The Uncomfortable Starting Point
According to JP Morgan Private Bank, since mid-2008, the S&P 500 has returned an annualized 11.9 percent. The MSCI EAFE Index, which tracks developed markets outside the US and Canada, has returned 3.6 percent over the same period. That gap of roughly 8 percentage points per year, compounded across a decade and a half, has reshaped how investors think about international stocks.
The data is not subtle. Investors who held a US-only portfolio over this period have done dramatically better than investors who diversified into developed Europe and Japan. Those who held emerging markets have done worse still, with EM equities trailing the S&P 500 by roughly 9 percentage points annualized over the past decade affected by currency devaluation.
This is not how the textbook is supposed to work. Diversification, the textbook says, reduces risk without sacrificing return. The past fifteen years suggest that diversification has reduced returns substantially while doing little to reduce risk during the moments that mattered most. The conclusion many investors have drawn is that global diversification is a relic of an earlier era, useful when the US economy was less dominant but unnecessary now. That conclusion has a problem. The same data that makes diversification look like a bad idea today made it look like an obviously good idea fifteen years ago, before the rotation began.
This article takes the question seriously. What does the data actually show about diversification across global markets? Why has it failed recently? And what does that history mean for an investor deciding how to build a portfolio in 2026?
What “Global Diversification” Actually Means
Global diversification means holding equities across multiple countries or regions, in proportions that differ from a single-country concentrated allocation. The simplest version is the MSCI All Country World Index, often referenced as ACWI. It holds approximately 2,500 companies across 23 developed and 24 emerging markets. Its weights are determined by the market capitalization of each country’s listed equities relative to the global total.
As of early 2026, the United States accounts for roughly 63.17 percent of MSCI ACWI. Japan is the next-largest single country at 5.05 percent. The United Kingdom sits at 3.4 percent. Emerging markets in aggregate make up about 12 percent. Developed markets outside the US carry the remaining 25 percent or so.
The first thing this composition reveals is how much “global” already means “American.” A market-cap-weighted global portfolio is heavily a US portfolio. The decision to “diversify globally” using ACWI is more accurately a decision to add roughly 38 percent non-US exposure to a 62 percent US base. Whether that constitutes meaningful diversification depends on what comes next.
Other forms of diversification
Sector diversification is the mix across technology, financials, industrials, consumer staples, and so on. The US and the rest of the world differ here in ways that are easy to miss. Technology is roughly 26 percent of MSCI ACWI overall but accounts for 32 percent of the S&P 500 alone, and as much as 45 percent if the market is broadened to include all “tech-related” mega-caps. Outside the US, technology is closer to 14 percent of equity allocation. The sector mix differs because the geographic mix differs.
Currency diversification is the exposure to non-dollar revenues and earnings. A US investor in MSCI EAFE earns part of the return from local equity performance and part from changes in the euro, yen, pound, and other currencies against the dollar. Over the past fifteen years, currency has been a drag on EAFE returns for dollar-denominated investors, taking roughly two percentage points off the annualized total.
Asset-class diversification is the mix between equities, bonds, real estate, commodities, and other assets. This article focuses on equity diversification across geographies, but the broader question of asset-class allocation is just as important and arguably more so for managing risk.
The article centers on country and region diversification because that is what most investors mean when they talk about “diversifying globally,” and because the long-running underperformance of international stocks is the question that needs honest engagement.
What the Data Actually Shows
The temptation when looking at the recent fifteen years is to extrapolate. US wins, international loses, this is the new normal, position accordingly. The temptation should be resisted because the longer history shows something different.
The first decade of the 2000s tells a different story than the recent decade. Between 2000 and 2009, the S&P 500 returned negative 0.95 percent annualized including dividends. That is a real, full-decade loss for US equity investors. Over the same period, the MSCI Emerging Markets Index returned 9.8 percent annualized. The MSCI World ex-US Index returned 1.6 percent. International stocks did not just match US stocks during that decade. They beat them by margins comparable to the US’s recent dominance.
Looking even further back compounds the point. From 1988 through 2019, MSCI Emerging Markets returned 10.7 percent annualized. The S&P 500 returned 10.8 percent over the same period. The MSCI World ex-US Index returned 5.9 percent. Across more than three decades, US equities and emerging markets produced almost identical long-term returns, while developed-international trailed both by a meaningful but not catastrophic margin.
