• US equities have consistently outperformed global peers, cementing their dominance in portfolios worldwide. But past outperformance is becoming less bankable, especially when policy risk, concentration and valuation premia are taken into account. Overall, US equities account for nearly two-thirds of global portfolios, with the rise of tech giants ramping up concentration of market capitalization: The Top 6 tech companies now account for one-third of the S&P 500. US equities also enjoy a 50% valuation premium that reflects past dynamics rather than ongoing trends of political, economic and financial fragmentation, amid rising concerns about Federal Reserve independence and policy uncertainty. Looking ahead, while the AI boost is already visible  the US economy’s growth potential is likely to remain capped well below +3% (close to historical average) by 2030 because of rapidly deteriorating demographics. 
• Even a small move away from US dominance – in terms of returns and correlations – could prompt significant repositioning. Fragmented financial markets mean the traditional risk-return portfolio will need to be balanced with downside protection (CvaR). We find that if history repeats itself, Sharpe ratios favor the US, but in an "end of US exceptionalism" scenario, downside risk is minimized when allocating significantly less to the US and a bigger share to Europe and Pacific, reflecting their lower standalone volatility and friendlier cross-correlations. [AH1] [TT2] In fact, all of our models suggest higher resilience when US allocations are at least 10% below market weight, balanced by particular increases in Europe and the Pacific.
• Beyond pure regional allocation away from the US and AI, robustness can be added without dramatically impacting performance. Emerging markets also offer some relevant solutions and we see three options depending on desired portfolio construction choice: 1) markets with low tech exposure and correlation to US stocks (value call), 2) markets with strong exposure to high added-valued segments of the AI supply-chain (value/growth mix call) or 3) rebalancing positions from mature firms to early-stage ones in terms of AI expansion (growth call).
• International portfolios are always exposed to currency risk. While hedging has historically reduced returns more than volatility, recent events show that tactical hedging can be essential as rapid currency shifts may outweigh hedging costs. Active currency management should be considered more often in a more political market.

US growth exceptionalism continues but risks are rising

The AI buildout is already visible in US GDP data. ICT industries, which encompass both manufacturing (mostly computers and semiconductors) and services (software and data processing, computer systems design) contributed to around 50% of US GDP growth in 2025, close to double the average contribution during the past decade. Investments in computer hardware and data centers are growing very strongly, while upticks in software and R&D investment have been more modest. But since those components are much bigger, they are still contributing almost as much to overall GDP growth. Most of the ICT growth is driven by the AI revolution gaining traction.

Historically US GDP growth has averaged +2.4-2.5% per annum over the past 35 years. While actual GDP growth fluctuates with business cycles, underlying potential GDP growth is determined by four supply-side components: the working-age population, the employment rate, hours per worker and labor productivity (i.e. output per hour worked). The working-age population and labor productivity are the two main growth drivers by far. General purpose technologies, like steam, electricity or more recently desktop computers and the internet  boost the overall economy’s trend productivity growth, boosting potential GDP growth and, as a result, actual GDP growth. Many other factors are also at play, including innovation, the degree of competition in markets, the skills of the labor force and corporates’ organization. In modern history, the US economy has boasted the highest potential growth rate of all large developed markets because US companies are generally at the technological frontier, and US firms in others sectors are quicker to absorb and adopt technological changes than in other markets. US GDP growth averaged +2.4-2.5% per annum over the past 35 years. Over the past couple of years, potential and actual GDP growth has hovered around its historical average; it was lower during the post-GFC recovery (less than +2%) and higher during the internet revolution from the mid-1990s to the early 2000s (above +3%). We use this supply-side framework to project the most likely path of US potential growth through 2030.

