319. Megatrends, AI supercycle and the future of investing

Zehrid ‘Zed’ Osmani, manager of the Martin Currie Global Portfolio Trust, elaborates on his three-step investment process that ensures only the most promising companies make it into the portfolio, before diving into the significant themes driving this strategy, including demographic changes, future technology, and resource scarcity. Additionally, Zed provides insights into the valuations of industry giants Nvidia and Mastercard, and touches on the implications of interest rate changes on quality growth companies.

Zed has shown himself to be an excellent manager of high-conviction strategies. This trust has the ability to tap into a series of long-term themes – such as the rise of electric vehicles, growth of the emerging market middle class and the onset of artificial intelligence – which have the potential to deliver strong outperformance for investors. The highly-driven research approach has proven to be extremely successful over the longer term across a range of portfolios.

What’s covered in this episode:

  • Three-step process to building a high-conviction portfolio
  • The trust’s valuation framework
  • The three megatrends targeted in this trust
  • The benefit of thematic investing
  • Targeting seismic thematic shifts, for example, an aging population
  • Why the market is underestimating the AI opportunity
  • The cross-section of AI and the stated megatrends
  • The “techno-industrial revolution”
  • The investment case for Atlas Copco
  • How to evaluate a company like NVIDIA
  • The investment case for long-term holding Mastercard
  • Do macro changes influence underlying holdings?

27 June 2024 (pre-recorded 20 June 2024)

Below is a transcript of the episode, modified for your reading pleasure. Please check the corresponding audio before quoting in print, as it may contain small errors. Please remember we’ve been discussing individual companies to bring investing to life for you. It’s not a recommendation to buy or sell. The fund may or may not still hold these companies at your time of listening. For more information on the people and ideas in the episode, see the links at the bottom of the post.


Staci West (SW): Welcome back to the Investing on the go podcast brought to you by FundCalibre. Today’s episode provides a detailed explanation and deep understanding of the intricacies involved in managing a high-conviction, concentrated portfolio in a dynamic global market.

Chris Salih (CS): I’m Chris Salih, and today I’m delighted to be joined by Zehrid Osmani, manager of the Elite Rated Martin Currie Global Portfolio Trust. Zehrid, once again, thank you very much for joining us.

Zehrid Osmani (ZO): Thank you. Chris,


CS: Let’s start with, I mean, your portfolio, in my opinion, is as active as they come. You’ve got 25 to 40 of the best companies. I mean, for a listener wondering essentially how you get from A to B, maybe just give us a bit of insight in terms of what sort of companies you’re looking for, how you sort of spot those gems in a sort of huge universe of companies.

ZO: Yeah, it’s a good question.

So, we have an investment process that’s a three-step process. And the way we narrow it down is in the first step, which is screening. Here, what we are doing is we are looking for companies that have high returns on the invested capital and an attractive growth profile. We also like companies to have solid balance sheets, so that’s another area of assessment. And when we look at return on invested capital, we include goodwill which really means that we like companies that grow organically rather than through acquisitions. If they go have a lot of acquisitions, goodwill will increase. At that point, we’ll also assess some of their capital allocation decisions to assess whether they’ve created or destroyed value as part of that and we are able to then filter down to companies that genuinely are able to create value for shareholders. So that’s how we then narrow down the universe at that screening stage.

We then take that bunch of companies and sit down in what we label research pipeline meetings, which is where we decide which of those companies are screening favourably across those measures. Are we going to move to step two? And that is a combination of harnessing our analyst knowledge on the sectors that they cover, and therefore their preference for which companies we should work on, as well as our needs at the portfolio level in terms of different areas. And then we move to that step two, the in-depth fundamental research which we do iteratively. That’s where we then understand all the aspects that we want to assess on the company at the fundamental level – and I’m sure you’re going to have questions around that valuation in that step two is critical and we’ve got detailed structured valuation framework with discipline on that front.

