How much US tech do we own?
An MSCI World ETF, an S&P 500 tracker and a handful of American tech stocks is a familiar portfolio. It leaves us owning many of the same companies several times over.
At the end of October 2026, MSCI World held 1,248 stocks, yet the US accounted for 72.77% of its value. Ten positions held 26.6% of it: Apple, Nvidia, Microsoft, Amazon, two Alphabet share classes, Broadcom, Meta, Tesla and Eli Lilly. Japan accounted for 5.72% of the index and France about 2.2%, a copio calculation from the MSCI France index's market value. 1 1
As those companies outgrow the rest of the market, their weights increase automatically. Years of US tech outperformance have left a world tracker heavily dependent on the same businesses continuing to lead. Adding a Nasdaq fund or buying those companies directly increases that dependence.
We own a large part of the AI investment boom, much of it at high valuations, with substantial dollar exposure. We own much less of the businesses that could benefit from the next phase of growth in Europe and Japan.
Where the last decade's returns came from
American companies increased their profits, and investors paid higher prices for each dollar of those profits. Both contributed to the return on US shares. 2
Russell 1000, USD, through 31 December 2024. Source: FTSE Russell / LSEG. Rounded contributions total 13.7% for 2011-2024; the published total return is 13.6%.
Earnings growth contributed 7.5 percentage points a year in 2003-2010 and 7.9 points in 2011-2024. The much larger change came from valuations: they went from subtracting 2.3 points to adding 3.8. Dividends contributed 2 points in each period.
That makes repeating the past decade's returns harder. Profits must keep growing, and investors must keep paying more for them, or one must compensate for the other.
Take a company earning €1 a share, bought at €20. If profits double over ten years and the market still pays twenty times earnings, the share reaches €40. If the market pays fifteen times earnings, it reaches €30. The same profit growth produces annual price returns of about 7.2% or 4.1%, before dividends.
At October's prices, US equities traded at 18.7 times expected earnings, against 14.6 for developed markets outside the US and 14.1 for Europe; at the end of September the figures were 19.4, 14.9 and 14.2. 1 1 1 We expect US equities to underperform other regions over the next decade as that gap narrows and earnings elsewhere catch up. Europe has room to improve margins through AI adoption and new investment; Japan has room to improve what companies earn on their capital.
US companies could continue to justify their premium through faster growth. We still want to own them, with a smaller share of our future wealth depending on that outcome.
The risks have changed too
Higher rates put pressure on both company profits and share valuations. Refinancing costs more, and investors can earn interest from bonds that paid very little a few years ago. Companies whose value depends on profits far into the future have more to prove at those rates.
Government spending adds demand for equipment, infrastructure and labour. It also adds borrowing and can keep inflation and interest rates higher. The same policy can help an industrial supplier while hurting a heavily indebted business.
Geopolitical shocks can make several parts of a portfolio fall together. An energy shortage raises costs, squeezes customers and gives central banks less room to cut rates. Stocks and government bonds both fell in 2022. For a euro investor, a weaker dollar can add to the loss on US holdings.
We need more sources of return and a protection plan that can handle different kinds of crisis. Replacing some US shares with European shares broadens the earnings exposure, but both markets can still fall in a global sell-off. Bonds, gold, cash and rebalancing each have a role to consider alongside the regional allocation.
Where we see the next decade's opportunities
Fiscal spending and rising debt
Governments around the world are spending on infrastructure, defence, energy and strategic industries. That money supports demand and creates orders for businesses. When it is financed by borrowing, it also adds debt that must be serviced for years.
Germany's €500bn infrastructure fund is one example. The NATO spending commitments extend across Europe and North America. 4 3 The cost of carrying debt is already substantial in the US: federal net interest reached about $1.1tn in fiscal 2026, roughly 19% of revenue. 14
Borrowing costs have risen across the major markets. The chart follows ten-year government bond yields in the US, UK, Germany, France and Japan on the same monthly basis. 13
OECD monthly averages, January 2015-September 2026. Nominal yields in each country's currency. Download data.
At the end of 2020, all five yields were below 1%, with Germany and France below zero. By September 2026, they ranged from 3.05% in Japan to 5.24% in the UK. Inflation and central-bank policy affect these yields alongside government borrowing and concerns about debt.
Higher yields feed into the public interest bill as old bonds mature and governments refinance. They also raise the cost of mortgages and business loans. If the spending lifts growth and tax receipts enough, it helps pay for itself. If debt keeps growing faster than the economy, more of the budget goes to interest and lenders can demand a higher return to keep financing it.
For our portfolios, fiscal expansion creates opportunities in industrials, engineering and infrastructure, while higher financing costs put pressure on indebted companies and existing bond prices. New bond investments earn more income at those higher yields. We need to account for both the spending and the debt it creates when choosing the allocation.
Companies using AI
Banks, insurers and service companies can use AI to process claims, review documents, answer customers and handle routine administration. For businesses with large payrolls, doing more work with the same staff can improve margins.
