They have huge amounts of profits and yeah, they have money that they can spend. And it's, it's up to them to spend whatever they want to spend that money on. They had about $1.4 trillion worth of debt on their balance sheets, which sounds like a huge number, but compared to the profits that they're making, it's sustainable. Then we learn that actually there's $1.7 trillion worth of debt that is off of the balance sheet that is not being reflected in the reporting and the earnings reports from these companies. The $1.7 trillion worth of debt is being accumulated in these financial instruments called special purpose vehicles, which are now the main way by which data centers are built. You essentially create a shell company, an spv, a special purpose vehicle, and then you find debt from somewhere else. It's usually provided by a thing called a private credit fund, which is what it sounds like. It's a fund that goes and loans out money in the private markets. And this has become the principal way by which data centers are built. And I should be clear here that's not debt that is being accumulated by the big tech companies. For the most part they are off the hook. So it's not, we shouldn't pull our hair out and say, oh my God, the world's going to end. But what we should recognize is that big tech is making an intentional decision to offload the risk of the data centers, throw it off their balance sheets and pass it on to someone else. And that someone else is in most cases private credit funds, which has exploded into a huge industry. The point being this AI buildout is accumulating vastly more amounts of debt than people are willing to acknowledge. It's not just the debt that's on the balance sheets of the companies, it's the debt that's buried away in these no name SPVs that no one has much clear insight into. They don't really know what it actually is. They don't really know who's actually financing it. That's where the AI boom is happening. And the less clarity you have, the less transparency you have, the more likelihood there is that people aren't going to get their numbers right, they're not going to do their underwriting properly, they're not going to do their homework. And that's how bubbles rise. This is what we saw in 2008, it's what we saw in 1999. Going into 2000. It's very likely that this could happen again.
[34:43]
Rob Guenther
Well, help me understand, if we use maybe one data center as an example, I'm thinking of last year Oracle made this $300 billion data center deal with OpenAI. And I think you alluded to this earlier, but deliveries and commitments on that deal, they're supposed to kick off next year. Like what happens, what happens when that deadline happens?
[35:04]
Ed Elson
So in Oracle's case, Oracle is actually the odd one out here because Oracle is one of the few, one of the, it is the only big tech company that is accumulating the debt on its own balance sheet. So that's an example of a company where you can actually see all of the risk they're inking these deals with with OpenAI and all of these other AI companies. They're spending more on AI data centers currently than they actually generate in revenue. So they are, talk about leverage. They are levered to the hilt. They are like South Korea's retail investor times a billion. It's reflected in the price because Oracle stock has been absolutely pummeled.
[35:47]
Rob Guenther
So it's baked in already.
[35:48]
Ed Elson
And the price, it's baked in the price of their credit default swaps. Which is basically insurance on the possibility that Oracle just defaults on its debt and therefore just goes bust. Those have been soaring recently as investors price in the increasing likelihood that Oracle could literally just go bust.
[36:05]
Rob Guenther
So it's not that Wall Street's ignoring the debt, it's sort of that they're aware of it and they're trying to like, bake it into their projections.
[36:11]
Ed Elson
They're aware of it because Oracle is reporting it. They're not taking the SPV route. And that's why I get anxious.
[36:17]
Rob Guenther
Oh, because of the SPV route, There's lots of vehicles that you have no idea what's going on.
[36:21]
Ed Elson
Yes.
[36:21]
Rob Guenther
Okay.
[36:23]
Ed Elson
That's why I get worried with Oracle, because I look at what's happening to Oracle right now. I look at how investors are reacting to what they're doing with their debt, and then I think, well, where else? Why isn't that being priced in elsewhere? I mean, there's huge amounts of debt being accumulated everywhere else. Why aren't we seeing it? And the answer is it's being stuffed into the SPVs and the private credit funds, which are not publicly traded. And so we don't get a proper markup and we don't get transparency into what's actually happening there. So this is the problem with bubbles that I should emphasize. It's not necessarily that people are doing something illegal or they're trying to pull a fast one on us. What happens in financial markets is people come up with these complicated financial derivatives of derivatives of derivatives, which makes the entire market increasingly more complicated, which means that fewer investors are actually doing their proper homework as to what's happening. That's how markets become disjointed from reality. And we're increasingly starting to see that in these financial markets.
[37:24]
Rob Guenther
And Elson, thank you so much for joining us.
[37:27]
Ed Elson
Thank you very much for having me.
[37:28]
Rob Guenther
Ed Elson is a co host of the Prof. G Markets podcast and that's our show. What Next is produced by Evan Campbell, Madeline Temzu Charme and Patrick Fort. Paige Osborne is the senior supervising producer of what Next and what Next tbd. Mia Lobel is the executive producer of podcasts here at Slate. Ben Richmond is the senior director of podcast operations. And I'm Rob Guenther filling in for Lizzie O'. Leary. You can find me on BlueSky. I'm Blue One, Rob Gunther. Thanks for listening. I'll talk to you soon.
[38:00]
Capital One Advertiser
This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.
[38:29]
Rob Guenther
If you're looking for something predictable and
[38:31]
Ed Elson
low friction over the long term, public
[38:33]
Rob Guenther
markets may be a suitable option. But the moment you want different investment characteristics, it may be time to consider the private markets. See how your wealth can work smarter@creativeplanning.com access.

