"I Would've Made Tens of Billions": Seth Klarman's Biggest Miss (and What He's Buying Now)

The opportunities that happen in the worst environments can lead to returns for years afterward from deploying more capital into those markets.

What does that translate into in this environment that we're in right now? This is a stretched environment, but it's an unusual circumstance. People have asked me, is this a bubble? I think it has characteristics of a bubble.

Thank you, sir.

I think the way we were founded probably affects how we operate day to day. Over that whole time, we were formed as kind of an expanded family office. Several families were selling an interest in TV station Channel 5 in Boston. Another friend of theirs was selling a computer publishing and consulting business. So that's how our founding $27 million of invested capital came together. Our mandate wasn't really to grow an investment business; it was to deliver good returns with limited downside to those clients. So that's been a driving force. The way we protect capital is we focus on downside as much as upside. We do meticulous, fundamental research on every company. We will hold cash in the absence of immediate opportunity. We don't leverage the portfolio at all. We'll buy senior securities like debt for a structured investment in the private markets, so that we're structurally senior in a lot of what we do, which protects on the downside. And finally, we do macro hedges. So it's cross-asset, asset agnostic. We are in many different asset classes. Originally we were in equities and credit, but over time we got a little bit involved in commercial real estate during some of the blips in that business and realized that real estate is as big a market as equities.

I think that most participants in the market, and understandably, the media are going to go with whatever the direction of the market is. It's a downer to talk about downside. Nobody wants to hear it when the market's going up every day, but there'll be a moment when things will correct. There'll be a lot of pain, and usually in those environments, if we've protected well, which so far we have, we'll be able to put capital to work when our competitors are maybe standing back a little bit. So the opportunities that happen in the worst environments can lead to returns for years afterward from deploying more capital into those markets. What does that translate into in this environment that we're in right now? This is a stretched environment, but it's an unusual circumstance. People have asked me, is this a bubble? I think it has characteristics of a bubble. It's an optimistic tone around a technology, around a new era, this kind of new era thinking. And you certainly see it in places like when Allbirds, the shoe company, added AI to their name, the stock did well. And that's crazy. That's reminiscent of the era. But AI seems like a technology that could be so game-changing that it would be hard to dismiss it or call anything in particular around it a bubble. What's hard is that it creates uncertainty. I think none of us can know with any confidence how it's all going to pan out. What are the winners today? Are they going to still be the winners? Will it be a winner-take-all technology, or will there be room for a lot of different winners? Where's it all going? And what's the right price to pay? Are these good businesses that are forming? Are they great businesses? What will the damage be? So I think the market is tending to think about AI winners. Everyone wants to own those. AI losers, everyone wants to avoid those. And AI agnostic, which people are also not interested in because the returns are not as exciting as AI winners. So we think we're spending a lot of time on AI agnostic, where people are just not paying attention and the price is drifting lower. We're looking for some perceived AI losers that we're not sure really will be losers. For example, in the credit market, there's a lot of software-related credits that are getting pretty clobbered, and we're not sure. I mean, they're trading at very, very low multiples of cash flow, and through the debt layer, even lower multiples. So at least it's something as credit investors that we're taking a look at.

Yes.

I think there's a temptation by a lot of people, and especially by academics, to think of value as cheap, paint-by-numbers. The lowest multiple stocks. We realized really at the beginning of Baupost that that's not the right definition. The right definition is to think about what a business is worth. Obviously, a growing business is worth more than a stagnant business. I think that the melting ice cubes of today's businesses are melting faster than ever. If you've got a business problem, you may be eroding really quickly. So we stay away from those. We don't care how, you know, four times cash flow, we don't care, because that's not ultimately going to have value that you can get your hands around. And what we want with our value approach is to think about what do we know for sure, especially at a time that's this heady, that people are thinking so many years out. When you're paying 40 times multiple or in some cases infinite multiples, you've got to have some degree of conviction about a very distant future. And I just don't see how we can do that. We don't know if AI is going to turn into AGI. We don't know what that means for the economy, for employment, for inflation. So I think if you throw this much uncertainty into a market, you'd say maybe the market should be at a lower multiple to accommodate all this uncertainty. And instead the multiple keeps going up. So that's what makes me think there may be bubbly aspects of it, even if the technology proves to be on the higher end of expectations.

Yeah, I would say about 10% of our book is probably going to benefit from the immediate rollout of AI. We have some Amazon, we have some Google. And what we like about those companies is they are such enormous cash flow machines. They are very versatile and flexible. So they have found ways... Google is obviously making their own chips. They both are players in various aspects of AI. The Amazon data centers are a really very quickly growing, much faster than people thought even a year ago. And so if you look at those companies, first of all, you don't have to buy them the second that there are blips. In almost every company, you get windows of opportunity. We put on Google a few years ago. I'm not supposed to talk specific names, but it was really at a below-market multiple. So you don't have to be making heroic assumptions to like stocks like that from time to time. We also own some raw land that we think could be sites for future data centers. It has adjacent power, and we have paid very small dollars to create valuable optionality with data center land.