About one in five EU enterprises with at least ten employees used AI in 2025, rising to 55% among large enterprises. 5 In a deployment covering 5,172 customer-support agents, AI assistance increased productivity by 15% on average. 6 There is room for those tools to reach more businesses and more of the work within them.
We expect European services companies to benefit as adoption spreads. They can buy AI services without financing the largest computing projects themselves. The profit improvement will depend on implementation costs, software fees and how much of the saving competition passes to customers. The Europe article examines the company results so far.
Europe and Japan put more capital to work
Europe has large savings pools that could finance more of its companies' growth. The EU's Savings and Investments Union aims to make investing easier and direct some of the €10tn sitting in bank deposits towards businesses. If it succeeds, companies gain access to capital and savers gain more opportunities to participate in their growth. 7
Japan's opportunity includes money already sitting inside companies. Listed businesses hold about ¥118tn in cash and deposits. 10 Business corporations also owned 17.7% of the listed market in fiscal 2025, including cross-shareholdings often maintained for decades to preserve relationships and management control. 15 11
The Tokyo Stock Exchange is pushing companies to explain how they will earn a competitive return on that capital. Managers face pressure to sell unproductive holdings, invest in profitable operations and distribute cash they cannot use well. The share of Prime-market companies trading below book value, the accounting value of their net assets, fell from 50% in July 2022 to 27% in March 2026. 11
Japanese companies announced a record ¥22.3tn of share buybacks in fiscal 2025. 12 Completed buybacks at sensible prices give the remaining shareholders a larger share of future profits. Selling long-standing cross-shareholdings also gives outside investors more influence over how a business is run.
We see scope for higher shareholder returns as this reform continues. Europe's efforts to finance growth and Japan's efforts to improve capital use give us two reasons to allocate beyond the US.
Electrification and climate adaptation
Data centres, electric vehicles and industrial electrification all need power. The EU estimates that its grids require €584bn of investment by 2030, with about 40% of distribution grids already more than forty years old. 8 That creates demand for cables, transformers, engineering and utility investment.
Climate damage adds spending on flood defences, water systems and cooling. Natural catastrophes caused about $224bn in losses in 2025, of which $108bn was insured. 9 Property owners, insurers and governments all carry part of that bill.
We want exposure to businesses that can earn from the investment: equipment suppliers with profitable orders and utilities whose allowed returns cover their financing costs. Higher claims can hurt insurers, while expensive adaptation can reduce property income.
A broader allocation
Copio estimates returns, volatility and correlations for more than 50 asset classes in euros over ten to fifteen years. Here is how four allocations compare, before product fees and personal taxes.
| Portfolio | Estimated annual compound return | Estimated annual volatility |
|---|
| 100% US large caps | 7.0% | 17.0% |
| World equities, including emerging markets | 7.0% | 15.5% |
| Regional equity mix | 7.7% | 15.1% |
| 60% regional equities, 30% bonds, 10% gold | 7.0% | 9.5% |
The regional equity mix puts 40% in US large caps, 20% in euro area large caps, 10% in euro area small caps, 20% in Japan and 10% in emerging markets. It remains fully invested in shares, with more weight in Europe and Japan than a world index holds.
The 60/30/10 illustration holds 30% US large caps, 15% euro area large caps, 15% Japan, 30% US aggregate bonds hedged to euros and 10% gold. Bonds add income with lower expected volatility, and gold adds an asset with a low estimated equity correlation.
The regional mix raises estimated return from 7.0% to 7.7%, helped by Japan's 8.6% estimate. Our model has euro area large caps at 6.8% and broader European large caps at 6.5%, so it does not put every overseas market above the US.
Adding bonds and gold brings estimated volatility down to 9.5%, with the same rounded 7.0% return estimate as US equities alone. That is about 44% less volatility, which measures fluctuations rather than the size of a future crash. If we move every cross-asset correlation halfway towards 1, estimated volatility rises to 11.4%. The calculation appendix shows the full assumptions and sensitivity results.
Decide how you will handle a sell-off
The money you need soon should be available even if markets fall. Set it aside before deciding how much of the rest belongs in equities.
Then consider which shock would hurt both your investments and your income. A tech employee holding mostly tech shares, or a French homeowner with domestic investments and a salary paid in France, can face several losses at once. Use the whole picture when deciding what to own elsewhere.
Choose protection for those risks and account for its cost. Government bonds can help in a recession, commodities during a supply shock, and gold when confidence in money weakens. Each can also lose money, as we examine in the crash-protection article.
Write down your target weights and when you will rebalance. For example, you could review annually and rebalance when a major allocation moves five percentage points from its target. That gives you a rule to follow when a rally or a sell-off pulls the portfolio away from the risk you chose.
This article is our dated view of the decade ahead. The forecast-free principles behind it, from market leadership to rebalancing, are in How do you build a long-term portfolio when the US is no longer a given?
Information only; no personal investment advice. Long-term estimates and illustrative allocations can differ substantially from realised results. Past performance does not guarantee future returns.