We expect US growth potential to remain roughly steady, hovering at around 2.4-2.8%, below the 3% witnessed during the internet revolution, with AI benefits being offset by worsening demographics. US growth is currently at a crossroads, facing both strong headwinds and tailwinds. The most significant headwind is tight immigration policy, which is increasingly weighing on the growth of the working-age population (and the labor force): We expect it to slow down sharply, from +0.7% in 2023-25 to only +0.1% by 2028. On the other hand, the AI revolution is clearly gaining steam, which should increasingly support the overall economy’s trend labor productivity in coming years – beyond the ICT sectors at the forefront of the AI rollout, which are already experiencing a productivity boom – through a combination of higher investment spending leading to a higher stock of capital per worker (capital deepening) and increased Total Factor Productivity (TFP) growth, i.e. more efficiency in production processes. We find that trend productivity growth (i.e. productivity corrected for cyclical influences) has already accelerated over the past couple of years, to above +2% in 2025, supported by capital deepening. The OECD estimates that, under a medium adoption scenario, AI could boost US labor productivity by 1pp per year over the next 10 years. In our baseline scenario, we expect a relatively moderate pick-up of trend productivity, from +2.1% last year to +2.4% by 2030 – driven essentially by capital deepening (investment). The US stands out as the economy that would benefit the most from AI adoption. In particular, the US has a stronger specialization in highly AI-exposed knowledge intensive services such as finance and ICT, and more widespread adoption. The productivity and potential growth acceleration could be faster, fueled by corporate tax cuts rolled out recently by the Trump Administration, which are pushing down the cost of capital.  However, the effect of tariffs on (potential and actual) GDP could become increasingly large over time. In particular, non-AI capex has been struggling recently in the face of policy uncertainty and tariffs. In all, we remain positive but cautious on US growth prospects: wesee US potential growth accelerating just a bit by 2030 to +2.8%, but to be weaker than during the 1990s internet revolution because of much less dynamic labor input growth and less dynamic TFP growth.

Private sector investment exuberance is not an issue at this stage. General purpose technologies typically trigger exuberance and ‘’FOMO’’ behaviors in financial markets. But despite warnings of a debt-fueled AI boom, we do not see signs of debt-fueled investment at the aggregate level. At +2.5%, total non-residential (business) investment is close to its sustainable trend level and much lower than it was in past episodes. For example, during the dot-com bubble, debt-fueled investment crashed in the early 2000s, contributing to the 2001 recession. Today, whilst pockets of vulnerabilities may be emerging, they are confined to some sectors, mostly in the AI space. Besides, the experience from the 1990s shows that a recession does not necessarily knock out productivity gains: following the 2001 recession, labor productivity continued to power ahead.       

US policy and global fragmentation cloud the outlook for foreign investors

The US economy and financial markets have delivered superior growth to US investors but at a risky level in allocation and valuation. US equities have consistently outperformed global peers, cementing their dominance in institutional portfolios worldwide. The rise of the “Top 6” tech giants (Apple, Microsoft, Alphabet, Amazon, Meta and Nvidia) has added to this, leading to a significant concentration of market capitalization within US and global indices. The “Top 6” now make up one-third of the S&P 500, and US equities make up almost two-thirds of global market capitalization. AI investments may drive US economic growth but dilute returns for US megacap tech companies where capex has become a key concern. The US valuation premium rests on the pillars of high Return on Equity (RoE) as well as free capital markets. But the US RoE advantage has started to retreat from its highs, and the capex-spree of major tech companies may dilute returns further.

The US’s historic valuation premium may be put into question as global financial markets undergo increasing fragmentation and foreign investors become even more attuned to escalating risks. Concerns are mounting regarding the Federal Reserve’s independence and the emergence of regulatory interventions that could be viewed as repressive. Recent policy proposals – including imposing ceilings on credit-card interest rates for banks, limiting dividend distributions in the defense sector despite robust profits and encouraging oil companies to pursue investments in Venezuela – have contributed to an environment marked by regulatory uncertainty. This unpredictability poses a tangible threat to future corporate profitability and investor returns. Geopolitical tensions are also finding more direct pathways into financial markets, with debates around European investors potentially reducing their exposure to US assets as a form of retaliatory response and “Section 899”, a proposed US tax targeting foreign holders of US assets. Although this measure was ultimately withdrawn following Congressional opposition, its possible re-emergence remains a salient risk should transatlantic tensions resurface. This context suggests that the global financial architecture long dominated by the US is now shifting towards a more fragmented, multipolar order. In this new landscape, prudent risk management and dynamic allocation have never been more critical for safeguarding returns and ensuring resilience. Investors need to confront the risks of political, economic and financial separation not only via tactical positioning but also in their structural portfolio construction.

