Transcript
Speaker: Welcome to HSBC Global Viewpoint, the podcast series that brings together business leaders and industry experts to explore the latest global insights, trends, and opportunities.
Speaker: Make sure you're subscribed to stay up to date with new episodes.
Speaker: Thanks for listening.
Speaker: And now onto today's show.
Speaker: This is a podcast from HSBC Global Research, available on Apple Podcasts and Spotify.
Speaker: However you're listening, analyst notifications, disclosures and disclaimers must be viewed on the link attached to your media player.
Speaker: Hello from Hong Kong, I'm Harold van der Linde, Head of Asia Equity Strategy here at HSBC.
Speaker: And I'm Fred Newman, Chief Asia Economist.
Speaker: You're listening to Under the Banyan Tree, where we put Asian markets and economics in context.
Speaker: Absolutely, and on today's show, we're going to look at the rally in AI stocks in Asia.
Speaker: And we're going to try to put a little bit of historical context around it.
Speaker: Stocks have been on a tear, but every run has succumbed to an end sometime.
Speaker: We're not here to tell you when that will exactly happen, but history has a thing or two to tell us about what to watch out for.
Speaker: Exactly.
Speaker: Demand, competition, regulation, they all play a part, and they're all part of the discussion right here under the Banyan tree.
Speaker: Let's start with some numbers to set the scene.
Speaker: HSBC's Asia AI Index, which tracks the performance of the key AI stocks in the Asian region, is up 200% since November 2022.
Speaker: Since the start of 2023, Taiwanese stocks have rallied 70%, led by TSMC, the chip maker that manufactures components for AI devices all over the world.
Speaker: TSMC itself is up about 110% in that time.
Speaker: Now, Harold, we talk a lot about AI.
Speaker: TSMC is, of course, a big company that's very well known.
Speaker: It's one of the producers of these advanced chips that go into actually doing AI in these massive server centers.
Speaker: But what other companies are involved?
Speaker: What are we talking here?
Speaker: Is it the guys who actually write the software?
Speaker: Is it the hardware?
Speaker: What are some of these companies and sectors that are really affected by the AI boom?
Speaker: There are companies who do the software and stuff, but it's really hardware that we're talking about.
Speaker: So it's the nuts and bolts of AI, if you want to put it like that.
Speaker: Now, you already mentioned the chips.
Speaker: So they do the fast computing, if you want to keep it very simple.
Speaker: But AI uses enormous amounts of data that's stored on servers.
Speaker: So people have to run these servers.
Speaker: So there are companies that build these servers or components for these servers, or they build components for the chips that...
Speaker: that are being used as well.
Speaker: You need, for example, particularly laminates and these sort of things.
Speaker: And you can go even further and say these servers, AI, use an enormous amount of energy.
Speaker: So we've seen actually stocks in, for example, Malaysia, energy stocks rallying because the demand of energy for these servers is so enormous that...
Speaker: that business is going to double.
Speaker: And that reminds me actually, I think on a previous podcast with your great statistics, we actually did mention that about a dozen servers, data servers or centers being built in Malaysia right now.
Speaker: Yeah, it was 40%, I think, of existing power capacity would be needed in order to service those data centers.
Speaker: To run those servers, yeah, which means that the Malaysian grid would have to expand by 40% capacity.
Speaker: in order to just run these data servers.
Speaker: So when we talk then about the AI boom, is it really just those advanced semiconductor companies that are benefiting from this?
Speaker: Everybody heard about NVIDIA and TSMC.
Speaker: Or do you also see this already being reflected in rallying stock prices of these other more esoteric companies that really provide the electricity or the data centers or the lamination or the chemicals or the software?
Speaker: What exactly is that?
Speaker: All of these stocks have gone up.
Speaker: All of these stocks have gone up.
Speaker: Nvidia has gone up, I think, seven times already since AI and JetGPT was announced.
Speaker: TSMC, you just mentioned, has gone up a lot.
Speaker: Some of the other stocks have gone up even more because they're even more exposed to it.
Speaker: And of course, you have some of these power stocks.
Speaker: They haven't gone up as much, but they've rallied as well.
