AHNIM STUDIO | Guru statistics | Reviewed October 26, 2026 (KST)
Did the 17 disclosed portfolios become more diversified from Q1 to Q2 2026? Mostly—but a larger list of holdings was not enough by itself. Effective diversification, measured as 1 divided by issuer-weight HHI, rose in 11 of 17 reporting groups and fell in six. The equal-manager median increased from 13.26 to 16.71 effective holdings, while the median Top 10 weight fell from 77.33% to 67.06%. Yet Dalal Street added one eligible issuer and became slightly more concentrated, while Tiger Global disclosed seven fewer issuers and became more evenly weighted. Form 13F name counts therefore describe breadth; they do not fully describe concentration.
Three findings from the matched Q1–Q2 panel
- Broadening was common, not universal: effective holdings rose for 11 groups, and the equal-manager mean increased from 19.26 to 23.31.
- More names usually helped, but did not guarantee diversification: 10 of 11 groups with more eligible issuers also gained effective holdings; Dalal Street was the exception.
- Fewer names did not guarantee concentration: Tiger Global’s eligible issuer count fell from 53 to 46, but effective holdings rose from 16.02 to 17.69 as its Top 10 weight declined by 2.43 percentage points.
The Guru research and statistics hub tracks the same 17 institutional reporting groups. Our earlier breadth study asked which issuers appeared in more portfolios. This study asks a different question inside each filing: how evenly was eligible disclosed value distributed across issuers?
HHI turns portfolio weights into an effective holding count
For each reporting group and quarter, we calculate every issuer’s weight in eligible disclosed operating-company equity, square those weights and sum them. That sum is the Herfindahl–Hirschman Index: HHI = Σwi2. A lower HHI means value is spread more evenly. Its reciprocal, 1/HHI, is the effective number of equally weighted holdings that would produce the same concentration.
A filing can contain many small positions and still be dominated by a handful of issuers. Conversely, a portfolio can remove small names while distributing value more evenly across its remaining core. The effective count is not the literal number of issuers and should not be read that way. It is a concentration equivalent.
| Panel statistic | Q1 2026 | Q2 2026 | Change |
|---|---|---|---|
| Mean eligible issuer count | 103.24 | 105.35 | +2.12 |
| Median eligible issuer count | 29 | 35 | +6 |
| Mean effective holdings | 19.26 | 23.31 | +4.05 |
| Median effective holdings | 13.26 | 16.71 | +3.45 |
| Mean Top 10 weight | 72.61% | 68.33% | -4.28 pp |
| Median Top 10 weight | 77.33% | 67.06% | -10.27 pp |
Bridgewater’s very large disclosed list lifts the mean issuer count far above the median, so we report both. The direction is consistent across the median effective count and median Top 10 weight: the typical group in this selected panel was less concentrated in Q2. However, six groups moved the other way, preventing a blanket “gurus diversified” conclusion.
All 17 reporting groups: name count, effective holdings and Top 10 weight
| Reporting group | Eligible issuers Q1 → Q2 | Effective holdings Q1 → Q2 | Effective change | Q2 Top 10 |
|---|---|---|---|---|
| Icahn reporting group | 12 → 12 | 3.10 → 2.81 | -0.29 | 98.58% |
| Dalal Street | 3 → 4 | 2.94 → 2.87 | -0.07 | 100.00% |
| Situational Awareness | 25 → 23 | 7.17 → 5.63 | -1.54 | 92.10% |
| Berkshire Hathaway | 26 → 26 | 8.40 → 8.14 | -0.27 | 91.04% |
| Pershing Square | 10 → 13 | 7.08 → 10.13 | +3.05 | 94.30% |
| Altimeter | 12 → 19 | 6.36 → 10.37 | +4.01 | 83.28% |
| Appaloosa | 29 → 24 | 13.26 → 10.48 | -2.78 | 83.07% |
| Greenlight / DME | 43 → 45 | 15.68 → 16.63 | +0.95 | 53.26% |
| Light Street | 23 → 26 | 13.25 → 16.71 | +3.46 | 68.20% |
| Tiger Global | 53 → 46 | 16.02 → 17.69 | +1.67 | 67.06% |
| Third Point | 28 → 35 | 10.79 → 18.43 | +7.64 | 61.64% |
| Whale Rock | 35 → 35 | 23.38 → 20.34 | -3.04 | 59.27% |
| Coatue | 58 → 63 | 19.22 → 20.37 | +1.15 | 62.52% |
| Fundsmith | 34 → 41 | 16.52 → 22.35 | +5.83 | 52.45% |
| ARK | 173 → 175 | 37.60 → 38.68 | +1.08 | 40.05% |
| Soros Fund Management | 214 → 222 | 54.26 → 60.38 | +6.12 | 31.62% |
| Bridgewater | 977 → 982 | 72.35 → 114.16 | +41.82 | 23.12% |
Dalal Street and Tiger Global are the useful counterexamples
Dalal Street’s eligible issuer count increased from three to four, but effective holdings slipped from 2.94 to 2.87. Its Q2 eligible snapshot remained concentrated: Warrior Met Coal represented 43.32% by our issuer-level value calculation, and the Top 10 measure was necessarily 100% because the portfolio had fewer than ten eligible issuers. The individual Dalal Street Q2 report provides the filing context. Adding a small fourth issuer did not offset the weight concentration in the core positions.
