When you buy an index, you're not buying diversification — you're increasingly buying concentration into one trade (AI/semiconductor).
| Index | Const | Eff-N | Top-10 | AI-core | Top-10% | Largest |
|---|---|---|---|---|---|---|
| KOSPI (Korea) | 828 | 7.0 | 64.5% | 52.6% | 90.4% | Samsung Electronics 28.8% |
| DJIA (US, price-wtd) | 30 | 21.2 | 55.1% | 18.2% | n/a | Goldman Sachs 10.5% |
| Nifty 50 (India) | 50 | 24.3 | 52.9% | 0% | 36.4% | HDFC Bank 9.9% |
| NASDAQ-100 | 101 | 31.7 | 46.8% | 57.1% | 46.8% | NVIDIA 8.4% |
| EURO STOXX 50 | 50 | 30.1 | 43.4% | 16.7% | 27.4% | ASML 10.8% |
| S&P 500 | 504 | 43.5 | 39.1% | 39.0% | 64.4% | NVIDIA 8.4% |
| MSCI EM | 1,153 | 27.4 | 39.0% | 37.4% | 66.7% | TSMC 15.6% |
| MSCI World | 1,239 | 83.7 | 27.8% | 29.7% | 61.2% | NVIDIA 5.8% |
| TOPIX (Japan) | 1,636 | 105 | 23.7% | 6.1% | 79.0% | MUFG 3.9% |
Concentration measures computed by us from each index's own full constituent file. MIXED BASIS — read the per-row date: S&P 500, EURO STOXX 50, DJIA, MSCI World, MSCI EM, NASDAQ-100 and KOSPI at 30-Sep-2026; Nifty 50 and TOPIX at 31-Aug-2026 (September primaries unpublished at capture). Eff-N = how many equal-weight stocks the index really behaves like. AI-core = share of the whole index in the AI complex — chips + hyperscalers, tracked as a live series per index.
The reading — and its limits
Three exceptions prove the point — Japan (TOPIX) is genuinely diversified (behaves like 105 stocks, AI-core just 6.1%); India (Nifty) is concentrated but in the old economy (AI-core 0% — its "tech" is IT services, the AI-disrupted category, not AI hardware); and the Dow (DJIA) sits near the top by top-10 (55.1%, second only to Korea) but this is a price-weighting artefact — its weights track nominal share price, not company size, so two high-priced blue-chips (Goldman, Caterpillar) dominate and the AI names (Nvidia, Microsoft) are muted. The Dow's low AI-core (18.2%) does NOT mean it escapes the AI trade — it holds the same names, down-weighted by share-price accident. Everywhere else, buying the index increasingly means buying the same handful of AI names — and at September's close NVIDIA was back as the largest single holding in the S&P 500, NASDAQ-100 and MSCI World.
Method Effective-N and concentration measures computed by us from each index's own full constituent file (index-owner or full issuer-holdings files). AI-core on a named-basket definition (semiconductors + hyperscalers) that crosses GICS sectors.
Computed 30 September 2026