Speaker: Now, when we say stocks rally, right, so they rise in price, what does this really signify?
Speaker: You're an equity strategist.
Speaker: Does that signify that investors expect, therefore, enormous profits in the future and that's reflected in the price today?
Speaker: And is that the way to think about this, that maybe some of the profits have not necessarily come through, but investors are rebuying the expectation of a big, big shift in profits?
Speaker: Yeah, Fred, you're becoming a strategist here.
Speaker: I learned from the best.
Speaker: There you go.
Speaker: No, but that's absolutely right.
Speaker: And I'm going to pose a question to you on how realistic that is in a second, because we have these expectations that these earnings are going to grow significantly.
Speaker: And we know that there is demand for, for example, chips, right?
Speaker: So we know that the order books are full.
Speaker: We know that these data servers are being built.
Speaker: We know that the power capacity is going up.
Speaker: The question is, how long will that then continue?
Speaker: How big will that eventually be?
Speaker: And...
Speaker: That's something then the markets don't know over time.
Speaker: So these stocks can rally up, but they can at some point in time also come down and have a bit of a reality check.
Speaker: We can talk about this maybe a little bit later, but let's take a step back.
Speaker: You as an economist, how then do you think about how new technology in general, maybe AI impacts the economies?
Speaker: So I think there are two ways of looking at this, and I'm going to come back to you with a follow-up question actually on this.
Speaker: One is that when you have such a boom, it's about the investment that's being spent.
Speaker: It's about building the data centers.
Speaker: It's about building the infrastructure, and that often drives economic growth.
Speaker: And you see that with...
Speaker: exports from Taiwan and for South Korea doing extremely well because people buy these semiconductors.
Speaker: That's what we can see at the moment, right?
Speaker: And that's what you can see now.
Speaker: And that's, I guess, to some extent reflected then in the profits of the companies that build the hardware already.
Speaker: They're already seeing the order books increase.
Speaker: And of course, that's economic activity.
Speaker: And that is, as long as that boom lasts, that's positive for growth.
Speaker: The question from the economic perspective is, does that investment over time yield a return?
Speaker: That is, can these data centers actually drive further activity?
Speaker: So is it not just building the data center itself, but is there something, once we use AI, is it actually profitable?
Speaker: Are we going to be more productive, for example?
Speaker: Are we going to be more productive?
Speaker: Are companies going to make more money in other fields?
Speaker: Are we going to revolutionize how we do retail, how we do finance, how we do research, right?
Speaker: All of that.
Speaker: And to some extent, it feels from an economic perspective, the proof is yet to come whether AI is really that transformative for all other industries and therefore whether the investment is ultimately justified.
Speaker: And from an economic perspective, what it's all about is the rise in productivity.
Speaker: Can AI drive productivity?
Speaker: And the answer as of now is we don't know.
Speaker: We have to see the use cases.
Speaker: We have to see the improvement in corporate profitability in sectors that are not building the AI infrastructure but are using the AI.
Speaker: And that only time will tell.
Speaker: That we'll see in the future.
Speaker: But for you then, I want to draw a parallel.
Speaker: We touched on this already earlier.
Speaker: in previous episodes, which is during the gold boom in California in the 19th century, it wasn't necessarily the gold miners who made a lot of money because in the end there wasn't that much gold in the ground.
Speaker: It was the people who sold the gold miners to shovels who made the big bucks, the infrastructure for it.
Speaker: So what I wanted to ask you is whether there's a similar analogy here for the equity market where
Speaker: the companies who've built the infrastructure are the ones who are now primarily the benefactors and the ones who will actually use AI.
Speaker: Well, we haven't really kind of seen the use case yet and therefore there is not yet a clear case to be made.
Speaker: I'll give you a nice example.
Speaker: I have on my phone an app, which is an AI app.
Speaker: And sometimes I put in questions and see what comes out of it.
Speaker: Sometimes, to be honest, it's a bit gibberish.
Speaker: I think sometimes it's quite interesting.
Speaker: So I use it as a sort of fast way of getting access to data.
