Stephen Rudman: Good morning and good afternoon. My name is Stephen Rudman. I am a member of the Ashmore New York-based client-facing team. Welcome to all. Hope all are well. Thank you for joining today's Ashmore webinar, ‘Emerging Markets: Beyond the AI Trade’. Our speakers today will be Gustavo Medeiros, Ashmore's Global Head of Research, and EM Equity Fund Lead Portfolio Manager Dhiren Shah. 

The topics are straightforward. First, obviously: AI, are the fundamentals already priced in? Two, geopolitical and climate risks, Middle-East tensions, and El Niño. What else can we throw in there? China: hegemon, peer, or laggard? And finally, and maybe most importantly, opportunities after this strong emerging market (EM) rally that we've all experienced. With that said, let's get to it because we want the show to be Gus and Dhiren. I'm going to hand it over to Gus to get us running. Gentlemen, thanks for joining, and to you.

Gustavo Medeiros: Thank you very much, Steve, for the kind introduction. Happy to be back with one of my favourite guests, Dhiren. As always, I'm going to set the scene very quickly before firing the questions away to him, the tough ones. 

Obviously, there are three dominant factors in the global macro landscape today, the war of the Strait of Hormuz, which remains closed despite this soft ceasefire that sometimes holds, sometimes doesn't. There is still plenty of US policy uncertainty. Yesterday President Trump announced more tariffs, for example, to replace some of the tariffs that were deemed illegal by the US Supreme Court, and the ubiquitous AI capex boom that we're going to discuss.

On the war front, both sides quite clearly want the Strait reopened and a lasting ceasefire. The skirmishes that we've seen over the last days should probably be seen as an attempt to get into a better negotiating position. And the severe drop in oil supply, some 14 million barrels a day by most estimates, have been compensated by a drastic decline in consumption from China, which has plenty of reserves, and poor Asian countries, also as well as a large drawdown of inventories. 

And we're living through, in my view, a one-off inflationary spike. It's not the start of a sustained inflationary regime like what we've seen in 2021, 2022. That is simply due to the fact that the pass-through to non-tradable goods and services should be limited because real disposable income in most large countries in the world is falling. There has been some fiscal expansion to compensate for the higher energy costs, but they've been relatively small, particularly when compared to the 2021 policies. Consumers have been absorbing the fuel shock and have less to spend elsewhere, which is why the pass-through is going to be limited. So, the inflationary impulse is real, but, I'll argue, self-limiting. Inflation should have peaked in sequential terms in April and is likely to decline sequentially, in terms of month-over-month in May and June. And in year-over-year terms, inflation is likely to peak between May and June this year. 

If the Strait reopens reasonably soon, the path back to the ‘Goldilocks’ set-up that we described in our 2026 Outlook – of disinflation, a softer Dollar, and allowing for some policymakers, particularly in emerging markets, to consider easing monetary policy further – will reopen. Plausibly, it's going to happen towards the fourth quarter of the year, Q1 2027. 

So, a lot of our 2026 Outlook is at least postponed, some people would say challenged. We don't think we're there yet. But against this backdrop, again Trump announced further increase in tariffs, an attempt to increase lost tax revenue after a good part of the tariffs were deemed illegal by the Supreme Court. US policy uncertainties likely to remain, particularly as the Republicans lose the House in the first week of November in the midterm elections. 

Underneath this, the AI capex is still the engine of the global economy, as we anticipated in our outlook. This is already the largest investment cycle in history except for the 19th-century railways when you look at capex as percentage of gross domestic product (GDP), total capex, or many other metrics, and the numbers keep on increasing. For EM specifically, this is the cleanest transfer of capital from developed world to emerging markets in modern history. The hyperscalers, in a nutshell, depend on EM suppliers for semiconductors, silicon memory, and assembled systems. In 2026 numbers, four of the world's 10 most profitable companies will be EM listed: Samsung, SK Hynix, Aramco, and TSMC. So, the debate that we were having in October last year was: are we in an AI bubble? The debates that we have been having since then, particularly in Q1 and April to June this year is: where are the bottlenecks? And the answer is, as we published in our recent Emerging View a couple of days ago, semiconductors and energy and energy in particular in the United States. So those constraints slow the US build-out, potentially in favour of data centres in EM countries. So that is an interesting opportunity for many countries. 

