The Stocks That Could Survive an AI Bust

Merryn Talks Money

In this episode, Merryn Somerset Webb interviews Stephen Yiu of the Blue Whale Growth Fund to discuss the sustainability of the AI-driven mark

Key takeaways

• Nvidia is currently more exposed to an 'earnings bubble' than a valuation bubble due to its high profitability relative to P/E ratios.
• The 'picks and shovels' strategy involves investing in companies providing the essential infrastructure for AI development.

Main topics

• AI valuation vs earnings bubbles
• Investment diversification strategies

Notable quotes

The biggest risk to NVIDIA... is actually quantum computing.

Conclusion

Investors should look beyond pure-play AI stocks toward fundamental se

Transcript preview

Speaker 7 (0:00) We don't just invest in cutting-edge companies. We look at companies with a history of steady growth. And companies whose growth cycle has come round again. Because in the real world, you have to look at growth in three dimensions. Monk's Investment Trust. Speaker 5 (0:14) Hey, it's Alec Baldwin. This season on my podcast, Here's The Thing, I talk to actor Stephen Root. Speaker 6 (0:21) I'm a people watcher. You see that guy going down the drain and go... Yeah, I can use that. Speaker 5 (0:25) And Open Igloo co-founder Alia Mohamed. Speaker 8 (0:28) I love staying on top of what is going on in our city, what is on renters' minds, and taking all of that knowledge to build a platform that hopefully is going to make New York City better in the long run. Speaker 5 (0:39) Listen to Here's the Thing on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts. Speaker 4 (0:47) You're not incurable. You're not broken. You're having a normal human experience, and we have a solution for that. Speaker 6 (0:52) And that solution is our podcast, Free Therapy. I'm Ben. And I'm Kevin. And every week, we talk to leading therapists in the country about relationships, work, family dynamics, and all of life's headaches. If Speaker 8 (1:02) you're trying to get rid of anxiety, you're actually fueling more anxiety. Speaker 6 (1:07) So listen to Free Therapy because therapy is expensive. But our podcast is free. Listen to Free Therapy with Ben and Kevin on the iHeartRadio app, Apple Podcasts, or wherever you listen. Speaker 3 (1:24) But Speaker 1 (1:24) if you do believe that AI is very powerful, that it's going to bring some positive to the human population across the board, I mean, obviously, there's going to be some negative as well. Then what you really want to do is to invest into companies that are receiving the spending or investment, which we call them the picks and shuffle companies. But then when you are... thinking about the company that investing into AI to make an AI model, to come up with different use cases and all that, that's going to be a trial and error process in a way that a lot of them are not going to survive, but some of them are going to make it to be very successful. So it doesn't mean that you can't pick them, but then I think this is more of a venture capitalist market, isn't it? You invest into probably a hundred companies and then some of them worked out, most of them doesn't, but then you still make a lot of money. Speaker 2 (2:23) Welcome to Maren Talks Money, the podcast in which people who know the markets explain the markets. I'm Meryl Somerset Webb, and this week I'm speaking with Stephen Yu, Chief Investment Officer of the Blue Whale Growth Fund, which manages more than $4 billion. Stephen is a major investor in technology and has built something of a reputation for identifying the likely winners, and indeed the losers, of the AI boom. His record makes him a voice worth listening to, and we have had him on the podcast a few times before. From his launch in September 2017 to the end of August, the fund returned over 320%. That's again, just under 130 % for its benchmark. So far this year, it has outperformed the index by more than 40%. But with so much of the global market now riding on enthusiasm for AI, can investors really diversify their portfolios enough to survive a disappointment? And if so, which companies might still prosper if the boom begins to fade? Stephen, welcome to Merriam-Dawg's Money. Thank Speaker 1 (3:22) you. Speaker 2 (3:22) Now, you are not really an AI fund. I know that you've had so much of your investments in AI-related stuff over the last couple of years that people look at Blue Whale and they go, well, that's an AI fund, or that's a tech fund, or that's just a growth fund, whatever it is. And of course, you are none of those things. You're a fundamental stock picker. However, before we move on to talking about stocks and how you pick them, do you mind if we just have a chat about the AI environment at the moment? Speaker 1 (3:51) Yes, of course. Can Speaker 2 (3:52) we start with the dangers and why we're talking so much about the dangers at the moment? Speaker 1 (3:59) I think that's a good starting point, just because if you think about the pace of how AI is, the AI cycle has been gone through over the last few years, has been very much on an exponential trajectory, which means like everyone is now feeling a part. feeling part of AI, whether you have ended up investing in the AI or not, that's a different story. And of course, if a technological advancement is taking shape very, very quickly without all the other gut rails around it, then we could end up in some troubles in time. And hence, I think we now realize as a community that we need to start probably putting some policies in place and thinking about the implications of what this might bring is not. all positive, there's certain negative. But unfortunately, I think maybe the policies that are still very much behind the curve and obviously thinking about the committees, how many meetings they need to do until they make some changes in terms of what AI is about or how we're going to deal with it, it's going to take a bit longer. There's been a narrative to suggest that we're going to see a lost generation. And what it means is by the time that we would have something public. that we can actually use and control AI, it might take longer than expected. Speaker 2 (5:18) Yeah, but controlling AI is a global business, right? You can't just have one country controlling AI. You can't put regulation in place in the UK and the EU or just put regulation in place in the US. So when you hear these calls for regulation, it's very hard to understand how that could possibly work. It's probably not going to work. This is Speaker 1 (5:39) a