At the end of that step two, we have a good sense of upside on any company we’ve researched, but also good understanding of the risks that that company brings. And then we therefore move to step three, which is the portfolio construction.

Portfolio construction for us is capturing the best ideas that come out of our research process. and delivering them in this high conviction concentrated portfolio that you mentioned of 25 to 40 names, but importantly, with no unintended risk exposures. And I would argue that many stock pickers focus on step one and two, screen for ideas, research them if you like it, put it into the fund. But for us, step three is just as important as step one and two, because of that aspect of ensuring that we do not have any unintended risks.

CS: Okay. Well, let’s quickly, I mean, there’s no time at the present – you talked about the valuation there. Maybe let’s just touch a little bit on what you look for in terms of valuation, and maybe just give us an idea of how many companies you sort of go really deep into before building that final portfolio.

ZO: Yeah, good question. On the latter part of that question, typically we go through 90 plus stocks per annum in terms of in-depth research, that’s step two of the process. And the valuation framework is based on three tools. One is discounted cash flow, which accounts for 50% of our price target assessment. The other one is economic value-add, for 25%. And the last one is target multiples, and for that it’s 25%. But instead of using 12 months forward multiples, we use year five multiples, so arguably also longer-term valuation tools. Ultimately the aim is to put a fair value on the business that we look [at], rather than where the market might be willing to pay over the next 12 months.

We forecast for companies over 15 years; it’s a long forecast period in order to make sure we capture the full growth potential of any company we research, but also forecast the stage of maturity of those businesses by year 16 onwards. And yes, forecasting is difficult, but we always say that if you’re not aiming to forecast, you’re having even less potential to understand the shape of growth and returns profile of any business that you’re looking at.

And then those forecasts at the valuation level. We also don’t just rely on one set of forecasts, which would be the base case scenario. We run different scenarios: we run a bull case, a bear case, a blue sky, and a dark sky scenario so that we capture different potential outcomes because of that forecast risk that I mentioned at the start. We ask the analyst to put a probability on each scenario, which means that that gives us additional information about upside/downside risk on any company as we look at the valuation. And we’ll always aim to favour companies where the skew is to the upside. But it’s by having that very detailed assessment, very structured framework that we’re able to then have that sense of which companies give us the most interesting upside, which companies have we got the most favourable risk profile [for], and which companies have we got the strongest conviction in.

CS: There’s sort of a, you can sort of paint in a picture of sort of this search for sort of quality growth, but also the valuation argument of sort of, you know, you can see the story of the company and how it evolves to a point where perhaps it becomes more mature. But there is another string that I wanted to talk about, which is the megatrends that drive your research. So could you maybe talk about those trends and how they’re evolving, please?

ZO: Yes, absolutely. And I’m glad that you mentioned the quality growth because that’s how we define the type of companies we invest in a good balance of high return invested capital for the quality side and a good growth profile typically.

So, in terms of the thematic lens, we’ve put in place a framework with three megatrends that we’ve identified, which are demographic changes, future of technology and resource scarcity. And if you think about it, Chris, those megatrends are here to stay in 10 years, 20 years, 30 years’ time – hopefully we’re still talking to each other and still in the industry – we’ll still be talking about this. There are themes within those three megatrends, and those themes can evolve over time. But what we do is we list the themes that each megatrend is capturing. When we research a stock, we assess which themes is that company facing on a forward-looking basis, and to what extent, which means we can then aggregate the data at the portfolio level, which permits us to achieve three things.

Firstly, we know the portfolio exposures more accurately in terms of themes that run through the portfolio and to what extent they run through the portfolio. Secondly, we can manage diversification more efficiently. So, we can look at how many themes have we got, and are we sufficiently diversified given that we are running typically 30 stock portfolios. And thirdly, it gives us visibility on themes we might not have researched, in which case we point the research into those, or if there are themes, we want to gain more exposure [in] – again, pointing the research efforts there. But we would only invest in a theme – or companies in a theme – if they’re attractively valued. I always say we don’t invest in attractive themes, we invest in attractively-valued themes. But importantly by this framework, we put our clients’ assets, our shareholders’ assets, on long-term structural growth opportunities, looking ahead into the future rather than looking into the rear view mirror.