Summary6 min read

What Next: TBD – "Artificial Intelligence, Very Real Debt"

Podcast: What Next: TBD by Slate Podcasts
Host: Rob Guenther (filling in for Lizzie O’Leary)
Guest: Ed Elson (co-host, the Prof G Markets podcast)
Date: August 7, 2026

Episode Overview

This episode plunges into the volatile world of artificial intelligence-driven markets, focusing on whether the AI boom is propped up by unsustainable debt, risky financial practices, and investor optimism untethered from fundamentals. With Ed Elson as guest, the conversation critically examines the mechanics behind AI company valuations, massive losses at leading firms (like OpenAI), the dangers of market bubbles, and how similar factors recently triggered a historic crash in South Korea. The discussion also warns of systemic risks posed to the broader US economy if parallels continue.

Key Discussion Points & Insights

1. The Current AI Market: Volatility & Optimism

AI Boom Drives Markets:
- Despite recent turbulence, stock markets (especially the S&P 500) hit record highs, powered by optimism in AI tech companies.
- However, just a week prior, pessimism prevailed when the South Korean market crashed, dragging down the NASDAQ.
- Quote:
  - “This feels to me like the most volatile market of all time.” – Ed Elson (01:48 & 04:32)
The Prize vs. The Peril:
- Chasing AI’s multi-trillion-dollar promise brings wild swings.
- The biggest companies are “massively money-losing,” yet their future value is hotly debated.
OpenAI’s Finances:
- “OpenAI is one of the most unprofitable companies in the history of companies. This is a company that lost more than $20 billion last year. That is based on their leaked financials. This is a company that is spending more than the US Government spends on the FBI and NASA combined.” – Ed Elson (01:54, 04:32)

2. Time Horizons and Market Bubbles

Optimism Cannot Last Forever:
- AI startups and their investors are betting that future returns will materialize, but there is a finite runway for these hopes to deliver.
- OpenAI paused IPO plans due to investor wariness about its business fundamentals—highlighting growing caution even among insiders.
- Quote:
  - “There is a timeline and there's gonna be a deadline on this. The optimism cannot exist forever.” – Ed Elson (07:53)
SpaceX as a Cautionary Tale:
- SpaceX’s IPO is referenced as an example—stock soared on hype (~$225/share), only to lose over half its value when fundamentals came into question.
- “They were sold a dream. A dream that didn't really have any basis in reality.” – Ed Elson (11:23)

3. Defining and Identifying the Bubble

AI Bubble Debate:
- Despite persistent use of the word "bubble," markets have remained high, but this does not mean there’s no bubble—bubbles can deflate slowly.
- “Just because the markets have increased since we started hearing murmurs…doesn't mean that asset prices have become inflated to a level that is not based on the reality on the ground.” – Ed Elson (16:45)
Localized Bubbles:
- Speculation is more intense in some sectors (AI, chips) and regions (like South Korea).
- “The nice thing about this bubble… the markets aren’t totally clueless here.” – Ed Elson (18:21)

4. South Korea: Microcosm and Warning

Crash Illustrates Dangers of Concentration and Leverage (19:56):
- South Korea’s market plummeted 44% in 40 days ($2 trillion wiped out), mainly due to Samsung and SK Hynix dominating the index.
- AI demand caused chipmakers’ values to surge, encouraging both concentration and reckless leveraging.
Leveraged Betting Spirals Out of Control:
- Retail investors poured into leveraged ETFs (borrowing to gamble on AI stocks). When markets dipped, 360,000 accounts were wiped to zero; 1.2 million margin calls impacted 3%+ of the adult population.
- “This is a nation that became addicted to a financial instrument that is really akin to gambling.” – Ed Elson (22:35)
Socio-Economic Roots:
- Widespread loneliness, economic stagnation among young people, and aggressive ETF marketing intensified the problem.
Implications for the US:
- US markets exhibit similar signs: high concentration in few tech stocks, rising use of leveraged ETFs, and a generation of young, disaffected investors.
- “The top 10 stocks in America now make up 40% of the entire stock market.” – Ed Elson (23:24)

5. Risky Financial Practices Fueling the AI Boom

Leverage as “Scaffolding” for AI Stocks:
- Returns in Korea were juiced by leverage, creating systemic risk (28:01).
Circular Financing Explained:
- Tech giants fund AI startups, then startups spend that money back on the tech giants’ hardware/services, creating the illusion of revenue on both sides without new value created.
- Quote:
  - “On the surface, it looks really good. The trouble is the money's just moving in a circle. There was no actual new value that was created.” – Ed Elson (29:54)
SPVs and Hidden Debt:
- Special Purpose Vehicles (SPVs) are used to accumulate debt off the main company balance sheets, masking real obligations.
- $1.7 trillion in tech sector debt is now off-balance-sheet—almost unnoticed but potentially destabilizing.
- “The less clarity you have, the less transparency… the more likelihood there is that people aren’t going to get their numbers right.” – Ed Elson (31:55)
Oracle the Exception:
- Unlike others, Oracle keeps debt on its own books and its stock price has suffered accordingly, with rising credit default swap prices reflecting default risk.

