超额收益播客——杰里米·格兰瑟姆谈泡沫、人工智能与均值回归（全文转录）

Full Transcript: Jeremy Grantham on Bubbles, AI, and Mean Reversion

Excess Returns

Subscribe Sign in

Full Transcript: Jeremy Grantham on Bubbles, AI, and Mean Reversion GMO Founder on Markets, Monopolies, and Long-Term Thinking

Excess Returns May 16, 2026

3

Share

Justin: Jeremy, thank you very much for joining us today, and welcome to Excess Returns. Jeremy: It’s a pleasure. Justin: Earlier this year in The Wall Street Journal review of your book — the book title is The Making of a Permabear — James Grant described you as a principled, value-minded, dogmatic nonconformist. And I have a feeling that brought a smile to your face. Jeremy: It did. Yeah, I like the dogmatic part in particular. Jim Grant never takes any unnecessary prisoners. Justin: No, absolutely not. We’ve had you on the podcast a few times. And you’re exactly right. And that’s what we’re gonna talk about today — the book is gonna be the foundation of the discussion. We were fortunate enough to be forwarded copies of that. Thank you very much. And the title, The Making of a Permabear, is your story.

It encapsulates your beliefs on investing, the stories of success in your career, the times at GMO where there was uncertainty and difficulty. And also, I think the book highlights what you think is important to successful long-term investing and also in life. So we look forward to spending the next hour or so with you to kind of work through a lot of these topics and a lot of the principles that you outlined in the book. I’m guessing that the title, even though many people that maybe listen to you might not think it’s tongue-in-cheek, but the title, The Making of the Permabear, was sort of a tongue-in-cheek, joking-around title in the sense because some people do kind of paint you in that corner. But the first question I wanted to ask you is, why do you think that people sort of paint you with that brush?

Jeremy: I voted, by the way, for putting it in inverted commas, and I was outvoted. Because I thought at least in inverted commas, people could tell it was tongue-in-cheek. And this way, I think quite a few do not. And that’s a shame really because the permabear comes out every super bull market. And the rest of the time, and in the bear markets, no one calls me a super bear. And my counterargument, of course, is that it’s been a very bullish 50 years. But there was a pretty decent decline in 1982. In 1982, the whole market sold at seven times earnings, and that was my first quote ever in a relatively short-lived publication called The Wall Street Letter. And in that edition of about mid-June ‘82, pretty close to the low, I said I thought the market was close to an unprecedented, or unusual, rally in both the stock and the bond market.

So that was my one bullish claim. It turned within a few weeks and then of course never went anywhere near seven or eight times earnings for the rest of our lives, up till now. And the other one was Reinvesting When Terrified, which was one of only two things I ever posted on our website that was not a quarterly letter. That was just a one-pager saying, “Dudes, get a plan together. Start to get your money back into the market. Of course you’re paralyzed. Everyone’s paralyzed. Who understands what’s going on?” And the banking system came very close to the edge, and everyone was suitably frightened, but the market is now cheaper than it’s been in 22 years. And on our infamous seven-year forecast, we see double-digit returns from the S&P and handsome returns from the rest of the world in equities, so get back in.

You will not call the low, but that doesn’t matter. It’s now cheap enough to justify rolling the dice and get in there. Even a bad plan is better than no plan at all because it gets you thinking, it gets your committees listening, and you start to develop a serious plan. So they’re my two... I’m afraid that’s all there is. That’s my entire bullish output really. But they were very good timing because the second one was not a few weeks ahead like 1982. It actually was published on the day the market hit its infamous low on the S&P of 666, which is of course famously the devil’s number. And now it’s over 10 times that. And you can’t say the period from then until now has been without its problems, so it is a fairly remarkable performance.

Justin: In the book, you talk about making the decision to be louder when you are bearish. What’s behind that idea? Is it because, to your point, the bullish calls don’t come out very often because you don’t see the opportunities as clear as when they are really truly opportunities? Or how do you think about that? Jeremy: I think that’s actually, if you’ll forgive me, a misreading. What I’m trying to say anyway is that it’s important when you’re bearish to differentiate between when you have really high confidence that the pain is imminent and desperate, and when you’re merely bearish. “Hey, guys, you invest today, you’ll probably be at least moderately disappointed with the returns you get at the end of 10 years.” That’s a kind of typical bearish comment.

And you’ve got to make sure you differentiate between those ordinary bear market calls and the real McCoy. And frankly, the average reader has a bit of a problem. So if you say the market’s overpriced, they’ll clock you down, and apparently they’ll equate that with abandon ship, the end of the world is around the corner. They’re not really good at hearing shades, and therefore you better give that up and ignore the idea of shading the opinion, make it absolutely clear. So in the great financial crash as it built up, one of my quarterly letters said just that — that it’s very important to make it clear when you’re really serious, and I’m now really serious. How do I express that? And I’ve spent the last week thinking about it, and this is it.

