Is the AI bubble about to burst?

Added: 2026-05-11

[00:00:00]Equity markets, so stocks and shares are absolutely mesmerized with AI and tech stocks and AI stocks now account for almost half of the S&P 500, which is the val the combined valuation of the 500 most important um listed publicly traded on stock markets companies in America. And even though they're relatively small in number, they now account for you almost half of the total valuation up from a quarter less than 3 years ago. And and some companies, their valuations are absolutely massive. So Nvidia, which obviously makes chips, it's both a software and a hardware company, what we call a full stack tech company. Their chips are absolutely pivotal. Their technology is very very important of course. But Nvidia's market cap, their market capitalization, their to if you get all their shares and add them and time add them up times them by the value that what company's worth uh is now bigger than the GDP of Britain, bigger than the GDP of Japan, bigger than the GDP of Germany.

[00:01:23]you know only two countries in the world the US and China have a GDP that's bigger than the value of one company now I'm not saying that and there are other companies like that with monstrous valuations I'm not saying AI is not important you know I'm not saying I'm not a lite about these things at all um but the market is priced as if all these companies a bit like the.com boom just cuz they had a you know com in their name and you know the CEO wore a t-shirt um suddenly they they had massive multiples. What does that mean? Massive multiples. The value of the company as judged by the stock market uh divided by the average earnings, the annual earnings, right? Uh and often a PE the long-term historic average of the S&P 500 is a multiple of 17. So, a company is worth 17 times its current annual earnings, which kind of kind of makes sense, right? Yeah. Um if somebody builds a small company and sells it, they might hope to buy sell the company for a multiple of two or five or you know if you're a big listed company obviously you got a massive culture a massive brand systems they built over many years in most cases a multiple of 17 well if you get the combined all the all the multiples of all the companies in the S&P the composite multiple of the S&P 500 as measured by something called the the cape Schiller index which is the cycllically adjusted price earnings multiple composite and it was an index founded very influential index founded by among others a professor from Yale called Robert Schiller who's done very well for himself but rightly so it's a very interesting composite index at the just before the global financial crisis the Cape Schiller got up to 28 so the average company was worth 28 times its annual earnings which is a lot. So in other words, shares are really, really expensive. So they're, you know, gunning for a fall. There's a lot of hype. A lot of people are borrowing to invest in those shares. That's called leverage. When you borrow money to invest in shares or you borrow to do an investment, it's called a leveraged investment. The cyclically adjusted uh price earnings ratio. Uh cyclically adjusted means you take out like the booms and the bust. you're trying to get a sort of common smooth path. We're now well above 40 uh which is you know well into.com territory and I think that's the biggest danger the fact that equity markets you know in all that these stocks are being massively hyped and hyped and hyped and I'm not saying AI is import isn't important but a lot of these companies are going to fail a lot of these companies aren't going to work out you know history doesn't happen in a straight line as I said in the column AI is really controversial.

[00:04:20]You know, that's a statement of fact.

[00:04:22]I'm not arguing that it's wrong. I'm just saying the reality is somebody that does economics and finance but does politics, too. A lot of politicians are going to try and stop this because a lot of people are worried that their kids can't get a job, the regulation is going to come. Often a lot of these AI companies, they're not factoring in the reality that the data centers, the chip plants that drive all this stuff need massive amounts of electricity. And a lot of Western countries, particularly our country, won't have the electricity, right? It's going to take time. So, a lot of the companies are going to fail because they're just going to consolidate, be overtaken. You won't always have the electricity.

[00:05:07]the regulation will styy a lot of sort of commercial visions and kind of rightly so you know I mean it's it's a it's a bargain right everything's a bargain so you got those three things happening which means the fact that the market is priced for perfection because everybody's assuming it's just going to go to the moon forever because I think a lot of people in finance you know geopolitics over here sovereign bond crisis over here they're going la AI AI AI I I'll keep investing, keep investing, keep investing. I want my bonus. I want my bonus. I got to build a conservatory. I got to pay the school fees. Right? That's what's happening.