6. Systemic Risk & Consequences

Financial Engineering Layers Risk:
- As risks are moved away from public scrutiny through complex derivatives, SPVs, and private credit funds, markets become “disjointed from reality.”
- “That's how markets become disjointed from reality. And we're increasingly starting to see that in these financial markets.” – Ed Elson (37:22)
Final Warning:
- The same risky tools and social conditions that destroyed wealth and confidence in South Korea are now embedded in the US market—potentially setting the stage for a far broader reckoning if present trends continue.

Notable Quotes & Moments (with Timestamps)

“This feels to me like the most volatile market of all time.” – Ed Elson (01:48, 04:32)
“OpenAI is one of the most unprofitable companies in the history of companies.” – Ed Elson (01:54)
“There is a timeline and there's gonna be a deadline on this. The optimism cannot exist forever.” – Ed Elson (07:53)
“They were sold a dream. A dream that didn't really have any basis in reality.” – Ed Elson (11:23)
“This is a nation that became addicted to a financial instrument that is really akin to gambling.” – Ed Elson (22:35)
“On the surface, it looks really good. The trouble is the money's just moving in a circle. There was no actual new value that was created.” – Ed Elson (29:54)
“The less clarity you have, the less transparency… the more likelihood there is that people aren’t going to get their numbers right.” – Ed Elson (31:55)
“That's how markets become disjointed from reality. And we're increasingly starting to see that in these financial markets.” – Ed Elson (37:22)

Important Segments & Timestamps

| Segment Topic | Start Time | |-------------------------------------|---------------| | Market volatility & AI optimism | 01:11 – 04:19 | | OpenAI’s financial reality | 04:19 – 07:35 | | Public markets, IPOs, and SpaceX | 07:35 – 12:33 | | Bubble debate and campaign | 16:19 – 18:21 | | South Korea’s AI meltdown | 19:47 – 23:09 | | Leverage and social context | 23:24 – 26:42 | | Financial engineering - circular financing, SPVs | 29:35 – 34:42 | | Oracle’s on-balance-sheet risk | 34:42 – 36:16 |

Conclusion

This episode offers a stark, informed analysis of the AI investment mania unfolding in both tech markets and broader society. By drawing direct lines between South Korea’s market disaster and present US dynamics, it serves as both a warning and an explanation—arguing that opaque debt, rampant leverage, and frenzied optimism are an unstable foundation for an AI-driven economic future.

The tone is urgent but reasoned, with Ed Elson supporting his claims with statistics, historical parallels, and accessible analogies—making this episode essential listening for anyone concerned about tech, finance, or systemic risk in the digital age.

