Home bias is the documented tendency of investors to hold domestic securities in shares far above those securities' weight in the world market. French and Poterba's 1991 American Economic Review study put domestic ownership at roughly 94 percent of U.S. equity portfolios, about 98 percent in Japan, and roughly 82 percent in the United Kingdom, far above capitalization weights.
Horison publishes information, not investment advice, and the geographic split of any portfolio depends on individual circumstances — currency obligations, tax residence, and jurisdiction — that a reference publication cannot know. This explainer documents the evidence for the pattern, the explanations proposed in the research, and the diversification cost the literature has measured.
What is the evidence for home bias?
The pattern was named in the early 1990s from ownership data. Kenneth French and James Poterba computed domestic ownership shares for major markets around 1989: roughly 94 percent for U.S. equity, about 98 percent for Japan, and roughly 82 percent for the United Kingdom, with similarly dominant levels in the large European markets. No market-capitalization benchmark implied weights anywhere near those figures.
The gap persists in modern data. The International Monetary Fund's Coordinated Portfolio Investment Survey records cross-border holdings far below what market weights would imply, decade after decade. By late 2024 the United States represented more than 60 percent of global equity capitalization in MSCI's flagship all-country index weights, while plan-level data such as Vanguard's How America Saves series shows U.S. participants' equity allocations remaining overwhelmingly domestic.
Because the tilt appears in every large market rather than only one, the academic literature treats it as a structural puzzle — the equity home bias puzzle — rather than as a local accident.
Measurement matters when reading the numbers. Researchers typically quantify the gap as the difference between the domestic share of a country's equity holdings and the domestic market's share of world capitalization — a number near zero for a fully weighted portfolio. The same construction applied to bonds shows an even stronger tilt, since debt portfolios remain overwhelmingly home-denominated in the recorded data.
How do taxes and institutions shape the tilt?
Institutional plumbing contributes its own push toward domestic holdings. Foreign dividends attract withholding tax in many jurisdictions, which U.S. accounts can partly reclaim through the foreign tax credit in taxable accounts but which remains inside most tax-advantaged wrappers, per Internal Revenue Service guidance on the credit's scope.
Plan defaults matter at least as much. Retirement menus in the United States long offered a domestic equity fund as the recognizable core line, with international options appearing later and less prominently, so a participant following defaults inherits a domestic-heavy allocation. Vanguard's plan-level series documents how slowly those inherited tilts move even after inexpensive global options appear on the menu.
Which explanations have researchers proposed?
Four families of explanation dominate the published record. Each has documented support and a documented weakness, summarized in the table below; none has closed the question alone.
| Explanation | Documented basis | Documented weakness |
|---|---|---|
| Information advantage | Ahearne, Griever, and Warnock (2004) linked home bias to information costs: foreign firms are less visible to local investors | Institutional investors with research access show the same tilt |
| Consumption hedging | Domestic assets can hedge local purchasing-power risk when real returns co-move with local conditions | Estimated correlations between local equity and local consumption are too weak to justify observed tilts |
| Explicit costs | Withholding taxes, custody fees, and historical trading costs raise the price of foreign holdings | Costs fell for decades while the bias barely moved |
| Familiarity (behavioral) | Huberman's 2001 study of regional telephone companies found investors favored their local Bell over otherwise comparable ones | A preference for the familiar is not a complete account of magnitude across all markets |
Contemporary work typically combines the families: modest information and cost advantages, amplified by behavioral familiarity and inertia. The weights assigned to each remain unsettled in the literature.
What does home bias cost a portfolio?
The reference point in most studies is the global market portfolio, from which a domestic-only portfolio deviates by construction. Goetzmann and Kumar's 2008 analysis of individual investor accounts documented that under-diversified portfolios carried higher idiosyncratic volatility without compensating return — a risk cost paid in volatility rather than in a visible fee.
An illustrative calculation shows how the size of that cost depends on co-movement. Assume a domestic equity market at 16 percent volatility and a global portfolio at 15 percent; if their correlation is 0.85, a blend barely lowers volatility below either holding, but at a correlation of 0.60 the same blend lowers it materially — all assumptions stated, arithmetic only, no forecast. The benefit of crossing borders scales with how independently markets move, which is itself unstable over time.
The return cost depends on the window measured. In 2023 and 2024 the S&P 500 returned 26.3 percent and 25.0 percent, per S&P Dow Jones Indices, and any foreign allocation lagged a domestic-only holding in those two years. Earlier stretches — including long runs when non-U.S. markets led — reverse the comparison. The published record contains both directions and projects neither forward.
Correcting the tilt has its own documented price: currency volatility. Unhedged foreign equity returns fluctuate with exchange rates, and hedging that exposure costs the interest-rate differential between currencies, a cost that varies over time. Geographic diversification is therefore not free of its own risks, which is part of why the puzzle persists.
Why has the debate refused to close?
Direct costs of foreign investing collapsed in the 2010s and 2020s — broad international index funds now charge single-digit basis points in many cases — yet measured allocations moved slowly. Inertia of that size is itself a documented finding: defaults in retirement plans and the familiarity of domestic brand names both anchor portfolios where they start.
The open question is whether the tilt is a mistake with a measurable cost or a rough hedge for liabilities denominated in the home currency. The research contains evidence for both readings, and the site presents them side by side. Readers' own currency obligations sit on one side of that ledger, and those circumstances vary too widely for any single answer.
For more context, read Glide Paths: Age-Based Allocation and the Debates.
For more context, read measuring diversification.
For more context, read The 60/40 Portfolio: The History and the Debate.
