Global Perspectives: AI – Invest with patience
Global technology Portfolio Managers Alison Porter and Richard Clode join Matthew Bullock, EMEA Head of Portfolio Construction and Strategy, to discuss the implications of demographic and geopolitical factors for tech investing, and why patience is needed on AI opportunities.
22 minute listen
Key takeaways:
- Each generational cohort is unique, with its own preferences in terms of sustainability, tech consumption, pace of adoption, and perception of the risks and opportunities of AI.
- Patience is required to invest in emerging technologies like AI as they develop in phases, starting with infrastructure buildout, followed by platforms, applications, and services.
- An active, global, diversified approach to tech investing is vital to navigate this exciting, evolving high-growth sector.
Alternatively, watch a video recording of the podcast:
* Paul Redmond is Director of Student Experience and Enhancement at the University of Liverpool and one of the UK’s leading experts on generational change and the future of work.
There is no guarantee that past trends will continue, or forecasts will be realised.
Bond proxy: a security/stock that is perceived to pay safe and predictable income with low volatility – characteristics that are more commonly associated with bonds. Typical examples include utilities, consumer staples and pharmaceutical stocks.
Bottom-up investing: an investment approach that focuses on the analysis of individual securities, rather than broader macroeconomic or market factors in order to identify the best opportunities in an industry or country/region.
Capex/capital expenditure: company spending to acquire or upgrade physical assets such as buildings, machinery, equipment, technology etc. to maintain or improve operations and foster future growth.
Diversification: a way of spreading risk by mixing different types of assets/asset classes in a portfolio, on the assumption that these assets will behave differently in any given scenario. Assets with low correlation should provide the most diversification.
Gen AI: generative AI refers to deep-learning models that train on large volumes of raw data to generate ‘new content’ including text, images, audio and video.
GPU cluster: a GPU-enabled cluster refers to a network of interconnected computers (nodes) that, in addition to traditional CPU (Central Processing Unit) capabilities, include GPUs (Graphics Processing Units) to enhance their computational power. GPU clusters facilitate the training of complex AI models by processing large datasets at unprecedented speeds.
Hyperscalers: companies that provide infrastructure for cloud, networking, and internet services at scale. Examples include Google Cloud, Microsoft Azure, Facebook Infrastructure, Alibaba Cloud, and Amazon Web Services.
Large caps: larger companies can be classified as publicly traded, well-established companies with a relatively larger market capitalisation (value of a company’s shares); typically more than $10 billion.
Leverage: (in the context of this article) is the use of debt to increase exposure to an asset/market.
Small caps: publicly traded, less established companies with a relatively smaller market capitalisation. They tend to offer the potential for faster growth than their larger peers, but with more volatility.
Volatility: the rate and extent at which the price of a portfolio, security or index, moves up and down. If the price swings up and down with large movements, it has high volatility. If the price moves more slowly and to a lesser extent, it has lower volatility. The higher the volatility the higher the risk of the investment.
Yen carry trade: borrowing the yen at very low interest rates to buy currencies with better yields. The trade involves buying the higher-yielding currency with the borrowed yen in order to buy bonds or other money market instruments in that currency.
Matthew Bullock: Hello, and welcome to the latest recording in the Global Perspectives podcast series. My name is Matthew Bullock. I’m the EMEA Head of Portfolio Construction and Strategy here at Janus Henderson Investors. And today, we’re fortunate to be joined by two guests. Normally, we have one, but we’ve got two exceptional guests today, we’re joined by Richard Clode and Alison Porter, both portfolio managers within the technology team here at Janus Henderson. So great to have you both here.
Richard Clode: Thank you, Matt.
Bullock: Alison, I’m going to start with you to begin with. So, this is a technology podcast and we’re going to go into some of your views on the technology space. But I want to go into demographics to begin with. And the reason that I ask that is that I’m aware that Janus Henderson and yourself was very much involved in a study that was done with the University of Liverpool, which focused in on changing demographics and the impacts on society. And I’m going to sort of probably be a little bit simplistic in my summary here because I went through the paper looking for what was some of the key points. And, you know, the sort of areas it covered was defining the different working age generations, looking at the strengths and weaknesses within each generation and their tolerance to things like AI, and then how comfortable people were with the adoption of AI and the implications of that. So I mean, I’ve just been very simplistic there, but from your perspective, what would be the key takeaways that you got from that study?