We have still a largely supportive political backdrop in emerging markets. The rightward market-friendly tide across Latin America is still in place. In Colombia, we had the first round of elections over the weekend, and unless a disaster unfolds, it's very likely that the far-right candidate, Abelardo de la Espriella, will win the second round of elections. In Peru, we have a runoff over this weekend. Keiko Fujimori is in the lead. It's a tighter race than most would hope for, but it's likely that the right will keep on edging forward in Latin America. In Argentina, Javier Milei is passing reforms after a good showing in the half-term elections last year. And in Europe, we've seen a political transition to a pro-European Union (EU) party in Hungary, which ended the Orban era, which is likely to unlock tens of billions of euros for Hungary. 

But the question today is not whether EM is structurally attractive. The valuation gap to the S&P 500 is still very wide. EMs are still trading at roughly 12 times forward-looking earnings against 21 times for the S&P 500. The question at the moment is whether the AI trades still have room to run and if the opportunity set looks strong once the dust settles. 

So, the first question to Dhiren after this scene-setting, is on the semiconductor space. TSMC and Korean memory companies are now pretty well known stories since the massive run-up in memory prices in mid last year. And, if you look at the free cash flow for memory companies, it's set to overtake the free cash flow of all the hyperscalers combined in both 2026 and 2027. So, these companies are looking cheap if you look at a discounted cash flow basis. It's looking cheap if you look at a price-to-earnings basis. And, if you assume that memory remains undersupplied beyond 2027, these stories are still attractive. And again, it remains a major bottleneck. So, Dhiren, how are we positioned in that space today?

Dhiren Shah: Yes, thanks Gus, and it's good to speak again. And so just to set the scene, and you've alluded to a few of these things, so, first of all, from a usefulness of AI perspective, I think those debates have gone away. And this obviously leads into where we are in the market cycle. With Claude and all the other various products from coding to more general use, it's now going mainstream. So that debate has gone. The second one is on the capex side. It's essential to all the cloud service providers, from Amazon to Meta and so on, that they continue to build out. As we've seen, some of them are now free cash flow negative, and you've seen equity raises and so on. So, this is something to be mindful of. 

You know, who's the ultimate winner there? I think it still remains unclear, There will be multiple winners, but they're competing amongst each other, but tokens are in short supply, and while they're all making very good money today, they are spending a lot of capex. And whilst we think about some of the intermediate parts of the supply chain, you know, is it Nvidia, or is it the ASICs (Application-Specific Integrated Circuits) who are winning, what we see very clearly is, the EM beneficiaries. And so here, the competitive landscape is stable. Companies that two years ago we thought were great companies from a technology perspective, AI was maybe 5–6% of revenue, small. Fast-forward with all compounding you have, AI is now 30, 40, 50% of revenues. The operating momentum is incredibly strong, and this speaks effectively to the pricing power of these companies. 

And so there are a limited number of companies that can do what they do from a semiconductor standpoint, multiple parts of the value chain. So, you talk about TSMC, you talk memory companies, but there's a long, long list that goes beyond that. And increasingly what we're seeing is a lot of these companies, as demand booms 50%, 100%, as the intensity of semiconductors increases, it affords massive pricing power. So, the operating environment is very, very strong. We think it sustains for multiple years. Whilst there can always be short-term gyrations in the market, the initial work we've done in terms of these are unique companies, very strong outlook, the outlook is much stronger than we anticipated just because of the usage of various chips, and so on, is much, much higher. So, from all that, from a quality and growth standpoint, things are great. 

