The measurable benefit of international diversification is set by three quantities: the correlation between home and foreign returns, the share of global market capitalization the home market represents, and the dispersion of returns across countries — and on U.S. data, long-run correlations of monthly developed-market returns with U.S. returns have run roughly in the 0.5 to 0.9 range depending on decade and market pair, high enough to blunt but not eliminate diversification's effect. UZU NEWS publishes information, not investment advice, and reports the numbers with their windows attached.
"Should I hold international?" is an allocation question that belongs to the reader. The measurable questions underneath it are narrower and answerable: how correlated are the markets, what fraction of the opportunity set does the home market miss, and what has the realized dispersion between markets been. The answers have all moved over time, which is itself the main finding.
What does correlation evidence show?
Rolling correlations between U.S. and developed international equity returns, computed on monthly data across the past half-century, have oscillated in the 0.5-to-0.9 corridor: lower in the 1970s and 1980s, rising through the 2000s, spiking toward the top of the range in crises — 2008 being the canonical case, when diversification within equities failed precisely when needed, as nearly all equity markets fell together. The pattern has a name, correlation asymmetry: correlations rise more in down markets than in up markets, documented across multiple studies since Longin and Solnik's work in 2001. The practical consequence is exact: international equities diversify the ordinary weeks and disappoint in the worst ones.
What does market-cap coverage show?
The United States has represented roughly 40 to 60 percent of global equity market capitalization across recent decades, with the exact figure moving with relative performance — the U.S. share rose through the 2010s as U.S. markets outperformed. A U.S.-only portfolio therefore holds a large fraction but not all of the world's listed companies, and the missed fraction has historically included whole markets — Japan in the 1980s, energy-heavy indexes in commodity booms — whose cycles differed from the U.S. cycle. Coverage is the least controversial argument for international exposure because it is arithmetic, not a forecast.
What does return dispersion show?
Realized long-run returns across national markets have dispersed widely. The canonical cautionary pair: Japanese equities returned spectacularly through the late 1980s and then spent decades underwater from their 1989 peak, while U.S. markets compounded steadily through the same decades — a dispersion any home-market-only investor experienced or avoided entirely depending on which home they had. More routinely, year-to-year leadership between U.S., developed ex-U.S. and emerging markets has rotated with multi-year runs in each direction: the 2000s favored non-U.S. broadly, the 2010s favored the U.S. broadly, in both cases by wide margins. Dispersion is measured fact; which side of it any future period delivers is not.
What costs and frictions offset the benefit?
Four, all quantifiable. Currency: unhedged foreign holdings carry currency swings that add volatility at the portfolio level; hedging reduces it at a cost, and the hedging decision is itself regime-dependent — a large literature with no settled one-size answer. Taxes: foreign withholding on dividends, with credit mechanics depending on account type and treaty. Costs: international funds historically carried higher expense ratios and trading frictions than domestic broad funds, though the gap has narrowed. Tracking and structure: some markets remain expensive to access through index vehicles. Each friction is smaller today than in 1990; none is zero.
| Quantity | Recent evidence range | Behavior over time |
|---|---|---|
| U.S.–developed monthly correlation | ~0.5–0.9 by decade and pair | Rises in crises; asymmetry documented |
| U.S. share of global market cap | ~40–60% | Drifts with relative performance |
| Annual market leadership rotation | Multi-year runs by region | U.S. in 2010s; ex-U.S. in 2000s |
| Crisis correlation | Toward the top of range | Diversification within equities weakens |
Where does diversification talk mislead?
Three recurring slips. First, quoting a single average correlation without a window: the number has spanned 0.5 to 0.9, and any conclusion built on one value inherits that fragility — the same window discipline this site applies to every rolling statistic. Second, treating international as an asset class rather than a collection of markets: Japan, the U.K., Germany and Korea have themselves dispersed widely; the aggregate hides the dispersion that was the point. Third, promising crisis protection from equity diversification: the measured fact is the opposite in the worst sessions — correlations converge upward — and honest commentary says so before the next crisis, not after.
What can a reader verify directly?
Correlations and dispersion from public return series — the Federal Reserve Economic Data service and index providers publish the underlying series — and market-cap weights from index fact sheets. The Securities and Exchange Commission's international-investing materials at investor.gov catalog the risks and frictions in standardized form. What no source can verify is the future path of any of the three quantities, which is why this article reports their history and their movement, and stops there.
For more context, read What is sequence-of-returns risk, measured properly?.
For more context, read fund turnover ratio.
For more context, read Why doesn't your index fund match its index?.
