Global market tilt towards AI

AI-driven gains are reshaping global markets, boosting emerging tech and shifting leadership, but rising concentration risks are increasing reliance on a few firms and challenging diversification across equities and fixed income.

The emerging markets technology sector has delivered truly outsized gains since the end of 2024, dwarfing broader emerging markets and even the strong advance in US technology. That strength has persisted despite tariff frictions in 2024 and the US-Iran war, underscoring the power of the AI-driven structural growth theme. We still see room for upside for the overall AI theme, but recognize that returns at this pace are unlikely to continue indefinitely.

Weekly highlights

Evolving market leadership and its portfolio implications

The impacts of artificial intelligence (AI) are not occurring in a vacuum. While much of the public discourse has focused on productivity gains, corporate earnings, and innovation within individual companies, the broader implications for investors are increasingly visible at the market and asset-allocation level. AI is not simply a sector story, it is reshaping the structure of global equity markets and, in doing so, altering the opportunity set and risk profile faced by allocators.

Coming into the year, we advocated exposure to AI-related growth both in the US and in emerging markets. While the US narrative is well understood—dominated by large technology platforms, capital expenditure cycles, and productivity optimism—some of the most profound changes tied to AI have occurred outside the US.

One of the clearest manifestations of this shift is visible in global equity index composition. Taiwan has become one of the largest country weights in global equity markets, recently surpassing the UK and Canada and now trading places with the UK between the third- and fourth-largest positions. Notably, it has also overtaken China, which only a few years ago represented more than 40% of emerging markets indices. This evolution is tightly linked to Taiwan’s central role in the global AI and semiconductor supply chain.

South Korea has undergone a similar transformation. Its weight in global benchmarks has risen meaningfully and is now larger than several major developed markets, including France, Switzerland, and Germany. These changes underscore an important point for allocators: AI is increasingly influencing country-level exposures, not just sector tilts or individual security selection.

For global investors, this evolution challenges long-held assumptions about geographic diversification. Countries historically viewed as “smaller” or “peripheral” in benchmark construction have become structurally significant drivers of global equity performance. As AI continues to shape capital flows, earnings growth, and equity valuations, these country-level reallocations are likely to persist.

Concentration risks: Systematic and idiosyncratic pressures converge

While shifts in country composition are notable, they are occurring alongside a broader rise in concentration risk across equity markets—both at the security level and through systematic exposure to the AI theme.

What stands out immediately is that concentration is rising across both developed and emerging market indices. In each case, the summed weight of the largest companies has moved steadily higher, illustrating that this is not a phenomenon isolated to one region or market classification. Equity performance is increasingly being driven by a small number of very large firms. 

That said, the issue is even more pronounced in emerging markets, where individual companies—most notably Taiwan Semiconductor Manufacturing Company (TSMC), which now represents 14.4% of the index—account for an outsized share of index weight.

As a result, returns could be increasingly sensitive to company-specific outcomes rather than broad macro or country-level dynamics.

This presents a challenge for active managers. Many operate under position-size constraints or risk guidelines that limit exposure to a single issuer. At the same time, managers are rarely inclined to materially overweight securities that are already extreme contributors to benchmark performance. This combination can create persistent tracking error and raises the risk of structural underperformance if leadership remains narrowly concentrated.

From an investment-risk perspective, rising idiosyncratic risk is an important consideration. When performance is driven by a handful of companies, firm-specific events—earnings surprises, regulatory actions, geopolitical developments, or supply-chain disruptions—can have an outsized impact on index-level outcomes. This dynamic is increasingly present across both developed and emerging markets.

At the same time, concentration is not purely a bottom-up issue. From a systematic standpoint, equity indices are becoming structurally tilted toward growth and AI-related exposures. Capitalization-weighted indices, by design, allocate more weight to companies and sectors that have already appreciated. As the AI theme continues to dominate earnings expectations and investor flows, benchmark exposure becomes increasingly correlated to that single growth profile.

These dynamics can be beneficial when markets are trending and leadership remains intact. However, history suggests that environments characterized by extreme concentration can also amplify drawdowns when sentiment shifts or growth expectations are challenged. As a result, allocators should consider not only how much AI exposure they have, but how concentrated that exposure has become, explicitly and implicitly, across portfolios.

The renewed case for diversification

Against this backdrop, the question naturally becomes: Where does diversification come from today?

Within equities, the simplest answer is also the most uncomfortable—diversification increasingly means being underweight the dominant AI and mega-cap concentrations, and overweighting areas that have lagged or are less central to the theme. While such positioning can be difficult in a momentum-driven market, it remains one of the few ways to reduce equity-specific concentration risk.

But equities are only one part of the equation. Traditionally, bonds have served as the primary diversifier for equity risk. That role has become far less reliable.

Over recent decades, correlations between equities and both Treasuries and corporate credit have risen to their highs. In environments where inflation volatility is elevated and policy uncertainty remains high, bonds may no longer provide the consistent ballast investors became accustomed to during the 2010s.

During that period, structurally low inflation allowed the Fed to focus on the growth side of its mandate, enabling bonds to absorb equity shocks effectively. Today, deglobalization forces, supply-side constraints, and persistent inflation risks suggest that investors cannot rely on bonds to “do the heavy lifting” for diversification in the same way.

As diversification within traditional public markets diminishes—both across equities and between stocks and bonds—investors increasingly need to look elsewhere. Real assets stand out as one potential diversifier, offering exposure to inflation-sensitive cash flows and different underlying economic drivers.

Beyond real assets, a broader toolkit is becoming more relevant. These include liquid alternative strategies, regime-aware approaches designed to perform across varying macro environments, and derivative-based strategies that reshape risk and return profiles without relying solely on directional equity or fixed income exposure. While implementations vary, the common theme is reducing dependency on a narrow set of growth and capital market assumptions.

The key takeaway for long-term investors is straightforward: The need for diversification has always existed, but it becomes most critical when concentrations—both systematic and idiosyncratic—are high. Today’s market structure reflects exactly that condition. Moreover, the ways investors historically achieved diversification may no longer function as expected, as correlations and market relationships continue to evolve.