I think at least one major bank will fail. And that is in the July ‘07 edition before the failures started. And by the end, I think they should have let Citibank go, but of course several banks basically went belly up and AIG was bailed out with government money, and so on. And of course, Lehman and Bear Stearns. Justin: It really was an amazing time how those major institutions, many of them came to their knees and went out of business. So you were right, clearly right on that call in mid ‘07. And the — Jeremy: The other time — I haven’t made many ultimate calls, but the next one was in the middle of July of the following year, ‘08. And we’d encouraged people who had still the edge to be bullish to own emerging markets, and they had gone up and up and up, and frankly, they had peaked out around the end of the year.

So they were already declining a bit. But the quarterly letter that was published on the 15th of July ‘08 — we always published on the 15th — basically said, literally, “Abandon ship.” The old French expression, sauve qui peut, anyone who can save themselves, save themselves. And I even included the old nursery rhyme, “Don’t be brave, run away, live to fight another day.” In other words, “This is it, guys. Sell all your emerging... everything you can possibly sell, get the hell out. This is going to be disagreeable.” And I can say with a clear conscience, it was disagreeable. And the emerging market index from that day fell 50% in four months, which I think is the sharpest decline of a broad index ever. And we actually got to buy back our position in about October of ‘08, because it had gone from dangerous to very cheap.

A 50% decline will do that. Justin: One of the core concepts that you talk about in the book repeatedly is this idea of mean reversion, and this is from the book. You said, “Early in my career, I came to the conclusion that everything important in the markets is mean-reverting.” Our audience is very... we have investment professionals that listen to us, but then we have beginning investors. When you think about mean reversion — and I think this plays into being bullish or bearish on overall valuations in the markets and the way that we look at future returns — how do you think about mean reversion, and why is it so important, in your mind, in the markets? Jeremy: I think the abstraction that people might get is that mean reversion is kind of shorthand for history matters. You go back into the past and you see what happened, and you learn from that.

You take lessons, and you make investment rules. And they’re never gonna work perfectly, but the past, if it replays somewhat, if there are recurrent patterns, then you will make money doing that. And if it’s always completely new, then you will not. And one of the things we had noticed for the prior 200 years was that P/Es — price earnings ratios — tended to rise and fall. And fashions between small stocks and big stocks, they tended to rise and fall over several years. The same with emerging. Emerging market equities would have eight years of massive outperformance of the S&P, and then the S&P would have 12 years of massive outperformance of emerging markets and so on. There are a lot of cycles. And what a cycle says, it’s mean reverting. Otherwise it wouldn’t cycle like that.

So when emerging has an incredible run, they almost certainly end up overpriced, and eventually they roll down. And history was pretty clear. Asset classes mean revert. Sectors within an asset, like small versus large, they mean revert. And even companies, in the end, mean revert. And then to be more detailed, mean reversion at the corporate level says: if you make abnormal profits, you will receive competition. If you make obscene profits, you’ll get ferocious competition. If you’re having a slump and you’re not making much money at all, or even losses, you will frighten away capital and you’ll have no competition at all. And eventually, as the market slowly grows or rapidly grows, you will reach a period of shortage, and everything will recycle and your stock will go up, your profit margins will go up.

And there was a very clear history of that happening. And in recent years, one has to say, that is not as clear as it used to be. The bottom 90% of the market seems to fairly clearly still mean revert. And yet you have a novel emergence of a kind of elite, a few handfuls of stocks that seem to have gone from strength to strength. And you have to ask the question, why isn’t the money flooding in to compete these ridiculous returns and drive them down? And one is the winner-take-all nature of software. The guys who get there first have such an advantage it’s hard to break in. The other is the attitude of the government, of the administration, to monopolies. And you can look back over this interesting different phase — and of course this time is different.

You look back to about 2000 and you say, “In what way is it different?” And one of the ways, at the top of the list, is a steady concentration in every industry. There are fewer companies and bigger companies and more dominant companies in every industry. In most of them it’s not that big a deal. In some of them it’s massive. And what did the Justice Department, et cetera, do about this? And the answer, uniquely in this time period, is nothing at all. So companies were able to quickly develop not just domestic monopolies, but global monopolies. The Mag Seven were not universally seven global monopolies, but there were several in that group that were. And there is no better way to make money than to have a near monopoly or a complete monopoly, to be a price setter.

So ask yourself, are the Mag Seven setting prices? And the answer is mostly yes. They fulfill the characteristic of a workable, profitable monopoly, and they’ve been tolerated. In other eras, the government would get in there and say, “No, Standard Oil is too big. We’re gonna break it up into nine pieces.” And so on. And that’s pretty effective. It’s pretty effective at certainly breaking the monopoly. Then you have more competition from the pieces, and the prices tend to be lower, and the competition tends to be higher. The unfortunate characteristic of increased monopoly is that the growth rate of the system tends to slow down. So the profit margins of the monopolists go up, the share going to workers goes down, inequality increases. But the growth rate of the GDP tends to slow down.

