Do You Really Need International Exposure? (VXUS Explained)

Photo by Vito Natale on Unsplash

The question keeps coming back: if the US has carried global markets for the last fifteen years, is there any point holding international equities at all? It is a reasonable thing to ask, and the answer deserves something more than "diversification is good." This piece works through what VXUS actually contains, what the realized numbers look like through May 2026, and where the cases for and against a meaningful international weight both hold up.

Context: what VXUS actually represents

Vanguard Total International Stock Index Fund ETF (VXUS) tracks the FTSE Global All Cap ex US Index. That index covers approximately 8,500 stocks across developed markets (Europe, Japan, Canada, Australia, UK), emerging markets (China, India, Taiwan, Korea, Brazil), and a thin sleeve of frontier markets — effectively everything investable that is not a US-listed equity. The fund launched in late 2010 and the underlying strategy now manages roughly $629B across all share classes (Vanguard issuer page, May 2026), making it one of the largest non-US equity vehicles in the world.

The expense ratio is 0.05% — three basis points wider than VTI's 0.03%, and a rounding error in any practical horizon. The bid-ask spread is also tight enough that retail implementation friction is minimal. This is not where international ETFs get expensive; the cost argument against VXUS rarely has any teeth.

VXUS by the numbers (through May 2026)

Sources: yfinance (price and total-return series, pulled 2026-05-18), Vanguard VXUS product page for fee, AUM, and inception. Macro reference: 10-year Treasury 4.47% and Fed Funds 3.64% (FRED, asof 2026-05-14).

The realized gap with US large-cap

Over the same five-year window, US large-cap equities posted materially higher returns. That part nobody disputes. Where the discussion usually breaks down is what to do with that fact going forward. Two readings deserve attention.

The first is that twelve-month or even five-year return gaps tell you very little about the next decade. Rolling ten-year regressions between MSCI EAFE and the S&P 500 since 1970 show alternating leadership — international led the 1970s and most of the 2000s; the US led the 1980s and the 2010s through the mid-2020s. Each regime felt durable while it lasted. The lesson is not that international "will" catch up, but that single-regime extrapolation is a well-documented statistical trap.

The second reading is more uncomfortable: even after the recent US run, VXUS's 9.7% ten-year CAGR is not a bad number in absolute terms. Investors who hold international expecting it to "win" are usually disappointed. Investors who hold it expecting roughly equity-like long-run compounding with different correlation properties have generally gotten what they paid for.

The currency layer — what most readers underestimate

VXUS is unhedged. That means a meaningful fraction of its measured volatility and a meaningful fraction of its return is currency, not underlying business performance. When the dollar strengthens (as it did through much of 2021–2024), international equities denominated in EUR, JPY, and CNY translate into fewer dollars regardless of how the underlying companies performed. When the dollar weakens, the same translation works in reverse.

This is a feature, not a bug, for a US-based investor. The currency exposure is also exposure to economies whose policy cycles, inflation regimes, and rate paths differ from the Fed's. With the 10-year Treasury at 4.47% and the Fed Funds rate at 3.64% (FRED, asof 2026-05-14), a US-only equity portfolio is linked to those domestic rate paths through nearly every channel — discount rates, dollar strength, corporate financing costs. VXUS exposes the portfolio to a different mix.

Whether that currency exposure is desirable depends on the investor's liability structure. Someone who will spend exclusively in USD has less obvious need for foreign-currency assets than an investor whose long-run plans involve cross-border spending. The point is to make the choice explicitly rather than assume the dollar-denominated portfolio is the "neutral" position. It isn't — it's a 100% USD bet by default.

What diversification actually does to the return distribution

The case for adding 10–30% VXUS to a US-heavy core is not "VXUS will outperform." It's that the joint distribution of (US-only) versus (US + ex-US) over multi-decade horizons has tighter tails and similar means. Historically, the correlation between MSCI EAFE and the S&P 500 in monthly returns has hovered around 0.75–0.85 — high enough that the diversification benefit is real but limited. Adding emerging markets, which VXUS does, lowers that effective correlation slightly more.

The non-obvious effect: international exposure mostly helps in the regimes where US-only portfolios feel worst — extended sideways periods, dollar weakness, US-specific policy shocks. It tends to look like dead weight precisely when the US is leading, which is exactly when investors are most tempted to cut it. This is the same pattern that breaks most factor allocations: the diversification works on the schedule the math predicts, not the schedule investor patience predicts.

For a more granular look at how this plays out alongside a US core, the editor previously worked through the math in Beyond a One-ETF Equity Core and the direct head-to-head in VXUS vs VOO.

International exposure mostly helps in the regimes where US-only portfolios feel worst — extended sideways periods, dollar weakness, US-specific policy shocks. It looks like dead weight precisely when the US is leading.

The honest case against a large international weight

An honest analyst has to acknowledge the counter-arguments, because they are not trivial.

First, US-listed multinationals already derive significant revenue from outside the US. The top 100 S&P 500 companies generate roughly 40% of revenue internationally. A US-only investor is not 100% domestically exposed in economic terms — they are heavily exposed to US corporate governance, US currency, and US regulatory regimes, but the underlying earnings have substantial global breadth.

