Transcription
If this exuberance doesn't produce a money-losing bubble, it'll be the first. As Buffett says, it's only when the tide goes out that we find out who's been swimming naked.
And there is Howard Marks, an incredibly successful Buffett-style investor, founder of Oaktree Capital, one of the world's most successful asset management firms. And recently, he went on the Prof G podcast to explain exactly what he thinks of the IPO frenzy we're currently seeing. So, I'm talking SpaceX, Anthropic, and OpenAI. Honestly, it was a really fantastic interview. I really encourage you guys to check out the full thing. I'll leave it linked in the description, but in it, we get a really good sense of what Howard thinks about not only these IPOs, but also what he thinks about the AI bubble more broadly, whether the stock market is too overvalued to participate, and also what to do about it as rational long-term investors. You know, people that aren't just in it for a bit of a gamble. So, in this video, let's take a look at exactly what Howard has to say, and we start with his thoughts on the IPO frenzy that we're seeing. And he does have some thoughts.
If somebody will tell me what they think Anthropic net earnings will be in 2036, I'll bet them that they're not within 50% of the truth. Of course, we have to wait 10 years to find out, but if I'm right and you make an investment in Anthropic stock in the IPO, you have to accept the likelihood that what you're doing is closer to speculating, and I don't say that word pejoratively, than analytical investing.
This is what I spoke about a lot of in our most recent SpaceX video. We have to remember that IPO timings are decided by the company. The company picks the time, which means they generally look for the frothiest, most speculative market that they can possibly find, because it means they can raise a lot of money much easier, right? That's why SpaceX, a company with just 6.8 billion in operating cash flow, just raised 85.7 billion like it was nothing. And the whole company is currently valued somewhere in the region of two and a half trillion dollars. Well, at least it's valued that at the time of this recording. Who knows where it sits when this video comes out. But the point Howard's making is that this sort of stuff, it's not grounded in reality and we have to recognize that. It's not the sort of thing that Warren Buffett would teach.
You know, the value investor, the old-fashioned investor like me and and my fellow travelers, what we do is we look at what it makes today and what it's potential earning power it's building. We try to figure out what its earnings will be in five or 10 years. We put a what we think is a reasonable valuation on those earnings largely related to the the earnings potential in the subsequent decades. And then we look at the price today and we try to figure out whether today's price is fair relative to that earnings power. I don't think I've ever seen an industry [music] or companies where that is less feasible. So, I don't think there's a an analytical or what we call a value-based way to decide whether or not to participate in these IPOs and if so, at what price.
This takes me back to the valuation I ran on SpaceX in my other video, right? What Howard is describing there is a discounted cash flow analysis which by the way, I teach in this book. But what we're trying to do is we're trying to imagine we own the whole company, then grow the business's current cash flow out into the future based on our estimated growth rate for the company. Then we imagine we sell the business in 10 years time. And then for all of those cash flow events that we model that we will receive in the future, we then discount them back to what we would be comfortable paying for each of them today based on our desired annual rate of return. Now, doing a very, very optimistic calculation for SpaceX, a very forgiving calculation, assuming a massive 20% growth rate each year for 10 years, imagining we could sell the business for 20 times free cash flow down the track, only asking for a 10% annual return, and also completely ignoring all their current CapEx, so using their operating cash flow as their free cash flow. I know that's completely stupid, but just doing that came to an intrinsic value of 440 billion, no margin of safety. And that was being extremely generous. But what is SpaceX currently trading at? A valuation six times that amount. Yes, six times. But as how it is about to talk about, these forecasts might not even be the investors' biggest problem in dealing with these IPOs.
In dealing with the future, the way most people deal with the future is by coming up with a forecast. I argue strenuously that if you want to deal with the future, you need two things. That one, you need a forecast, and you need a judgment regarding the probability that your forecast is right. So, you can make a forecast about the future of AI. You can make a forecast which is optimistic. But I I just think if you say, "This is my judgment about the future of AI, and by the way, I'm highly confident that I'm right." I think you're probably making a big mistake. Uh you know, I've never met anybody who who thinks they can tell me what this world is going to look like 5 or 10 years from now. And and So, why should any the young person you described who's starting who's laying the foundation for his investment portfolio, why should he conclude that he's probably right?
