Gartner, Inc.

The full register, from 13F filings

The Register tab above is hand research: a dozen holders, each traced to the filing it came from. This asks the same question of the whole market for the quarter ended 31-mar-2026, where 699 managers reported a position. Broad where the research is deep, and shallow in the ways a 13F always is.

Built from 13F filings, which report US listed, long, US custodied positions only. A register from this source is a floor rather than a level: a holder's true position can be far larger, as BlackRock's ASML stake shows at 1.28m shares in a 13F against 26.3m in its 13G/A. Holdings held through derivatives, outside US custody, or below a manager's reporting obligation do not appear. Where a holder is marked as a family, its total sums several separate filing entities and describes something no single filing describes; the constituent entities and their share counts are listed so the total can be checked. That roll-up can legitimately exceed the family's own Schedule 13G, because a 13G is filed by one legal entity: Vanguard reports Alphabet through ten advisory entities, and The Vanguard Group Inc itself reported zero after its January 2026 disaggregation.

What a reader needs to know to read these numbers

• Ticker confirmed exactly 'IT' against EDGAR's own submissions record for CIK 0000749251 (tickers: ['IT'], exchanges: ['NYSE']). All filing searches used the issuer name 'GARTNER' or CUSIP 366651107, never the bare string 'IT'.
• Holdings side: searched the newest 10-Q (period 30 Jun 2026, filed 4 Aug 2026, which postdates the FY2025 10-K since Gartner's fiscal year ends 31 Dec) and the FY2025 10-K (filed 12 Feb 2026) for 'equity method', 'equity investment', 'non-marketable', 'unconsolidated', 'joint venture', 'minority interest', 'cost method', 'investment in', 'privately held', 'equity securities'. Neither filing contains any equity-method, cost-method, joint-venture or non-marketable-security disclosure; the only 'unconsolidated' hit in each is the standard boilerplate 'has not entered into any material off-balance sheet arrangements or transactions with unconsolidated entities.' Exhibit 21.1 of the FY2025 10-K (a20251231-gartnerincsubsid.htm, as of 31 Dec 2025) was checked: it is a bare two-column list (Entity Name, State/Country of Delaware/UK/etc.) of roughly 75 subsidiaries, every one a recognizable wholly owned Gartner acquisition or brand (CEB, Capterra, Dataquest, Digital Markets, L2, Software Advice, UpCity, Valtera, etc.); there is no ownership-percentage column and no joint-venture entity anywhere in it. Conclusion: as of the current reporting date, Gartner holds no equity-method, cost-method, joint-venture or minority stake in any other company; holdings is a genuine, sourced empty array, in the same shape as Labcorp's and Darden's Exhibit 21 (a bare name list, not a percentage-column exhibit).
• Capital Group family checked and found to hold NOTHING in Gartner as of Q2 2026 (30 Jun 2026), despite being a >5% holder as recently as year-end 2025. All 8 Capital Group 13F-filing registrants were read directly for CUSIP 366651107 and none has a row: Capital Research Global Investors 0001422848, Capital World Investors 0001422849, Capital International Investors 0001562230, Capital International, Inc./CA/ 0000895213, Capital Group Investment Management Pte. Ltd. 0001939970, Capital Group Private Client Services, Inc. 0001857666, Capital International Ltd /CA/ 0001065350, Capital International Sarl 0001065349 (all Q2 2026 13F-HR, filed 12 Aug 2026). Three further Capital Group entities filed 13F-NT (no holdings) for the quarter: Capital Group Companies, Inc. 0000732812, Capital Group International, Inc. 0000949308, Capital Research & Management Co 0000017283. This is independently corroborated by Capital International Investors' own Schedule 13G/A history on Gartner: 4,085,852 sh/6.1% as of 31 Dec 2025 (13G filed 13 Feb 2026, quoted in the 2026 proxy's stale 5%-owner table), reduced to 2,022,839 sh/2.9% as of 31 Mar 2026 (13G/A filed 14 May 2026, matching this project's own Q1 2026 register snapshot exactly), then to zero as of 30 Jun 2026 per the fresh 13F sweep above: a genuine, three-quarter exit, not a data artifact. Recorded here as a note per the brief's instruction that a holder found at zero is a note, not a null row.