And the third part of the question is now: "What's priced in" right? This is where I initially opened on. So, there've been a lot of doubts, been a lot of sceptics on the sector for the last few years, and, you know, those voices have almost gone now. And with that, the share prices of many of these companies in the global wider industry have increased significantly. That obviously starts to make us slightly concerned in some selective areas. What that's meant for us is we've reduced our exposure. We're still fundamentally very positive, but what it means is for selective companies, whether we think they're pricing in too much now, we've exited. And so, this is evaluating it from a two, three-year cycle. There are certain parts of the value chain which we think no longer make sense. Having said that, there are still companies which have a multiple-year positive outlook. They're still generating lots of cash, and they're growing incredibly fast. So, we still like selective parts of the value chain, but prior to this was our largest overweight. Within our portfolio, it was one of the largest areas of our portfolio risk, and it was very deliberate. For us today, the set-up has somewhat changed to effectively taking profits as stocks have gone up substantially and recycling that capital elsewhere. In terms of some of the risks, in terms of the things that we look out for, are very mindful of, so clearly capex is increasing significantly. That increased the cost of tokens. But you know, this is the headline, but the underlying, the software developments behind tokens, are still growing exponentially so that the value of the usage is increasing. But just to be mindful of the absolute capex numbers, you have 1,000 billion going to possibly a trillion next year. These are becoming quite large numbers, and there is a cost from this, and these are just things we're mindful of. And these could obviously cause short-term gyrations in markets, which, within our framework, we'll look to take advantage of.

 

Gustavo Medeiros: That's great, yes. Clearly, hardware has been the AI winner then. As you mentioned, this won't be a straight line, but it is hard to see how these companies won't be doing well in terms of free cash flow, in terms of profits, over the next two years. There will be opportunities on these gyrations. But within the AI revolution, one thing that I've been mentioning to clients is that you need to be very focused on winners and losers, right, because there will be many companies that will be disrupted. And so, what are the sectors, and where are the places that we think companies can be severely challenged by this revolution?

Dhiren Shah: It's obviously a good starting point in terms the winners we've covered, and we have been taking profits. I guess one of the questions which you may have, but I guess I'll answer anyhow, is: "Where are you putting that? Where are you redeploying that money?" So, some of it is going into the perceived losers, right? We've been very clear. We've got a clear framework on it, which we've adopted for a number of years, in terms of identifying the industries and sectors, the companies, which will lose but also will win. They're ones that can defend their existing moats and moreover, the ones that are using AI to actually enhance their competitive industry. 

What you've seen in markets has been, it's been pretty phenomenal in terms of the bifurcation of markets. And so, the narrowness of markets has been very, very acute in the last quarter in particular, in terms of there have been very, very few winners. The winners, AI winners, have been sucking liquidity out of the rest of markets, especially with retail investors, but as well as institutional. And so, they've been chasing into that winning space, and it's been very easy just to have a perceived loser or a company people aren't really sure about, and just use it as a funding source. 

For us, this is a very interesting set-up, as there are clear losers, and we've identified, as an example of that, a very clear loser is IT services, right? And so, these have sold off, but we think this will continue to play out over multiple years, and you know, we're not interested. It doesn't really fit our framework. But there are plenty of other kind of significant moats, significant companies out there, which have been using AI, enhancing what they're doing, enhancing industry position. We obviously continue to see how the landscape evolves and changes, and our degree of confidence around the enduring nature of these moats has actually increased. At the same time, the stock prices of these companies have actually gone down 20–25%. So, the set-up, the starting point, was these companies were cheap anyway given where EM has been. They've actually continued to grow their businesses this year. They've actually become better companies, and the stock price has gone down. So, for us, this is presenting the next era of stocks that we think will be substantial winners over a three to five-year view. 

Clearly, within what I've said, there's a big need to be selective. It's very stock-specific, very stock idiosyncratic. So, you're buying management teams as well as the industry analysis that goes within that. Even within an industry like e-commerce, which is viewed as potentially being disrupted from AI, there will be companies with weaker industry positions that potentially will be challenged. But for certain companies, where their requirements are ad tech, infrastructure, logistics, and many other things, they'll actually utilise AI to become winners. And so, it's very stock specific. It's industry specific. But there are a lot of interesting opportunities within here, and as these companies continue to deliver, and the earnings are coming through, and potentially you'll get earnings upgrades, I think this will force the market to pivot to these companies at some point, and there'll be a significant returns potential in these companies.

Gustavo Medeiros: That's great. Yes, for sure, this is one of the factors that I keep on mentioning. Every client I meet on the road, it's very hard to find the AI winners and losers, and it's very hard to time it. We're doing that 24/7. It's hard enough for us. Don't try to do it at home without the tools and the team and we have at our disposition, right? Dhiren, how are you seeing other tail risks and challenges? We've been discussing at our macro meetings the impact of higher fertiliser costs and potentially a harsh El Niño, which, the probability of a harsh El Niño is increasing over the last couple of months. How are you seeing these, and who are the most challenged in these scenarios?