CS: I mean, you’ve got those three themes that come to mind, you know, the future technology, the resource scarcity and those demographic changes. I just want to sort of maybe try and sort of bring them to life a bit earlier. So, I mean, we should talk about some of those underlying companies.

I wanted to touch on healthcare, which sort of covers a few of across all of those themes, really. I mean, it makes up a quarter of the portfolio and, you know, people are clearly aging. I think there’s like 1.3 billion people are predicted to be sort of 65 or older by 2050. You tend to favour companies that tie into that sort of theme of futuristic technology or have more efficient means of delivering care rather than those more traditional pharmaceutical companies. Could you maybe just sort of explain why and sort of give an example perhaps?

ZO: Yeah, absolutely. And you’re touching on one of our three seismic thematic shifts, as we call it, which capture the eight themes that we favour. And so I’ll very quickly mention those three seismic thematic shifts. One is energy transition, the other one is aging population which is the one you’re asking about; and the third one is artificial intelligence. And so as we zoom in, maybe we’ll go on to energy transition and artificial intelligence later on, but on the aging population, it is about healthcare infrastructure.

So your question is, well put, Chris. The population is aging over 16% of the world population will be 65 years or over by 2050. And sadly, as a result, incidences of diseases will increase with age. We have a theme called the 21st century diseases which are cancers, obesity and diabetes, which are prominent lifestyle-related diseases and sadly they’re comorbidities, so they’re linked to each other.

There’s estimates that show that incidents of diabetes will increase by 49% by 2040; cancers will increase by 31% over that period. And then on the obesity side, so we can mention things like heart failures increasing by 92%, or breathing disorders that are somewhat related to obesity increasing by 37%. So, a lot of reasons why there will be a significant need to invest more in healthcare infrastructure. And as you observe from our positions, we like to be in companies that are capturing the spending that will go into that healthcare infrastructure, so companies in the medical technology space it’s a significant part of the portfolio.

And we’ve got a combination of companies that do benefit from the drug developments and production and the outsourcing that goes around that. We’ve got companies that are exposed to the genomic space, which touches on the theme of bespoke healthcare, so more targeted therapies as a result of the advances in genome sequencing. So it’s an exciting area, that one.

CS: Okay. Obviously, your biggest exposure is in the IT space. I mean, it’s the largest single sector exposure. Maybe just talk us through that as well. I mean, is that heavily linked to AI in terms of your focus on long-term trends? Do you sort of have primary and secondary beneficiaries? For example, I think we talked before and you mentioned Atlas Copco as an example of a company that perhaps is a sort of secondary beneficiary of that. Maybe just give us a run through on that and also the idea that perhaps it’s may be, you know, some of the pessimists would argue it’s a bit too early to sort of monetise AI as an investment theme. Could we get a bit of insight on that please?

ZO: Yeah, sure. So our view is that the market is underestimating the AI opportunity, both in terms of size and in terms of speed at which AI is going to be taken up. Four of our eight midterm thematic opportunities capture AI. The first one is metaverse and quantum computing, and this is about big tech spending on these areas. The second one is robotics and automation which has been accelerated by the advance of AI. The third one is cloud computing as cloud providers are racing each other to upgrade that cloud infrastructure to make it AI ready. And related to that, the increased spend in cybersecurity. And the fourth theme that touches on AI is technological and geopolitical fragmentation. Here we’re talking about China, we’re talking about the Taiwan focus and the tensions between China and the US around Taiwan. The fact that the US is limiting access to leading-edge semiconductor technology by China which creates that technological fragmentation and geopolitical fragmentation. So there’s going to be winners and losers from that.