Varying macro regimes and dynamics favor different regions and make a strong case for building breadth into regional allocations. Different equity markets respond to different macro engines: Emerging markets (EMs), for instance, tend to benefit from oil price increases when these reflect stronger global demand rather than supply disruptions – as in 2003–2007’s commodity upswing amid USD weakness and easy global liquidity – and they also react positively to declines in US interest rates, which ease financial conditions and support capital flows. In fact, emerging markets had a significant higher return within this period compared to the US (30% vs 6% annualized), with only moderately higher volatility, effectively leading to a four times better Sharpe ratio. Developed markets (DMs), by contrast, load more visibly on global growth factors as their corporate sectors are more globally integrated and therefore more sensitive to shifts in global GDP – notably during the 2014-2017 synchronized upswing, when DM cyclicals and exporters rallied, especially in the US with double returns, lower volatility and hence a four times higher Sharpe ratio (5 vs 1.5 for EM). This heterogeneity in macro exposures is clearly reflected in long-term correlations (e.g. US-EM and Europe-Asia with each only 66%) and creates a multi‑engine return structure, where regions do not rise and fall for the same reasons at the same time. A portfolio that draws deliberately on these differentiated exposures naturally smooths its return path, because weakness in one macro regime is often offset by strength in another.

The impact of a fragmented world on portfolios

Market-cap weighting is not a neutral stance for asset owners. It embeds a concentrated US bet, so a robustness lens argues for more balanced regional equity allocations. Portfolio construction is dependent on return expectations that are often anchored in historical data reflecting US supremacy and a US-dominated financial market. The message for allocators is not that US leadership vanishes, but that yesterday’s outperformance is less bankable – especially when policy risk, concentration and valuation premia are taken into account. Today’s market-cap structure embeds a heavy tilt toward the US – and therefore a concentrated bet on a single macro regime in which US-led growth, profitability, dollar strength and large-cap tech dominance persist. Recognizing the diversity of macro exposures across regions makes clear that market-cap weighting is itself an active stance. There is a structural sensitivity of optimal regional equity allocations to estimation inputs and portfolio targets, where small deviations can cause large jumps in the ‘optimal’ mix. With macro uncertainty high, structural allocations should be more robust to reflect macro resilience. The current policy and market regime makes any single ‘optimal’ regional equity mix unusually fragile, challenging assumptions that underpinned the past decade’s US-centric portfolio construction and inviting a more robust approach than relying on past data. Optimal portfolio allocations should therefore be interpreted through a robustness-oriented lens of different metrics and scenarios rather than as a singular solution.

To understand the impact of a fragmented world on portfolios, we perform portfolio optimization across potential regimes. We acknowledge how quickly an optimizer can ‘jump’ when assumptions shift and aim for more robust guidance across regimes. For this reason we focus not on a single optimal portfolio but the top 10. We focus on regional equities – US, Europe, Pacific and EMs – which are most responsive to macro shifts. We run 50,000 Monte Carlo portfolios under distinct scenarios and evaluate them using two complementary indicators that reflect a fragmenting world. In addition to traditional optimization metrics such as Return and Risk, typically measured by Sharpe Ratios, we include Conditional Value at Risk (CVaR) to integrate increasing risk aversion in a more fragmented, high-risk environment. Both objectives guide our optimization. Our emphasis is on long-term regional allocations that are robust to fragmentation, while short-term developments are handled through tactical positioning. For a comprehensive assessment of robust asset allocation, we apply this two-engine portfolio construction on multiple scenario simulations, ranging from “history repeats itself” and “end of exceptionalism” to scenarios where major regions individually excel and achieve 1% outperformance. Table 4 presents a heatmap illustrating the allocation variations and related outcomes.