Loading summary

Transcript52 lines

[00:00]
Capital One Advertiser
Think of a toilet as just an everyday object. Your entire perspective shifts the moment you experience a Kohler Smart toilet. Design changes everything, transforming a basic routine into something extraordinary with standout aesthetics and intuitive controls. These are functional works of art. Stunning design that never sacrifices performance. Enjoy customizable features for elevated comfort and convenience. It's more than a fixture. It's a cleaner routine and a more refined space. Experience the difference of Kohler Smart Toilets. Find more@kohler.com this episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love. It helps schedule a test drive, get pre approved for financing and estimate trade in value add. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.
[01:11]
Rob Guenther
It has been a wild week for the stock markets. The S&P 500 reached record highs, fueled by what sounds like good news coming out of the Strait of Hormuz. Although it does feel like we've been here before, but okay. And tech companies posted some strong earnings, so. So that's reassuring for any investor worried that the AI boom is going to stop booming. But if you go back not even that far, like a week ago, the picture wasn't so rosy. South Korea's stock market tanked, which pushed the NASDAQ way down toward corrections territory. This pattern has been on repeat all summer. Up, down, up and down. What's going on?
[01:48]
Ed Elson
This feels to me like the most volatile market of all time.
[01:53]
Rob Guenther
Ed Elson Co host, the Profit Markets
[01:55]
Ed Elson
Podcast I think what's happening here is that because the prize of AI is so gigantic, but also because the possibility of of the destruction if the AI thing doesn't work out, because that possibility is so gigantic as well, you have huge amounts of volatility. Well, these are massively money losing businesses. OpenAI is one of the most unprofitable companies in the history of companies. This is a company that lost more than $20 billion last year. That is based on their leaked financials. This is a company that is spending more than the US Government spends on the FBI and NASA combined. So you have to start asking the question, what happens if the money runs out?
[02:41]
Rob Guenther
Today on the show, did somebody say AI bubble? Don't tell the markets. I'm Rob Guenther filling in for Lizzie o'. Leary. And you're listening to what Next? Cbd, a show about technology, power and how the future will be determined. Stick around. So, Ed, we're talking on Wednesday morning this week. We've seen the S and P reach new highs. What's driving this growth?
[03:08]
Ed Elson
It's hard to fully understand at this point because we're kind of reacting in real time here. I think this is a lot of optimism coming off of a lot of pessimism that investors were feeling in the last couple of weeks. Basically the past month, a lot of concerns related to AI, related to the amount of leverage that was being poured into the AI trade. And we saw the ramifications of that in what was a pretty brutal market for the Nasdaq and also for the South Korean stock market, which is heavily leveraged on the AI trade. So we can maybe get to that in a moment. But I think what we're seeing now is investors are looking for any reason that they can to feel optimistic again and to believe that this AI thing is going to work out. Because, yeah, there are a lot of warning signs and there's a lot of instability right now. But the other thing that investors know is that if it does work out, if things happen the way they'd like, well, this is a multi trillion dollar opportunity. So I think investors are looking for reasons to feel optimistic again. They're looking for reasons to believe that this will work out over the long run. And I think that's what we're seeing this week.
[04:19]
Rob Guenther
You're talking about the NASDAQ going down recently. To me, that screams volatility. Do you see this market as somewhat volatile? And if so, do you think investors are starting to price volatility into their business?
[04:33]
Ed Elson
This feels to me like the most volatile market of all time. But I have to be cautious when I say that because most markets feel like the most volatile market of all time. But this market especially feels that way. I mean, if we just look at big tech earnings, which we saw just last week and the week before, and we look at the way that some of the largest names, some of the largest, most valuable companies in the world, the way they whip soared up and down. We looked at Amazon, for example, which literally rose 15% in a single day. It's kind of unheard of for a company that's worth literally trillions of dollars. We saw similar things with Meta. We saw similar things with, with Google and Microsoft. I think what's happening here is that because the prize of AI is so gigantic, but also because the, the possibility of, of the destruction if the AI thing doesn't work out, because that possibility is so gigantic as well, you have huge amounts of volatility. You do not have a real sense of consensus in the markets right now. Investors say, I think AI is going to be worth X trillions of dollars. I think it's going to be worth nothing. And no one really knows right now because we're all just making bets on the future. We're all speculating at this point. We haven't seen a genuine ROI on AI investment so far and people will debate me on that. But if we just look at some, some of the end products, if we look at OpenAI, if we look at anthropic, if those are sort of ground zero for what the AI future might look like, well, these are massively money losing businesses. Open is one of the most unprofitable companies in the history of companies. This is a company that lost more than $20 billion last year. That is based on their leaked financials. This is a company that is spending more than the US Government spends on the FBI and NASA combined. They spent $34 billion last year. And even that isn't enough. And we're seeing this now. We saw just this week that Anthropic is looking to receive a $36 billion loan from Blackstone because the nearly $170 billion that they already raised in the private markets, that's not enough money. That's not enough money to cover the cost of simply selling their product. So what we're starting to see here is, yeah, on the front end of AI, this business has not been proven whatsoever. Sure, you can, you can build data centers and you can sell the compute to those companies, but that business only works so long as an OpenAI and an Anthropic still exist. And a lot of these big tech companies are becoming increasingly reliant on those two AI startups for their growth. So you have to start asking the question, what happens if the money runs out for OpenAI? What happens if they're no longer able to go to the debt markets and raise tens of billions of dollars at a time? That's a real concern.
[07:35]
Rob Guenther
Well, I guess my question based on what you said is what is the timeline there? Because from my perspective right now, these companies are, they're chugging along on the faith that something good will happen in the future. Do we have a sense of how long that Runway is? How long can these companies maintain the status quo just by promising something shiny in the future?
[07:54]
Ed Elson