Alison Porter: So, a few things. I mean, I think it really defined how every generation is its own people. And *Paul Redmond and the University of Liverpool have done a lot of work on that. But sometimes we oversimplify on trends for young people or older people, actually it really brought out some, you know, significant differences between, you know, Millennials and Gen Z and very, very significant differences between much older generations. And that’s something that we’ve talked about for a long time in terms of digital adoption, how younger people, you know, now tend to know much more about artificial intelligence. And they’re also less fearful about adoption of artificial intelligence. Whereas actually older generations that tend to know less are also more fearful about what that might mean for them in the future. I think there was also some marked differences in terms of attitudes towards sustainability, and but particularly, actually, Millennials are the most focused generation on sustainability, where it really impacts how they think about their purchase choices from cars to housing, possibly investing longer term it’s a key focus for them.
Bullock: Right. OK. So, one extreme, you’ve got the Millennials with the sustainability side, and then you’ve got Gen Z, which is more sort of comfortable with the AI side of things. Richard, I’m going to guess you’re somewhere in the middle of all of that. You can answer that if you wish or you can just move on. But from the investment perspective, then I sort of want to come back into that AI piece in particular, and the sort of opportunity/fear associated with AI. From an investment perspective what does that actually mean? So when you think about demographics, does that change the way you invest? Does the attitude to AI change the way you invest?
Clode: Yeah, I think, you know, I’ll come back to that Warren Buffett quote about sort of patience and how important it is to be patient in the stock market. And ultimately, the rewards of the stock market come to those who are patient, not impatient. And we always think about that when we think about investing in new technologies and, and just thinking about how the phases of that investment in a new technology take place that, you know, you always start with, with infrastructure, then you end up with the platforms, then you end up with the applications, the software and, and the services.
And I think, you know, we saw that through the internet age over the last 20 to 30 years. We’re now seeing that in very early innings in that AI phasing where it’s infrastructure. But even to, you know, the last couple of years, it’s all been about training and, and that’s all been about, you know, NVIDIA and we’re seeing that sort of broadening out. But then to think about what we’re going to see in years to come in terms of actually seeing those, you know, bottom-up built gen AI products and services and who’s going to be taking those on, and using those, and the companies that are going to sort of benefit from that. You know, that Allison was thinking about it from that survey work that’s still years away. And I think, you know, what we’ve seen, particularly in the last few months has been a little bit of that impatience of investors. You’re thinking about that return on the huge amount of capital spending that we’ve seen, and wanting to see that now. Now, you know, that’s just not the way things work, you’re always going to have the capex put in and then you get the revenues and profits down the line. But you do want to see some of those kind of signposts along the way to give you comfort that you are going to get that return. Otherwise, you know, you start thinking, contemplating that, you know, ultimately that spending might be turned off at some point. So, everyone’s read that, that Sequoia [Capital] note or that Goldman Sachs note about that return on that capex. And, and we’d, you know, as a team, I think we’ve seen this movie before. We’re being sort of quite patient about always thinking that sort of software benefit would still be some years away. And I think you’ve kind of seen that more recently, in some of the [company] updates that we’ve seen. But some good proof points. And I think that’s what we’re looking for, looking for the proof points from the companies, but more importantly from CIO, CEOs of management teams that we’re talking to, about as to, you know, what are the pain points of adoption, the use of friction? Where does your data infrastructure need to be to be able to adopt this? You know, how are, how important are sort of AI advocates in the organisation or consultancy firms to come in to, to tell you about best practice to push you forward?
And we see some companies like Klarna that have a founder who, you know, was one of the first adopters with OpenAI that sort of incredibly kind of powerful in pushing forward gen AI usage across a firm. And it’s a younger firm and it’s much easier to do. That’s very different to a big bank or, or a sort of much more established company with tens of thousands of employees. So, patience I think is is where it comes back to when we think about investment, the actual impact to financials of companies.
Porter: And can I just add to that as well on the point of demographics, because one of the areas where we are seeing, you know, AI have a significant impact is on health care, you know, in terms of the pace of drug development, how we’re diagnosed and able to diagnose conditions much earlier. And that really is going to help with longevity and people living much longer. We’re already seeing that demographic time bomb hit play out in many Western markets. And with much of the ratios, you know, people who work, to people who are retired really starting to stretch, have really put a focus on productivity gains. And that’s where, you know, in addition to what AI does to health care is what AI can do to productivity. And that’s many of the areas that, you know, Richard was talking about with Klarna and having patience and seeing those productivity gains come through. But AI really, you know, hitting on both sides of the demographic answers and stretching out the issues.
