Why the emerging markets trade is nowhere near its end
Naomi Waistell, co-fund manager of Carmignac Portfolio Emergents, argues the artificial intelligence (AI) build-out is not a single bottleneck but a series of them, with the opportunity migrating through cooling, power, connectivity and packaging as each constraint in turn hits its limit.
Is the AI trade getting too crowded?
Taiwan and South Korea now account for around half the MSCI Emerging Markets index after the share prices of the likes of SK Hynix and Samsung Electronics have each more than doubled over 20261 so far. The concentrated nature of that move has prompted a predictable question: is this trade getting too crowded?
Naomi Waistell, co-fund manager of the Carmignac Portfolio Emergents fund, recently attended the Computex technology conference in Taiwan, where Nvidia chief executive Jensen Huang delivered the keynote. What she observed there was a striking divergence in mood between Asian and European investors. "People in Asia have a mentality of: regardless of what happens, just buy," she said. "Investors in Europe are a bit more concerned."
Caution has a rational basis. Memory markets have been through this before: heavy capital expenditure commitments from manufacturers flood the market with supply; demand disappoints; prices fall sharply and stocks follow. The fear is that current AI enthusiasm is setting up a repeat of a cycle that has caught investors out more than once in the past.
AI build-out: a series of bottlenecks, not a single trade
However, Waistell argued: "This time is different. AI is not just one theme. Everything you need to support the whole AI infrastructure is in short supply. So even if memory becomes oversupplied, you will then need more cooling, because as compute power increases and requires more and more data centres, you need more cooling solutions."
The fund manager’s argument is that the AI build-out is not a single bottleneck but a series of them. As one constraint eases, another opens elsewhere in the infrastructure stack. Cooling systems hit capacity limits, then power supply, then connectivity chips, then back-end packaging. The opportunity migrates through the value chain rather than ending when any single component catches up with demand.
Two structural features reinforce the case for holding companies that have already risen sharply. The first is valuation. "If you look at the underlying numbers on the companies, their earnings growth is exceeding 200%2," Waistell said. "The price has gone up a lot, but it has followed at the same pace as earnings growth, so valuations are still reasonable."
The second is the nature of the commitments underpinning those earnings. "For the first time in history, these players are executing long-term agreements, which is good because it is providing good visibility," she explained.
SK Hynix and Nvidia have jointly committed to a co-investment arrangement, with Hynix chips built exclusively for Nvidia's AI hardware. That contractual certainty was absent from previous memory cycles and materially changes the risk profile of these positions.
Recycling capital across the AI value chain
Nearly half – 45%3 – of Carmignac Portfolio Emergents is invested in the AI build-out theme. Despite their stellar performance, most of these names still trade on attractive valuations. SK Hynix itself trades at a significant discount to US peer Micron on a P/E basis and EV/EBITDA basis.
What Waistell describes as the ‘Magnificent Three’ of emerging markets – TSMC, SK Hynix and Samsung, a counterpart to the Magnificent Seven that dominate US equity portfolios – accounts roughly for half of the tech exposure of the fund. The remainder of the Tech/AI exposure sits in less-followed parts of the value chain: liquid cooling systems, uninterruptible power supply hardware, co-packaged optics and back-end semiconductor packaging. Given this concentration, diversifying and managing our tech/AI exposure is key.
"What we do in our fund is try to concentrate on where the next needs are in the AI build-out — where you see shortages — and we try to reorient our investments accordingly," Waistell said. "We don't play AI through only one single theme or one memory name."
Managing a position that has more than doubled requires a clear-eyed approach to when to act. "All the way up, we have been trimming positions in some of these names and taking substantial profits," Waistell said. "We try to recycle that capital into other parts of the value chain, where more investment is needed to support the build-out, or even to other regions".
In practice, that has meant reducing SK Hynix after sharp appreciation and redeploying the capital into companies such as, semiconductor equipment suppliers across Asia . The fund has also moved into Chinese names more broadly, where the market that has lagged year to date and where Waistell and co-fund manager Xavier Hovasse see selective value.
"It is all about balance," Waistell said. "48% is AI build-out, but it is very diversified — Taiwan, Korea, China — and we are very mindful about position sizing, taking profits and being careful about valuations. What we look for is mispriced stocks. We have broad exposure but we target where we see mispricing. We believe this is how we can generate alpha in the portfolio."
Emerging markets: the new geography of innovation
The non-AI allocation provides further balance to the portfolio. CATL, the world's largest EV battery manufacturer and leading Energy storage solutions provider and a top-five holding, sits at the intersection of energy security and the energy transition, a theme the fund has been building since the Middle East conflict accelerated investment in renewables and storage infrastructure.
In Latin America, the fund holds Brazilian utility companies with long-term government contracts, offering yields in the 10-15% range4. These are defensive, bond-proxy positions designed to cushion the portfolio in a high interest rate environment in Brazil.
As for where this cycle ultimately leads, Waistell's view extends well beyond the current portfolio construction: "The next decade won't be like the last. In the last decade, emerging markets were the laggards, the followers. But in terms of the next phase of innovation, they are building it. Emerging markets are the new geography of innovation."
1Source: Bloomberg 15/06/2026.
2Source: Bloomberg, company data June 2026.
3Source: Carmignac, as of 31/05/2026.
4Source: Bloomberg, company data 2026.
Carmignac Portfolio Emergents
Carmignac Portfolio Emergents F USD Acc Hdg
- Recommended minimum investment horizon
- 5 years
- Risk indicator*
- 4/7
- SFDR - Fund Classification**
- Article 9
*Risk Scale from the KID (Key Information Document). Risk 1 does not mean a risk-free investment. This indicator may change over time. **Sustainable Finance Disclosure Regulation (SFDR) 2019/2088. The SFDR classification of the Funds may change over time.