[00:05:44]And on top of all that's known, right?

[00:05:46]That's all that is kind of out there.

[00:05:48]That that sense is out there. You can hear that at any investment conference.

[00:05:52]Someone will stand up and say electricity and someone else will stand up and say, you know, my kid can't get a job.

[00:05:59]But but the new thing, not new, but some an idea that isn't out there is what I wrote about on Sunday and in the Telegraph and maybe you can put the link in the in the show notes to your episode. Is this what we're seeing is massive multi-billion dollar investment in AI, right? In all the data centers, they're really expensive.

[00:06:25]um the chips are really expensive that you need.

[00:06:31]Now you had a massive boom in the 19th century on both sides of the Atlantic in the railways, right? And it was it was an investment mania.

[00:06:42]You had massive investment in the dotcom boom with you know lots and lots of fiber optic cable was laid in the ground that was very very expensive. The cable was expensive. You glass cables encased in plastic. You got bury them under sensitive parts of the town, cities, long distances. So the the do the railway boom and the.com boom, they've, you know, fostered the railways and the internet, which you know, two massive industrial revolutions, two completely transformational things.

[00:07:16]What's happening with AI, the fifth industrial revolution? If you want to know what the other four were, you can read my column. Um, the fifth industrial revolution, it will be transformational.

[00:07:29]It's that means that massive investment is going on to reap the benefits of that transformation.

[00:07:36]Both the railway boom and the dotcom boom, the massive investment that you had in the infrastructure was driven by leverage, by people borrowing to invest.

[00:07:46]This AI boom is the massive investment in the data centers is driven by leverage and how absolutely huge amounts of borrowings going on. A lot of it from so-called um shadow banks. So these are basically hedge funds pretending to be banks. Hedge doesn't mean that you're going to see a branch on on the high street anytime soon or they're going to give you a loan to start your plumbing company. No, the they're just lending big money to AI developers.

[00:08:16]Um and it's complete almost completely unregulated right so the risks that are being taken are huge without any kind of well with very minimal regulatory oversight. So what's the difference between the railway revolution, the dotcom boom on the one hand and this the leverage the loans in the railway revolution and the dotcom boom were collateralized often uh largely in some cases on the infrastructure that was being created.

[00:08:49]Right? So okay if my company goes past you my creditors own the infrastructure and the railway lines were in the ground. you know, all the digging had been done, all the rolling stock was there, the stations have been built. And in the internet revolution, that the fiber optic cable was in the ground. It's still really, really valuable. And in both cases, Brendan, that infrastructure was there, remained valuable, and had economic use and worth. I mean, the line, the railway lines are still being used. The fiber optic cable is what's make that was laid in the dotcom boom with massive overinvestment and massive mania is what is now facilitating these data centers. Right? So the loans were collateralized against infrastructure that was being built that had demonstrable ongoing sustainable value whoever owned it.

[00:09:47]Right? So the loans could be transferred, they could be refinanced.

[00:09:52]You know, obviously you have to take risks. Obviously when technology is at a cutting edge, the risks need to be taken quickly and people of you know um without the coones to do it need to get out of the way, right? That is what human progress is about. Yeah.

[00:10:08]But this boom and of course the railways and the internet even though they were both massively important there were mania panics and crashes linked both even though the technology was obviously important. No one's saying the technology wasn't important then. No one's saying AI isn't important. You can't dismiss what I'm saying by trying to convince me by by trying to uh present me as some kind of lite that doesn't understand. So here's the thing.

[00:10:37]The collateral for these AI investments are the chips that are being used. And the chips by definition have a very short lifespan because they're constantly being outmoded by better chips.