Second, governance, accounting transparency, and capital-allocation quality differ across markets in ways the headline numbers do not fully capture. The argument that "US companies have outperformed because they are systematically better-run" has some empirical support, even if it is hard to prove will persist.

Third, for investors with limited capital or tight tax-advantaged space, the marginal complexity of holding VXUS alongside a US core may not be worth the small expected variance reduction. A simpler portfolio that gets actually held through stress beats a theoretically optimal one that gets abandoned.

None of these arguments make VXUS a bad fund. They make the right allocation question "how much" rather than "yes or no" — and the right allocation almost certainly is not zero, but it almost certainly is not 50/50 either for a US-based investor with USD liabilities.

What this analysis can and can't tell you

Five-year and ten-year windows ending in May 2026 cover the post-2015 US dominance regime and a portion of the post-pandemic period. They do not include the 2000–2010 decade when international substantially outperformed US large-cap, nor the 1970s currency regime. A reader who weights the available window too heavily will underestimate international's variance — both upside and downside — relative to a longer-run view.

The data also cannot tell you what currency regime you will live through. The framework here is conditional: if continued, structural US dominance is the base case, the marginal case for VXUS weakens. If leadership rotates on cycles that are not reliably predictable, VXUS becomes a hedge against a regret state that is otherwise difficult to insure against cheaply.

Scenarios where VXUS fits — and where it doesn't

A reader in their 30s building a long-horizon equity core in a tax-advantaged account: a 15–25% VXUS weight alongside a US large-cap base is defensible, with the percentage scaling toward the higher end the longer the horizon and the lower the conviction in continued US leadership.

A reader in their late 50s already drawing income: the case is weaker. Sequence-of-returns risk dominates, and the marginal diversification benefit from VXUS is smaller than the marginal benefit of reducing equity risk overall. International here usually belongs inside a smaller equity sleeve, not added on top.

A reader with a US-only 401(k) menu and no IRA: this is the genuinely hard case. Forcing VXUS through a taxable account creates real friction (foreign tax credit paperwork, a higher share of non-qualified ordinary distributions). Sometimes the cleanest answer is to accept the home-country tilt rather than create implementation complexity that may not survive a stress period.

Frequently asked questions

Is VXUS hedged to USD?
No. VXUS is unhedged, which means returns reflect both underlying equity performance and currency movements relative to the dollar. Hedged international funds exist (HEFA, DBEF, etc.), but at higher expense ratios and with their own embedded forward-rate risk.

How does VXUS handle emerging markets versus a developed-markets-only fund?
VXUS includes both developed and emerging markets at roughly market-cap weights — historically about 75% developed, 25% emerging. Investors who want explicit control over that ratio sometimes pair VEA (developed) with VWO (emerging) instead.

What is VXUS's dividend yield, and how is it taxed?
Trailing yield is approximately 2.8% as of May 2026 (yfinance, 2026-05-18). A meaningful portion of VXUS distributions arrives as non-qualified ordinary dividends because of source-country rules, which makes the fund somewhat less tax-efficient than a pure US large-cap ETF in a taxable account. Many investors place VXUS in tax-advantaged space for this reason.

Has VXUS underperformed US equities consistently since inception?
Over the single window from November 2010 to May 2026, yes — VXUS has trailed the S&P 500 by a wide margin. Whether that gap reflects a structural shift or a regime that will mean-revert is the entire debate. Pre-inception data for the equivalent indices shows multiple decade-long periods when international led.

Can I just hold US multinationals and call it international exposure?
Partly, but not fully. US multinationals capture global revenue but not foreign currency exposure, not foreign governance regimes, not foreign monetary policy paths. The diversification you get from US multinationals is real, but narrower than holding direct international equities.

Key takeaways

• VXUS is a low-cost (0.05%) vehicle for roughly 8,500 non-US equities across developed and emerging markets — the operational case against it on fee, breadth, or liquidity grounds is weak.
• Realized 10-year return of 9.7% trails US large-cap but is not a poor absolute outcome. The useful question is allocation size, not inclusion versus exclusion.
• Unhedged currency exposure is a feature for investors who want to reduce single-currency dependency, and friction for investors with strictly USD liabilities.
• The diversification benefit shows up in regimes that are uncomfortable to sit through. That is the price of the insurance, not a sign the position is broken.
• For most long-horizon US-based investors, a VXUS weight somewhere in the 10–25% range of the equity sleeve is defensible; the exact number should reflect horizon, account type, and conviction in continued US leadership.

The editor holds VXUS as part of a long-term equity allocation at the time of writing.

Methodology: price and return data via yfinance (daily total-return series); expense ratio, AUM, and inception via the Vanguard VXUS issuer page; macro context via FRED. Data pulled 2026-05-18. Windows analyzed: trailing 5 and 10 years through May 2026.

This article is for educational purposes and does not constitute personalized financial advice. See Disclaimer.