Speaker 2 (5:39) competition. It's Speaker 1 (5:40) not going to Speaker 2 (5:40) work. Speaker 1 (5:42) Okay, Speaker 2 (5:42) right. Okay, fair enough. So we're not going to have a regulatory effort that slows AI down, but might it be the case that we've reached a point where it naturally slows because after a very fast shift, we're hitting a couple of bottlenecks, a couple of places where things have to slow down. So we hit that in terms of power, in terms of resources, in terms of the capacity of what LLMs can really do when we haven't made the next leap. Might it be that we're about to see a kind of natural slowdown in the move of AI anyway while what's happened so far consolidates inside the economy? I think Speaker 1 (6:20) that's definitely probably a given in time. Just think about in any technological advancement, think about the penetration of internet, the penetration of smartphones. how we have evolved through the digital transformation, through COVID, shopping online, working from home and all that. It took a long time to get there. And obviously, there's always this thing about the rate of change, like it's how quickly that would have happened, how long does it take. And I think when you look into AI today, that there's a lot of capital that has gone into the space, which is the Quebec cycle. There's a lot of money that has gone into developing the model to make the model more advanced. we have not seen yet, which takes a bit longer than just investing into a model to make it better, is the penetration is taking time to get it penetrated within the consumer space, within the enterprises space, within the government space. So then what it means is in time when the technological advancement slows down, then the penetration can start to take shape. Obviously, you want that cost of computes to be quite low, so it's affordable for most people. And hence, it's still a question about for the technological advancement of models to slow down doesn't mean AI is unreal because you still have a lot of use cases that we are not yet adopting. Just think about all the productivity that we are yet to realize across functions. Speaker 2 (7:53) And then we also, I mean, at this point in a technological revolution like this, it's unclear what form the penetration will take, right? So in the early parts of say the data center build out, there was a general assumption that it would be LLMs across the board. But now we see people talking about using different types of models, using more of the Chinese type of model, using SLMs, et cetera, small language models, and so they can keep things more contained. So already the way that it's being adopted and penetrating through ordinary companies. It's slightly different from what might have been expected even a few years ago. So it's very hard to predict how this unfolds Speaker 1 (8:30) in Speaker 2 (8:30) real Speaker 1 (8:30) life. Speaker 2 (8:30) AI is very much Speaker 1 (8:31) of an unknown or uncharted territory in a way that we don't know how this is going to end up. But if you do believe that AI is very powerful, that it's going to bring some positive to the human population across the board, I mean, obviously, there's going to be some negative as well. then what you really want to do is to invest into companies that are receiving the spending or investment, which we call them the picks and shuffle companies. But then when you are thinking about the company that investing into AI to make an AI model, to come up with different use cases and all that, that's going to be a trial and error process in a way that a lot of them are not going to survive, but some of them are going to make it to be very successful. So it doesn't mean that you can't pick them. But then I think this is more of a venture capitalist market, isn't it? You invest into probably 100 companies and then some of them worked out, most of them doesn't, but then you still make a lot of money. But I think from our perspective that we invest into a publicly listed company for a high conviction, only about 30 companies, then you probably need a bit more certainty in terms of what the direction of travel is going to be. Speaker 2 (9:38) And one of the big conversations at the moment when looking at the big AI companies, the listed ones, the hyperscalers, et cetera, is you look at the valuations and you're like, well, those aren't bubble valuations, but is there maybe a bubble in earnings expectations? So, you know, the PE looks relatively low, but well, not low, but not bubble levels. But is that because the E number is grotesquely over-optimistic? As you say, we can't really predict the future with any of this stuff. So is this more of an earnings bubble? Yeah, I think you also Speaker 1 (10:07) need to differentiate amounts of a few probably major companies, because if you're referring to the likes of SpaceX, OpenAI or Anthropic, I would definitely say that they are in a valuation bubble, right? Two of them yet to be listed in the stock market. They're not profitable at this moment in time. Their revenue is growing very quickly. The projection for them to be profitable would be in a few years' time. And obviously you're paying up for that. But then when you look at the other side of the fence, let's say talk about Nvidia, because obviously it's the biggest company in the world. Nvidia is currently trading at probably 16, 17 times earnings. I mean, it's the biggest company, $5 trillion. Then you can say that, okay, Nvidia is not exposed to valuation bubble, but it's exposed to earnings bubble. Okay. But the challenge with earnings bubble is a lot unclear. So it's less clear about... versus valuation bubble. Because when you think about valuation bubble, it's all about future expectations. It's about, okay, what is my payback period? But when you think about earnings bubble, earnings bubble means that the companies are actually earning a lot of money as we speak. Most companies want to buy the product from NVIDIA or alongside Broadcom because there's only two companies that could probably give them the product they want to build AI infrastructure. And ultimately, you have to ask the question, is more to do with whether AI is going to be more game changing or not. Whether these companies who have invested money are going to get a return on investment. If they do get a return on investment, then the earnings bubble is actually not a bubble. It's actually right because this is the product they're selling that people actually want. And then I think also when you think about the valuation