Home Bias: How Much of Your Equities Are Foreign

Key takeaways

• On the FTSE Global All Cap Index at 30 June 2026, the US was 62.09% of world equity value — so market-cap weighting implies a US investor holds about 38% abroad, an Australian about 98%.
• Vanguard's June 2024 study found Canadians held 50% of their equities at home against a 2.6% market weight — over 18 times overweight; its 2021 paper put UK investors at 4.9 times.
• Minimum-variance analysis does not land on market weight. Vanguard's 2021 study found volatility fell most at 35%-55% international; its 2024 Canadian note put the volatility-minimising home allocation at 25%-60%.
• Currency matters less for equities than bonds. Vanguard's May 2026 paper hedges fixed income and leaves equity unhedged, with modelled hedge ratios falling from 70%-80% at 20% equity to 10%-20% at 80%.
• For some investors the framework runs out: the entire MSCI Frontier Markets Index was worth $203.6 billion at 30 June 2026 against $101.5 trillion for MSCI ACWI — about 0.2%.

Market weight gives one answer, and where you live decides how strange it sounds

Two people read the same sentence — "you're far too concentrated in your home market" — and reasonably reach opposite conclusions. One lives in Boston. The other lives in Sydney. Whose portfolio needs the bigger change?

Here's the arithmetic that settles it. The FTSE Global All Cap Index held 10,139 stocks worth $114.9 trillion at 30 June 2026. The United States was 62.09% of that. Japan was 5.82%, Taiwan 3.38%, the United Kingdom 3.07%, Canada 2.96%, Korea 2.78%, China 2.60%, Switzerland 1.95%, France 1.93%, Germany 1.75%, India 1.67% and Australia 1.56%. The chart plots those twelve.

Weighting by market capitalisation — holding each market in proportion to what it's worth — therefore means completely different things depending on your passport. A US investor at market weight holds about 38% of their equities outside America. A British investor holds nearly 97% outside Britain. An Australian holds over 98% outside Australia. Same rule, opposite lives.

That asymmetry is why most writing on home bias is quietly useless outside the United States. Your Boston reader is being asked to move a portfolio from roughly 80% domestic to 62%. Your Sydney reader, from the same sentence, is being asked to move to 1.6% from a home weight that Vanguard's 30-times-overweight multiple implies is roughly half the portfolio. Those are not the same conversation, and the second one has objections the first one never has to answer.

The two largest index providers disagree by 1.5 points, which tells you how precise this number is not

MSCI's ACWI factsheet for the same date, 30 June 2026, put the United States at 63.63% — 1.54 percentage points above FTSE. Neither is wrong. They're measuring slightly different universes.

MSCI ACWI holds 2,461 large- and mid-cap stocks across 23 developed and 24 emerging markets, covering roughly 85% of the global investable equity opportunity set. FTSE's Global All Cap holds 10,139 stocks, of which 5,874 are small caps. Add the world's small companies and the American share of the total shifts by a point and a half. How much those emerging markets are worth holding is a separate question from how much of the world they represent, and the record on emerging markets returns is thinner than the growth figures imply.

So when a chart tells you the US is "63% of the world", the useful question is: on which index, at which date? A point or two matters if you are building a mechanical rule. It matters not at all if the question is whether 95% at home is a lot. It is.

For a US investor this is a small question; for everyone else it sets the currency of the whole portfolio

Vanguard's May 2026 currency research, The FX dimension, makes the point better than a cap-weight table can. Take a market-cap-weighted global equity portfolio and ask how much of it sits in a currency that is not the investor's own, as at 31 December 2025. For a US investor: 30%. For a euro-area investor: 91%. Japan: 95%. UK: 96%. Canada and Switzerland: 97%. Australia: 98%.

Put the same investor in a 60% stock, 40% bond portfolio with the bonds fully hedged and the numbers become 18% for the American and 55%-59% for everyone else.

An American who goes global is making a diversification decision. Everyone else is making a diversification decision and a currency decision at the same time, and can't separate them. As Vanguard's authors put it, the only way to fully eliminate foreign-exchange risk is to avoid international investing altogether — which is precisely what a heavy home tilt does, and part of why it persists.

Home bias is measured in multiples, and outside America the multiples get strange

The standard way to size the gap is to compare what domestic investors actually hold at home against their market's weight in the global index. Vanguard's 2021 global equity paper did this using the IMF's Coordinated Portfolio Investment Survey as at 31 December 2019. Here's how far each country's investors sat above their own market's weight in the world index.

The multiple is small for Americans only because the denominator is enormous. In absolute terms the US is the most home-biased large market of all: Vanguard's 2024 Canadian study, tracking the same measure to 2023, reported US investors holding 81% of their equities domestically, down from 83% in 2012. Canadians moved much further over the same period, from 67% to 50%. Australian and Japanese home bias fell by 4 and 6 percentage points.