We know that there's, at the moment, and it could change, insatiable demand for data centers. The key thing is: will it get built? Will it be permitted politically? And do you have power, which is the most important thing? And we have an experienced operating partner who is finding sites and who's banking the land until we can figure out if there's enough demand. So we're not sure we're going to build any data centers, but we can make them available to people who are needing more capacity. We also own a private investment that is non-China Asia data centers, and it was spun out by a Chinese company. We have a stake in it on a public basis, and participating in that private offering has given us a position at a really significant discount from what public multiples are. So I can't vouch for whether the public multiples are right, but I think what we've created in the private market is probably 40% of what you'd pay in the public markets, which seems like it might be valuable.

We're uninvolved. I would say the question I would ask as an investor is: A, is it winner-take-all? Are you sure that those are the two companies? Those are the only companies? B, those companies seem to eat a lot of cash to keep the models trained, keep the models updated. If they ever fall behind the curve and aren't the best model, I think those companies could have real problems. So they have to keep investing. That's not Warren Buffett's definition of a great business. And those are enormous valuations for companies that are still years away from bottom-line profit. That said, and I'm trying to heed what Eric Schmidt said a couple of years ago, which I think is largely true, probably: don't make the mistake of underestimating AI. It could be a bigger winner than anybody's discounting. And if it is, though, you still want to think it through: is the right way to play it through the model companies, or is the right way to play it through agents or some ancillary businesses that are selling into it? And I think that's still yet to be seen.

AI agnostic would be a company that you think just isn't going to be disrupted, that AI will not have much effect at all on their business. So maybe they'll have some chance to cut costs, like roofing, like housing supplies. Sure. I mean, there may be companies. I think travel largely... there may be ways for people to cut costs, but that's not the primary benefit of AI. If AI is able to actually help you discover drugs with higher certainty, much faster than ever before, that's going to change mankind. That could make all of us survive diseases that would kill us right now. So the highest-end uses of AI are not cutting costs; they're figuring out things humans can't figure out, like reading our X-rays more accurately. So I think that, and there obviously are uses that we can imagine even better than that. So I'm not... this is not how I'm preoccupying most of my time. I decided early on in this: A, I'm not going to be at the cutting edge of AI. I'm not a tech guy, not at the cutting edge. B, I think that there are going to be... I have to know about it. I can't be at an information disadvantage, even though I'm not at the cutting edge of the whole thing. And I need to stay focused. And what I really don't want to do is whiff when a pitch comes right down the plate. So whether that's a stock that's not related to AI, whether that's a distressed debt deal—there's more distressed debt all of a sudden in the last few months, and we're seeing more to do there—or whether it's a private investment like commercial real estate that we think right now the logjam is starting to break. Fundamentals are starting to improve in a lot of real estate markets, and we're seeing opportunities to deploy capital at significant discounts to replacement cost, at very attractive returns without heroic assumptions. And so that's one of our favorite areas.

I think people are burned on real estate. I think most people that have exposure to real estate are like, I have enough. People haven't done well in real estate for a decade or more. And I also think it's a lot of work to go find an individual building that you'd like to buy at a price where you can't deploy that much capital. So I love being below the radar of the big gorilla firms, and I love being able to find individual investments that are $50 million, $100 million, whatever it is, that I can put to work with confidence to get returns that are disproportionate to any risk we're taking.

The largest areas of distress, I'd say, are idiosyncratic. We're seeing a company in Brazil that has been forced to do a restructuring. We have a position. They just announced what the restructuring looks like, and I think they're going to be securities coming out of that restructuring that probably will trade reasonably poorly, making them a potential source of further opportunity. We saw a large private equity deal where the debt was impaired. They did an exchange offer. And again, there are securities that have come out of that that we think have a chance of being quite mispriced. And the business probably is... they probably caught and cinched it and took all the write-downs they're going to take, and it's probably about to turn. And so those securities, also preferred and equity securities, could be quite interesting. So credit is inherently interesting because of the transition of ownership. Bonds are bought to get paid at par and earn your coupon. When they're not going to do that, people dump when they get downgraded, people dump when they file for bankruptcy, people dump. But then even when the distressed guys swoop in, they maybe don't stick around for the last dollar because they're moving on to the next distress situation. And new capital is needed to absorb the securities to wait for the turnaround of the business. So there's always something to do. And inherently, as you know, value investors are patient.

Yeah. It's been a very strange environment where we really haven't seen significant downside volatility since the GFC. We haven't seen a lot of corporate bankruptcies. As I said, we're starting to see more. We're seeing them in private credit too. So there's a lot of trouble out there. I don't have a market forecast. We really never make market forecasts. I don't know what's going to happen, but I think we are due for a credit cycle. You can imagine scenarios where we're in a sort of inflationary boom environment. We're onshoring a lot of companies. There's a lot of demand within our real estate book for industrial land, for warehouses, for cold storage. Even so, things are starting to heat up. And I think that with all the demand, all the money that's going to build data centers, you kind of feel that that's inflationary. The building of it, demand for land, demand for equipment, demand for electricians, it's through the roof. Inflationary short term, deflationary long term it might be. But those are also... once the genie's out of the bottle, will people be happy to not be getting their wage increases? And it's such a political hot potato. And of course, increasingly AI is a political potato. Are we going to build all these data centers? If we're going to build them, where are we going to build them? Because a lot of citizens are pretty upset.