That is the multitrillion dollar question that Literally no one knows the answer to. But what I can tell you is that there is a timeline and there's gonna be a deadline on this. The optimism cannot exist forever. We've seen this with asset bubbles throughout the history of markets. There are a lot of people, what I can tell you, there are a lot of people who say, don't worry about it, they'll get there eventually. Just give them some time, it's gonna happen. Don't worry about the timeline. That's not a problem. Those people are for sure, certainly 100% wrong. There is a deadline on this. They cannot continue to raise hundreds of billions of dollars ad infinitum. And by the way, their next, their next place to go in order to get the money to continue building this dream of an AI future is the public markets. But what we are starting to see now is that they're actually getting anxious about that themselves. OpenAI had plans to go to the public markets, to go public, to issue an IPO and raise money there. But then they table the. Those plans because I think they probably realized, actually, we're not sure that we're going to be able to convince enough investors in the public markets that this is worth investing in. I think they probably saw some of the reaction to their leaked financials. They probably saw the way people were talking about OpenAI and the anxiety about it. And they said, you know what, let's, let's hold off for now. Let's continue searching for money in the private markets. Let's maybe go and figure out some debt financing as Anthropic is doing with Blackstone and Apollo. And they said, let's table this for now. And I think the real sort of canary in the coal mine here when it comes to AI existing in a fully disclosed world of public markets. If you want to get a sense of how an AI company like an OpenAI might trade on the public markets, how the public markets would receive that company, just look at SpaceX, which was cut in half in less than two months. And we're recording this on Wednesday, August 5th. SpaceX has lockup periods that expire tomorrow, which means that a lot of the insiders are going to be allowed to sell legally for the first time ever since the company went public. Which means we're about to see even more selling pressure on this stock. Billions of dollars of SpaceX shares that will likely be sold by investors who invested a long time ago and want to cash in because they want, they want their paycheck.
[10:25]
Rob Guenther
Yeah, I want to drill down on that because you had a guest on recently who said that even if everything goes right for SpaceX, which includes, you know, building these AI compute centers in space, which seems like an unrealistic proposition so far, even if that comes to light, you say he, your guest told you that the valuation should be like a quarter of what it is right now. It's just so overvalued.
[10:45]
Ed Elson
I mean, this is the guess you were referring to as Nicholas Owens. He's an equity research analyst at Morningstar and his fair value estimate. And this is not an ideological and biased person, this is someone who has simply run the numbers, who actually has a pretty middle of the road view on what AI could do. I mean, I'm someone who thinks that OpenAI and anthropic are running into trouble. His belief is that they're going to be okay. And yet he believes that the fair value estimate for SpaceX is $62 per share.
[11:23]
Rob Guenther
And it was up as high as
[11:24]
Ed Elson
it went to 225 was its high. And when that company, when that stock went out and when I saw the IPO prospectus, the original IPO price was $135 per share. It went out immediately at around 150. It went up on the first day, hitting 225. My prediction before the company went public was that the stock was going to rise 25%. That happened. My second prediction before the company went public was that then it would get cut in half. That did happen. It went down to around 107. We've risen back up a little bit more because they reported earnings recently which were decent ish. I mean, behind some of those decent ish numbers are some scary numbers related to the amount of money that they are spending to build the data centers to build the AI. But that stock has gotten hammered and a lot of retail investors have lost a lot of money on that thing. And it's largely because they were sold a dream. A dream that didn't really have any basis in reality.
[12:34]
HeyGen Advertiser
This episode is sponsored by HeyGen. HeyGen turns a script, photo or presentation into a polished video of you in minutes. No camera or crew required. It's how real estate agents, financial advisors, attorneys and creators show up consistently on every feed and build a brand that brings in clients without spending their whole week making content. Rated the number one AI video platform for small businesses on G2, with the most realistic AI avatars in the industry. Over 30 million people use Heygen, including 85% of the Fortune 100. It's built to work in over 175 languages so you can reach your audience wherever they are. Here's the wild Record yourself for just 15 seconds and Heygen builds an AI avatar that makes professional videos for you on demand so you can post everywhere your audience is as yourself without filming every time. Your first three videos are free@heygen.com pod that's H E Y G E N.com pod starting your own business is never easy. Starting your own podcast? That seems easy, but actually there are a ton of landmines to step on along the way. Finding producers, selling ads, and connecting to WI fi. Oh, does that sound straightforward? It's not. I'm talking about sitting in coffee houses for hours after buying one scone. I'm talking about sitting in hotel lobbies and pretending your backpack is luggage. It's torture. I spent so much time making my home office look professional, but my connection didn't get the memo. The last thing you want during a major interview is for your guest's voice to turn into a stutter. When your bandwidth can't keep up with your ambition, your home office starts feeling like an amateur operation pretty fast. And for a podcast, the Internet is key because the Internet is how we talk to almost everyone. And no matter the guest, a laggy connection can ruin an exclusive interview. Great connectivity isn't a bonus, it's the whole game. And ATT Business is here to help. They've got the tools, team and expertise you need for a stable network you can rely on. And when you can rely on the network, you can get back to thinking about the more important stuff, like nabbing that great guest and getting back to work at AND T Business. Built to Work get at AND t business@business.att.com. This episode is brought to you by Bill, the intelligent finance platform that helps businesses and accounting firms scale with proven results here at what Next TBD we know business. And in businesses across America, smart people are stuck doing the grunt work. You know the drill. Those hours when you could be brainstorming big ideas, you're instead filling in spreadsheets, filling out invoices, or hunting down somebody else's signature. Bill wants to change that. With AI powered automation, Bill removes the busywork from your account's payables workflow. They handle capturing invoices or routing approvals and syncing with your accounting software so that your team can focus on growth instead of paperwork. Bill is so reliable, 98 of the top 100 accounting firms in the US trust it to simplify and secure their bill payment processes. Bill's handled over a trillion dollars in secure payments and is ranked number one overall on G2's 2025 list of best accounting and finance products. So stop the guesswork and start scaling with the proven choice. Go with a company whose financial infrastructure is trusted by nearly half customers. Ready to talk with an expert? Visit bill.comproven and get a $150 gift card as a thank you. That's bill.comproven. terms and conditions apply. See Offer page for details.