So a lot of people think because there’s so much profit being made by the top 20 firms that somehow everything in the garden is great and growth rate must be higher. No, it isn’t. Growth rate in GDP is actually slower for the last 20 years than it was for the prior 20 and the prior 20 to that. Jack: Does AI exacerbate this or make it better? I could argue AI is gonna be a wind at the back of these big companies, but I also could argue it gives smaller companies a better chance to compete. If you think about Google having a monopoly on search, you could argue maybe they don’t anymore with everything that’s going on with AI. So does AI make the situation better or does it make it worse? Jeremy: So first, a few things.

AI is infinitely complicated. It’s the first time in history that I can remember where every group disagrees. People with a modest knowledge of AI all disagree. Is it going to destroy mankind or is it going to make us all sit on the beach and be waited on? And then you go up to the Nobel Prize winning level and they disagree. At every level of sophistication, there is vast disagreement on how AI will work out. Second point, AI is incredibly important. It will change our lives. Now, whether it will kill us all or make us all rich is part of that guessing game. But it is obvious to everybody that it’s an important idea. And there are several things to be said about AI. But the most interesting one to me is that when a new technology comes through, the early adopters often make considerably more profit than normal.

Their profit margins become wider than normal. The medium adopters maybe have a slight edge. The late adopters bring the market into balance again. And if I go back and look at prior cycles, what I have to conclude is that when the smoke clears, any new technology is merely a cost of doing business. Let me focus on the asset management business, which I know a lot about. I remember not so long ago when we had to dig deep and buy our first Prime computer, a mini computer. It filled the room the size of the one I’m in, it generated enormous heat, and it processed the data pretty quickly, and we had a competitive advantage for two or three years until everybody gritted their teeth and paid up for their mini computers. And then it became a cost of doing business.

Everyone in the investment management business had a computer, right? I can assure you everybody was not having higher profit margins than they used to. The profit margins settled back down to normal. The return on capital, which is the central driver in capitalism — how much money do you make on your investments — has been remarkably stable for a couple of hundred years. And after the computers had all been bought, all you got was a modest return on the cost of the computer. That was it. You had to put more assets into the game of money management. You made the same average typical return on it, and the early adopter thing was gone. Well, think of AI. The guys who adopt it professionally first have a huge advantage. But fast forward a few years, everybody is doing AI, aren’t they?

Everybody is paying for the service that they need. It is now, in that point in the future, the cost of doing business. It will not move aggregate profit margins or aggregate profits notably higher than they are typically. Only in the early adoption phase — where we are now — does that effect occur. Once everybody has settled in, this is just another cost of doing business. It’s obvious. I think it’s incontrovertible. But you would never guess it from the conversations of the day. There are many other problems and possibilities with AI that we haven’t talked about, but that’s the one I think I understand the most. Jack: Do you think it makes any difference, like, relatively in the market? You mentioned we’ve had this set of huge companies that’s been dominating the market for a long time.

Do you think it benefits the average company relative to them, or if it sort of reasserts their dominance over everybody? Have you thought about how that works? Jeremy: Yeah. I just sit here thinking all the time about these things. But I don’t think it will... it will not be quick. The average company will not catch up easily or quickly, and there can be a lot of pain along the way. But this is clear. If you back up five years even, you see a world where the Mag Seven basically represents seven monopolies, seven quasi monopolies, half monopolies, or at the very worst, duopolies and oligopolies, right? A handful of firms dominating the industry. You got a whiff of competition in the cloud, where you had three of them converging. But they got on very happily.

They either got on the telephone to each other, or more likely, they just deduced that if they behaved in a discreet way, the enemy would not get in a price war, and they didn’t. And so all three of them became a well-behaved oligopoly, and they made tons of money. That’s okay. But now we are faced with an utterly different world, where all seven of these guys, plus another 15 around the edges, are showing their hairy chests and beating their hairy chest and saying, “My 120 billion this year is better than your 82.” Every one of them means to win the AI war. They realize the first people to break through into a new level of technology will become trillionaires, et cetera — or they think so — and they mean to be there first. So we have gone from a monopoly world to a brutal competitive world, and we will stay there for years, and there will be blood in the streets because these are aggressive, take-no-prisoner type companies with lots of money, and they are heading right into each other’s teeth, and it will be exciting to watch.

But this is unlike, dear viewer, unlike anything we have been seeing for the last 20, 30 years as we watched Amazon inherit retailing and Tesla inherit EVs, et cetera. Jack: Just one thing on mean reversion before I lose that thread. How do you think your average investor should think about mean reversion with respect to valuation? If you look at the historical valuation data, you’ve got a period in history where it was much lower. You’ve got a new period — whatever it is, post-1990 — where the average is much higher. And then one of our guests pointed out recently that even inside this new period, you’ve got a mean that continues to rise. How do you think people should think about mean reversion and where we mean revert back to and how this whole process plays out?