Canada is the cleanest illustration because Vanguard put both numbers side by side: 50% of Canadian equity portfolios in Canadian stocks, against a 2.6% share of the global market at 30 April 2024. That is the over-18-times figure, and it's the shape of the problem almost everywhere outside the United States.

When your home market isn't in the global index at all

There's a group of investors this literature simply doesn't cover, and pretending otherwise is the main failure of articles on this topic.

FTSE Russell reclassified Pakistan from Secondary Emerging to Frontier market status, effective from the open on 23 September 2024, after Pakistan failed the Minimum Securities Count requirement on data as at 28 June 2024. Pakistan doesn't appear anywhere in the FTSE Global All Cap country table. Nor do Bangladesh, Kenya, Vietnam, Nigeria or Sri Lanka.

MSCI still indexes them, in a separate universe. The MSCI Frontier Markets Index held 244 stocks across 28 countries with a combined float-adjusted market value of $203.6 billion at 30 June 2026. MSCI ACWI, at the same date, was $101.5 trillion. Every frontier market on earth is roughly 0.2% of the investable world. MSCI publishes a combined index, and over the year to 30 June 2026 MSCI ACWI plus Frontier Markets returned 24.18% against 24.16% for ACWI alone. Adding every frontier market moved the global index by two basis points.

So what does market weight tell an investor in Karachi or Nairobi to do? Hold essentially nothing at home. That isn't a portfolio, it's the framework running out of road. Capital controls, currency convertibility, access to foreign brokers, and the plain fact that rent and school fees are denominated in rupees do most of the work in that decision. The market-weight argument was built for investors in markets an index provider can see, and it should be read with that boundary attached.

A home tilt has a concentration cost you can measure — but so does the global index now

Vanguard's Canadian study quantified what a home tilt buys you in concentration. At 30 April 2024, the ten largest holdings in the Canadian equity market were 36.87% of it. The ten largest in the global market were 14.64%. Canada was 15.9 percentage points overweight financials and 14.1 points overweight energy, against 13.2 points underweight information technology and 10.6 points underweight healthcare. A Canadian holding only Canadian shares owns a leveraged bet on banks and oil, whether or not that was the intention.

The obvious rejoinder is that the global index is no longer the diversified thing it was. That rejoinder has numbers behind it. On the same FTSE Global All Cap Index, the top ten holdings went from 14.64% in April 2024 to 20.70% at 30 June 2026. On MSCI ACWI the top ten were 23.28%, with NVIDIA alone at 4.55% and information technology at 32.09% of the index. Buying the world today buys a portfolio that is 62% American and roughly one-fifth concentrated in ten companies. Diversifying away from your home market's concentration means accepting a different concentration, not escaping concentration.

The case for a home tilt is stronger than the market-weight argument admits

Three of its four legs hold up under the evidence.

The diversification benefit plateaus, and it plateaus a long way from market weight. Vanguard's 2021 study ran its capital markets model across the US, euro area, Canada, UK and Australia and found portfolio volatility generally began to rise once international allocations passed 35%-55%. Its 2024 Canadian note put the volatility-minimising home allocation across those same five regions at 25%-60%, and for Canada specifically found volatility rising beyond a 70% international weight. So where does Vanguard itself land for Canadians? At 30% at home — against a 2.6% market weight. The firm making the global diversification case sits an order of magnitude away from market weight, and says so in print. Weight is only half the question in any case: risk contribution is where the equity sleeve turns out to supply 76% of a five-asset portfolio's variance.

Correlations have risen. The ten-year rolling correlation between US and non-US equities went from 0.51 at the end of 1989 to 0.86 by September 2020 in Vanguard's data. Each extra unit of foreign exposure buys less than it used to. Vanguard's own 30-year forecasts still show imperfect correlations — 0.62 for Australia, 0.64 Canada, 0.65 UK, 0.72 US, 0.76 euro area — so the benefit hasn't vanished. It has shrunk.

Tax is real and jurisdiction-specific. Say you're a top-bracket Canadian with a taxable account, and $100,000 of distributions lands in it this year. Vanguard's Canadian analysis worked exactly that case. Hold Canadian stocks and you owe $37,813. Hold global ex-Canada stocks and the identical income costs you $53,492 — a gap of $15,679, on the same money, driven by Canada's dividend tax credit. Vanguard's conclusion was that on tax grounds alone a Canadian holding a taxable account is better off in Canadian stocks. That advantage disappears inside a sheltered account, and its size and direction differ in every jurisdiction. Anyone reasoning from a US article about foreign tax credits is reasoning from the wrong tax code.