Sharpe favors the US if history repeats itself, but CVaR points to Europe and Pacific as drawdown anchors when tail risks cluster. In a ‘history repeats itself’ scenario, the Sharpe ratio captures the historical reward-for-risk structure and therefore points toward regions that have delivered persistent carry – most notably the US, with EMs as the preferred diversification mate (US–EM correlation about 66%). By contrast, Conditional Value at Risk (CVaR) focuses on the average loss in the worst 5% of scenarios and highlights where tail risks concentrate when correlations spike or shocks are asymmetric. Using both indicators is essential in a world where return distributions diverge: Sharpe shows what works if history repeats itself, while CVaR shows what survives if markets struggle. This dual-lens prevents regime myopia and produces allocations that remain functional across both continuity and fragmentation. Downside risk is minimized when allocating a bigger share to Europe and Pacific, reflecting their lower standalone volatility and friendlier cross-correlations; EMs are nearly absent and the US weight shrinks materially. The lesson is straightforward: mean-variance alone will overemphasize the past ‘carry’ of US equities, while a tail-aware lens reveals the stabilizing role of Europe and Pacific when stress concentrates. A less US-dominated, multi-polar world may further lower US correlations, reinforcing this picture.

When US exceptionalism is stripped out as possible scenario in a fragmented world, volatility and correlations dominate, shifting weights toward Europe and Pacific. A plausible fragmentation scenario is that the US growth/return advantage softens while correlations remain elevated given its scale. We remove the historical return dispersion that favored the US – our ‘end of US exceptionalism’ scenario. The optimizer behaves as intuition suggests: volatility and correlation dominate, pulling the Sharpe solution toward the CVaR solution and shifting weight into Europe and Pacific while penalizing higher volatility in the US and EMs. This mirrors practice whenever valuation discipline or uncertainty compresses return differentials: the risk manager takes the wheel.

Regime changes can swing ‘optimal’ weights dramatically, so bands and scenario-averaging are preferable to point solutions. Tactical positioning gains importance but should not impair structural calibration. In ‘conviction regimes’, we rotate a modest +1% annual return edge among regions – well within forecast error or near-term dispersion. Even such incremental edges trigger large reallocations and dominance of the favored region (often 50–60% for US, Europe or Pacific). Interestingly, a +1% edge in EMs does not justify a significant rotation or even a balanced 25% share given its higher volatility. Conversely, a -1% shock to expected US returns can drive its weight toward only 14–9% despite its long leadership. Hence we read any single ‘optimal’ output through a robustness lens and prefer predefined bands to point estimates; we report the average of the 10 most optimal portfolios within a narrow Sharpe/CVaR window (e.g., ±0.01).

Fragmentation lowers correlations and lifts diversification value. With deeper fragmentation, correlations have already shifted over the past two years – most notably a very low Europe–EM dependence and, in general, lower EM correlations across regions. Combining these shifts with baseline mid-term returns (slightly higher EM, slightly lower Pacific) yields a fairly balanced, near-equal-weight optimum with a clearly higher allocation to Europe; under CVaR, the tilt to Europe is even stronger. The sweet spot for robustness is a diversified core rather than extremes; a market-weighted global portfolio is less suitable when the US separates and European sovereignty rises alongside China’s influence.

For resilient long-term strategy allocation, our combined Sharpe and CVaR lens points to capping the US below 54% and lifting Europe/Asia to about 40%, delivering efficiency with stronger drawdown protection. Across scenarios, the US allocation sits well below the ~66% market weight; even the highest is ~55% if based on past market data, more than 10pps lower. The single best allocation (by combined Sharpe and CVaR attractiveness, via Z-score) appears in the ‘+1% US outperformance’ Sharpe-optimal portfolio, capping the US at ~54% and allocating ~40% to Europe and Asia as risk diversifiers. Although Sharpe/CVaR are calculated from historical inputs that allocation is robust to a significant smaller US exceptionalism than history shows (only 1% assumed vs more than 2.5% vs Europe and Asia historically). Sharpe is ~0.44 (upper-range) and CVaR the lowest recorded outside direct CVaR optimization, which otherwise compresses returns. The heatmap shows that Europe and Asia Pacific receive consistently a higher allocation than the market portfolio across scenarios, providing robustness across various outcomes and especially adding downside cushion. Obviously each strategy would excel in the scenario it is optimized for but even under past returns, equities would earn a satisfactory risk-premium and Sharpe ratio.