Jeremy: Yeah. Well, these are each topics you could run a podcast on. I noticed that GMO’s emerging market fund in the last 12 months is up 70%. And this is not a normal occurrence. The S&P is not up 70% — it’s up, you know, 25. That’s a big gap, and that is the spirit of mean reversion. Last January — not the January of this year, but the January of the prior year — in my podcast, I said, “The good news is that outside the US, equities are pretty interesting, pretty cheap, and they’re about as cheap relative to the US as they have ever been. They’ve only reached that point two or three times, approximately the same level, and from that level the rest of the world outperformed the US handsomely, typically for multiple years.” And happily, I got that bit right.

What I got wrong was I implied that the US would not be up 25%. On the other hand, I wasn’t suggesting for a second that emerging and international value would be up 50 or 70. So everything was up more than one thought. But there was a massive mean reversion from an all-time low of the value of the world versus the S&P, and the world has handsomely outperformed. And typically, this goes on for a few years, and I expect it will. So mean reversion is alive and well in certain areas, even as the Mag Seven have made life very difficult for the study of mean reversion. Jack: Do you think there’s a case, though, that the mean should rise over time? You could argue we have better companies. The mix of the market is more towards high-margin companies.

Do you think there’s kind of an in-between here where we’re going to get mean reversion but maybe the mean should be rising to some extent over time? Jeremy: For a couple of hundred years, the return to capital was pretty even. And is there a chronic systematic shortage of capital? I don’t think so. So the reasons for moving return on capital above normal should tend to be idiosyncratic, short-lived events. And when I say short-lived, even the emergence of the Mag Seven will only be a blip in history, of course. Is there anything that feels like it’s permanent? Why would the required return to invest a dollar in the economy suddenly go up after 200 years? It ebbs and flows, of course. You have economic cycles. You have cyclical peaks and troughs.

But is there any reason why the central tendency should go much higher? And the answer would be yes, if there’s a chronic permanent shortage of capital — if there are more people bidding for capital and nobody is saving. There’s not a hint of that. Seen through my eyes and other people’s eyes, there does not seem to be a permanent chronic shortage of capital. And therefore, in the long run, if capitalism is working effectively, high returns are competed down. If you get in the way and you allow monopolies to start — like Rockefeller back in the days where he arm-twisted the railroads and said, “I am doing so much business with you. If you don’t increase the rates for my competitors, I will leave and go to another railroad.” And so they did.

And whether that was breaking the laws of the day or not had to be decided. And Teddy Roosevelt decided it was breaking the laws, and slashed and burnt. So these are very important, very difficult issues to deal with. But history is pretty clear. The return on equity is fairly stable in the long run, and if capitalism is working, there should be competition to push the return down. And for a while, new technology can create a monopoly feeling. But in the end, they all converge — as they appear to be about to do. They’re in the early stages, are they not, of investing vast sums of money to attack the same market. This is the kind of process that ends up in the long run creating regression to the mean. Jack: One of the topics you talk about in the book, which many people would argue is very relevant right now, is the idea of bubbles.

But before we talk about that in relation to AI, I just wanted to talk about what you think a bubble is. You talk about that in the book. You say there’s two conditions for a bubble to form and that they’re quite simple. How do you define a bubble? Jeremy: A bubble occurs typically when economic conditions are nearly perfect and there’s plenty of money around. If you get those two things together you will have market prices rise. If they stay in that condition for several years the market will rise and rise and rise until it’s in a bubble. But we asked ourselves 27 years ago, we better define a bubble so that we can know what we’re looking at. And we defined it in nerdy statistical terms as a two-sigma event.

And a two-sigma event in statistics is the kind of thing that happens every 44 years in a random series. So imagine yourself putting numbers down in a little green book from a roulette wheel that you’re spinning. That’s a random series. And a two-sigma event is the kind that would happen every 44 years in the stock market. On the roulette wheel it’s equivalent to, let me guess, eight or nine reds in a row. So you get an arbitrary measure — two-sigma — based only on the price. So we’ve kept it unbelievably simple. Give me the price series and we will tell you when it’s two-sigma. And we went through all the asset classes and we had well over 200 two-sigma events. And there were a few that were paradigm shifts — the other group.

But just a handful. And they were all commodities. Oh, except one or two emerging countries as they became developed countries — they went from non-serious to serious and so they had a permanent move up in price earnings ratios and so on. I think India was in that place many, many decades ago. But, and this is the key, if you limit your two-sigma events to serious developed country stock markets, every single one of them — when we did our opening study — every single one of them always had gone back to the preexisting trend. That is to say, the trend that existed prior to the bubble. By the end of the bubble of course the trend line is a bit higher, but they more than met the standard. They went all the way back to the preexisting trend, every single one of them.