Bullock: So, and I understand the sort of excitement around AI, but if I just go back to one thing that Richard you mentioned before, about, so the last few months there’s been quite a bit of volatility because expectations were probably way ahead of themselves. How much then is the market really already pricing a lot of this upside in? Is the market getting away from itself or is there still sort of upside there?
Clode: I think that there have been a lot of narratives sort of painted through the posts, you know, what, what we’ve seen through the summer, I think in the summer, you know, so people still do take some holidays now and again. And it’s a thinner market as a result, you do get some exacerbation of kind of, you know, will probably be some natural digestion after very strong [tech sector] returns in the first half. And we saw that, you know, again, you know, last year, I think what was different this time was I think there was an exposure of a huge amount of leverage that was in the system generally, you know, whether that be funded by yen carry trades, but you know, the end outcome was a lot of leverage being built up in areas that were working on the investment side. And you know, a lot of those areas were related to, you know, areas like AI. And you saw that in particular, some of the AI semi[conductor]-related names, you know, some exacerbated kind of drawdowns that seem to dislocate from fundamentals.
Because actually, you know, this was sort of coincidental with, you know, major sort of reporting season from the big customers of NVIDIA and AMD or Broadcom. And actually, those customers were saying some very positive things about not just, you know, capital spending on AI now, but your capital spending into next year. And so from a fundamental point of view, you still felt kind of comfortable and then you saw that dislocation. And obviously as active investors and hopefully as experienced investors, we’ve seen that movie before and we can use those dislocations that we think from fundamentals to situations where we think that there is a an outsized move that doesn’t really reflect where things are and ultimately that stock has just got cheaper.
Bullock: And so we saw the [US] Federal Reserve cut rates just recently. So then what sort of impact does that have, because technology can be quite sensitive then to rate rises, then also rate cuts?
Porter: And I think if you look at the long-term outperformance of technology, it’s actually been much less sensitive to interest rate moves than you might think, you know, especially getting, you know, the sharp outperformance that the sector has had over the last two years and a rising rate environment. And we will see that actually the technology waves that we see are much more impactful than interest rate waves in terms of how the sector acts. And actually, you know, part of those technology waves as we move into this fourth wave of investment, is much more impactful, you know, both from the perspective of capital spending and also from the perspectives of you know, discount rates and how the longevity of growth rates, how sustainable people think that they are long term, and we have seen even into this interest rate cut, you know, this actual shift into, you know, bond proxies over the last three months, which was quite curious because we do see some of these consumer staple and defensive-type names actually trade at significant premiums to some of the large cap technology stocks. You look at a Walmart, that you know, was significantly more expensive than Amazon here, despite much lower margins and slower growth. So, you know, we think that interest rate cuts can help to accelerate some of the cyclical [economically sensitive) side of the economy, but that structural piece of investing in the fourth wave, and investing in this AI infrastructure we think will, you know, continue pretty much unaffected by interest rate shifts.
Bullock: And then so just sticking to the AI story right now, are you seeing sort of a level of protectionism to some extent from countries as far as when you look at the evolution of AI, it’s quite a powerful tool, but also it can be a very dangerous tool. So, does that mean that, you know, is there a bit of an arms race going on in AI? What’s the implications of that, you know, bigger picture implications to society, but then you know, from an investment perspective, what does that, what does that mean?
Clode: Yeah, absolutely. It’s certainly that, you know, the question of our times. You know, I think and you know, to Alice’s point about, you know, the technology waves being more powerful than the interest rate waves. I mean, you know, we heard from the CTO of Microsoft at the Goldman [Sachs Communacopia +Technology] Conference last week. He made probably the most important comment of the conference, which is, you know, we’re just not seeing diminishing returns from making these large language models bigger and more powerful, throwing more compute at them, throwing more parameters at them, throwing more data at them. And, you’ve got to think about how early we are in the innovation of this technology still. So, the transform models have basically taken the shackles off. You can make these models bigger, you can make the multimodal, you can throw more data at them and compute at them. And we’re just coming to terms with actually, you know, how you scale these models up. Actually, we’re nowhere close to sort of thinking that actually, if you make these models bigger, we’re not going to get more exponential capability or innovation out of them. It’s just, it’s, you know, we’ve only gone from 10,000 GPU clusters to 100,000 GPU clusters. And now people are thinking about, you know, we’re still on a pathway to a million GPU clusters. And, you know, we’ve gone in the space of less than two years from, you know, a text prompt from ChatGPT to, you know, Adobe Firefly, being able to do text to image to an OpenAI Sora being able to do text to video. And, you know, we think that ultimately that has huge, huge implications across a lot of white-collar work, which will have some demographic impacts, but also has some huge kind of security, national security, cyber security productivity, health care implications. And so, this is very different to the internet wave where, you know, you saw in China that technology did develop, you know, very domestically, they could access all of the chips they needed from the US and the domestic champions were built up, domestic ecosystems were built up. And you could invest in those companies and do very well as well as the FAANG stocks. But then you could also buy a proxy in South Africa, like a Naspers or a process in, in, in Europe. There are many ways to access that and to take advantage of the internet wave in China.