Arguably, AI touches on all areas of the economy ultimately. So even our theme of energy transition will benefit from AI. At this stage, the market debate will be about what is the size of the opportunity and how quickly will that opportunity be realised. So, our view is it touches all agents of the economy; governments, corporates, and households. Corporates as are most likely to harness AI faster because AI makes a corporate more productive and/or more creative, and therefore either keeps it competitive or makes it more competitive in what is an ever more competitive world. So harnessing AI will be critical. And we think governments will increasingly be spending more on AI – Nvidia calls it the sovereign AI opportunity. It started to come through, but from an early stage. And here it’s about national security, it’s about defence, and clearly AI will be able to enhance the ability to deliver on both of those fronts.

There’s an interesting report by Goldman Sachs that highlights [that] AI has a potential to increase global GDP growth by seven percentage points and boost productivity by 1.5 percentage points per annum. So, an increase in productivity over a decade of over 15 percentage points: that’s staggering! But it also has a potential to displace 300 million jobs worldwide, which is why it’s going to impact households, it’s going to impact governments from the point of view of needing to put policies in place to retrain workforces. And there’s going to be potential shifts and migration in displaced labour, which itself creates its [own] political challenges. But AI will impact not just the services, but also the industrial economies, whether it’s manufacturing, construction, transport, autos, agriculture, aerospace, energy.

So, we are with AI living through the start of what we think to be a techno-industrial revolution. Jensen Huang, the CEO of Nvidia, called it an industrial revolution. We amend that quote by saying it’s a techno-industrial revolution that we’re on the cusp of living through. And it’s going to speed up breakthroughs. I can maybe give you a few anecdotes.

There’s about 300 million known proteins to mankind. It typically takes a student, a whole PhD, so five years, to unfold and map in 3D terms, one protein. Google, through its AlphaFold programme using AI, has unfolded all known 300 million proteins in record times and made that data available to humanity. So, technically it saved over 1.5 billion PhD study years through that process. And we’re already getting proteins being developed that have the potential to make plants disease resistant using that knowledge. We’ve got some proteins that have been developed that eat into plastic waste, which tackles some of that. And there’s so many other examples like that of breakthroughs that have been accelerated, which makes it a very exciting era to be in.

As far as Atlas Copco is concerned, Chris, the way we look at it is about a quarter of its business is in vacuum technology. And vacuum technology is used in the semiconductor industry to keep the semiconductor chambers free of any dust. The bigger the chambers, the more complex vacuum technology you need to use. So that’s where Atlas Copco is somewhat exposed to the semiconductor ecosystem. And on the semiconductor ecosystem, we believe that we are facing a supercycle. There’s going to be a need for more semiconductors generally across devices. And we are now saying that AI is turbo boosting the semiconductor supercycle. So, where we’ve been forecasting an 8% annualised growth in the semiconductor market to 2038, so over 15 years, we are now increasing that estimate by 2.5 percentage points so we’re now looking at a 10.5% annualised growth in the semiconductor market as a result of that turbocharged AI.

CS: I wanted to follow up on Nvidia because it’s timely, well, it’s been timely for a few months, but particularly because it’s now the world’s largest company*, it’s also the, as of speaking, the biggest holding in your portfolio, I believe. How does a valuation of a company like that get assessed in your portfolio? Because as we’ve said that, you know, it is AI is being monetised, but it’s such a huge theme. How on earth do you sort of evaluate, you know, a company like Nvidia for a long-term growth story?

ZO: Yeah, so it’s the importance of forecasting over the long term to really capture the potential addressable market that AI brings. And as an aside, even a company like Nvidia back in March 2021 was estimating AI to be an addressable market of $300 billion. Then 12 months later, increased that figure to $1 trillion. And this March has actually started talking about a $2 trillion market potential by the addition of sovereign AI, so government spending on AI, as we mentioned. So, it’s still going to be an ongoing debate, but for us, valuation is about we forecast the addressable market, we forecast how much of that Nvidia can capture through its competitive positioning, we assess revenues, profits, and cash flows through that 15-year period. We feed that into our valuation framework, and we stay disciplined.