• Proxy 5%-owner-table check (Allegion-style gap): the 2026 DEF 14A's 5%-owner table (record date 2 Apr 2026) lists exactly four names, BlackRock 10.8%, Baron Capital Group 6.4%, Capital International Investors 6.1%, The Vanguard Group 14.2%, all sourced to Schedule 13G/A filings dated between Jun 2025 and Feb 2026. Vanguard is NOT missing from this proxy the way it was from Allegion's: it is present, just stale. What the proxy's table does miss are Independent Franchise Partners LLP (8.84% by the fresh Q2 2026 13F) and AQR Capital Management (6.61%), both of which crossed the 5% threshold and filed their FIRST Schedule 13G on Gartner in May 2026, after the proxy's 2 Apr 2026 record date and 15 Apr 2026 filing date. That is a timing gap, not a reporting-structure omission bug.
• Implied per-share price check: every institutional Q2 2026 13F row (Vanguard's 8 entities, BAMCO, Independent Franchise Partners, BlackRock, AQR, Harris, State Street, Invesco, Morgan Stanley, T. Rowe Price Associates after converting its thousands-denominated value) divides to a tight $129.5-129.62 per share, all dated 30 Jun 2026, a strong internal cross-check that every row is the same CUSIP/instrument at the same date with no units error. Gartner's own current market price ($192.53-192.81, 19 Aug 2026) is about 49% higher than this 30 Jun 2026 implied price; the 52-week range on file (124.25-265.85) confirms this is a genuinely volatile stock and not a pricing error, and an 8-K was filed 30 Jul 2026 (likely the Q2 earnings release) consistent with a post-quarter rally.
• Self-gate, computed from this final file per the chart's own arithmetic (members at full value, the aggregate reduced only by the one member that rolls into it): 13 register rows. Raw shares sum: Vanguard 8,515,132 + BAMCO 7,601,009 + Independent Franchise Partners 5,582,962 + BlackRock 5,217,146 + AQR 4,175,036 + Harris Associates 2,846,799 + State Street 2,808,781 + Geode 2,603,296 + Invesco 2,061,112 + Morgan Stanley 1,867,660 + directors/officers aggregate 1,760,321 + Eugene A. Hall 1,219,897 + T. Rowe Price Associates 102,699 = 46,361,850. Chart total after reducing the aggregate by its one member (Hall, who rolls up into it, so the aggregate counts at 1,760,321-1,219,897=540,424 and Hall counts separately at full value): 45,141,953 shares, against shares_outstanding 63,148,976 = 71.48% of the company. This is above the roughly 25-45% typical band and above the 55% flag level, so it was investigated rather than shipped as-is: (1) every institutional row's implied price is a consistent $129.5-129.62/sh, ruling out a units or wrong-CUSIP error; (2) AQR's and Independent Franchise Partners' 13F share counts are EXACT matches to those same holders' own fresh Schedule 13G/A filings, an independent confirmation outside the 13F system; (3) this file's 13 rows are a SUBSET of this project's own independent Q1 2026 register (data/registers/IT.json, 773 filers, unrelated to this file's sourcing), which is the direction the check has to run: a top-13 total of 45,141,953 sits inside that wider population, so nothing here is counted that the wider data does not also see. CAVEAT ADDED BY THE COORDINATOR, 19 Aug 2026, correcting this note: that register must NOT be read as confirming ownership is near 100 per cent. Its `holders` array is a TRUNCATED top-40 summing to 63,307,092, which is what an earlier draft of this note quoted, while its `total_shares_reported` across all 773 filers is 75,069,477, or 118.9 per cent of the 63,148,976 shares outstanding. A figure above 100 per cent is impossible and is evidence that the 13F-derived register OVER-REPORTS, most likely because an adviser and a sub-adviser report the same shares under two different CIKs, which the pipeline's (cik, cusip) dedupe cannot catch. That is a known open item recorded in STATE.md and it is not a defect in this file; it simply means the register can bound this file from above but cannot corroborate a level. The load-bearing checks here are (1) and (2), which are independent of the 13F aggregate entirely. The high reading is reported rather than suppressed, per the brief's own guidance for a verified high concentration.
• Gartner has been repurchasing stock very heavily: diluted weighted-average shares fell from 77,606 thousand (H1 2025) to 68,288 thousand (H1 2026) per the Q2 2026 10-Q, and the cover-page count used here (63,148,976 as of 31 Jul 2026) is lower still than either weighted average, confirming the buyback has continued past quarter-end. This denominator, not an older or proxy-stated one, is used throughout this file.