That’s just not going to happen in an AI sort of wave because of the restriction on US semiconductor exports, which is, you know, today you can, you know, train a model in China on a chip that’s, you know, half as good as the one that Microsoft or Google is training on. Next year, that will be five times or 10 times a couple of years out that will be 50 times, you know, in a very exponential kind of curved way. You know, within a few years, you know, something that would be trained at Microsoft in a day will take a Chinese company a year. And that means the innovation curve of AI in China is going to look so different to what we saw in the internet age and from an investment opportunity that’s going to make it very different as we look at sort of investing globally and accessing that innovation globally. But also, from a risk point of view, just given the geopolitical kind of nature of actually one thing the Democrats and Republicans actually agree on. And no matter who’s in the White House at the end of the year, they probably don’t want the AI capabilities of China moving forward at any great pace.
Bullock: But does that result in more winners and losers? Like, are you going to get a different spread? So, if you’ve got different countries and different companies going at different paces and it’s almost like this exponential curve of development, are you going to get this big differentiation between countries and their abilities to implement AI in its most effective form?
Porter: I think you know in technology it’s always been a sector that benefits from economies of scale. And we’re definitely seeing that in terms of the rollout of AI infrastructure where we do see the large hyperscalers who have the capital available to invest in this early stage. And even at the semiconductor level, you do see a concentration of both companies who have the R&D [research & development] capability who are benefiting from scale up and the number of winners. We’ve seen a lot of M&A [mergers & acquisitions] in semiconductors as well over the last 10 years. So, you really do naturally see some of that concentration and that focus on winners and losers if you looked back for the last 10 years. But then, you know, in terms of relative performance, only four of those 10 have continued to outperform versus technology. But it would be overly simplistic to say while, you know, let’s just buy small caps or let’s just buy large caps. Because actually when you look at, you know, the pace at which if you’d bought a company who’d hit a trillion dollars, then the time to it doubling, it’s actually been much faster than it is if you’re buying companies at one billion going to two billion, or 10 billion going to 20 billion. So there are scale advantages, but there’s also the leaders of yesterday that are not necessarily the leaders of tomorrow. And what that tells you is you just really need to be bottom-up and be an active stock picker to really think about these winners and losers and to really understand the supply chains of all of these companies as well. I think, you know, for national security reasons, for data security, for really controlling your supply chain, because COVID really shone a light on the importance of that.
Clode: And we’ve seen a kind of a difference in the way that these technologies are kind of populating globally and to your point in the internet era, you know, companies or countries were pretty comfortable with US companies dominating that and, you know, US kind of dominating the internet technology and dot coms and registries and, and all of that. And we’ve seen kind of in the sort of the AI era, a change in that thinking from, I think kind of a deglobalisation and geopolitical kind of point of view. But then also from a just what this technology can ultimately lead to. We’re seeing a lot more sovereign investment in this technology that countries want to have some control as to where this technology is going to be going, some control over their capabilities and not be reliant on an ally today that might not be an ally tomorrow. So, you know, NVIDIA is going to quantify this as low teens, billions of spend with them is coming from sovereigns and sovereign entities. That’s something we’ve not seen in the past. And that again, will create opportunities for us. Yes, we’ve talked about, you know, some potential risk if you’re investing in an AI company in China. But now there’s an opportunity to invest in a Middle Eastern company that’s developing AI, or we’ve seen that in Singapore, in Germany, in Japan. And so that as with all of these and the beauty again of being an active manager, there are risks but there are also opportunities from that AI being wanting to be controlled in certain countries, and hopefully fostering some domestic champions there as well.
