Michael Burry’s Warning, AI Math Reality, and the Hollowing Out of “Core”

In this episode of Lead-Lag Live, I sit down with Seth Cogswell, Managing Partner at Running Oak Capital, to break down Michael Burry’s warnings about depreciation manipulation, inflated AI earnings, and the concentration risk dominating today’s market.

From the speed of chip innovation to the widening gap between revenue and AI CapEx, Cogswell explains why valuations at the top of the S&P 500 may be far more stretched than investors realize — and why diversified, valuation-discipline strategies could be the antidote to extreme concentration.

In this episode:
– Why extending depreciation cycles may be inflating mega-cap earnings
– How today’s top 10 S&P names are valued above the peak of the tech bubble
– What investors misunderstand about investing in AI
– Why Running Oak has reduced tech exposure despite an AI revolution
– How passive flows have hollowed out the core of the U.S. equity market

Lead-Lag Live brings you inside conversations with the financial thinkers who shape markets. Subscribe for interviews that go deeper than the noise.

#RunningOak #Markets #TechValuations #AIBubble #PassiveInvesting #RiskManagement #Equities #Investing

Start your adventure with TableTalk Friday: A D&D Podcast at the link below or wherever you get your podcasts!
Youtube: https://youtube.com/playlist?list=PLgB6B-mAeWlPM9KzGJ2O4cU0-m5lO0lkr&si=W_-jLsiREjyAIgEs
Spotify: https://open.spotify.com/show/75YJ921WGQqUtwxRT71UQB?si=4R6kaAYOTtO2V

Chapters

• 0:00 Valuations And Concentration Risk
• 3:01 Burry’s Depreciation Warning
• 5:35 Faster Chip Cycles Vs Longer Lifespans
• 9:05 What If Earnings Are Overstated
• 12:35 Diversifying Beyond Megacap Tech
• 15:45 AI Capex Reality Check
• 19:35 Who Actually Benefits From AI
• 23:05 The Hollowing Out Of Core

Transcript

SPEAKER_01: 1:12

The valuations are higher. That means that you have further to fall if things don't go perfectly. If you do a cap weighted calculation of the top 10 holdings in the SP 500, it is roughly 39. According to Apollo, the forward P of the top 10 holdings during the tech bubble at the peak was 25. So that means that the forward P right now of the top 10 holdings in the SP is 50% higher, roughly 50% higher than at the peak of the tech bubble.

SPEAKER_00: 1:56

Big tech is under fresh scrutiny after Michael Burry warned that major AI and cloud computer companies may be manipulating depreciation schedules, extending the useful life of data center hardware to artificially boost earnings. And at the same time, markets are wrestling with what AI investing really means now that the hype cycle is colliding with rising capex and margin pressure across the megacaps. My guest today is Seth Cogswell, managing partner of Running Out. So, Seth, the last time you were here, we talked about quality growth versus what you call uh themes, memes, and dreams. Burry's now accusing big tech of gaming depreciation to inflate profitability. What's your reaction to that claim?