That's the other problem with these IPOs and indeed AI companies more broadly. It's predictability. Yes, you need to be able to model the future of the business, but you also need to do so with confidence. You need to have confidence in the growth rate that you slap on that business, right? This is what Buffett means when he says things like the internet isn't going to change how people chew gum. What he means here is that yes, you know, back in the tech bubble, the internet was the big new technology of the future, clearly it was going to be huge, but the winners and the profitability was highly uncertain. You know, in times like that, he prefers to own the chewing gum company that's had, you know, consistent 10% annual growth in its profits for the last 10 years. So, modeling out chewing gum sales or, you know, oil, mine, or lumber produced is going to be much more predictable than how much money will this AI company make in 10 years time. So, as Howard says, two things needed, you need, yes, a forecast, but you also need a judgment of how probable your forecast is of being correct. And his point is that in the world of AI, sure, it's a transformative technology, don't get me wrong, but to actually know with any level of certainty what OpenAI or Anthropic's operating cash flow will be in, say, 2036, it's impossible to even grasp. And that's also the problem more broadly with IPOs. Most of the time, they're pretty new businesses, and they only give you a couple of years worth of financial data. So, to be able to take that and model out 10 years with any level of certainty, I just don't think it can be done. It's much different to the chewing gum company that's posted a 10% annual rise in profits for the last 15 years. So, that leads us to a question, why are people buying these things? Well, Howard also gives us the flip side, too.
Investing in these companies today could be a huge error, but it could be great poetry. And the people who resist because it could be an error could miss out on the greatest thing in history. And that's what makes these decisions so hard.
And this is the tough thing to do because when you invert the arguments we've made so far, it sounds tempting. We are living through a revolutionary technological time. And while we can't predict the upside with certainty, one thing we do know is that AI will dramatically change the world that we live in. It already has. It's only been mainstream for what, 3 and 1/2 years? It will be economically game-changing. It could create unimaginable wealth as Elon Musk likes to talk about so much, but while that might be true at an investment level, these companies and their valuations, they might be overblown. And this is why so many people are comparing 2026 and what's going on right now to the tech bubble of the late 1990s. It's actually very similar.
Does it not seem like those previous cycles does it not feel to you like I don't know, dot com era?
It does feel like that. We have a technological innovation. I've seen several. I've read about many more over the last let's say 150 years. This may be the greatest. This may be the most powerful. It's also in many ways the least specifiable. Just for a starting point, let's say the railroads back in the 1860s and then radio in the 1920s, the automobile, computers in the 1950s and 60s, internet in 2000. It may be revision vision is history, but I think we had a much better view of what all of those could do. They didn't have this unimaginable, unlimitable upside that AI has or the, in my opinion, degree of uncertainty. We knew that the railroad would carry goods and people from coast to coast. We knew that radio would carry messages. But all the things I mentioned were accompanied by what we call bubbles. People got excited about developments which were unprecedented. They threw vast amounts of money about building the infrastructure for it. There was a winner-take-all race. There was excitement. There was exuberance. The capital flowed in like water. In every case, too much capital flowed in. I think it's fair to say too much infrastructure was built and prices were paid that were too high and a lot of the people who provided the capital for these bubbles lost their money.
And that's the tricky thing about where we sit today with AI. You know, the companies in the Meg 7 even say they don't really know what their end game is with all their AI investment. They're just building out the infrastructure to ensure that they don't get left behind. I mean, Goldman Sachs expects 5.3 trillion dollars of capex spending between Meta, Microsoft, Amazon, and Alphabet from fiscal year 2025 to fiscal year 2030. Google, Amazon, Microsoft, and Meta alone collectively plan on spending 725 billion in capital expenditures in just 2026. That would be up 77% on last year. It's incredible. This then sweeps up investors in optimism about the future as it has done in previous tech bubbles. And eventually, when true demand is understood, then the stock market wakes up and it corrects itself. And as Howard says, the people who put their money in these bubbles ultimately lose it.
I think it's fair to say that those comments are have been true in every case that I enumerated.
If this technological innovation with its exuberance doesn't produce a money-losing bubble, it'll be the first.
So, that then leads us to the question of what the hell do we do about it? Well, one strategy that Warren Buffett did back in the tech bubble is to just not participate, but let's put that aside for What if you felt compelled to? What if you're a rational investor, but at the same time you're also someone who doesn't want to potentially be left behind from what will probably be a pretty revolutionary technology? And we don't know, it may produce great returns in the market for years to come. So, what do we do? Well, in this next clip, Howard talks about three parts of the risk spectrum that you can play in depending on your appetite.