Given our near-term forecasts, a balanced portfolio need not lag a US-centric one. Including our baseline forecasts and 2024–25 correlations shows a more balanced portfolio need not be inferior to a US-centric one: Sharpe, volatility and CVaR are close to on par. This argues for higher allocations to Europe (lower volatility, improved cross-regional diversification, balanced macro exposures) and a modestly higher Pacific weight for diversification vs. the US cycle. The US can remain a large position but below market, maintaining exposure to the productivity impulse while acknowledging concentration, valuation and policy-risk premia. Finally, EMs offer oil/rates exposures that complement Europe and enhance total-portfolio diversification; one balanced variant overweights US and Europe vs. equal-weight, financed by Pacific and EM, and ranks third by combined Sharpe/CVaR across scenarios.

In rising policy stress, let CVaR set the course. If policy escalation intensifies and the financial channel remains in play, we would keep European allocation more pronounced, rotate more risk into Pacific as the cleaner diversifier, widen the US underweight and bias EM toward higher-quality Asia while watching the dollar and US rates (see next sections). In this scenario, we would focus on drawdown control (CVaR): the CVaR-optimal portfolio is notably stable across shocks (historic performance, equalized performance, regional outperformance and persistent short-term trends), with Europe and Pacific averaging around three quarters of the portfolio.

Opportunities beyond the US: diversification and risk-return 

Dispelling the underperformer tag. Over the past 15 years, emerging-market (EM) equities have significantly underperformed US equities, a trend driven by a confluence of macro and structural factors rather than any single cause. Since around 2010, the MSCI Emerging Markets Index lagged developed markets – particularly the US – by a wide margin. Commodity price depression was a powerful downward catalyst in the 2010s for resource-heavy EMs like Brazil, Russia and South Africa, which saw earnings and export receipts come under pressure when notably oil and base metals prices slid, weakening equity performance relative to more diversified US firms. Additionally, higher volatility and risk premiums in EMs – amplified by episodes of currency stress and regional policy uncertainty – discouraged capital inflows despite attractive yield offered by this asset class. Investors preferred parking their money into markets perceived as safer and more liquid like the US. More recently, the rise of US rates following the post-Covid expansion period turned out a double burden for EMs, notably those with a part of their debt labelled in USD, which saw debt-service costs increase, but also those whose yield-related attractivity was challenged by the US. More recently, hopes that rising technology exposure within EMs (like in the US, technology is by far the largest sector in the MSCI EM index benchmark, see figure 6) might shift the narrative have not materially changed the broader picture: Foreign investors have so far preferred mature US tech and hyperscaler growth stocks over early-stage “copy-cat” tech plays in EMs, whose earnings trajectories and governance standards remain less certain. This preference was reflected in the decade of outperformance by the US S&P 500, buttressed by mega-cap tech giants, compared with restrained returns and valuation discounts across much of EMs. However, today investors are humming a new song and looking for relatively safe but also profitable alternative to hedge rising concerns over likely AI-driven bubble in the US.

Strategy #1 – Hedge US by picking up opportunities out of tech segment. Historically, EM equities have traded at a significant valuation discount versus the US, with the EM index often around ~40 % below US multiples over the past decade – reflecting persistent growth and quality differentials. From an investment strategy perspective, this structural backdrop suggests that investors concerned about rising US tech volatility and concentration risk might consider rebalancing away from pure growth-tech beta toward assets with lower tech exposure and diversifying correlations. For example, reallocating a portion of equity exposure into emerging markets with smaller technology footprints and sector profiles more oriented toward financials, industrials or commodities can help shrink portfolio beta without fully sacrificing return potential. Markets such as Malaysia – with its large banking, telecom and consumer sectors and a more modest tech representation relative to the US – offer broad EM exposure at historically cheaper valuations (MSCI Malaysia constituents cover roughly 85% of the local market at subdued multiples). Similarly, Turkey’s BIST 100 and Saudi Arabia’s Tadawul feature indices dominated by industrials, banks and energy/resources rather than high-growth technology, providing a complementary profile with historically low correlation versus US tech-centric indices. In practice, blending these low-tech, value-oriented EM exposures with traditional US equities can act as a partial hedge against tech-led drawdowns while maintaining long-term alpha opportunities through relative value and diversification.