I think there was something like 27. And there were no exceptions. And so we look at a two-sigma, we see it as a complete outlier. We have enough faith in history to think they will go back. And therefore, back in the day, we thought Japan was horrifyingly out of line. We actually thought owning it was fiduciarily irresponsible. And so in our foreign accounts, even though it was then 45% of the benchmark, we went to zero. And we watched it rise to over 60% of the benchmark, and we watched it go from 45 times earnings to 65 times earnings. Japan had never sold over 25 prior to that great bubble, and it went to 65. That’s the kind of outlier that makes value managers — or should make value managers — wake up in the middle of the night screaming. Because that was long and brutally painful.

And we had to ride that out. We were out of Japanese stocks completely. We underperformed by 10 points a year for three years and came out when Japan had finished breaking far, far ahead on the round trip. Jack: It’s funny, I think that Japan bubble is under-followed by people. You mentioned 60% of the benchmark — that’s insane. It’s crazy to think about. Even if you look at the dot-com, that Japan bubble is many, many levels above that. Jeremy: Yeah. And it’s substantially ahead of today. And that’s the thing. You think this market has broken all the rules. Nah, it’s a piker compared to Japan. And then of course we had to go through the tech bubble, or the internet bubble, if you prefer it. And a similar story. The market had never sold over 21 times earnings, which was the peak in 1929.

It hit 21 again in 1972, the Nifty Fifty era. And it hit 21 at the end of 1997. We had perfectly good competitive performance. We had managed to fake it, even though the market was at 21 times earnings. As high as it had ever been, we were in the game. We were a player. Perhaps we were bending our rules a little bit, who knows, with hindsight. I’m a little suspicious that we had done that well. And then breaking through 21 times, it went, you know, 23, 25, 27... you must be joking. 31, 33, 35 times earnings. Nothing had been like that before. And along the way we made a value bet, we doubled up, we tripled up, we rolled the dice. And in two and a quarter years, we lost about half our book of business. Slightly over half our market share of the asset management business.

Pretty quick. Heartbreakingly quicker than I would’ve guessed beforehand. But it came back down, returned all our money, and then made us a ton of money beyond that. But none of the clients came back, by the way. Half the clients left, not a single one came back. Even though we’d made the right bets for the right reasons, and we had won. And in the decline of 2000, 2001, and 2002, we made very nice money as the S&P went down 50% and the NASDAQ went down 80. We made good money — double-digit money in 2000 and 2001, and half a percent in 2002, hanging on by our fingernails. And we were up over 30% by the low, and the S&P was down 50. So that was okay. And that took us through to the housing bubble, which is the only one we ever got right.

We wrote a series of quarterly letters saying, starting in the spring of ‘07, “Watch your tails, things are looking tricky.” And then, “This is like watching a slow motion train wreck,” unquote, as everything seemed to move through the paces. Market was continuing to go up, but the debt levels were continuing to look worse. The financial system was beginning to look shaky. And the more we looked into the quality of the subprime instruments, the flakier they got. And although that was not our specialty, we did understand that it was a drastic source of risk. And then we said, “Get out of everything, really, but emerging.” And then finally, slightly after the peak, we advised everyone to sell everything they possibly could, including emerging, as I said before, in July, on July 15th. And we made a huge trade out of everything we could.

In some accounts you have constraints, but we sold every emerging dollar we could before, of course, we published the letter for the rest of the world. So they’re the three big cycles until you get to ‘22. In 2021, we had every condition of a bubble in place, including my very favorite rule that I could or could not describe to you, which was flashing brightly. And that rule only occurs at the top of 1929, the Nifty Fifty of ‘72, and the tech bubble of 2000. It had only occurred three times in history, and the fourth time was late 2021. And that is when the market leaders of the second to last year — where they’re all up 70, 80% — the market leaders start to go down as the broad market, the S&P, continues to go up. So in 1929, the S&P was up 40%, but the S&P had an index there of, called the low-priced index, which were $5 flaky fallen angels typically.

And they had been up 80% in 1928, and come January ‘29, they started to go down. And before the market broke in October, they were down 40%. They were actually down 40% with the S&P up over 30. Oh my — that was what I call the biggest scream from the stomach of the market in history. And nothing like that happens again until 1972, where you get a significant but only a faint echo in comparison. The S&P goes up 17 in 1972 and the average of the S&P goes down 17 — so that I can remember it forever. And nothing like that happens again until 2000. As the tech bubble begins to break, the growth stocks go down 50%, the S&P continues to go up. The balance goes up another 14%, so that you have the same level on the S&P in September 2000 that you had at the growth stock peak of March 2000.