And for us, what’s exciting when we look at Nvidia is many aspects. And I went to their headquarters and R&D centre in March actually to get a better sense for the culture that runs through this company and to discuss how they see the future. And we think Nvidia is effectively in a similar position to where ASML was 10-15 years ago as a quasi-monopoly facing a super cycle. That’s the first point to mention. It’s got a significant market share in the GPU space, over 90% share. It’s got scale benefits, which means its R&D spend now in dollar terms outweighs some of its competitors like AMD by two to three times, which permits it to not just keep its technological leadership, but actually, in our view, widen the technological gap. And you see that from the fact that they’ve accelerated their innovation cycle from introducing new products every two years to now going to a one-year upgrade cycle. And what we’ve seen from this year is that they’re actually further consolidating their competitive positioning because they’ve presented this Blackwell chip**, which is combining GPUs, CPUs network components as well as software all into one. So, they’re really providing solutions to the cloud providers, and the cloud providers are actually spending an awful lot of money. So, when we’re looking at the three cloud providers and Meta in terms of CapEx spend and how much they’ve increased, we’re looking at $130 billion additional CapEx over the next three years, over and above what they were planning to spend prior to this AI excitement.

CS: Ok. I wanted to touch on one more holding quickly before we sort of round off here, and that would be perhaps one of your longer-term holdings in Mastercard. That’s probably a good example of you highlighting the story of what this fund tries to do. Could you maybe just talk us through, you know, just, just briefly on sort of what you, what it has achieved and why you expect it to remain so, I guess more importantly, why you expect it to remain so prominent holding going forwards?

ZO: Yeah, so we like the credit card [companies] generally because they’re in an industry that’s got favourable dynamics as an oligopoly. And they’re capturing effectively transactions globally because they’ve got this very dominant positions, they’re able to achieve scale economies and high returns. So, you’re looking at companies [that] are therefore able to grow at a reasonable rate and are able to achieve high returns. So, Mastercard, we’re looking at 79% ROIC [return on invested capital] last year which we think will go to 153% by year five of our forecast. And in terms of gross profile, we’re looking at a company that should be growing its top line at 14% annualised over the next five years and the earnings at 16% annualised over the next five years. So, a lot of reasons why like, it.

CS: I was going to say that that return on invested capital kind of gives you an idea of how, not dominant, but sort of secure position it has in the market if you catch my drift on that without sort of putting the mockers on them.

ZO: Absolutely, absolutely. And actually, we should maybe mention the same <inaudible> that this is a [company] that’s asset light. So, we’ve got 113% ROIC last year; we’re looking at 146% ROI by year five. And the growth profile of Nvidia is exciting. 33% annualised growth on top line 33% also on earnings over the next five years. All the aspects that really make this an attractive algorithm of high growth, high returns.

And what I should have mentioned on Nvidia as well is we like companies that can monetise the AI spend and what Nvidia has shown is that it is indeed able to monetise. So, when you look at the share price performance, yes, it’s been very strong, but it’s actually barely kept up over the last 12 months with the sales revision. And actually, earnings revisions have been significantly more than the share price move which maybe highlights Chris a point we want to highlight, which is in AI, what you must watch out for is a potential risk of froth.

So, we track a basket of over 50 AI stocks globally. Over the last 12 months the share price on average has gone up 44%; we’ve taken Nvidia out of that because its share price went up by over 200% so we didn’t want to skew that average. So, a 44% increase in price on average for that basket of 50 AI stocks, but the sales estimates have not moved. And the earnings estimates have only moved up by 7%. So, we believe that at some point over the next couple of years, if these companies do not show the market that they can monetise our AI opportunity, you could have an element of realisation that not all companies are well-positioned to harness AI.