[00:10:59]So, if you are collateralizing your loans on the Nvidia chips that you're buying with those loans to build your data centers and the Nvidia chips, you know, you're putting in your nice spreadsheet that convince you and your investors that it's you're all going to get rich, you know, you're saying, "Oh, these chips will be around for at least five or 10 years." But actually, what's happening in the real world is these chips are being outmoded after one or two years. the collateral you're holding, you know, you you you got a warehouse full of Walkman's in the age of an of the iPod to bring it back to our generation.

[00:11:39]Now, people who I really really respect in the AI world and I talked to lots of, you know, serious AI entrepreneurs. Some of them are my good friends. Um, so it's not personal. They're saying, "Yeah, but the chips will still be valuable." I'm like, but how valuable?

[00:11:58]And they'll be increasingly less valuable.

[00:12:01]And if you're collateralizing a loan that's looking increasingly marginal as the other things happen like the electricity isn't there, the the politicians go nuts and try and ban it or try and, you know, rein it in or all the other stuff, then your loans, you know, look less likely to be paid back and your collateral is depreciating.

[00:12:25]Houston, we have a problem. And that is why I've it's a long explanation, but I really think this is important.

[00:12:33]We've got to get this idea in in the thick skulls of our political and media class. Lots of them. There are exceptions. Don't at me. Um, unless we we get hold of this and realize that yes, AI is important, but we are now deep into bubble territory with a case shiller index above 40. with stock markets are record highs despite you know carnage in geopolitics carnage on western countries sovereign balance sheets including some very big countries we're not talking Greece and Thailand you we're talking the UK we're talking France big countries so I think we are facing multiple dangers to financial stability um I think the prices of equities have completely lost touch with any kind of reality when it comes to the underlying economics and politics of the actual geopolitics of the actual world. Um and you know remember the late queen um genius that she was in ter rhetorically I think you know I mean what better phrase is there than recollections may vary you know you bracket she's a lying bastard. Yeah, but the late queen when she came to the LSE and I was there during the visit got close associations with the LSC and she said why didn't anyone see this coming after the global financial crisis. Well, you know, some of us do see these things. You know, there were journalists like me that were warning about um you know, massive buildup of um synthetic debt in in in in uh housing markets in 2005, six and seven. There are people like me now saying this, not many, because people that have looked at it as closely as we have, they're too busy making money and they don't want the party to stop. And and there there has to come a point where um people who who have who who move in these circles who can see these things, you have to just say what you see. And of course I'll be lampooned.

[00:14:50]You know, I'm getting mullered on social media because of this because of the column I wrote on Sunday. Why? Because I'm I'm threatening people's hopes and dreams. And there's an awful lot of people trying to make serious money out of this stuff and just keep, you know, hyping the stocks up. But meanwhile, there's a there's a there's a silent majority out there that's, you know, tens of thousands of times bigger of ordinary people who are going to get absolutely waloped as and when there's another, you know, serious bout of financial instability. And we we're making a habit of this, Brendan, our generation, you know, you know, the dotcom collapse at the end of the '9s, you know, obviously the biggest financial collapse since the Great Depression in 2007. You know, the Euro zone almost destroying itself and the rest of the world in 20101. We don't hear much about that in the FT and the Economists these days as they're banging the drum to try and get us into the single currency back into the European Union.

[00:15:53]Then of course the the the massive policy error that was lockdown. I'm not saying CO wasn't important. Of course it was. But should we have locked down the kids? Really? Really? I don't think so.

[00:16:06]So we're we're at a state situation where I think if there's another massive financial crisis, a lot of ordinary people at a time when democracy is fragile and and and and political leaders in democracies are increasingly making fools of themselves and the rise of political violence even in democracies is clear for everybody to see. Then I think another big financial collapse at a time when governments across the western world haven't got the fiscal firepower like they had in 2008 because they're indebted to their gills because they printed hundreds of billions of dollars, pounds and euros of, you know, funny money. I think this is a a serious situation and, you know, shoot me down, but some people have to say what they think is going on. And I I wish I had a quid, Brendan, for every prominent person who's WhatsAppapped or emailed me and said, "You're right. Keep going." And I'm like, "Well, retweet it then.