Why does that happen? I think it’s because the players — you know, Mr. Prince once said, “If the music’s playing, I’ve gotta keep dancing.” And that is an ultimately important statement of how the professional money management business works. You cannot fight a major bull market. It is ruinously unprofitable. It’s not optimal behavior at all. So the big companies never fight the market and never tell you to get your ass out of the market, and they never have, they never will. If you are waiting, dear listener, to be told to abandon ship by the Goldman Sachses and the JP Morgans, you will have a long wait. They have never told you, they never will tell you because it’s simply bad business for them. They do very nicely being bullish all the time and trying to be a little quicker and slicker on the execution — on the way up and the way down — and it works very well.

Thank you. Jack: Well, I think it’d be a terrible interview if I didn’t at least ask you if that rule is triggered now. We’ve kind of seen a reversal — the tech stocks have taken off a little bit, but last year I believe only two of the Mag Seven outperformed. So have we seen any signs of the market still up, the leaders going down? Jeremy: Unless I’m missing it, and it’s a complicated business, I would say absolutely not. As far as I can tell, junky stocks are beating safe stocks. We have an index where we go long quality and short junk, and it’s being bashed around the ears. Okay? But go back to 2021, and it was not being bashed around the ears. In 2021, you had a perfect example where the Super Blue Chips were doing well, but the flakier ones had peeled off.

For example, Cathie Wood’s portfolio. She had been brilliant off the COVID low. And then starting a few months into 2021, Cathie Wood’s portfolio that had been brilliant — had tripled and doubled, I mean, doubled and tripled — and then it started to go off. It fell off. And it lost a lot of headway. And one by one, you remember the meme stocks? They were taken out and shot. As the market climbed handsomely through the year, they were shooting the meme stocks. And I inadvertently owned a meme stock, QuantumScape. A long story — I had a huge position. That was the only terms we could get the stock on, 5% of the company, and it was 5% or nothing. And that was far too big for the foundation that we run for the protection of the environment. So that was the only thing I owned myself.

And so I owned it, and I watched it come as a SPAC at 10 in late 2020. And while the market was still roaring upwards, it came as a SPAC at 10, four times my money. Not bad. Much better than a kick in the pants. And in three months it went to 131. At 131 it was worth more than General Motors. It was a research lab that was going to develop one day, we hope, a solid state battery. But today they still don’t have a commercial product on the market. But they will, I think, in a year or so, maybe even in six months. In any case, they did not then. It was an idea still, and yet it sold for more than General Motors. There was never an idea that scale in 1929, by the way. I mean, that you have to admit.

To be worth more than a seasoned firm selling millions and millions of cars when you have a research lab with years to go before you have any sales — forget profits. So that was the leading... when did it peak? It peaked, sadly for me, in December 2020. It was the first meme stock to peak, and then all the other meme stocks peaked in the first half of the year, and Cathie Wood’s portfolio peaked about mid-year, and so on and so forth. It was classic. It was just like the behavior in 1929, 1972. Better than 1972. Better than everything except 1929. And we do not see that definitive behavior. And I’ll tell you why it happens. I was talking about Mr. Prince. When the music’s playing, we’ve gotta keep dancing. But you don’t have to keep dancing with Puma Tech.

Puma Tech is my favorite example since it was the most advanced stock in ‘99 in that bubble. And mysteriously, in the next few months, people made sure they recycled out of the Puma Techs into the Coca-Colas. So Mr. Prince was still dancing, but he was veering towards the Coca-Colas. And why not? History tells you in 1929, et cetera, it is better to go over the waterfall with Coca-Cola, with the survivors, than it is with the Puma Techs, which quickly ceased to exist. And Coca-Cola of course did not. So it’s a strategy that makes sense, and you can imagine how it would happen. They say, “Yeah, I’m optimistic and the market’s been lovely, but this is ridiculous, or it’s getting ridiculous. I better make sure I play the game, but at least I will play it with survivors.”

So I understand the logic. I see why it only happens in extremes, and it clearly happened in 2021. And 2022 was a very nasty bear market in case people have forgotten already. S&P down 25, growth stocks down 35, Mag Seven down 40. The bond market — the worst year in the history of the bond market. And then I like to say my nice prediction, my wonderful bear market was rudely interrupted by AI introducing — what was it — ChatGPT. Jack: October of ‘22. Jeremy: October ‘22. ChatGPT. Wow. And that ruined the bear market. The following year, the following 10 months, all the other stocks on average were down, by the way. They continued down. They couldn’t quite believe that the bear market was over. But the Mag Seven doubled and dragged the S&P with it, and then 10 months later, the rest of the stocks threw in the towel and joined in the bull market.

