World benchmark comparison methodology
Published · Research based on rules reviewed 13 October 2026
This study compares six benchmark definitions and highlights FTSE World as an additional naming exception. The analytical framework uses two axes: eligible market classifications and company-size segments. Category similarity is not a claim of identical constituent membership or weights.
| Benchmark | Eligible markets | Size segments | Includes small caps |
|---|---|---|---|
| MSCI World | MSCI developed markets | Large + mid | No |
| FTSE Developed | FTSE developed markets | Large + mid | No |
| MSCI ACWI | MSCI developed + emerging | Large + mid | No |
| FTSE All-World | FTSE developed + advanced/secondary emerging | Large + mid | No |
| MSCI ACWI IMI | MSCI developed + emerging | Large + mid + small | Yes |
| FTSE Global All Cap | FTSE developed + advanced/secondary emerging | Large + mid + small | Yes |
Country examples and implementation dates
| Country | MSCI treatment | FTSE treatment |
|---|---|---|
| South Korea | Emerging | Developed |
| Poland | Emerging | Developed |
| Greece | Emerging as of research date; developed in May 2027 review | Developed effective 21 September 2026 |
| Vietnam | Not in the 30 September 2026 ACWI country list | Secondary emerging; initial 10% implementation 21 September 2026, further tranches March/June/September 2027 |
MSCI’s 30 September ACWI factsheet lists 23 developed and 24 emerging markets. FTSE’s classification matrix is dated 6 October 2026. Greece’s current FTSE treatment is confirmed by its annual announcement published 6 October, rather than inferred only from a previous intention. Vietnam’s FTSE transition begins at 10% implementation and continues in three tranches; classification status must not be confused with full index-weight implementation.
The FTSE matrix footnote calls Greece’s former category Secondary Emerging, whereas the annual announcement says Advanced Emerging. We rely on the current Developed classification and the annual announcement’s implementation date; we do not repeat the inconsistent historical label. The ground-rules cover says September 2026 while one running header says August; the retained document is identified as v14.4 with both observations recorded.
Coverage and weighting definitions
MSCI GIMI, August 2026, sets Large Cap, Standard and IMI reference targets at 70%, 85% and 99%, subject to market-specific construction and buffers (PDF pages 25–26). MSCI World describes approximately 85% of free-float-adjusted market capitalisation in each included country. MSCI ACWI and ACWI IMI profiles describe approximately 85% and 99% of their global investable opportunity sets respectively.
FTSE’s February 2026 GEIS FAQ describes a regional construction process: the top 98% becomes the regional index universe, with large and mid-cap bands of 70% and 20% within the eligible universe, subject to buffers (PDF page 5). Its product highlights describe All-World targeting approximately 90% and Global All Cap approximately 98% of the global investable equity universe. These are provider descriptions, not a common-denominator calculation. Do not subtract 85 from 90 and call the result measured extra global exposure. No coverage bar chart with a shared global denominator is supplied.
Both families use investability screens and free-float/foreign-ownership adjustments. MSCI’s country-based construction and FTSE’s regional size segmentation can classify securities differently. Listing venue, assigned country, corporate revenue geography and currency exposure are separate variables. Market-cap weighting is not equal weighting across countries or holdings.
Dated counts and what they do not prove
The MSCI profiles/factsheet report 1,249 constituents for World, 2,414 for ACWI and 8,036 for ACWI IMI at 30 September 2026. These are provider constituent counts, not ETF holdings counts or consolidated parent-company counts. The retained FTSE product-highlights table is dated September 2025, so its counts are excluded from a current cross-provider ranking.
Public aggregate data do not provide a complete, aligned cross-provider constituent reconciliation here. No security overlap percentage, incremental portfolio weight or causal return/risk comparison is claimed. Existing Rogha ETF snapshots remain useful product evidence, but sampling, dates, cash and implementation prevent treating them as authoritative full index lists. This rules-led article can stand on its documented definitions; a separate quantitative benchmark-overlap study needs aligned constituent data and applicable data-use rights.
Scope and status
No performance series is compared, so no mixed gross/net/price-return ranking is introduced. Index-currency variants and currency-hedged variants must be distinguished in future return work. Broad standard benchmarks are in scope; screened/custom versions, bonds, private markets and gold are outside the comparison.
Research does not amend the Irish ETF directory, the frozen holdings release or deployed feeds. Source files and hashes are retained internally. Public methodology links point to official documents; live URLs may change. Any public reuse of provider statistics or graphics should follow Rogha’s existing source-rights review. This pack is not a public download.
Official sources
- msci-world.html — retrieved 13 October 2026; retained with SHA-256 hash.
- msci-acwi.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- msci-imi.html — retrieved 13 October 2026; retained with SHA-256 hash.
- msci-gimi.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- msci-summary.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- msci-greece.html — retrieved 13 October 2026; retained with SHA-256 hash.
- ftse-ground-rules.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- ftse-country-2026.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- ftse-matrix.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- ftse-faq.pdf — retrieved 13 October 2026; retained with SHA-256 hash.
- ftse-highlights.pdf — retrieved 13 October 2026; retained with SHA-256 hash.