Strategy #2 – Play tech call out of US by focusing on added-value part of the AI chain in Asia. Historically, both South Korea and Taiwan have offered compelling ways to hedge US tech volatility by anchoring equity exposure in markets with real structural links to the AI and semiconductors value chain while trading at discounts relative to US tech multiples (-15% and -30%, respectively). South Korean and Taiwanese equities benefit from strategic positions in memory and foundry segments crucial for global AI build-outs – with giants like SK Hynix and Samsung driving AI-oriented memory chip revenues (e.g. high-bandwidth memory used in AI data centers) and TSMC dominating advanced logic chip fabrication for major global clients. However, Taiwan’s market, where almost 80 % of market cap has tech exposure and around 60% has revenue growth driven by AI-capex cycle, looks less compelling than its South Korean peer due to its hyper-concentration but also its already high valuation (PE at 24.5x) – in absolute and relative terms against the US – and the fragile geopolitical situation between China and Taiwan. Nevertheless, the structural advantages imply that tilting part of a growth-oriented portfolio toward South Korean and Taiwanese tech makes strategic sense as a partial hedge against US tech drawdowns, particularly if AI capex momentum persists but US multiples reset lower. However, South Korea’s relatively broader industrial base and slightly cheaper valuations can offer better risk-adjusted entry points while still benefiting from AI demand narratives. In terms of growth drivers, the tremendous increase of memory chip prices expected this year due to shortage risks should deliver a strong boost to manufacturers, while the rising competition on the wafer foundry business, with ongoing major investment made in new capacity in the US and China, could hammer growth potential and challenge the historical lead of Taiwanese foundries.  In contrast, the Hong Kong market offers exposure to AI-related growth without excessive concentration – its tech sector trades at historically lower PE levels than US counterparts but still supports earnings growth, making it an attractive diversification play if US tech volatility rises. By blending allocations to South Korea, Taiwan and Hong Kong, investors can maintain exposure to AI-driven revenue stories in alternative geographies, potentially reducing overall portfolio beta to US tech while capturing diversified alpha from different stages of the semiconductor and AI value chain.

Strategy #3 – Looking for new AI growth potential and targeting rising stars. In an investment context where US tech volatility risk is rising, one strategic hedge is to diversify into markets with high AI growth potential driven by active start-up ecosystems and strong state support, particularly China, which is positioning itself as a formidable alternative AI hub. Unlike the capital-intensive frontier model dominant in the US, China’s AI strategy emphasizes cost-efficient, open-source and broadly deployable frameworks – a narrative supported by rapid adoption of models like DeepSeek’s R1 and Alibaba’s Qwen series, which are being integrated across government, healthcare and enterprise applications domestically, and increasingly abroad in cost-sensitive regions. This state-backed momentum is further reflected in robust start-up financing and policy support: Cities like Hangzhou are investing billions in AI innovation initiatives and computing infrastructure, while exchanges such as Hong Kong have seen a surge of AI-related IPOs, including firms like MiniMax Group and memory interconnect specialist Montage Technology, which have delivered strong first-day returns, underscoring investor appetite for China’s AI ecosystem. China’s efforts to reduce external chip dependence – such as expanding SMIC’s advanced manufacturing capacity – and the proliferation of LLM developers like Z.ai and 01.AI also suggest a deepening value chain beyond software into infrastructure. However, this strategy carries valuation and execution risks: Chinese tech valuations, exemplified by the STAR 50 index, whose median valuation is twice higher than its US tech-led Nasdaq 100 peer (PE at 69x vs. 35x), imply high expectations that may be vulnerable if growth slows or geopolitics interfere. For investors seeking diversification and a hedge against concentrated US tech risk, selective exposure to China’s AI start-up boom and state-supported innovation platforms can offer access to a different growth archetype – one rooted in broad adoption, cost advantage and emerging global niches – while acknowledging the heightened valuation and political risks inherent in that market. In any case, in a fragmented world, it will be exactly about carefully selecting such themes and diversifying between regions with a tendency to cap high US exposure.

Does strategic FX hedging pay off for internationally diversified portfolios in times of geopolitical turmoil?