Tiger Global moved in the opposite direction. Its eligible issuer count fell from 53 to 46, yet effective holdings rose from 16.02 to 17.69 and its Top 10 weight declined from 69.49% to 67.06%. The Tiger Global Q2 report shows the underlying position changes. The shorter Q2 eligible list was less concentrated as a whole; this comparison includes new and continuing issuers and does not isolate the effect of removals.
Across levels, Q2 issuer count and effective holdings had a Spearman rank correlation of 0.91. Across changes, the correlation between the change in issuer count and the change in effective holdings was 0.82. The relationship is strong in this selected 17-group panel, but the sample is small and intentionally curated. More importantly, correlation does not turn raw name count into a substitute for portfolio weights.
The biggest broadening and narrowing moves
| Reporting group | Effective holdings Q1 | Effective holdings Q2 | Change | Top 10 change |
|---|---|---|---|---|
| Bridgewater | 72.35 | 114.16 | +41.82 | -7.81 pp |
| Third Point | 10.79 | 18.43 | +7.64 | -18.69 pp |
| Soros Fund Management | 54.26 | 60.38 | +6.12 | -0.85 pp |
| Fundsmith | 16.52 | 22.35 | +5.83 | -16.39 pp |
| Whale Rock | 23.38 | 20.34 | -3.04 | +7.87 pp |
| Appaloosa | 13.26 | 10.48 | -2.78 | +5.28 pp |
| Situational Awareness | 7.17 | 5.63 | -1.54 | -2.09 pp |
Bridgewater’s effective count rose by 41.82, by far the largest change. Its Q2 Top 10 weight fell to 23.12%, and no eligible issuer exceeded 4.58% of the calculated denominator. Third Point and Fundsmith also broadened materially as Top 10 concentration fell. At the narrowing end, Whale Rock’s effective count fell by 3.04 while its Top 10 weight rose by 7.87 percentage points.
Situational Awareness shows why Top 10 and HHI should be read together. Its Top 10 weight fell by 2.09 points, yet its effective count also fell because the distribution inside that Top 10 became more concentrated: Sandisk alone accounted for 28.13% of eligible Q2 value. HHI responds to every issuer weight, with especially large influence from the biggest positions.
Method: one consistent eligible-equity denominator
The preserved source set contains 4,269 rows from 35 Q1/Q2 information tables. We classify 4,073 rows as eligible operating-company equity and exclude 196 rows, including options, ETFs/ETPs, debt or principal units, warrants, preferred or depositary preferred, acquisition units and dated equity units. Six previously identified classification exceptions—two Grab warrants, two Oracle depositary preferred rows and two Global X Uranium ETF rows—remain excluded under the corrected derivative.
Within each reporting group and quarter, we aggregate security classes and related rows by the first six CUSIP positions in the preserved issuer bridge. We divide each issuer’s reported value by the group’s total eligible disclosed operating-company equity value, then calculate HHI, 1/HHI and Top 10 weight. If a group has fewer than ten eligible issuers, Top 10 equals 100% by definition.
The first-six CUSIP bridge is one-to-one in this eligible two-quarter derivative, but it is not a universal entity-resolution rule. Corporate actions, identifier migrations, different depositary structures and amendments can break that assumption. The next-quarter update must preserve this dated mapping and add explicit bridges rather than silently recomputing the historical series.