Speaker: I've also subscribed at home to an OpenAI sort of website and I can put a document in there and I'm asking to give me a summary or something like that just for me to get to know this.
Speaker: So I spent nothing on the app and 30 US dollars a year on that website.
Speaker: On the other hand, we have billions of dollars going into these investments.
Speaker: So there seems to me a kind of somewhere along the chain, I wonder who's making the money at this point in time, because I'm not really putting a lot of money in.
Speaker: So somebody's providing some of these services even for free on my iPhone, right?
Speaker: You're right.
Speaker: Is it the AI company that's providing it, that's making the money, or is it the picks and shovel kind of guy?
Speaker: And that's the chip maker and the server builder at the moment.
Speaker: Now, these stocks are rallying, so clearly that's the case.
Speaker: But in how far that will then lead to the profits that will be made, for example, by the companies that provide AI,
Speaker: That's open.
Speaker: That comes back to your productivity thing is how much of the work that we do will be basically done by AI in the next 10 years.
Speaker: And if that's 50%, that's a lot.
Speaker: So there's a lot of business there.
Speaker: But if that's only 5% of what we do, then that's maybe a bit of a disappointment.
Speaker: And then people will struggle.
Speaker: So the potential is certainly there, but we get to see the actual implementation, the returns on the investment to come through on that.
Speaker: But we should probably also not forget that the AI craze, quote unquote, has only started about 18 months ago.
Speaker: So we're very early in the process.
Speaker: Yeah.
Speaker: And there is a bit of an analogy with other cycles that we've looked at since about 2000 with other sort of, we call them big market cycles.
Speaker: We've had, for example, e-commerce rise in 2000.
Speaker: We had the Chinese consumer stocks rising and banks in 2006 and 2007.
Speaker: I mean, they also went up a lot on the expectation of that would be a lot of future profits, right?
Speaker: But they fizzled at one point in time.
Speaker: And that was because sometimes there's a couple of common factors that seem to be a place.
Speaker: If it's a really big market, what we very often misunderstand that other companies are coming in and compete.
Speaker: Everybody says, I want to do AI.
Speaker: So it might become more competitive in a crowded space.
Speaker: Or governments come in and say, I'm going to regulate this.
Speaker: This is an important market now.
Speaker: It needs to be properly regulated.
Speaker: That happened amongst others with China Internet, for example.
Speaker: There are these sort of curve balls that can come to these rallies that suddenly change the dynamics, sometimes even overnight.
Speaker: So we need to be a little bit careful.
Speaker: So is it fair to say there are two uncertainties or three uncertainties?
Speaker: One is how profitable the technology will be, how good it actually is.
Speaker: That's of course still the question.
Speaker: Then the second one is presumably what the timeframe is, how long of these benefits it takes to come through, even though they're to come through, is it three or four years?
Speaker: Is it 20 years or 30 years?
Speaker: And then I guess the third one is that is the companies that investors think about today also the ones who will gain actually in the future?
Speaker: Because if you think about the internet rally, for example, in 2000s, some of the companies that rallied back then don't exist anymore.
Speaker: No.
Speaker: And companies like Facebook, for example, only emerged five or six years after the e-commerce platforms was established in the late 90s.
Speaker: So...
Speaker: Yeah, the ones who dominate the industry now might not be the ones who dominated in the future.
Speaker: So we sometimes see companies making the wrong decision in certain places.
Speaker: We've seen this in China.
Speaker: For example, China consumers stocks had a big rally in 2007 and 2008, but some of them really didn't have the right, they grew too fast and stumbled literally over themselves.
Speaker: So, yeah, we've got to be careful just to make very easy sort of comparisons in this rally.
Speaker: So clearly an impressive rally so far, but a lot of uncertainty associated with that.
Speaker: Maybe it's time to take a quick break.
Speaker: And when I come back, I want to ask Harold how this current rally compares to one of the last big run-ups, which was the China Internet rally that we saw only a few years ago.
Speaker: So, Harold, before the break, we discussed, you know, the rally in AI stocks.
Speaker: We highlighted some of these examples in Asia that rallied quite a bit.
Speaker: We also had a big rally in Chinese internet stocks a few years ago, which is kind of reminiscent of what we see now in markets.