How to invest in AI. Like anything else, there's a spectrum, and at one end of the spectrum, we have ultra-high possible returns with great uncertainty. And at the other end of the spectrum, maybe we have somewhat lower possible returns with less uncertainty. Let me give you a couple of examples. You can invest in what we call the hyperscalers. They have established businesses with moats, enormous operating cash flow, and some diversity of business. These are, as I said before, without naming names, some of the greatest companies I've ever seen. So, you would think that investing in them would be maybe the low-risk way to invest in AI. Then you have established companies, as you said before, you know, we don't know their profitability, their finances, and maybe and they're one-product companies in the sense that they're all AI. So, maybe it's harder to specify their future, but Anthropic and OpenAI, for example, Nvidia, have a very high probability, I think, not being an expert, high probability of still being successful 5 or 10 years from now. So, there, depending on the price you pay and its fairness, they may be riskier than the hyperscalers, but they're not, uh, make it or break you know, they're they're already up and running. And then you have startups. You have startups where you don't know where they may not have revenues. They may have revenues but not profits. You may not even know what the product will be. But if you can get in at something called ground level and they turn and one of them turns into be a big winner, you can make an incalculable amount of money. And I described this in a recent memo as a lottery ticket. And so, at the at the riskiest end of the spectrum, you have lottery behavior. And if you think about the lottery, most people who buy lottery tickets lose all their money. A few people become incredibly rich. So, that's probably the profile of performance at the riskiest end of the spectrum. You can pick where to play on the spectrum. You can mix positions on the spectrum, and then you can decide how much should all of these companies on the spectrum be of your total portfolio.
I think that's really smart to think about. In many ways, it's not about not playing. I mean, even I have some risky businesses in my stock portfolio. It's just about recognizing where on that risk scale you are playing. And naturally, the riskier it gets, the lower percentage of your portfolio it should probably take up. You know, the Google's, the Microsoft's, the Amazon's, probably slightly lower risk due to the just the diversified nature of their business models. The startups and the AI pure plays on the other hand, much more risky because of what we're talking about earlier, that it's so hard to have confidence in your future models with any of these companies. But as Warren Buffett would say, it's also fine to look completely elsewhere. He sidesteps a lot of technology, and he is the world's best investor ever. And Howard isn't against this plan either.
Where do you find value right now?
I still think there is a more predictable part of the economy. It'll probably be a while before the energy business gets disrupted to the point where we use something in lieu of oil and gas. That's probably largely true of the food industry, probably the timber industry, and the home building industry. Transportation, it's probably going to be a while before we walk into a station, become dematerialized, and show up in another city. So, you can identify areas, you know, metals and mining, paper, chemicals. I guess I would say for the most part that things that have less intellectual content are less likely to be disrupted by AI, which is basically an intellectual problem-solver and productivity tool.
And this is more drifting into what Monish Pabrai would say, you know, boring businesses. They're great, you know, funeral homes or garbage collection or manufacturing, the list goes on. These industries, they're not sexy like tech and AI, but that works in the favor of the average investor because these aren't the companies that are on CNBC, they're not the ones that are being hyped across the headlines. And let's be real, outside of tech and the world's largest companies, there are still plenty of interesting opportunities in unloved sectors in our current economy. So, that's the last point. Don't assume it's just AI or bust. There's lots and lots of stuff going on out there, and you don't have to play in tech if you don't want to. Also, last thing, if you are interested in learning this approach and you'd like to join 300 other people in Atlanta in September for 3 days of in-person education with myself and Phil Town, I would love it. Please register your interest via the link in the description and the pin comment. There's no formal announcement yet, but we are taking expressions of interest. We're thinking September 18th to 20th. Spots will be limited to 300. That's the capacity of the venue, and it's also free. So, I'll put the link to sign up in the description, and I'll email that list before any announcement. So, if you really want to get in, make sure you put your name down on that list. But, apart from that, guys, thank you very much for watching. Also, yep, check out the new money strategy if you haven't done so already. I would really appreciate it. Like the video if you enjoyed. But, apart from that, guys, thanks very much for watching, and I'll see you all in the next video.