SPEAKER_01: 2:39

First, I would say that this really matters in that one, these companies that that Bury is somewhat accusing, he actually used the word fraud, but that he's you're at least kind of pointing out that they might be somewhat gaming their earnings. That's problematic because they are the biggest companies in the index. Uh they're a massive, you know, they might be the biggest percentage ever. And everyone, and I mean, other than maybe myself, but almost everyone has a massive amount of them, right? Whether it's uh passive portfolios, large cap growth, or as individual holders of stocks, almost everyone has a massive percentage of these companies. So this really matters. Um now one of the most interesting aspects of what he pointed out is how clear it is. It's one thing for any anything can be said on Twitter or whatever. And we all have a tendency to kind of lean toward believing it because we're like, why would anybody lie on on Twitter? Um but what he has pointed out really resonates for me, but also just if you just simply look at the numbers. Um, NVIDIA's chip cycle used to be two years, right? So they'd come out with like a new uh a new state-of-the-art chip every two years. Now it's down to a year or less. I think it's actually less than a year. Um, these data centers are evolving quickly. Uh, I recently heard a cool story about how one of the data centers kind of figured out how to use water in a way to really cool off the chip so it actually makes water usage and energy usage more efficient. But that means that you're having to basically reinvest a whole bunch of money into data centers again to improve them. And so the progress and the speed of change and the speed of innovation right now is crazy. Meanwhile, these tech companies are extending their depreciation cycle, right? So, which means that they're basically saying that the speed of change is getting slower. That's what extending the depreciation cycle means. Um, and so they're doing the exact opposite of what any person who observes what at all what's going on, they're doing the exact opposite. So it's worth paying attention to. It's it's really concerning. I'm not going to use the word fraud because I'm not, or certainly I'm not going to accuse it. I did use the word. I'll let Michael Burry throw that word around. But it's it's really concerning. And um, and one of the reasons why it's so concerning again is because these companies are so widely held, right? It'd be one thing if this is Enron or WorldCom or something like that, and that's sort of a one-off. It's another thing if it's pervasive across the largest percentage of holdings, you know, maybe ever, for set for let's say 10 holdings. Now, if you another reason why this really matters is one of the things I'm trying to really emphasize is risk. And and risk at its heart is just uncertainty, right? It's if I like to think of risk as the risk of loss, because that's what matters to most people. But many measure risk is standard deviation, which measures volatility, right? Which is uncertainty. So let the right now is an uncertain type. The issue is with valuations, if valuations are higher, that means that you have further to fall if things don't go perfectly. Right now, the if you do a cap-weighted calculation of the top 10 holdings in the SP 500, it is roughly 39 or so. According to Apollo, the forward P of the top 10 holdings during the tech bubble at the peak was 25. So that means that the forward P right now of the top 10 holdings in the SP is 50% higher, roughly 50% higher, than at the peak of the tech level. Maybe that's fine, right? There's we've got crazy innovation going on. Maybe things work out. But what if they don't? And and that that's the big that's the big question is just what if? Uh because if they don't work out perfectly, it's you know, there's a lot of downside. Now, again, if Michael Burry's correct, and again, common sense says that he's not just correct, but actually being conservative. So in Michael Burry's post, he said uh I think two to three year schedule. It's probably less. Uh these, you know, if the if NVIDIA's cycle is now down to a year, that's not two to three years. And these companies, the way in which they're investing and and and what their needs are is for compute to be as powerful as possible. Uh and so the odds that they're not, the odds are that they're very close to one year and very far from six. Bury's giving them a bit of a a little bit of uh room for error, where I'd say, I mean, by all appearances, it seems like it's far worse than Bury, but let's be conservative and we use Bury's numbers. And we say that earnings are being overstated by roughly 20%, which he said for Meta. I think he believes that by 2028, they'll be overstating their earnings by 27 to 28%. So let's say 20%. We'll be a little conservative. If that's the case, the current forward P of the top 10 holdings in the SP, if we kind of correct for what Burry is pointing out, is now 85% higher than the peak of the tech bubble. So it it starts to become pretty crazy. And yeah, that just means that if in order to, let's say all of this is true. Maybe it's not, but let's say it is. That means that if the market were to simply mean revert to the peak of the tech bubble, that's almost a 50% decline in these companies. That's to the peak of the tech bubble. That's not uh anything that was normal. That was one of the that was arguably the craziest time in the history of the stock market. And we're talking about a 50% decline just back to the craziest time in the history of the stock market. Again, I don't I've one of the great things about our strategy is it's built upon not predicting the future. I don't believe I can predict the future. Um, but I do believe in common sense. I believe in numbers, I believe in trusting oneself and uh an ability to analyze what's fairly obvious. And so again, this is a risk, right? And it may not come to pass. Um, but it's it's worth, it's just worth constantly reiterating what if. And the best way to uh approach that what if is to diversify, which so few people are have right now, right? So less of those companies, you can still have plenty of it and you're gonna have exposure in different ways, but just less. And and consider investing in ways that are very different because what if Bury's right?

SPEAKER_00: 10:45

Yeah, so I mean, talking um, I I guess if if Bury's wrong, he'll delete his uh X account eventually, right? But but it I want to talk about AI specifically. When you look at the AI narrative today and extreme concentration in the largest names, which you've talked about in terms of how much of the SPY or the SP 500 is is made up of those companies. What do you think investors are misunderstanding about how to truly invest in AI exposure then, if Furry is correct?