And my belief is that we would’ve gone into a minor recession. Why didn’t we? Because they stepped up the CapEx for AI to such a huge degree that if you back it out of that time period, it appears to account for all of the GDP growth. So we would have approached a zero-growing GDP. As Keynes explained, animal spirits is very important. As the market goes down and the growth rate of the economy goes down, animal spirits sometimes can get very pessimistic. They affect the investment rate, and you slide into a recession. My guess is that in 2023, we would’ve moved into a recession, and the market would’ve gone down another 25%. And AI headed it off. So now we’re in terra incognita. We’ve never been here before, where a major event — AI — with massive CapEx, unprecedented in any era, even including railroads as a percentage of GDP.

We have a bubble forming out of a bubble that was only halfway completed. Only halfway deflated, and then resuscitated by this magnificently complex and important business, AI. Justin: One of the things you had mentioned was in the late ‘90s, early 2000s, that was a pretty dark time for GMO as you were losing clients and underperforming just because everything was going parabolic. You obviously were positioned great when that party came to a standstill or came to an end. But I’m curious — and the question is rooted in investor behavior — because what I was thinking was, even if those clients, which you said none of them came back, but even if they would’ve come back and said, “You know what, Jeremy and GMO, we wanna put money with you,” would you have even taken them, if that’s a fair question?

And really what it gets at is the investor behavior aspect and the importance of believing in a strategy, a firm, and being able to see through those times and get the most out of a strategy. So there’s a couple different questions wrapped up in there, but I’m interested in your thoughts. Jeremy: I would very happily accept anyone who came back. Let me point out that we had 30 billion going into the tech bubble, and the market roared upwards, and we roared downwards to 20 billion. So a typically successful firm starting at 30 billion would’ve gone to 50 or 60, and we went to 20. And then on the way down, they went from 50 to 35 or 30, and we went from 20 to 22. But then the third phase, starting at the bottom of the market in 2002, we went from 22 to 165 in four years.

All new business, all thinking GMO is our kind of guy. They’re tough. They stood their position. They got it right. And very few of them really aware that had they been exposed to the stress of badly underperforming in a bull market, they too might have found that the committee was advising them to fire us. Committees, after all, are made up of illustrious, important members of the college — famous alumni who are in hedge funds and private equity and therefore of course know everything about investments. And these are powerful guys, and they tend to be much more short term than some of us. And they lean on the hired guns. The hired guns weren’t too bad in ‘99, 2000. They would call up on a Friday evening and say, “Oh my God, Jeremy, that last meeting was terrible.

They’re beating me up. Give me some ammunition for the meeting next week.” And I’d tell them the best data we had, and we’d spent a lot of time in ‘99 reviewing the data and improving it. So we’d hand over impeccable data about mean reversion, and he’d go there, and he’d call back a week later and say, “Oh, thank heavens. They bought it for the time being.” And then a couple of months later, he’d call back and say, “Oh, this was even worse this quarter. And I think it’s gonna be either your job or mine.” And then a couple of weeks later, “I’m sorry, Jeremy, it’s you.” And we’d be shot. But they kind of believed — a lot of them believed — that we lived in a mean-reverting world, and that these stocks were dangerously overpriced.

And they were being forced by their committees to fire the value managers and hire yet another growth manager. In a couple of cases, we managed the whole account. And in both cases, funnily enough, there was one important member of the committee common to both of them. And they were going to shoot us. And I made the most passionate plea of my investment career probably, and I said that firing a value manager under these conditions and hiring another growth manager was on the very border of fiduciary responsibility. And I knew I’d get fired, so I was thinking, “Well, I’ve got nothing to lose. At least I can say what I honestly believe.” But it turned out there were enough nervous Nellies on the committee that they persuaded the boss to go with shooting half of us.

So they shot half of us, and they did indeed put the rest in another growth manager. This was fairly near the peak. And in 21 months, their half was half of our half. They had wiped out a quarter of their endowment in 21 months. That’s a tough world we live in. Justin: Well, I guess that comment on the fiduciary aspect was more right than wrong. But yeah, interesting. Thank you for sharing that history with us. And Jeremy, there’s a lot — we really touched on a few aspects of the book, and we have a lot more to ask you. So maybe this conversation can continue at another day in the future, if you’re so kind to come back and join us at some point down the road. But I wanted to ask you two kind of standard closing questions here.

Well, one is not a standard closing question. One is more about how you ended the book, and this is talking about the purpose in your life, in your career. And for people that don’t know, you run a number of philanthropic efforts on a number of different things. So this idea of purpose, it’s not just talking about it, it’s actually putting your money that you’ve earned and your resources behind some of these things. So just I thought we could, at least on this point, share with us and our audience how you think about the purpose in your life that you would like to impart on others. Jeremy: Yeah. Well, I have come to believe that purpose is pretty darn important, that as you get older, you don’t wanna be sitting around thinking, “What a useless life.