CS: Okay. That’s very interesting. I want to just quickly finish with a sort of a minute on the economy. It seems like, well rate rises have already started in Europe, but they seem to be imminent everywhere else too. Does that play any impact whatsoever on your portfolio given that quality growth focus? Or are you looking at it as an opportunity given the underlying companies you invest in?

ZO: Well, our companies are long duration companies in terms of cashflow generation, compounding characteristics, high returns. So, they tend to have an element of sensitivity to rates in the similar way to bond-like characteristics. So, as we are heading into central banks starting to go into easing mode and interest rate cuts, we believe that should be favourable for the quality growth type of companies for those rate sensitivities.

The aspects we’ve been highlighting the backend of last year though has been aware of the view that inflation will be steering longer lasting. That’s now coming to be a realisation by the market. That feeds into our view that central banks for us, were not going to pivot towards cuts until the second half of this year, which is more or less what’s happening. It’s pleasing to see that the market has now adjusted its expectations because it [was] probably a bit too bullish on that front.

But it’s got to be said Chris, it’s an interesting period because central banks that have started to cut, and the Fed that’s signalling a cut in H2, are really talking about hawkish cuts. And that’s quite unusual. It’s effectively saying we’re cutting, but actually we’re not going to cut anymore. So, a hawkish cut is maybe a realisation by central banks that inflation is indeed stickier, and they might not be able to do much cutting if at all, but perhaps they’re doing one cut just to give to the market what they had signalled towards back end of last year in order to not be disappointed. Net net, we would say as long-term investors heading into a period where rates might get cut, should be supportive, but keep an eye on wage inflation because that’s the biggest contributor to inflationary pressures. And we think wage inflation still remains elevated in the US, in the UK and in Europe. Hence our view that inflation will be stickier. And what we wrote last year, Chris on the back of the Jackson Hole [Economic] Symposium is that central banks have gone into a data dependency mode now. So, any inflation print will be something that they’ll look in order to determine their next interest move. And therefore, the market will be somewhat more volatile because market with every inflation print will be speculating as to whether this means the central banks are going to be on hold or whether they’re going to cut.

CS: So, on that note, that very cheerful note, I’m going to say thank you very much for joining us once again.

ZO: Thank you very much, Chris, it’s been great to talk to you.

SW: This is a high-conviction, unconstrained portfolio of 25-40 companies from across the globe, with the top ten accounting for more than 50% of the overall portfolio. The trust has a strong focus on quality growth businesses that the team believe are leaders and innovators across long-term investment themes, such as the future of technology, resource scarcity and demographic changes. To learn more about the Martin Currie Global Portfolio Trust please visit FundCalibre.com

*Nvidia briefly overtook Microsoft as the world’s most valuable company on Tuesday 18 June

** The NVIDIA GB200 Grace Blackwell Superchip

This article is provided for information only. The views of the author and any people quoted are their own and do not constitute financial advice. The content is not intended to be a personal recommendation to buy or sell any fund or trust, or to adopt a particular investment strategy. However, the knowledge that professional analysts have analysed a fund or trust in depth before assigning them a rating can be a valuable additional filter for anyone looking to make their own decisions.Past performance is not a reliable guide to future returns. Market and exchange-rate movements may cause the value of investments to go down as well as up. Yields will fluctuate and so income from investments is variable and not guaranteed. You may not get back the amount originally invested. Tax treatment depends of your individual circumstances and may be subject to change in the future. If you are unsure about the suitability of any investment you should seek professional advice.Whilst FundCalibre provides product information, guidance and fund research we cannot know which of these products or funds, if any, are suitable for your particular circumstances and must leave that judgement to you. Before you make any investment decision, make sure you’re comfortable and fully understand the risks. Further information can be found on Elite Rated funds by simply clicking on the name highlighted in the article.