[16:20]
Rob Guenther
You had mentioned asset bubbles before. I didn't want to say the bubble word. It sounds a little, maybe a little cliche. Only because I feel like we've been hearing about an AI bubble for over a year now. Right. And I'm just thinking of. Paul Krugman recently talked about the end of this optimism that you're kind of describing right now. He said that this is like not the beginning of the end, it's the end of the beginning because these are such slow moving phenomenon. What are your thoughts on that?
[16:45]
Ed Elson
I think that's probably right. I mean, I can understand why someone would feel frustrated by the idea that we keep on hearing this word bubble. And you're right, we have been hearing it for, I would say it started to be thrown around in September of last year or in August of 2026. So it's been almost a year of people saying, is there a bubble? Is there not a bubble? Are we starting to see signs of a bubble? And so I'm sure that people look at the markets and they see the markets going up and they go, you guys are wrong. There is no bubble because the markets are up today and the markets are up from where they were a year ago. That doesn't negate the possibility of there being a bubble. Bubbles can inflate over long periods of time. They can also deflate over long periods of time. Just because the markets have increased since we started hearing murmurs of there being a bubble doesn't mean that there isn't a bubble. It doesn't mean that asset prices have become inflated to a, to a level that is not based on the reality on the ground.
[17:57]
Rob Guenther
So if we try to define like this bubble, right? The idea, I think we've already said it, but the idea is that we just have billions and billions of dollars going into this idea that AI is going to become this product that just generates way more money than it than it's showing the potential to right now. Let's say I'm not a retail investor. Let's say I feel like I don't have any skin in this game, should I be freaked out? Anyway,
[18:21]
Ed Elson
it's tough because I think the bubbles are located in different pockets of the market and the market is grappling with this question in real time. I mean, the nice thing about this bubble, and I believe that it is a bubble, the nice thing about this one is the markets aren't totally clueless here. If you go to 2008 and you look at the housing bubble that developed, there was such an amount, such a, a large amount of misinformation and obscurity that led the entire market in one fell swoop to believe something that was totally wrong. I don't think that's exactly what's happening here. I think that people are. I think investors in some corners of the market are somewhat aware of what is happening here. And then there are other corners of the market where they just have their blinders totally on and they don't really care about the fundamentals. They don't really care about what might be going wrong over at the Frontier Labs. And I think that you start to see that in some of the more obscure AI names that have become more popular over the past year or so. Some of the, The Neo Cloud names like Core, Weave and Nebius, some of the chip names like sandisk. There are a lot of these stocks which have just become synonymous with AI is going to the moon. And you have a lot of investors who are just covering their eyes and just saying, buy, buy, buy. I don't really care. We did see the ramifications of that, by the way, last week in South Korea because of course, the top two stocks there are.
[19:47]
Rob Guenther
Yeah, let's dig, let's dig into that. What happened in South Korea, I know that so much of the value of their stock market was wiped out. It sent the NASDAQ briefly toward correction territory.
[19:57]
Ed Elson
First off, the South Korean stock market got eviscerated. It fell about 44% in 40 days. It wiped out $2 trillion in market value. There were two main reasons for that. And the reasons are named Samsung and SK Hynix. And these are these two chip stocks which became extremely valuable in a very short amount of time because of their role in AI. Because everyone's trying to build a data center because there is a shortage of, of chips to go into those data centers. Samsung and SK Hynix make those chips so they are able to ratchet up their prices well while selling a lot of. A lot of chips. So their businesses are flat out crushing right now. The trouble is that a lot of investors are now extrapolating that into the future and going, oh my God, these businesses are going to crash for the next 2, 3, 4, 5, 10 years. They believe that this is perpetual, this is never going to end. Which is a bold prediction to say the least. So those two stocks absolutely skyrocketed and eventually they breached roughly half of the value of the entire South Korean stock market. They're worth around 55% of the entire stock market. So this is the most over concentrated stock market in the world. Suddenly we started to see some murmurs of how the AI trade might not be working out. We saw some sketchy stuff with some circular financing over in video. We saw some kind of shady levels of debt that were being raised. Investors got a little bit of the jitters and that sent the AI stocks back down again. The thing that was really the problem was leverage, and that is that the South Korean retail investor community had gotten so excited by this idea of leverage, which is where you basically just borrow money to amplify your returns. You, you can do 2x leverage, you can do 3x leverage, you can do 5x leverage. They employed this strategy of leverage. What they forgot about is that it also amplifies your losses. Yes, it can. To your return. But also.
[22:04]
Rob Guenther
Yeah, let's like. I would love to like translate this for like a non business audience. My understanding is that in May, South Korea was given the ability to do what, what you're talking about. These leveraged ETFs, they're allowed to borrow a lot of money to make a bet, and then if they win, which you know, the whole market had been winning on AI, they'll get a bigger, they'll get a bigger piece of the pie. Right? And what you're describing here is when you have too much leverage though, like when things turn on a dime, you could just like get wiped out, which is what happened.
[22:35]
Ed Elson
360,000 accounts in South Korea went to zero because they became so hooked on the leverage, they became so obsessed with it that as soon as the stock started to enter a downturn, suddenly they were getting margin calls from the bank and they had to give up all the money back that they borrowed because they had levered up so high. There were 1.2 million South Korean investors that received margin calls. It's more than 3% of the entire adult population. This is a nation that became addicted to a financial instrument that is really akin to gambling.
[23:09]
Rob Guenther