SPEAKER_01: 11:13

We reconstitute our portfolio generally three times a year, where we make additions and deletions. And the last time we did so, the result was less tech exposure than I ever recall seeing. Usually we have more tech exposure than the SP. Right now we have quite a bit less than I've seen since running the company. We have more industrial exposure. But to have more tech exposure during what appears to be a little bit of a, you know, well, maybe not a little bit, very much a tech revolution is was disconcerting, uh, certainly on its surface. But the more I've thought about it, the better and better I feel about it. And our our process is rules-based. So to have less tech exposure was not a conscious decision. It's not based on opinion. It was based upon our the traits that we invest in, which is we want to maximize earnings growth, but we're really disciplined around valuations, very focused on avoiding overvalue companies in particular. And that is what led to us having less tech exposure than any time I recall. So I've I've thought about this a great deal and why our process would take us this route. And it led me to some interesting conclusions. So, right now, and this is gonna, these are gonna be broad strokes, but they're pretty easy to verify. So that AI is estimated, we'll say a trillion dollars is estimated to be invested in AI CapEx within, let's say, over the next year or so. AI native companies, so, or native AI companies, uh, such as um OpenAI and a number of the you know, the big names, currently have about 20 billion in revenue. So we've got a trillion dollars in capex, 20 billion in revenue. So that's that's 2%. It's hard to see how these companies have a path to profitability given their current cost and revenue structure. But one of the interesting things about AI is it can be applied to any technology. It can make almost anything either a little or a lot better. And that includes, again, anything is anything, right? So that includes industrials, financials, healthcare. I mean, there's some amazing potential in those areas, real estate, utilities. And right now, investors are piling into tech. So not just big tech. They're piling into tech across the board, anything that even remotely, you know, rhymes with AI, people are piling into, regardless of valuations, regardless of whether they're profitable or not. And that means that that money isn't going into other things. And our strategy is designed to lead us where money, at least according to our numbers, should be going, but isn't, and it creates this relative valuation. And so all this money is piling into tech, and therefore our strategy has led us elsewhere. And again, the the interesting thing about that is that AI can be applied not just to tech. It's AI is not a tech-specific thing, it can be applied to anything to make it better. Um, and so you know, if you really think about that trillion dollars of CapEx,$20 billion in revenue. This doesn't even account for the fact that that to provide that$20 billion in revenue has a lot of costs associated with energy and compute and whatever else. Regardless, that$20 billion in revenue has an associated$20 billion in cost, right? So some other people are spending$20 billion in order for these companies to have$20 billion in revenue. So other companies are getting this$20 billion worth of products that's costing a trillion dollars plus to provide. And so other companies, not tech, the the clients of these tech or these AI companies are getting potentially something of great value at a massive discount, right? We're talking basically at a 98% discount. And so that's one of the things that I again, we we did talk, we talked about Michael Burry and and the exposure to big tech and the risks there. And so few people are investing elsewhere, and there's so much more opportunity there. There's profitable companies that are in a position to immediately apply AI and make their companies better, and people aren't investing in this. So I'm slightly talking to my book, but certainly with regard to diversification, there are some amazing opportunities outside of tech. It is not an AI-specific thing. Uh and again, it's a hell of a deal that these companies are getting.

SPEAKER_00: 16:19

Yeah, Seth and I mean, I I think this is what you mean by the term hollowing out of core. But for our listeners, can you explain what that means in practical terms and why you're concerned about the foundation of the whole market beneath the megacaps, um, even as we've seen the SP hit new highs?