Two important facts emerge from this longer view.
The first is that returns leadership rotates. The decade-by-decade pattern is not a steady US dominance with occasional noise. It is a genuine rotation, with periods where international stocks beat US stocks substantially and periods where US stocks beat international substantially. The recent fifteen years are one of the longest US dominance stretches on record. They are not the whole record.
The second is that long-term returns across major regions converge more than the short-term data suggests. An investor in 1988 who held the S&P 500 and an investor who held emerging markets ended up at similar places in 2019 even though their journeys looked nothing alike. The path mattered enormously. The destination, less so.
This is the insight most generic articles miss. Diversification is not primarily about producing higher returns over the long term. It is about smoothing the path of returns. The investor who holds only US equities accepts the full force of US underperformance whenever it comes. The investor who diversifies accepts lower returns during US-dominated periods in exchange for protection during US-trailing periods. Whether that trade is worth making depends on a judgment about whether the next fifteen years will look like the last fifteen years or like something different.
Why the Last Fifteen Years Looked the Way They Did
US dominance over this period is not a mystery. It came from a specific combination of factors that compounded.
Earnings growth led the way. According to JP Morgan Private Bank, the S&P 500 grew earnings at an annualized 6.3 percent between mid-2008 and the end of 2024. The MSCI EAFE Index grew earnings at 1.6 percent over the same window. US companies expanded profits roughly four times faster than developed international counterparts. That alone explains a large share of the return gap.
Valuation expansion added another layer. The price-to-earnings multiple of the S&P 500 expanded from roughly 12.8 to 21.7 over the same period. EAFE multiples rose from 11.3 to 14.0. Investors paid more dollars per dollar of US earnings than they did fifteen years earlier, which boosted returns mechanically. EAFE saw modest multiple expansion. Emerging market multiples actually contracted.
Currency provided the third tailwind. The US dollar strengthened against most major currencies over the period. This was a headwind for dollar-denominated investors holding international stocks and a meaningful component of the return gap. Of EAFE’s 3.6 percent annualized return, roughly negative 2.1 percentage points came from currency drag.
Underneath these factors sat a single dominant structural force: technology. US tech and tech-adjacent companies grew earnings dramatically faster than the rest of the global market. Within the S&P 500, technology stocks alone grew earnings per share by 571 percent over the period. Including the seven largest tech-adjacent mega-caps that dominated late-cycle returns, the S&P 500’s tech-related weight climbed from a fraction of the index to roughly 45 percent of it. EAFE tech, by comparison, grew earnings about 24 percent total over the same window.
The artificial intelligence cycle accelerated this dynamic. From 2023 onward, a handful of US companies captured the bulk of capital and investor enthusiasm tied to AI infrastructure, compute, and applications. Nvidia, Microsoft, Alphabet, Meta, Amazon, and Apple together added trillions of dollars in market capitalization while continuing to grow underlying profits at extraordinary rates. The international market has no equivalent companies at that scale. Europe’s leading technology firm by valuation, ASML, is critical infrastructure to the AI cycle but trades at a fraction of Nvidia’s market cap. Japan, Korea, and Taiwan have important component manufacturers, but no platform-scale beneficiaries.
The combination of faster earnings growth, valuation expansion, currency strength, and AI concentration produced returns that nothing outside the US could match. None of these factors were obvious in advance. None would have been correctly predicted by an investor in 2008. The pattern looks inevitable in hindsight only.
The Risk Hidden Inside the Returns
The same forces that produced US outperformance produced concentration. That concentration is itself a risk most retail investors underestimate.
The S&P 500 today is more dependent on a smaller group of companies than at almost any point in its history. The top ten stocks account for roughly 35 percent of the index. Seven of those companies are technology or technology-related, sharing exposure to similar customers, similar capital cycles, and similar regulatory risks. A drawdown in any one of them moves the entire US equity market. A coordinated drawdown across two or three of them moves it dramatically.
This concentration is not visible in the returns figures. A 12 percent annualized return looks identical whether it comes from 500 companies pulling roughly equal weight or from seven companies dragging the rest behind them. The lived experience of holding the index, however, is very different. An investor in the 1990s S&P 500 was diversified across hundreds of meaningfully sized businesses. An investor in the 2026 S&P 500 is making a concentrated bet on the continued dominance of a handful of mega-cap technology firms.