What was it all about?” In the end, it’s about doing your best or trying to improve the system in some way that you can. And so my recommendation to anyone who’s young and graduating is to make sure that your career is likely to be useful. Be an engineer. Be a scientist. Be in the health business. Be anywhere where you know that what you’re doing is useful. Start a new company. I think venture capitalism is the very best part of modern capitalism. I think capitalism is a bit fat and happy and too monopolistic, but I do think that the venture capital world generates very important new ideas. If you look at the Mag 7, ask yourself this: how many of them came out of the venture capital business, 50 years ago, 30 years ago, 25 years ago?

GMO was started a month or two after Apple and Microsoft. It’s stunning, and they’re the old fogies. The other guys are all younger than we are as corporations. It’s a really good idea to have a target and keep going. And it’s particularly important when you’ve retired, otherwise you sit around playing golf and wondering what it’s all about. And there are so many really important risks — risks that could bring culture, civilization, safety, growth to a screeching end in the not-too-distant future unless we address them. There’s the chemicals that we swim through every day and we eat, the food we eat covered in toxic chemicals. The fertility rate of our species is going to hell in a handbasket. This is all a matter of public record. Our sperm count is down to less than 30% of what it must have been in the hunter-gatherer days, and down well over half since 1972 when the academics took a brilliant survey of the topic.

So we’re busy going out of business as a species. The trouble with capitalism is killing off its own market by resolutely defending their right to produce toxins. And they were thinking of going to the Supreme Court to make sure that fossil fuel companies and chemical companies could not be sued for the damage they create. So fine — go ahead. But at least toxicity is regional. If they wanna behave pretty well in the EU, they live longer and healthier. If you wanna behave badly in the US, the reverse. And we’re the only country — 15 years ago, our life expectancy was longer, and our healthy years have gone even worse. The gap between us and Sweden has gone from two years to six years in the last 40. And as I said in the book, or in a quarterly letter, my estate would be willing to bet you a considerable amount of money that come back in 50 years, that gap will be eight years or more.

Let me give you one example. In cosmetics, there are 10,000 chemicals. The EU has banned 1,500. If they banned carefully the worst ones, that’s a pretty good dent in the danger of cosmetics. Canada has banned about 550, and the US has banned 12. And as a Brit, my tendency is to say, “Fine, you guys wanna kill yourselves, go ahead.” But it seems like a bad idea, doesn’t it? And then we have climate change. If unaddressed, that will make life very unpleasant. The last two years, for the first time, it’s about a half a percent of global GDP preventing climate disasters, fixing them, et cetera. Half a percent of GDP. That is gonna get steadily worse, and everyone in the world realizes this is a serious topic now except the US. It’s not a good place to be.

So these are really serious topics, and there are plenty of others. Even worrying about AI and trying to decode some of the problems would be pretty worthwhile. But many and most jobs, in fact, are just making a living, and I recognize someone has to do them. But if you have a choice, pick something useful. Justin: The last question that we always like to ask all of our guests is, based on your experience in the markets, if you could teach one lesson to your average investor, what would that be? Jeremy: The really impressive thing about Homo sapiens is it doesn’t like to deal with long term, and it doesn’t like pessimism. I suspect we’re optimistic because it helps you to survive. If you look at the millions of years as we developed, I think pessimism is probably not very helpful, and that the willingness to see the bright side and keep on driving through any dangers that came along was helpful.

Who needed the long term for 99.9% of our history as a species? Who needed to worry about the long term? Eventually, no doubt painfully, we learned to try and save enough for the winter, to stash away enough potatoes or whatever, roots, berries. And that was it. So as a species, we do not do unpleasant news, and we do not do long term. And my advice to the investor is try and fight those two tendencies. Try and be realistic rather than systematically looking for the good news, and try and focus on, say, a five-year horizon. What is going on? So since the war broke out with Iran, the S&P is up 5%. Oh, I get it. That war must be good for us, right? Somehow, somewhere. I’ll tell you, it’s very, very good for the profits of US oil companies.

Terrific. Not very good for the people who buy their product, but terrific for the oil companies. But can you tell, dear listener, that since the war, this has not been very good for the world, for Europe, for anybody who needs oil? The only two who look even semi-reasonable are Russia — our good allies here — and the US. The US is oil and gas neutral. We don’t need to import oil and gas. All that’s happened in the US is as the prices went up, the oil companies made the money. The people buying at the pump lost an equal and offsetting amount. But together we did okay. In Europe, everyone is bleeding. In the Far East, in China, Japan, they’re all bleeding. But Russia is making money, and the US is merely shipping money from the poor to the rich.

So try and overcome these two tendencies, and you will be one of the better investors. Justin: Thank you, Jeremy. Jeremy: You’re welcome.

3

Share

Discussion about this post Comments Restacks

Top Latest Discussions

No posts

Ready for more?

© 2026 Excess Returns · Privacy ∙ Terms ∙ Collection notice Start your Substack Get the app Substack is the home for great culture