The fourth leg — that domestic markets are too illiquid or too costly to leave — did not survive Vanguard's own testing in Canada, where the conclusion was that Canadian markets are sufficiently liquid, and so liquidity isn't a reason to reduce the home allocation.

Currency is the objection people lead with, and Vanguard's own work says it matters less for equities

Over long horizons currency has no intrinsic return — no yield, no coupon, no earnings growth — so it affects volatility rather than expected return. Vanguard's 2021 paper measured the effect of hedging international equity across five home markets and found the hedging effect relatively marginal, and in some cases volatility-reducing. For Australian and Canadian investors over 2000 to September 2020, hedging slightly raised volatility, because their commodity-linked currencies tend to move with global equities.

So how much of it should be hedged? The May 2026 paper turns that into a portfolio rule: hedge fixed income, leave equity unhedged. Its modelled hedge ratios fall from roughly 70%-80% for a 20%-equity portfolio to 10%-20% for an 80%-equity portfolio across five domiciles. The logic is that as a portfolio gets less volatile, currency becomes a larger share of what is left, so hedging matters more; equity volatility swamps currency volatility going the other way. The rule is stated per asset class, but the number a holder actually carries is the portfolio currency exposure left once both sleeves are weighted and added together.

None of which means currency is small in any given year. Picture two people holding the identical global fund through 2025, one counting in dollars and one counting in francs. The MSCI World Index returned about 21.6% in US dollar terms and 6.3% in Swiss franc terms. Same stocks, same year, a gap of more than 15 points, entirely exchange rates. Hedging isn't a free fix for that either: Vanguard estimates annual hedging costs of roughly 0.25%-0.50%, and the realised ten-year volatility effect across the markets it studied ranged from a 7.1% reduction for a Japanese investor in the decade to 2016 to a 3.5% increase for an Australian in the decade to 2010, averaging a 1.7% reduction. You can pay for a benefit that doesn't arrive for a decade.

What this evidence can't tell you

Market-cap weight is a starting point, not a law, and the people who publish the research say so. Vanguard's 2021 conclusion is that global market-cap weight "serves as a helpful starting point" and that in practice many investors will hold international equities well below it. Whether that weighting still buys a measurable diversification benefit is a separate question, and across 433 months of regional returns the answer came to 0.40 percentage points of annualised volatility.

The weights are not stable. Vanguard's own long history shows the US share of global equity market capitalisation as low as 29% in the 1980s. It was 58.3% in September 2020 and is 62%-64% now. Anchoring hard to today's number is anchoring to something that has moved 30 points within one investing lifetime.

Most of the forward-looking figures here are simulations, not history — 10,000 runs of Vanguard's capital markets model, calibrated on the past. The volatility-minimising ranges are outputs of a model, and models understate extremes. The home-bias multiples come from the IMF's voluntary portfolio survey, which Vanguard itself flags as imprecise against index market values.

And Vanguard is not a neutral referee: it sells globally diversified index funds. Worth noting in the other direction, though, that its Canadian arm argues for 30% at home, which is not the marketing-convenient answer. Finally, none of this literature knows your pension rules, your withholding-tax treaties, your account wrappers, your currency controls, or the fact that your job, your house and your future spending are already a large undiversified bet on your home economy.

What would change the conclusion

If the US share of global market cap fell sharply. An American's "38% abroad" only looks modest because the US is 62% of the index. At the 29% of the 1980s, a US investor at market weight would face the 70%-abroad question a British investor faces today — and would discover the currency and tax objections that everyone else already lives with. The asymmetry in this piece is a fact about 2026, not about America.

If correlations reached one. At a correlation of 1.0 between home and foreign equities, geographic diversification would buy nothing but currency exposure and costs. The move from 0.51 to 0.86 is most of the way there; Vanguard's 30-year forecasts of 0.62-0.76 say the remaining gap is real. If realised correlations settle above 0.9, the case for pushing past the volatility-minimising range weakens considerably. Just how wide the flat stretch already is, when the sweep runs from 0% UK weight to 100%, is measured in our guide to global equity allocation.

If your tax treatment changed. The $15,679 Canadian gap is a creature of one country's dividend tax credit inside one type of account. Move the same portfolio into a sheltered account and it goes to zero, and the home-tilt case loses a leg.

If you are in a market the index can't see. Frontier-market investors are outside this evidence base entirely, and no amount of Vanguard research fixes that.

So forget Boston and Sydney for a second. What share of your equities sits in one country and one currency — and did you choose that, or inherit it? That's the number worth tracking, and it isn't the index's. How often you check that matters less than knowing it, and the same is true of any other single-sleeve weight you carry. LedgerTouch reports geographic weight continuously; a spreadsheet reports it once a year. Either beats the common case, which isn't knowing at all.