What is home bias and why is it costly?

The short answer

Home bias is one of the most documented anomalies in international finance. Despite the clear theoretical case for global diversification, investors in nearly every country hold disproportionately domestic portfolios. The bias has weakened modestly over the past three decades but remains substantial.

The traditional argument against home bias is straightforward: global diversification reduces variance for any given expected return. But this argument has weakened over time as global equity correlations have risen — particularly in the 2008-2024 period when major equity markets often moved together with correlations of 0.7-0.9.

The counterintuitive insight is that the diversification cost of home bias is highest precisely when global correlations are low, and lowest when they are high. This means the structural underweight to international assets hurts investors most in calm regimes and least in crisis regimes — the opposite of what diversification narratives typically suggest.

→ New to international investing? Investing for Beginners Hub

What the data shows

Home bias measurement combines holdings data, market cap weights, and correlation analyses across decades.

The numerical context (French-Poterba 1991, Coeurdacier-Rey 2013, IMF CPIS 2000-2023) :

• 1989 — U.S. investor allocation to U.S. equities: ~94%; U.S. market-cap weight in global equities: ~48% (French-Poterba)
• 2023 — U.S. investor allocation to U.S. equities: ~78%; U.S. market-cap weight in global equities: ~62% (IMF CPIS)
• French investor allocation to French equities: ~55-65% over 2000-2020 vs French market-cap weight ~3-4% globally
• Average global equity correlation 1990-1999: ~0.4; average 2010-2024: ~0.7
• Theoretical Sharpe ratio improvement from full global diversification: ~0.05-0.15 in low-correlation regimes; ~0.02-0.05 in high-correlation regimes

The exception : during currency crises and idiosyncratic local shocks (Asian crisis 1997, Greek crisis 2010, Turkish lira 2018), domestic concentration produces severe localized losses that global diversification would have avoided. Home bias is most costly precisely when local idiosyncratic risk materializes.

→ Dataset: U.S. Dollar Index Dataset

Why it happens — the macro mechanism

Three mechanisms explain the persistence of home bias despite known costs.

Familiarity heuristic channel. Investors prefer assets they perceive as familiar — domestic companies they read about daily, local economic conditions they understand intuitively. Huberman (2001) documented that this familiarity preference operates even within countries — investors over-weight stocks of their state, their employer, and the companies they encounter as customers.

Currency mismatch channel. Domestic investors typically have liabilities denominated in domestic currency (mortgages, retirement spending, education costs). Foreign equity exposure introduces currency risk that has no offsetting liability hedge. Some portion of home bias is rational from this asset-liability matching perspective.

This second channel connects directly to global correlation dynamics.

Information cost channel. Foreign equity research is costlier to produce, slower to access, and harder to verify than domestic research. The cost differential creates a rational allocation tilt toward domestic markets even for investors aware of the diversification trade-off. Globalization of information access has weakened but not eliminated this channel. Anchoring effects reinforce the asymmetry by making domestic price levels feel more meaningful than foreign ones.

Synthesis by regime : in low-correlation regimes (1990s when emerging markets and developed markets moved largely independently), home bias produced substantial diversification opportunity costs of roughly 50-100 basis points annually in Sharpe ratio terms; in moderate-correlation regimes (2000-2007 with average correlations near 0.5), the cost falls to roughly 20-50 basis points; in high-correlation regimes (2008-2024 with correlations frequently above 0.7), the cost narrows further to roughly 5-20 basis points. The transition between regimes is governed by global liquidity conditions and the dominance of synchronized monetary policy — when central banks move in unison, equity correlations rise globally, and the home bias penalty shrinks.

Home bias hurts most in calm markets and least in crises — the opposite of what diversification stories tell.

→ Framework: Portfolio allocation pillar

What it means for different economic actors

Retail investors. Home bias is most pronounced in this group, with allocations frequently reaching 70-90% domestic regardless of investor sophistication. The cost is highest for investors in small markets where idiosyncratic concentration risk is largest.

Institutional asset owners. Pension funds and endowments display measurable home bias but at smaller magnitudes. Regulatory frameworks (e.g., currency hedging requirements, prudential rules) often institutionalize partial home bias even when investment philosophy favors global allocation.

Sophisticated global investors. Sovereign wealth funds, large endowments, and global multi-asset managers operate with the smallest home bias coefficients — but at the cost of substantial currency-management infrastructure and operational complexity. Loss aversion in domestic-currency terms remains a documented constraint even in this group.

A common error is to treat home bias as a binary choice between full global cap-weighting and full domestic concentration. The optimal approach historically falls between these extremes and varies with the correlation regime, currency hedging costs, and investor-specific liability structure.

Practical observation

Go deeper

Related questions

• What is loss aversion and how does it affect investment decisions?
• How do anchoring effects affect price perception?
• How do asset class correlations shift across regimes?
• What is strategic vs tactical asset allocation?

Frequently asked questions

Is home bias always irrational?

No. A portion of home bias reflects rational asset-liability matching: investors with domestic-currency liabilities hedge currency risk by holding domestic-currency assets. Information cost differentials and tax considerations also justify some allocation tilt. The component that is genuinely costly is the residual — the over-allocation beyond what these rational factors explain. This residual is consistently documented across countries and decades, but its magnitude has shrunk over time.

Why does home bias matter less in high-correlation regimes?

Because the diversification benefit comes from low-correlation co-movement. When all major equity markets move together with correlations of 0.8-0.9, the additional Sharpe ratio gain from global diversification is mathematically small — perhaps 5-20 basis points. In low-correlation regimes (1990s), the gain was 50-100 basis points. The economic case for global diversification weakens precisely in the regimes investors associate with crisis, which creates the counterintuitive timing problem.

Has home bias decreased over the past 30 years?

Yes, modestly. U.S. investor allocation to U.S. equities has fallen from ~94% in 1989 to ~78% in 2023, but remains well above the global market-cap weight of ~62%. European investors have similarly reduced home bias, though variation across countries is substantial. The persistent gap suggests that information access alone cannot close the bias — institutional, behavioral, and asset-liability factors continue to support meaningful home concentration.

Last updated — 22 May 2026