Dhiren Shah: In terms of one of the areas that we look for in terms of risks within that macro framework, one of the things we see on the horizon to potentially change the macro landscape, is El Niño, as you say. So, the probabilities, and we track that data, are increasing, and it's potentially going to be a super El Niño, so the impact will be more significant. What this means, and I'll go into the details in a second, the setup beforehand was we get a number of EM central banks cutting. And I think potentially with El Niño and compounding what's happened, and hopefully it's transitory with Hormuz, but we'll have to wait and see, the inflationary pressures will potentially build. And this will be true for EM but then obviously DM as well, right? So, this is not just an EM-specific issue. 

What this means in terms of the harvest cycle, from soft commodities, but also causing bottlenecks in terms of hydrology, it's an important part of the energy matrix across a number of countries. And what that does to production? Does it cause shortages in particular goods, particular commodities? So, there are a number of different variables to consider, and there are particular countries, parts of Southeast Asia, parts of India, which we see potentially being more impacted as this works itself through over the next six to nine months. 

And there are other parts obviously to note which are broadly unimpacted: Mexico, Poland. And there are countries such as Argentina, which more than likely has quite a positive harvest cycle from this. These are some of the details we're thinking through and what it means in terms of the EM rate cycle and particular countries and sectors. We think it is broadly manageable, but it's certainly a headwind that wasn't one that we were thinking about or one the markets were thinking about, I'd say, three months ago.

Gustavo Medeiros: Great, thank you. Maybe we should switch gears a little bit to talk about China, another topic that was mentioned. Obviously, when you think about China and AI, it's pretty clear that Chinese enterprises have been working on adopting AI in a different way. They've been trying to use AI to optimise existing processes in manufacturing and other areas rather than this massive race towards AGI that you see in the West, right? China has advantages. It owns a very large choke-hold, particularly in strategic metals, that is likely to remain with us for at least two to three consecutive years, but it's also behind the cutting-edge chips. When we look at the Chinese macro, the consumption remains very soft. We've had some improvement on tier-one city real estate, but it's very, very specific there so far. And you know, you have still this dichotomy between very strong manufacturing and very strong, you know, supply chain advantages, a strong technology sector on paper but with a very weak macro. How do you square this structurally deflationary soft domestic economy with an equity market which is close to two-thirds of the index? And are you adding or trimming exposure to China here?

Dhiren Shah: I'll answer the last bit of the question first. We have been adding more recently in the last few months, and I'll go into detail why and what specifically. This obviously comes from the backdrop of the middle of last year, we've been reducing exposure as we saw the macro slowdown come through in China. And so there are few things to cover. First of all, I think China's showing the world its power. You know, this is a raw kind of manufacturing technology power, and with that, it gets a lot of control, and it puts it in a strong geopolitical position. But it's just something worth bearing in mind because it is meaningfully better, stronger, than where it was even, you know, two, four years ago. I think with that comes, in a way, less risks, but that's maybe for a separate discussion. 

In terms of the macro indicators, we're now into, as you say, a multi-year real estate down cycle. Yes, new starts are down 80%, 50% below reoccurring sales. So, when you have just a normalisation of the real estate cycle, a number of metrics will double in terms of year-on-year from an activity standpoint. So, there's a huge potential tailwind there, as you say, and we track this data very frequently. There are parts of the real estate market which have shown some life: rental markets, housing sales, primary, secondary, but we still need to see this play out for a few more months to get conviction that this has turned genuinely. 

This is also a function of the wealth effect. There are strong parts of the AI value chain that is creating wealth within China, as well as industrial profits, which have also improved. And so, as money is created, naturally over the time some of it will flow into real estate, where prices have been down for a long, long time. The consumption environment remains weak, and this is one of the headwinds we've had for a number of years. We've tagged it. We flagged it. It remains one of those large optionalities, right? So, the cash is there. Consumption is significantly below trend, but for it to materialise, you need that real estate cycle to turn. So, yes, this will happen at some point in the next three to five years. As you said, the questions are on the timing aspect, but we're not quite convinced we're there yet. We have some exposure in the portfolio, but we could certainly own a lot more when we see that particular turn coming. 