What the statistics do not measure
Form 13F reports selected long positions at quarter end. It omits short positions, most non-13F assets, many private holdings, purchase prices, transaction dates and investment motives. This analysis measures concentration only inside each group’s eligible disclosed operating-company equity. It is not total-portfolio diversification, risk, turnover, performance or a ranking of manager quality.
Reported values can also move because security prices changed, not because a manager traded. We use value weights because concentration is inherently a value-allocation question, but we do not label a weight change as a purchase or sale. Likewise, Bridgewater’s 982 eligible issuers do not prove 982 independent risk exposures; business correlations, factors and derivatives sit outside this calculation.
The next test is the first complete matched quarter after June 30, 2026. We will update only after the relevant filings and amendments are available, preserve Q1/Q2 as a dated baseline, and keep the same eligibility and issuer-resolution rules. This is disclosure research for informational and educational purposes, not personalized investment advice or a performance forecast.
Sources and original documents
U.S. SEC — Frequently Asked Questions About Form 13F, Form 13F Data Sets and Form 13F instructions, reviewed October 26, 2026. These establish the quarter-end fields, fair-value basis, as-filed data limits and filing timetable; they do not supply motives or whole-portfolio holdings.
Altimeter — Q1, filed 2026-05-15, CIK 1541617: information table / cover summary; Q2, filed 2026-08-14, CIK 1541617: information table / cover summary.
Appaloosa — Q1, filed 2026-05-15, CIK 1656456: information table / cover summary; Q2, filed 2026-08-14, CIK 1656456: information table / cover summary.
ARK — Q1, filed 2026-05-12, CIK 1697748: information table / cover summary; Q2, filed 2026-08-14, CIK 1697748: information table / cover summary.
Berkshire Hathaway — Q1, filed 2026-05-15, CIK 1067983: information table / cover summary; Q2, filed 2026-08-14, CIK 1067983: information table / cover summary.
Bridgewater — Q1, filed 2026-05-15, CIK 1350694: information table / cover summary; Q2, filed 2026-08-14, CIK 1350694: information table / cover summary.
Coatue — Q1, filed 2026-05-15, CIK 1135730: information table / cover summary; Q2, filed 2026-08-14, CIK 1135730: information table / cover summary.
Dalal Street, LLC — Q1, filed 2026-05-14, CIK 1549575: information table / cover summary; Q2, filed 2026-08-13, CIK 1549575: information table / cover summary.
Greenlight / DME — Q1, filed 2026-05-15, CIK 1489933: information table / cover summary; Q2, filed 2026-08-14, CIK 1489933: information table / cover summary.
Fundsmith LLP — Q1, filed 2026-05-15, CIK 1569205: information table / cover summary; Q2, filed 2026-08-14, CIK 1569205: information table / cover summary.
Icahn reporting group — Q1, filed 2026-05-15, CIK 921669: information table / cover summary; Q2, filed 2026-08-14, CIK 921669: information table / cover summary.
Light Street — Q1, filed 2026-05-15, CIK 1569049: information table / cover summary; Q2, filed 2026-08-14, CIK 1569049: information table / cover summary.
Pershing Square — Q1, filed 2026-05-15, CIK 1336528: information table / cover summary; Q1, filed 2026-05-15, CIK 2026053: information table / cover summary; Q2, filed 2026-08-14, CIK 2026053: information table / cover summary.
Situational Awareness — Q1, filed 2026-05-18, CIK 2045724: information table / cover summary; Q2, filed 2026-08-14, CIK 2045724: information table / cover summary.
Soros Fund Management — Q1, filed 2026-05-15, CIK 1029160: information table / cover summary; Q2, filed 2026-08-14, CIK 1029160: information table / cover summary.
Third Point — Q1, filed 2026-05-15, CIK 1040273: information table / cover summary; Q2, filed 2026-08-14, CIK 1040273: information table / cover summary.
Tiger Global — Q1, filed 2026-05-15, CIK 1167483: information table / cover summary; Q2, filed 2026-08-14, CIK 1167483: information table / cover summary.
Whale Rock — Q1, filed 2026-05-15, CIK 1387322: information table / cover summary; Q2, filed 2026-08-14, CIK 1387322: information table / cover summary.
AHNIM STUDIO’s About Us and research policy explains source responsibility, AI assistance and dated corrections.