The fragility this creates is asymmetric. When the US economy or US technology sector suffers a shock, the US equity market falls farther and faster than markets that are less concentrated in any single sector or country. The 2000 dot-com bust is the most relevant historical reference. The S&P 500 fell roughly 49 percent peak to trough between 2000 and 2002. The MSCI EAFE Index fell roughly 47 percent. Emerging markets fell less in some currencies and more in others. But the S&P 500 then took until 2007 to fully recover, while a portfolio diversified across regions and styles recovered considerably faster.
The point is not that US concentration always produces worse outcomes. It plainly has not over the past fifteen years. The point is that the same concentration that produced the recent gains amplifies the downside when conditions reverse. An investor holding only the S&P 500 in 2026 is exposed to a specific kind of risk: a sustained underperformance of US technology, a regulatory shock to the largest US companies, a sudden weakening of the US dollar, a productivity disappointment in the AI cycle. Each of these scenarios is non-trivial. None of them is the base case. Together they form a real risk profile that the headline return numbers do not capture.
This is the argument for global diversification that survives the past fifteen years of disappointing data. It is not that international stocks will outperform US stocks over the next decade. It is that they will probably perform differently, and that difference is the thing of value. The investor who needs to draw on a portfolio over the next twenty or thirty years has more to lose from a single-country shock than from a few percentage points of foregone return during good times.
What Practical Diversification Looks Like in 2026
A market-cap-weighted global index fund is the simplest starting point. Holding MSCI ACWI gives you the 63 percent US, 37 percent international split by default. That is not a neutral position; it is a bet that today’s market capitalizations correctly predict tomorrow’s returns. Given current US valuations and concentration, that bet deserves scrutiny.
The valuation gap is real and significant. As of early 2026, the MSCI ACWI ex-US trades at a price-to-earnings ratio of roughly 13.75 versus the S&P 500 above 21. International developed markets and emerging markets are cheaper by almost every valuation metric. That does not guarantee they outperform. But it does mean their expected return for a long-horizon investor is mathematically higher than the headline figures of the past decade suggest.
A practical framework for most investors:
Start with a global market-cap-weighted index. This delivers diversification across countries and sectors at minimum cost and decision-making. It is the baseline against which any other choice has to justify itself.
Tilt only when you have a view you can defend. Investors who believe US tech concentration is a risk can hold more MSCI ACWI ex-US relative to a US-heavy default. This is a deliberate choice, not a default. It also has to be sustainable through the years when the choice looks wrong.
Account for currency. International equity returns for non-US investors carry currency exposure in both directions. A strengthening dollar took 2.1 percentage points annually off EAFE returns over the past fifteen years. The dollar will not strengthen forever, and when it weakens, that drag reverses.
Treat emerging markets as a considered allocation, not an emotional one. EM has produced long-term returns comparable to the US with substantially higher volatility. Institutional portfolios commonly allocate 5 to 15 percent. The hard part is holding the position through the years when it disappoints, not selecting it.
Common Mistakes and Conclusion
The most costly mistake in global diversification is abandoning it during the periods when it underperforms. International stocks have trailed the S&P 500 for fifteen years. Investors who sold their international allocation in 2018 or 2022 because it “wasn’t working” locked in the underperformance permanently.
The second mistake is confusing diversification with a returns strategy. It is not. It is a risk management strategy. Its value is not visible in good times. It becomes visible when one market suffers a shock that others do not share.
The third mistake is over-engineering. A single global market-cap-weighted index fund achieves most of what elaborate portfolio construction attempts. Layers of country-specific funds, factor tilts, and currency hedges introduce complexity and cost without proportional benefit for most investors.
The honest conclusion is this. The past fifteen years make diversification look like a mistake. The past fifty years make it look like the only sensible approach. US equities have produced extraordinary returns, but those returns have become increasingly concentrated in a small number of technology companies, in a single currency, and in a single country’s regulatory and monetary environment. That concentration is not a problem until it is. Investors who diversify globally accept the cost of forgone US-specific gains in exchange for protection against a US-specific shock. Given where valuations, concentration, and geopolitical tension sit in 2026, that is not a trade-off to dismiss lightly.