So South Korea, everybody's pissed off. They've, they've halted trading. My understanding from watching your content is that the government has banned these leveraged ETFs. Right. But you've called this crash a warning for U.S. investors. And I want to, I want you to articulate how that is so.
[23:25]
Ed Elson
Well, if we look at what brought down the stock market in South Korea, it was an over concentration into a handful of AI stocks. It was a prolifera proliferation of leverage because they got obsessed with the gambling. And there is one other aspect to this which I think is relevant, but it's more of a cultural sociological point, which is that South Korea has a population of young adults who are aimless and lonely. And we are seeing this in the fertility rates in South Korea, which are, I believe they're 0.7 right now. It's the lowest fertility rate of any OECD D nation. Their population is entering into a state of structural decline. Their marriage rates have fallen 40% over the past year. There is a very serious loneliness crisis that is developing in South Korea. And it's also related to the fact that housing prices in South Korea have risen. Inflation is also rising. Young people are generally struggling economically. And so there is a large swath of the population, the young population specifically, that is sitting at home, especially if you're a young man, because young men are more interested in these and more have more of a penchant towards these gambling tendencies. They're sitting at home and they're gambling on these 2x and 3x and 5x leveraged ETFs. And when you look at the, the demographic makeup of the accounts that were wiped out, a significant portion of those accounts were young people. It's people under the age of 30. So I believe that that is a big contributor to the problem, which is that you have a lot of young people. They don't feel that they have economic prospects. They don't feel that they have much of a purpose, much of a mission in life. And so they see these opportunities for 200% returns, 2000% returns. This is how I'm going to do it. This is my lottery ticket out to make my life better. And they get it hooked on these strategies. And it's not entirely their fault, by the way, because also they're being sold these strategies. They're being, it's predatory dollars that are going in to putting these leveraged ETFs out there and getting people to believe that this is your ticket out. This is how you're going to make money. Now, why is that a warning to the U.S. well, I look at the U.S. and I see all of the same trends. I see a population of young people that is unprecedentedly lonely. One in five of us young people, I'm Gen Z. One in five of us say we have zero close friends whatsoever. I look at a market that is unprecedentedly over concentrated into a handful of names. The top 10 stocks in America now make up 40% of the entire stock market. 30 years ago that number was 20% given. Less concentrated than South Korea, but still very very concentrated. And then we're also seeing a rise in leverage, specifically leveraged ETFs, which have jumped from $120 billion in AUM in April to more than $200 billion today. It's up nearly 70% in just a few months.
[26:42]
Capital One Advertiser
This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One. This message is brought to you by Cologuard. If you recorded songs off the radio for a mixtape, it might be time to screen for colon cancer. The Cologuard test is non invasive, requires no special prep or time off, and you start screening right from your home. It's just three simple setup, sample and ship. If you're 45 or older and at average risk, ask your doctor about the Cologuard test, available by prescription only. Learn more or request a prescription@cologuard.com podcast.
[27:47]
Rob Guenther
All this behavior that you're identifying, all this predatory behavior and our sort of being habituated to participate in it, is this the type of activity that's sort of providing the scaffolding for the AI boom here in the US
[28:01]
Ed Elson
that's the scarier implication and that is the scarier implication of South Korea, because that is what created those phenomenal returns in South Korea. Yes, there was AI optimism, but at the same time there was huge amounts of leverage that was juicing the trade upward when things were going well. And it's, you know, if, if people are making 2000% on Dogecoin or Comerocket and then losing all of their money like it's not a good thing. But at least it's not really systemic to the wider stock market. At least that's just happening out in crypto land and people are kind of messing around with those nonsense coins. But in the case of, of the AI trade, which is not just juicing the stock market, but it's juicing the economy itself. I mean, this is not Samsung SK Hynix. These are not meme stocks per se. Yes, they have meme like qualities, but they also make up literally half of the entire Korean stock market. And they're also inking deals with some of the most important systemic companies in the world. They're inking deals with Nvidia, they're inking deals with the big tech companies. So these companies and these stocks actually matter. And so when we use financial products and we use leverage and we use all of these different derivatives and instruments and vehicles to make the line go up higher, we're starting to play a very dangerous game which could result in the evisceration of a lot of portfolios and retirement accounts and pension funds, all of the things that actually do matter in the real economy.
[29:36]
Rob Guenther
So you've identified leverage, you've identified this gambling on your programing. You've also identified a couple of other phenomena that seem to be playing a big role. One is through these SPV special purpose, special purpose vehicles. And then the other one is called circular financing. Maybe let's start with circular financing. What is that?
[29:54]
Ed Elson
So circular financing is a phenomenon that arose actually more than a year ago. And it was the reason that a lot of people started to get worried about there being a bubble in AI, and rightly so. Essentially what you do say, I'm, say I'm Google. We'll use a recent example. I'm Google and I want to increase my revenues this year. What I'll do is I will invest $10 billion into an AI startup called Anthropic, and then a few months later, Anthropic will turn around and they'll buy $10 billion worth of my AI chips. So what happened? There is one. Google looks good because they're investing in Frontier Technologies and they're participating in the upside of Anthropic. And they also look good because their revenue went up by $10 billion. They sold $10 billion more worth of chips. So on the surface, it looks really good. The trouble is the money's just moving in a circle. There was no actual new value that was created. I'm just sending you the money and then you're wiring the money back to me, which creates a false sense of growth, a false sense of business, because
[31:01]
Rob Guenther
the numbers do go up and people do get returns. Is that the idea?
[31:05]
Ed Elson
They don't look enough into the numbers, they don't register the fact that actually the money is moving in a circle. And so how sustainable is that revenue growth? Yeah, I can just send you money and then you can send it back to me and I'll say, look, my revenue went up, but if you're going to keep on doing that, then I mean, eventually the revenue is going to run out and eventually people are going to realize that actually we're not creating new economic value here. So this was the concern and it was a strategy that was employed by many big tech companies.
