Overlap · from SEC filings
FXI vs KWEB: what actually overlaps
FXI and KWEB: 21.02% of a dollar ends up in the same companies either way (SEC Form N-PORT, 2026-04-30 and 2026-03-31), across 5 securities the two funds both hold. Every free comparison tool can list the tickers in common; this weights them, which is the part that decides whether owning both changes anything.
The short version
Held alone, FXI behaves like 21.7 equally weighted positions and KWEB like 19.7. Held together in equal size they behave like 31.9 — 10.2 more than FXI on its own, even at 21.02% overlap.
Overlap here is the sum, over securities both funds hold, of the smaller of the two weights: the share of a dollar that sits in the same company either way. Matching is on CUSIP. Figures are as of each fund's own reporting date (2026-04-30 and 2026-03-31), which do not always align.
Not 50/50, or holding a third fund alongside these two? The portfolio calculator runs this same math for any weights, across up to 12 funds.
Side by side
as filed| FXI | KWEB | |
|---|---|---|
| Assets | $6B | $5B |
| Holdings | 52 | 31 |
| Effective holdings | 21.7 | 19.7 |
| Top ten weight | 58.83% | 60.21% |
| Largest position | 9.41% | 10.47% |
| Priced at market (Level 1) | 1.36% | 100% |
| Reporting date | 2026-04-30 | 2026-03-31 |
Where the overlap actually is
5The ten shared positions that account for the most of it, with what each fund holds of its own. Shared is the lesser of the two — the part that is held in common either way. Where the two columns diverge, one fund is making a larger bet on that company than the other, and holding both does not average it away.
| Company | FXI | KWEB | Shared |
|---|---|---|---|
| Tencent Holdings Limited | 7.717% | 10.099% | 7.717% |
| MEITUAN | 4.787% | 7.364% | 4.787% |
| JD.COM, INC. | 3.001% | 4.919% | 3.001% |
| BAIDU, INC. | 2.831% | 4.284% | 2.831% |
| TRIP.COM GROUP LIMITED | 2.687% | 3.568% | 2.687% |
What the difference is made of
78.98% · 69.77%78.98% of FXI sits in companies KWEB does not hold at all, and 69.77% of KWEB in companies FXI does not hold. The largest of each, by weight in its own fund:
- Alibaba Group Holding Limited
- CHINA CONSTRUCTION BANK CORPORATION
- INDUSTRIAL AND COMMERCIAL BANK OF CHINA LIMITED
- XIAOMI CORPORATION
- Ping An Insurance (Group) Company of China, Ltd.
- Alibaba Group Holding Limited
- PDD Holdings Inc.
- NETEASE, INC.
- Ke Holdings Inc
- JD Health International Inc
By country of issuer
2026-04-30 · 2026-06-30Where each fund's money sits by the issuer country its filer reported (incorporation, not revenue). Two funds can share companies and still differ here, and a pair that looks diversified by name can be one country twice.
| Country | FXI | KWEB | Difference |
|---|---|---|---|
| Hong Kong SAR China | 0.77% | 72.25% | −71.48 |
| China | 53.38% | 0.00% | +53.38 |
| Cayman Islands° | 44.30% | 0.00% | +44.30 |
| United States | 0.06% | 27.75% | −27.69 |
| Switzerland | 1.49% | 0.00% | +1.49 |
What each fund did last quarter
Each fund's newest filing against its previous one. A pair whose overlap is rising or falling does so because of these moves; the figures are the same ones on each fund's page and in its feed.
Written about this pair
1This page is the measurement. These essays argue about what it means, with the figures as they stood when each was written.
Related comparisons
4Other pairs involving FXI or KWEB, ordered by how much they overlap.
| Pair | Overlap | Combined effective |
|---|---|---|
| FXI vs MCHI | 52.76% | 28.4 |
| KWEB vs MCHI | 19.18% | 32 |
| EWT vs KWEB | 0% | 36.9 |
| EWY vs KWEB | 0% | 19 |
Sources and limits
Computed from SEC Form N-PORT (NPORT-P), the portfolio report every registered US fund files. Positions below 0.001% of net assets are dropped; weights are normalised to each filing's own total.
Unit investment trusts including SPY and DIA do not file N-PORT, so no overlap can be computed for them. Where a comparison is missing from this site, that is usually why rather than an oversight.
Overlap says what is held in common, not what will move together. Two funds with no shared holdings can still fall together, and the effective holding count takes no account of correlation. All 275 pairs are in the concentration dataset, and every fund's own history is on its fund page.