Ways to play the AI revolution
We have touched on some of the opportunities and changes out there. But how are you looking to tap into these opportunities when it comes to funds, ETFs and public companies? What’s your criteria for selection?
Clients often ask if they can invest in a certain fund. We tend to favour those with the diversification element that I alluded to before. Other times, we may invest in certain companies that can give similar exposure should the client prefer direct exposure.
Robotics is one of the themes we’re looking at. We looked at other themes such as cybersecurity and healthcare, and we have tried to find suitable funds or ETFs to play these themes. The supply is there, but the demand tends to be more for stock-specific opportunities, where clients are excited about one particular area or development. At the end of the day, that’s often the same names that they may be invested in anyway, be it Amazon, Facebook or Apple. But when there is demand for niche names, we’ll look into it too.
For us, the starting position is actually no different from any other fund. At Standard Chartered Private Bank, we really hone in on the three Ps: process, people and performance. That’s what matters most to us, regardless of whether it’s a robotics fund or a US growth fund.
On our platform, however, we do have a number of funds that specifically target robotics. During 2018, robotics was an actively promoted theme that investors evaluated and validated as something that was attractive. Now of course, a lot of the names are still those traditional tech names we’ve talked about. But there are other companies beyond the obvious ones that these funds are investing in, and they can give you some differentiation and alpha.
The criteria for us is similar to what Clive just mentioned. Typically, our fund selection team in Zurich won’t look at a fund unless it has a three-year track record and $100 million in assets under management.
But because these one-off thematic funds are in demand, sometimes we can make exceptions. I think if you have to wait for a three-year track record, you are not going to be one of the earliest platforms distributing this particular fund. There might not even be any capacity left. A lot of these funds do close.
You do have to put a few of those issues to the side. You look at the people, you look at the process, you look at the performance. So yes, for some of these thematic ideas, you have to put some of those traditional metrics aside because you have to get in line and be part of that wave.
I agree with John. On these specific themes, you have to be a bit more flexible than for other investments. And because many of the big names are covered in traditional funds or ETFs, we like to focus more on funds that provide something specific, especially if they’re more focused on small- or mid-cap opportunities.
Daniel, you’re an index expert. How do you design and build an index based on AI and robotics?
This one was quite tricky. As all of you have said, if you take the top 10 AI companies, you will find Alphabet, Apple, Baidu, Alibaba and so on. But creating an index in these names means you’re adding very little value.
When I came back with this idea of creating an index, I thought, ‘I have to change my mind. I cannot focus only on the top names.’ So I contacted Martin and we discussed the key issues. We adopted a ‘quantimental approach’ when it came to finding the potential.
That is to say, partly based on a quantitative approach and partly based on a fundamental approach. The fundamental approach was quite simple: we focused on those companies that were investing in AI. You reach a large number of companies, and then, doing some homework with Martin, we went into greater depth to create a universe of companies. We reached a list of 228 companies worldwide.
It’s quite diverse. There are a lot of US companies, as well as some Chinese, Japanese and European ones too. Then we did back tests on different criteria, mostly on research and development. We want to capture companies that are investing for the future, and we don’t care about the size of the company.
We touched on clients wanting single names and some thematic funds. Where else are they looking?
I think they like the big themes that are interesting. But this presents a challenge, in that the most interesting themes are often the newest, which makes it harder to find a track record for three years of performance. It’s important to manage clients’ expectations. You want to access those themes with an appropriate investment horizon.
This is definitely an area with huge interest and demand – you can see from the results in terms of some of the fundraising that has happened out there. The difficult part is that you can have that conversation with someone and you can talk about the theme. But you have to be willing to stick around longer-term, because it won’t always be an instant success.
I know that AI and robotics go hand-in-hand, but what is the right exposure? Should there be more exposure to AI or robotics?
If you look at the few ETFs listed in this theme, the first one that was listed – all the way back in 2012 – had more robotics than anything else. Then you had three others listed and they were more diversified. I would say still two thirds are robotics, one third is AI. But that may be because it’s very difficult to define exactly what impact AI has on a single company. What is the impact of AI on Alphabet, for example? It’s much easier with robotics. But I think with the greater use of data, we’ll have more and more AI companies in the different indices.
That’s true. Often the robotics companies are more established than the AI companies.
In terms of the overall trend, I would say that AI is much more important overall, and really I would view robotics as one segment of AI. AI is much broader, and it’s where a lot of value is going to be created outside of robotics. For that reason, I think it’s a mistake to focus too much on robotics. It’s important to have that broad view about how these technologies are transforming other businesses and other industries.
I also think there’s a geographical element as well. Robotics is an area where you have more names in Europe and Japan than you have in AI, which is more focused on the US and China. So, in terms of portfolio construction, I think it’s something that you have to keep in mind. Do you want exposure to those very mature markets such as Japan and Europe, where valuations may be less demanding? Or do you prefer the US and China, where valuations are higher but growth expectations are better too?
That’s true. If you look at any fund factsheet in this realm or theme, in general, it’s usually 60% to 70% US-listed.
And if you had to pick a single stock to play AI?
It’s not really a pure play, but Google is the greatest, most powerful concentration of AI capability and talent that exists. I published a book recently interviewing the pioneers in this space, and some huge percentage of the people I talked to were working at Google or one of its divisions. Google is absolutely the leader.
But companies such as Facebook and Amazon are not too far behind. There’s not an AI company that has done an IPO and that is just doing AI as a pure play. That actually doesn’t exist.
Every company that is really leading the innovation in AI is also doing web search and lots of other stuff.