A key pillar of cross-border investment management is integrating a conscious view on currencies and their hedging – even more so in a fragmented world. Major currencies have seen sharp moves in recent months, including the USD after “Liberation Day” and the JPY on loser fiscal spending. The ECB is also analyzing how to boost the Euro’s role in the global financial system, while the Chinese government has already made its ambition very clear. One aim of the plan is to absorb any shocks if the US — which has backstopped the global financial system with dollars for decades — suddenly decides not to, or attaches unacceptable conditions to its support. The other goal is to underpin its foreign trade more actively and, ultimately, grab some of the benefits that the US has historically enjoyed from controlling the world’s reserve currency. An environment characterized by rapidly changing geopolitical conditions also makes the question of strategic foreign exchange hedging increasingly relevant. The central question is whether FX hedging can improve portfolio outcomes and to what extent its effectiveness depends on the investment horizon.

International equity investments are often held without currency hedging for multiple reasons. This reflected the view that exchange-rate fluctuations provided diversification benefits and were not systematically priced risks in equity portfolios. Moreover, exchange rates were widely assumed to mean revert over long horizons, rendering currency risk largely transitory for long-term investors. Consistent with this perspective, Warren Buffett has repeatedly emphasized that long-term investment outcomes are driven by underlying business fundamentals rather than short-term currency movements, reinforcing the historical tendency to treat FX exposure as a secondary consideration. Accounting treatments may also disincentivize currency hedging for equities as it would rather add volatility to the income statement than smoothen results.

We review this question in current context via a cross-country analysis comparing hedged and unhedged equity portfolios across EMs, Japan, the US and Europe from 2015 to 2025. Two model portfolios were constructed using equal regional weights in order to neutralize allocation effects and isolate the impact of currency movements. FX exposures were incorporated symmetrically across regions so that each currency contributed equally to portfolio risk, allowing for a clear assessment of exchange rate effects within a diversified portfolio.

Over the full sample period, FX hedging materially reduced portfolio volatility but at the cost of lower returns. The unhedged portfolio generated an annualized return of 6.3%, compared with 4.5% for the hedged portfolio. Volatility declined from 14.1% in the unhedged portfolio to 9.5% in the hedged portfolio. Despite this reduction in risk, risk-adjusted performance remained broadly similar, with return-to-volatility ratios of 0.45 and 0.47, respectively. Figure 12 (see pdf) illustrates these dynamics by showing the indexed performance of both portfolios over time. While the hedged portfolio exhibits a smoother return profile, the unhedged portfolio maintains a persistent performance advantage over the long run, indicating that FX hedging has historically acted as a structural drag on returns rather than a source of excess performance.

2025 shows a different regime where currency hedging improved both returns and volatility. The long-standing relationship reversed in 2025, a period characterized by elevated geopolitical stress and pronounced currency dislocations. During this year, the hedged portfolio outperformed meaningfully, achieving a return of 18.2% compared with 13.6% for the unhedged portfolio. At the same time, volatility declined from 12.0% to 9.1%, resulting in a sharp improvement in risk-adjusted performance, with the return-risk ratio almost doubling from 1.1 to 2.0. The drivers of this reversal are closely linked to asymmetric currency movements, most notably the sharp depreciation of the Japanese yen and, to a lesser extent, the US dollar. Figure 13 (see pdf) presents the performance of the individual regional equity indices in 2025, highlighting significant dispersion across regions. Figure 14 directly compares hedged and unhedged portfolio outcomes during the same period and underscores the stabilizing and performance-enhancing role of FX hedging under conditions of acute currency stress.

From an investment perspective, in a more fragmented world order with higher volatility, FX hedging should be considered at least tactically. Over long horizons, currency hedging has tended to reduce return potential without delivering commensurate improvements in risk-adjusted performance. However, in environments characterized by elevated geopolitical risk, policy uncertainty and disorderly currency adjustments, FX hedging can provide meaningful downside protection and enhance portfolio efficiency. The effectiveness of FX hedging is highly regime- and horizon-dependent. While it has historically acted as a drag on long-term portfolio performance, it can deliver substantial benefits during periods of heightened geopolitical and currency stress. For institutional investors, FX hedging should therefore be considered a dynamic risk management tool rather than a universally optimal or suboptimal strategy.