Yes. Investors should reassess the role of emerging markets (EM) equities because benchmark concentration has reduced diversification benefits and increased exposure to the same artificial intelligence (AI)–driven forces leading developed markets. That does not weaken the case for EM. It changes it. Investors should be more deliberate about what role they want EM to play and how they gain that exposure. Active management and manager selection remain as important as ever.
Historically, EM has helped diversify developed markets (DM) portfolios by offering exposure to different countries, sectors, currencies, and growth drivers. That remains true in part, but less so at the benchmark level. EM indexes have become more concentrated, with a small group of Asian AI semiconductor companies now accounting for a meaningful share of the index. TSMC, Samsung Electronics, and SK Hynix represented about 28% of the MSCI Emerging Markets Index by weight, double their end-2024 share.[1]Data are as of 31 July 2026. Those three companies have contributed roughly 69% of MSCI EM returns year-to-date. Similarly, the ten largest contributors to S&P 500 Index returns were AI-linked, accounting for roughly 64% of index returns. This dynamic has helped drive Taiwan and South Korea to account for nearly half of the MSCI EM Index (47%), more than the two largest EM economies, China and India, combined (33%).[2]These differences also vary by index provider: FTSE classifies South Korea as a developed market, leading to meaningfully different country weights and performance than MSCI. Broad EM exposure therefore increasingly provides another route into the same AI buildout theme that has shaped DM leadership, rather than a clearly distinct set of return drivers or proportional exposure to EM economies. High index concentration, combined with strong retail buying and leveraged positioning, have exposed investors to greater volatility.
These developments make implementation especially important. Active management has long mattered in EM because the universe is broad and heterogeneous, spanning very different governance standards, sector mixes, currencies, and policy regimes. Today, that challenge also includes benchmark concentration, rapidly changing AI competitive dynamics, and uncertainty over which parts of the AI stack will capture durable value. Just as important, not all active exposure is equally differentiated from the benchmark, as many managers also hold significant positions in the AI hardware names dominating the benchmark. Active management alone is insufficient to ensure thoughtful implementation. Investors must assess whether an approach truly offers the exposure they seek, including broader diversification or selective AI exposure, and improves their chances of capturing future winners.
That selectivity matters even more when recent winners already reflect elevated expectations. AI semiconductor demand remains robust, and supply remains tight in some EM-relevant parts of the supply chain, particularly advanced foundry capacity and AI memory. That has supported extraordinary earnings growth and strong recent share price gains. But rising expectations for future earnings suggest much of the upside is already priced in, making it harder for even strong companies to keep outperforming. At the same time, extraordinary profits can attract new competition in a rapidly evolving AI landscape, potentially eroding the economics of today’s winners. For investors, the challenge is distinguishing between businesses where upside is still underappreciated and those where much of the good news is already in the price.
Memory is the clearest example of why this matters. Samsung and SK Hynix are major EM exposures and beneficiaries of current AI demand, but memory remains a cyclical industry in which strong economics have historically attracted new capacity. It may therefore prove less durable than parts of the AI buildout chain that benefit from more structural scarcity, such as access to power, grid capacity, and other physical infrastructure bottlenecks. Strong current profits in memory may be less defensible over a multi-year horizon than today’s earnings suggest.
Even if AI continues as a major driver of markets, leadership is already beginning to extend beyond today’s chip leaders. Other beneficiaries of the AI buildout, including infrastructure providers and suppliers of enabling technologies, as well as end users that adopt AI most effectively, may capture a greater share of future gains. That broadening may not align with current benchmark exposures and is likely to further reshape relative opportunities across EM regions. We recently reflected part of these shifting dynamics by unwinding a tactical position favouring Latin American equities relative to broader emerging markets, with Latin America’s lower exposure to the AI buildout among the factors weighing on the trade’s forward return potential.
This is not a case for abandoning EM. It is a case for being more explicit about what role EM should play in the portfolio and how that exposure is implemented. If the goal is diversification, broad passive EM may deliver less of it than many investors assume. If the goal is AI exposure, investors should recognise that they may already be getting a concentrated version of it. And if the goal is alpha, selectivity still matters. EM still plays an important role in portfolios, but increasingly the case is less about broad beta than about clearly defining EM’s role and choosing the right exposure through thoughtful active management.
Max English, Investment Director, Capital Markets Research
Footnotes