And then in terms of the large technology companies, let's not forget, four or five years ago and even two years ago, these were viewed as some of the biggest technology winners, but more recently they've kind of been forgotten, as I mentioned, funding sources, or the market's gone to frothy parts of AI within China on these new IPOs which are loss-making. And sounds great, but we're not really convinced they're going to be around in three years' time. And, yes, the dominant franchises which have the people, have the technology, have the brands, have the ecosystems, some of them are trading at record low multiples. So, it's quite extraordinary, the valuation opportunity you're getting. Just talking through large stocks; Tencent was cheaper than China Mobile, which is extraordinary. It's not something I ever thought would come through in markets. And these companies which generate... They grow very attractively, generate significant free cash flow, good governance, and so on. But that's just an interesting reference point. 

So, we do see a lot of opportunities within this space. There has been a lot of technical selling in China. I think this is something worth thinking about. And you know, China’s National Team of state backed investors started to prop up the market over a couple of years ago, and what they've done over the last three to four months is they've sold 80% of their exchange-traded fund (ETF)position. That's been a strong technical headwind for the market. Obviously, this is multiple billion dollars, and we should be coming to the end of that in the next couple of months. So, what we've seen is stock prices which have been under pressure, and so with that, as well as some new holdings, we've been adding to exposure within China. You know, it's also very strong from a macro standpoint, all the intellectual property (IP), the manufacturing, and a host of other metrics leave it quite well insulated to the wider global economy. And sentiment's also quite negative towards China, which is handy, so it's a good opportunity to be adding exposure.

Gustavo Medeiros: Absolutely. The 'National Team' is very important and relevant. It's so big that if you don't exclude the Chinese ETFs from the EM flow analysis, it completely distorts the last three years' analysis, right? It would suggest that there has been much more inflows to EM three years ago until late last year. And then in 2026, it would suggest that investors are panicking and selling EM, and that's all the National Team selling ETFs in China, essentially. So, yes, I think that's interesting indeed.

The final topic, Dhiren, that I have in terms questions before you open to the audience is in Latin America. Obviously, they had phenomenal performance in 2025, did quite well also in Q1 2026. There is this path, you know, politically. The whole region is switching from far-left, centre-left regimes towards far-right regimes, and it's not as straightforward. And if you speak to any Latin American, they're half depressed about that because of the polarisation and because not necessarily they like the person that is on the ticket, and there are issues, but with that comes better policy. Colombia, Brazil, Chile, Mexico are up about 30% over the last 12 months, but the index is down about 12% since mid-April, partially on the Hormuz shock and partially on positioning and maybe partially in Brazil as well, as it no longer has a clear market-friendly front-runner heading into October. How are we positioned in the region in that environment?

Dhiren Shah: So, in terms of LatAm, I guess let's start where there hasn't been any political change or where there is no electoral cycle in terms of Mexico. Obviously, clearly, the Mexican economy is significantly tied to the US, and they continue to gain market share as the North American trade block is favoured, or at least Mexico within that block is favoured. The challenge for Mexico has just quite simply been awaiting USMCA renewal, and it seems like it's difficult to know whether we're going to move to annual negotiations rather than like a 10-year one. But let's see. 

For now, USMCA is working. It's likely to continue working, but companies clearly seek a 10-year, 20-year visibility when they're making capex requirements. Having said that, Mexico has the lowest tariffs out of all trade partners, and so it remains a good place to invest. And the government is proving to be more pragmatic, and they've got a growth agenda, but the economy is growing below potential. If USMCA does get ratified, growth can go to 5–6%, but the timing on that is not exactly clear. So, Mexico for us is selective, it's interesting, but given the current operating environment, it is not as appealing as it was, say, last year. 

We look into the other parts of the region in, say, where we've had obviously change. Argentina, that continues to be an attractive place. We have selective exposures, likewise, in Chile. And these things are multi-year, right? So, when you have a government that's doing the right thing, it sets up the correct operating environment for companies to flourish in. And so, we see that in a number of places, making the ability for companies to operate better. Likewise, Chile. Having owned both these markets for some time, we have exposure there. 