[31:33]
Rob Guenther
And it's one thing when we know what's happening, right, we can see these trades taking place. But there was another piece of news recently, and it was, it touches on the SPVs, which is the second thing that you, that I learned from watching your show. But it turns out that a lot of these companies had even more debt than we imagined. So it was debt that was hidden. It was a massive amount of debt. And they're all being done through these vehicles called SPVs. What's going on there?
[31:56]
Ed Elson
Yeah, so one of the reasonable bull cases in the markets right now is that you could say, okay, yeah, they're spending all of this money on AI and who knows if the AI is going to generate a return, but at least they're not taking on too much debt. And that was a fair argument that I subscribed to for a long time. Because the reality is these big tech companies, they are cash cows. They have huge amounts of profits and yeah, they have money that they can spend. And it's, it's up to them to spend whatever they want to spend that money on. They had about $1.4 trillion worth of debt on their balance sheets, which sounds like a huge number, but compared to the profits that they're making, it's sustainable. Then we learn that actually there's $1.7 trillion worth of debt that is off of the balance sheet that is not being reflected in the reporting and the earnings reports from these companies. The $1.7 trillion worth of debt is being accumulated in these financial instruments called special purpose vehicles, which are now the main way by which data centers are built. You essentially create a shell company, an spv, a special purpose vehicle, and then you find debt from somewhere else. It's usually provided by a thing called a private credit fund, which is what it sounds like. It's a fund that goes and loans out money in the private markets. And this has become the principal way by which data centers are built. And I should be clear here that's not debt that is being accumulated by the big tech companies. For the most part they are off the hook. So it's not, we shouldn't pull our hair out and say, oh my God, the world's going to end. But what we should recognize is that big tech is making an intentional decision to offload the risk of the data centers, throw it off their balance sheets and pass it on to someone else. And that someone else is in most cases private credit funds, which has exploded into a huge industry. The point being this AI buildout is accumulating vastly more amounts of debt than people are willing to acknowledge. It's not just the debt that's on the balance sheets of the companies, it's the debt that's buried away in these no name SPVs that no one has much clear insight into. They don't really know what it actually is. They don't really know who's actually financing it. That's where the AI boom is happening. And the less clarity you have, the less transparency you have, the more likelihood there is that people aren't going to get their numbers right, they're not going to do their underwriting properly, they're not going to do their homework. And that's how bubbles rise. This is what we saw in 2008, it's what we saw in 1999. Going into 2000. It's very likely that this could happen again.
[34:43]
Rob Guenther
Well, help me understand, if we use maybe one data center as an example, I'm thinking of last year Oracle made this $300 billion data center deal with OpenAI. And I think you alluded to this earlier, but deliveries and commitments on that deal, they're supposed to kick off next year. Like what happens, what happens when that deadline happens?
[35:04]
Ed Elson
So in Oracle's case, Oracle is actually the odd one out here because Oracle is one of the few, one of the, it is the only big tech company that is accumulating the debt on its own balance sheet. So that's an example of a company where you can actually see all of the risk they're inking these deals with with OpenAI and all of these other AI companies. They're spending more on AI data centers currently than they actually generate in revenue. So they are, talk about leverage. They are levered to the hilt. They are like South Korea's retail investor times a billion. It's reflected in the price because Oracle stock has been absolutely pummeled.
[35:47]
Rob Guenther
So it's baked in already.
[35:48]
Ed Elson
And the price, it's baked in the price of their credit default swaps. Which is basically insurance on the possibility that Oracle just defaults on its debt and therefore just goes bust. Those have been soaring recently as investors price in the increasing likelihood that Oracle could literally just go bust.
[36:05]
Rob Guenther
So it's not that Wall Street's ignoring the debt, it's sort of that they're aware of it and they're trying to like, bake it into their projections.
[36:11]
Ed Elson
They're aware of it because Oracle is reporting it. They're not taking the SPV route. And that's why I get anxious.
[36:17]
Rob Guenther
Oh, because of the SPV route, There's lots of vehicles that you have no idea what's going on.
[36:21]
Ed Elson
Yes.
[36:21]
Rob Guenther
Okay.
[36:23]
Ed Elson
That's why I get worried with Oracle, because I look at what's happening to Oracle right now. I look at how investors are reacting to what they're doing with their debt, and then I think, well, where else? Why isn't that being priced in elsewhere? I mean, there's huge amounts of debt being accumulated everywhere else. Why aren't we seeing it? And the answer is it's being stuffed into the SPVs and the private credit funds, which are not publicly traded. And so we don't get a proper markup and we don't get transparency into what's actually happening there. So this is the problem with bubbles that I should emphasize. It's not necessarily that people are doing something illegal or they're trying to pull a fast one on us. What happens in financial markets is people come up with these complicated financial derivatives of derivatives of derivatives, which makes the entire market increasingly more complicated, which means that fewer investors are actually doing their proper homework as to what's happening. That's how markets become disjointed from reality. And we're increasingly starting to see that in these financial markets.
[37:24]
Rob Guenther
And Elson, thank you so much for joining us.
[37:27]
Ed Elson
Thank you very much for having me.
[37:28]
Rob Guenther
Ed Elson is a co host of the Prof. G Markets podcast and that's our show. What Next is produced by Evan Campbell, Madeline Temzu Charme and Patrick Fort. Paige Osborne is the senior supervising producer of what Next and what Next tbd. Mia Lobel is the executive producer of podcasts here at Slate. Ben Richmond is the senior director of podcast operations. And I'm Rob Guenther filling in for Lizzie O'. Leary. You can find me on BlueSky. I'm Blue One, Rob Gunther. Thanks for listening. I'll talk to you soon.
[38:00]
Capital One Advertiser
This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.
[38:29]
Rob Guenther
If you're looking for something predictable and
[38:31]
Ed Elson
low friction over the long term, public
[38:33]
Rob Guenther
markets may be a suitable option. But the moment you want different investment characteristics, it may be time to consider the private markets. See how your wealth can work smarter@creativeplanning.com access.