Speaker: How does that compare?
Speaker: You did some analysis on that.
Speaker: Yeah, so the China Internet kind of rally that we saw 2016, 17, 18, 19 really, really rallied up is actually a very nice comparison.
Speaker: But it was more of a, you could say, a nearly traditional sort of rally that builds because it takes time for these market rallies normally to build as more and more people get convinced that, hey, there's a big market or a big growth opportunity.
Speaker: opportunity here.
Speaker: So these rallies build very slowly and then they gain momentum and as they get a momentum these stocks go up more and more and more.
Speaker: What that happened with China internet is that regulations came in place but also these companies started to compete with each other.
Speaker: So all these factors I mentioned earlier started to come in play and that then eventually turned the markets and if you go back and say what happened on that particular day that that peaked nothing particularly happened.
Speaker: It's the same with the 2001
Speaker: internet bubble.
Speaker: Nothing happened on that particular peak.
Speaker: It was just a kind of confluence of factors that started to put gravity to these stocks.
Speaker: Now what is different this time around is that we don't have a slow build.
Speaker: We had a moment whereby ChatGPT was announced.
Speaker: That was in November 2022.
Speaker: A couple of weeks prior to that you saw the stocks already moving because there was talk about something was happening.
Speaker: But there was a clear moment and the stock bang went off.
Speaker: But the rally we've seen now, although it's been much faster, is about the same magnitude of the rally that we've seen in China Internet.
Speaker: So we're getting now, if history is any guidance, to a limit whereby we say, hey, we've seen the same sort of ascend up the mountain.
Speaker: Yeah, of course, though one might also say that this is potentially much more transformative technology than maybe the Chinese Internet space, which was really just, you know, a consumer play, different companies coming through.
Speaker: So, you know, one doesn't necessarily imply that the other is not just the thing.
Speaker: We know that now.
Speaker: But at that time, the feeling was, wow, this is really big, right?
Speaker: So we know over time.
Speaker: And you and I might have a podcast in five years' time.
Speaker: And we say, well, you remember we talked about AI and everybody thought it was transformative, but actually it hasn't.
Speaker: Maybe that is going to be the outcome as well.
Speaker: That will be obvious at that time, but it's not obvious to us at the moment.
Speaker: Now, it's always, in hindsight, always easy to identify bubbles and then they burst because you can see it run up and it collapses.
Speaker: But we should also be mindful that not every rally ends up in a collapse.
Speaker: Sometimes it is justified, right?
Speaker: So we see some companies who maybe go down over the course of their lifetime, but really ultimately over 20, 30 year period, those rallies are justified, the valuations that investors put into them.
Speaker: What are some of these examples that you can think of with companies that have actually endured such tremendous rallies and were ultimately turned out to be justified?
Speaker: Yeah, in Asia, I think there are two sets of companies that, broadly speaking, have been able to do so.
Speaker: It is companies that have become global leaders and therefore have an enormous scale to benefit from.
Speaker: And actually, TSMC is a nice example.
Speaker: In the internet bubble in 2001, it peaked.
Speaker: TSMC rallied and then it came down a lot.
Speaker: And then over the years thereafter, it continued to recover.
Speaker: I think it was three years later, it was back at its peak and then now we are way beyond that multiple times.
Speaker: So that stock is actually a very nice example and it's a company, an Asian company that has gone completely global.
Speaker: Or you have companies that become domestic dominant players.
Speaker: So there's also scale, but in large economies.
Speaker: So for example,
Speaker: Indian shampoo makers or Indian toothpaste makers, they've had a really nice rally over the last 20 years because they were able to dominate a particular market.
Speaker: So it's that dominance in substantial markets and you come from a small base that really allows that to happen.
Speaker: And that reminds me when we had in one of the recent episodes where we talked about amazing statistics that you collected.
Speaker: They're actually companies that outperformed even some of the big AI players today.
Speaker: Exactly.
Speaker: And those were stocks that you wouldn't think that was a company that did motorcycle financing in India, for example, and has done phenomenally well over the last 20 years.
