How much international stock should you own?

Investors usually keep almost all their money in their own country. Economists call this home bias. A recent study tested a different path: a portfolio of stocks only, with about one third in the investor's own country and two thirds abroad, held for a whole lifetime.

This article sums up what the study tested, what it found and its limits. It doesn't say how much you should hold abroad: the targets are your choices.

Home bias

In 1991, a study showed that US investors held 94% of their stocks at home, Japanese investors more than 98% and British investors 82%. To the authors, this appeared to come from investors' choices, not from rules that kept them from investing abroad (French and Poterba, 1991).

One explanation is familiarity. In the US, shareholders of each regional phone company tended to live in the area it served. The author concluded that people invest in what they know, often setting aside the principles of diversification (Huberman, 2001).

What the study tested

The study is called Beyond the Status Quo, by Aizhan Anarkulova of Emory University, Scott Cederburg of the University of Arizona and Michael O'Doherty of the University of Missouri. The latest version is from August 2026, and it hasn't been published in a journal yet, which means it hasn't gone through peer review (Anarkulova, Cederburg and O'Doherty, 2026).

The authors used stock and bond returns from 39 developed countries, from 1890 to 2023. Each country only enters the sample once it became developed. With that data, they simulated 1 million lives of an American couple who start working at 25, retire at 65, receive US Social Security and can't borrow to invest. To keep the sequence of good and bad years, each simulation strings together stretches of history, 10 years long on average.

They compared five ways to invest:

• 100% in short-term bills.
• 100% in domestic stocks.
• 60% stocks and 40% bonds.
• A target-date fund, which holds fewer stocks as the investor ages.
• The best combination the model found.

What it found

The best portfolio was close to one third domestic stocks and two thirds international stocks, at any age, with no bonds. With fixed weights, the best split was 34% and 66%.

Compared with the target-date fund, it led to these results, as averages across simulations, in 2022 dollars:

In other words, by saving less, the couple reached retirement with more. By the study's welfare measure, staying in the target-date fund is like paying a 1.90% annual fee on the best portfolio.

The price: bigger drops

An all-stock portfolio falls further. On average across simulations, the largest inflation-adjusted drop during working years was 55% for the best portfolio and 52% for the target-date fund. In retirement, it was 48% and 40%.

In the worst cases, this flips. In 5% of scenarios, the largest drop was 77% or more for the best portfolio and 81% or more for the fund during working years. In retirement, it was 74% or more and 89% or more. The fund suffered more in those cases because high inflation eroded the value of its bonds.

The authors acknowledge the behavioral risk: large drops hurt, and many people tend to abandon the strategy when they happen. The study's result depends on holding stocks through good years and bad.

Why international stocks rather than bonds

In the study's data, long-term bonds returned 0.95% a year above inflation, and international stocks returned 7.03%. Over 30-year horizons, the per-year variance of bonds, a measure of risk, rises to 2.39 times its one-year level, while that of international stocks falls to 0.75 times. And the real return of bonds drops when inflation rises, with a correlation of −0.78, while that of international stocks barely relates to it, at −0.02.

Spreading across countries also lowers the risk that a single country has bad decades. In an earlier study by the same authors, published in the Journal of Financial Economics, across 39 developed countries, the chance that someone holding only domestic stocks loses to inflation over 30 years was 12%. With US data alone, it would be 1.2%. In Japan, from 1990 to 2019, stocks lost 21% after inflation (Anarkulova, Cederburg and O'Doherty, 2022).

In another paper of theirs, not yet published in a journal, domestic stocks lost to inflation over 30 years in 13% of cases, long-term bonds in 27% and short-term bills in 37%. International stocks lost in 4%, because exchange rates offset part of local inflation (Anarkulova, Cederburg and O'Doherty, 2023).

What the critics say

The study drew strong criticism. Cliff Asness, of the asset manager AQR, pointed out that not everyone can hold 100% stocks, because someone has to hold the bonds, and said the period in the data likely favors stocks. In his view, people who want more return can borrow, in moderation, to invest in a more balanced portfolio (Asness, 2024). The authors replied that most retirement savers can't invest with borrowed money (Fortune, 2024). The 2026 version tested that case, and even when borrowing at a moderate cost was allowed, the portfolio stayed almost entirely in stocks.

Larry Swedroe, writing for Morningstar, noted that losses weigh on people more than gains and questioned whether most would stick with the bigger drops without giving up (Swedroe, 2024).

These critiques were about the first version, from 2023, which proposed half at home and half abroad. Since then, the study has changed its method and its numbers.

What changes outside the US

The household in the model is an American couple, with US Social Security. And the sample only covers each country while it was developed.

"One third at home" also means something different in each country. The US is 64.63% of MSCI's world stock index, so for an American, one third at home is less than the country's weight in the world. Japan is 5.08%, the UK 3.12% and Canada 3.06% (MSCI ACWI). For investors in those countries, one third at home is many times their market's weight.

The study doesn't work out a split for each country, and its own result changes with the data: without the US in the sample, the best split was 49% at home and 51% abroad. So it doesn't say that someone outside the US should hold exactly one third at home.

Investing abroad also means exchange rates. In the authors' data, currency moves offset part of local inflation, but each country's history is different. The model also leaves out costs and taxes, which change from country to country: sending money abroad has currency costs, and the tax on foreign investments depends on where you live.

How much to hold abroad is your choice. What the study offers is numbers on a path few people take, with its risks.

Sources

When a journal article is paywalled, the link goes to a free version from the authors when there is one, or to the page with the abstract.

• Anarkulova, A., Cederburg, S. and O'Doherty, M. S. (2022). Stocks for the long run? Evidence from a broad sample of developed markets. Journal of Financial Economics, 143(1), 409–433. DOI: 10.1016/j.jfineco.2021.06.040.
• Anarkulova, A., Cederburg, S. and O'Doherty, M. S. (2023). Long-horizon losses in stocks, bonds, and bills. Working paper, SSRN 3964908. DOI: 10.2139/ssrn.3964908.
• Anarkulova, A., Cederburg, S. and O'Doherty, M. S. (2026). Beyond the status quo: a critical assessment of lifecycle investment advice. Working paper dated August 27, 2026, SSRN 4590406. DOI: 10.2139/ssrn.4590406.
• Asness, C. (2024). Why not 100% equities. AQR Capital Management.
• Fortune (2024). Billionaire investor Clifford Asness rips new report that tells investors to only buy stocks.
• French, K. R. and Poterba, J. M. (1991). Investor diversification and international equity markets. The American Economic Review, 81(2), 222–226.
• Huberman, G. (2001). Familiarity breeds investment. The Review of Financial Studies, 14(3), 659–680.
• MSCI (2026). MSCI ACWI Index. Factsheet dated October 30, 2026.
• Swedroe, L. (2024). Should long-term investors be 100% in equities? Morningstar.