SPEAKER_01: 16:35

Uh so I'm reading right now the Bogle effect, which is the most recent biography on Jack Bogle. And and one of the comments that's made on pretty regularly throughout the book is the hollowing out of core. Now, they don't mean it as a negative per se. It's more an observation of how passive is impacting uh say equity management or managers, right? So um the popularity of passive investing, which is now roughly 55 to 60 percent of all money invested in stocks or US stocks especially, is estimated to be in passive investing or cap-weighted portfolios. Now, there is an assumption or a view that those passive portfolios, let's say again, the SP, the Russell 3000, there's a view that those are a core holding, right? They're viewed as being diversified. They're viewed as providing broad exposure to the market in general. But the problem is those things are moving around, right? I mean, 41% of the SP now is in 10 companies, and let's say eight of those companies are very much tech and large cap growth. That's not a diversified exposure. That's not a core holding, right? So there's this misconception that passive is core. It's not. It is not the central, diversified, risk-focused, whatever. It's not balancing all the things, right? Core should be the central foundational holding that balances all things. That is not what passive is. Passive right now is wildly large kept growth. Um and again, going back to that comment, the hollowing out of core, as passive has gained in popularity, it is just it has killed most managers or many managers that are in that core space, that are in that sort of, you know, walk the line between growth and value, let's say Garp or mid-large, it's really it's hollowed it out, right? It's it's destroyed those managers, just put them out of business. And as the uh investment management industry has evolved around the um the popularity of passive, as you mentioned themes, memes, and dreams before, um, you know, a lot of the active managers have turned to the periphery, right? Because if passive is viewed as core and central, then the opportunity is around the periphery to provide differentiated exposure. So you've got a lot of thematic, you know, active ETFs. And and those can be really cool. They can provide kind of targeted, concentrated exposure to interesting themes. But one thing I'm focused on a lot is there's no buy, low, sell high component to themes. There's also no buy, low, sell component to passive. It's it's actually buy high. And if the price appreciates, they buy more because it becomes a higher percentage. And so it's it's buy high, buy higher, buy more higher, never sell. So whether it's the themes, whether it's the dreams, I call it dreams on um passive because it's a momentum construction. And so it's a dream that it just goes up forever. Because if it doesn't, you see the uh you know, you see things reverse. All that's to say that the central holdings or the central managers that balance growth and value mid and large that are meant to be that core, to be that foundation of a client's portfolio are gone. Except for us. Now there's some others too. Uh, but running oak, that's what we provide. We walk the line between growth and value, mid and large. It's focused on very focused on risk and valuations. It's common sense, it's rules-based, which means we're very disciplined. And so we are consistently kind of like a rock, delivering precisely what our clients expect because we do the same thing over and over. Um, and so I see this hollowing out of core as an incredible opportunity for us. I I if I'm gonna make a wildly outlandish claim, it's that five years from now, uh many passive investors will realize will realize that they were investing in portfolios with the wrong assumptions and that we are the solution that they've actually been looking for. So uh it's an interesting time, but it's uh I'm excited to see how it all plays out. And I believe that we can add significant value and diversification where needed. Going back to Michael Burry's you know tweet, the last thing that people need is more exposure to a small number of big tech companies that are burning cash like nothing ever seen before, that are at valuations that are conservatively 50% higher than the peak of the tech bubble. If Michael Burry's correct, they're actually 85% higher than the peak of the tech bubble. And and we can help clients or advisors really somewhat hedge that bet or diversify away from it.

SPEAKER_00: 22:09

Seth, you did an amazing job of uh sort of wrapping up our conversation for today. Um, before you go, can you let people know for and for anyone who's watching who wants to learn more about your work, uh your research, or how to follow running Oak Capital? Where sh where can they find you?

SPEAKER_01: 22:25

I have very begrudgingly, it's hence this video, begrudgingly been much more active as far as interviews go and content. So you definitely look me up on LinkedIn. There's a lot of content there, and a lot of it's a little more targeted, so you can kind of focus on different aspects that you that might be of interest. Um, feel free to email me, Seth at runningoak.com. And then please keep an eye out. I'm starting a new kind of short video series called Not So Passively Aggressive, which fits my personality really well. Uh, but the goal is to serve as many investors as possible, help educate, help them see the market in a through a lens of critical thought, um, see passive as it is, which is not bad. Uh, just simply it's it's it's neither good nor bad, right? It's it's a momentum construction. And and so really helping people to uh learn in any way I can. Uh so again, that's not so passively aggressive. Please keep an eye out for it. Um, but hit me up anytime. I I I love having these discussions. I love helping in any way I can.

SPEAKER_00: 23:43

That sounds extremely exciting. And Seth, it's always great to have you on. And thanks to everyone for watching. Be sure to like, share, and subscribe for more episodes of Lead Leg Live.