In terms of the big one, which is essentially Brazil, one of your obviously natural favourites, Gustavo, it remains still complicated. So, the message that we had last time around, it still remains unclear. It's an electoral cycle in Brazil. These are fabulously difficult to predict and volatile. Brazil obviously also has a relatively large current account deficit. So, these are reasons for us to have exposure because there are great companies there. Valuations are two standard-deviations attractive. You have some incredible companies trading at very, very attractive multiples. We're ensuring we're not having significant exposure given some of these uncertainties and risks. We'll see where we are in six months' time, when obviously we'll have more clarity. But there are no reasons to have exposure given the valuations and given the bottom up, but not significantly.

Gustavo Medeiros: That's great. Thank you, Dhiren. I think that concludes our plan in terms of questions. Any questions from the audience, Steve?

Stephen Rudman: Gus and Dhiren, thank you. We had a couple, but I'll sum it up into one, which is probably the general gist and the most important part of all the building blocks you guys just laid out. Simply, with 2025 and year-to-date 2026, obviously some very strong performance relative to benchmark. How do we see valuations today versus one and three years ago? In other words, I think the question was, "Am I late?", or "Am I just buying at the top?" which nobody wants to do. So, from a valuation perspective, maybe there's some confidence we can give people that this is not necessarily the end but maybe still somewhere towards the beginning. I'll leave it to you.

Dhiren Shah: Sure, I'll start with a few points. I'm sure Gus will have many more. As you say, it's been an 18-month, two-year cycle after a 10-year down cycle, right? And so, I think it's important to frame a significant underperformance during a period when companies were delivering. Valuations are still cheap from an absolute standpoint and especially when you look under the bonnet within the index. So, there are parts of the market which have worked. We've talked about this, IT. But there are large parts of the market that haven't, where valuations are almost at near all-time lows. And so there is a significant valuation opportunity, and against that backdrop, EM earnings are growing, right? And so, it's not just the tech earnings, but there are a broad range of companies in EM which continue to deliver consistent earnings growth. And that provides that compounding of earnings, that compounding of returns, in the long term. So, we still see plenty of opportunities, see plenty of really interesting opportunities. Gus, is there anything you want to add yourself?

Gustavo Medeiros: I mean, just in very simple terms, if you buy EM stocks today, you're getting 12 times earnings on a forward-looking basis. If you look at the average of the PE over the last five years, it's been 12 times as well. So, we didn't go anywhere in terms of valuations. EM outperformance over the last 18–24 months is not because we've re-rated, and we're now expecting the earnings to improve significantly. Earnings have been improving for three years now, about 27% per year. Obviously, it's gone parabolic recently, and we're now having the highest level of earnings per share in EM in the 20–30-year history that we have data because of the memory cycle. 

But it's not only about Korea. It's not only about Taiwan. Earnings per share are expanding pretty much across all countries in Latin America over the last 12–18 months from a forward-looking basis, earnings per share increasing in China, increasing in India, increasing in South Africa. There are a broad number of countries and sectors delivering healthy earnings per share growth. Some of that is the macro. The Dollar stopped appreciating like crazy and had a little bit of a sell-off. There's going to be a lot of chop on that. But again, looking at the earnings in Dollars, if the Dollar keeps on a steady depreciation trend, or not so steady, depending on the US policies, then obviously that helps as well. Or this was a massive headwind that doesn't exist anymore, and if anything, it became a tailwind. So, I think that, yes, it's been a long time that all you wanted was to have an underweight in EM, overweight in the US in terms of asset allocation. I think over the last two years, that turned on its head, right? If you had an overweight in EM, and you've been more cautious on parts of the market that have been more expensive, you've done well, and I think that there's a long runway for that.

Stephen Rudman: Thank you both. I think that's a good spot to end, both on an optimistic note. So, with that said, thank you so much for your time today. To all attending, again, we always appreciate your time. For those of you who are investing with us, we certainly appreciate your trust. So, thank you for that. If you do have any remaining questions or want to chat, please reach out to your Ashmore representative, happy to spend as much time as necessary to get you the data and the information that will help you make decisions on behalf of your clients. 

Finally, we will be sending out an email with the replay. Feel free to share that with your colleagues that you think will find valuable. Additionally, Gus and team just wrote a nice piece called ‘The Emerging View’ all about AI as well. We'll make sure that you receive that as well with the replay. So, again, thank you all for attending. Hope all have a great beginning of the summer and enjoy the rest of it as we get to it. This will conclude today's webinar. Thank you.