Bullock: So, linking this all together and going back to the study with University of Liverpool and that acceptance or reluctance with AI, does that sort of also change where you’re seeing the sort of growth opportunities? So, for example, if a population is younger, are they sort of adopting AI faster than a more, you know, an older population? Does that come into your thinking as well?
Porter: I think not yet going back to Richard’s point on patience. I think you know, certainly you know when you see these early [tech] waves, you tend to see, we have to see the infrastructure build out fully first, then some of the applications emerged over the longer term. And at the moment, many of the applications that we’re seeing are really much focused on cost savings rather than, you know, new revenue generation opportunity. You know, you saw with the adoption of ChatGPT, you know, hitting over 100 million users within a matter of weeks. You know, it’s happening at a record pace. Our kids are all using ChatGPT 4 for homework. There’s not the reluctance, the proliferation of these technologies is much faster than it has been historically.
Clode: But on the one hand, you know, you see, you know at one end, you know a TikTok or Reels, you know benefiting from recommendation engines based on large language models and that was skewed younger, but to you know your point about in Japan being very soon having one worker and one dependent. You’re going to need a very old generation to adopt AI and the productivity gains and the care potential of, you know, robotics there to ultimately solve for that much more aging demographic issue there.
Bullock: So, we’re almost out of time, but I just want to ask one other question to both of you. And yeah, we talked a lot about AI, but I also want to focus on broaden technology and ask you the question of unloved areas. So are there any areas in technology that people aren’t talking about that aren’t grabbing the headlines right now, but you’re thinking about or you’re seeing opportunities that others are not?
Clode: I think there are a couple of ways you can look at that, I mean, in terms of technologies and you know, I can go into that, but I can, Alison can talk to that as well. But you know, maybe start on, on geographies. So actually, if you kind of look, this year, I mean, there’s been so much focus in the e media on Mag[nificent] 7 and US equities. I think people have forgot that actually, you know, some other areas are actually performing and kind of when probably China got to its lowest ebb, actually, we’ve seen some pretty good performance out of some of the Chinese internet names. And you know, we’ve been thinking about, you know, keeping them in the portfolio and, and [potentially] being rewarded for that this year, and then yet you’ve seen that in other places too, say in Latin America we saw MercadoLibre and that [has] been performing well. We’ve seen some decent performance out of the Southeast Asian names too. So I think that reminds you that you want to stay global and that, you know, we very much focus on funding underappreciated earnings growth. And often, you know, in places that are feeling very underappreciated, those are good places to go and hunt. So I think geographically there been some good places to hunt outside of the US this year and it certainly has paid to stay global.
Porter: And and more specifically in terms of sectors, you know, as we’ve seen in the second and third waves of technology during the mobile internet, the first areas to perform really well have been on the semiconductor side and advertising. The two areas where the use case is advertising, improved targeting has really benefited early. And but once actually you’ve bought a lot of the silicon to enable those advertising, targeting and use cases. We also then have to think about the connectivity and we certainly see opportunities in areas like networking that has been left behind over the last 18 months or so as we’ve seen these hyperscalers build out their infrastructure. But now we have to think about connecting it up and particularly at an enterprise level for consumers.
Bullock: Great. Well, Alison, Richard, thank you very much for your time. We’re out of time, but thank you also to our audience for listening. Of course, if you have any questions or you want to know more about Janus Henderson’s views or strategies, don’t hesitate to contact your client relationship manager or visit our website. So with that, I’d like to thank you all very much for listening and wish you a very pleasant rest of the day.
Queste sono le opinioni dell'autore al momento della pubblicazione e possono differire da quelle di altri individui/team di Janus Henderson Investors. I riferimenti a singoli titoli non costituiscono una raccomandazione all'acquisto, alla vendita o alla detenzione di un titolo, di una strategia d'investimento o di un settore di mercato e non devono essere considerati redditizi. Janus Henderson Investors, le sue affiliate o i suoi dipendenti possono avere un’esposizione nei titoli citati.
Le performance passate non sono indicative dei rendimenti futuri. Tutti i dati dei rendimenti includono sia il reddito che le plusvalenze o le eventuali perdite ma sono al lordo dei costi delle commissioni dovuti al momento dell'emissione.
Le informazioni contenute in questo articolo non devono essere intese come una guida all'investimento.
Non vi è alcuna garanzia che le tendenze passate continuino o che le previsioni si realizzino.
Comunicazione di Marketing.