Speaker: It's a phenomenal track record that outpaces NVIDIA.
Speaker: Now, of course, the trick is always to know it before it rallies, not after it rallied.
Speaker: That's the job of the strategist.
Speaker: That's the job of the strategist, exactly.
Speaker: But Fred, I know you're reading a book about spices and spices, Asia.
Speaker: There was a bubble there somewhere in the distant past as well, right?
Speaker: There was, yes.
Speaker: There was the Spice Islands, of course.
Speaker: You know, this is the Malukkus in Indonesia, Banda, Indonesia.
Speaker: for example, where we saw nutmeg, for example, or clove, being spices that were really in the 16th century were desired by Europeans.
Speaker: And there was a real bubble in the sense that Europeans invested in kind of bringing these spices back to Europe.
Speaker: And these spices fetched enormously high prices.
Speaker: because that added flavor to the European cuisine and so forth.
Speaker: But during the spice trade, one of the products that was brought back from the East was actually tulip bulbs.
Speaker: And you being Dutch, I'm sure you know one or two things about tulips.
Speaker: That is absolutely right.
Speaker: Actually, look into this one.
Speaker: At one point in time, there was so much of a craze in tulip bowls that the price of a tulip bowl equaled the price of a small house in the city of Amsterdam.
Speaker: But imagine that.
Speaker: There must have been somebody who sold his house and got a tulip bowl for it.
Speaker: Now, the tulip bowl probably, I don't know, five years later would have been kind of...
Speaker: defunct and not dead.
Speaker: But even if that tulip ball would still grow a tulip today, what's the price of a tulip?
Speaker: It's maybe two euros.
Speaker: The price of a small house in Amsterdam, you're talking about half a million euros, if not more.
Speaker: So clearly, that was not always the right thing to go to, to follow the rally and the
Speaker: And of course at that time, that was the real market for tulips in the Netherlands in the 17th century.
Speaker: They bred different tulip varieties, different varieties, fetch higher prices.
Speaker: People were crazy about all the differences and people had an enormous amount of knowledge about the differences in tulip balls and these sort of things.
Speaker: But again, it is wealth that was created because of that trade with Asia that allowed these kind of accesses to emerge in consumption, right?
Speaker: We've seen this a bit in China 15 years ago in consumer stocks as well.
Speaker: So things haven't changed that much.
Speaker: One of the differences we should keep in mind is that the tulip bubble, as it's known famously in the 17th century Netherlands, that was purely on a product that was a luxury good.
Speaker: That is, people bought these tulips really to decorate their house.
Speaker: It was the first time when people actually brought flowers into the house for decoration.
Speaker: Fetch, you know, led to this speculative bubble.
Speaker: Social standing.
Speaker: You don't have a children in the house.
Speaker: That's right.
Speaker: Today it's maybe a handbag or it might be something else.
Speaker: Yeah, exactly.
Speaker: The difference, just bringing it back to AI, is that here, of course, it's not necessarily a luxury good.
Speaker: It's the belief that this is a transformative technology that will deliver economic returns over time and improve productivity.
Speaker: And that's why it's so powerful, I think.
Speaker: That is right.
Speaker: And that seems to be more likely with AI than it was with Chullibub.
Speaker: So maybe this is not going to be a Chullibub rally, but this is going to be a different sort of rally.
Speaker: Well, Harold, this is definitely a story we could keep talking about for hours, but I think it's time to wrap up the podcast for today.
Speaker: Indeed.
Speaker: But if you do want to read more and you're an HSBC client, head to the research website and check out our report.
Speaker: It's called Peak AI in Asia with a question mark at the end.
Speaker: I hasten to say...
Speaker: Remember to follow Under the Banyan Tree if you haven't already.
Speaker: We're on Apple, Spotify, or wherever you get your podcasts.
Speaker: And while you're there, listen, like, and subscribe to our other HSBC research podcast, The Macro Brief.
Speaker: Thanks for joining us and talk to you again soon.
Speaker: Thank you for joining us at HSBC Global Viewpoint.
Speaker: We hope you enjoyed the discussion.
Speaker: Make sure you're subscribed to stay up to date with new episodes.
