Workday, 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 924 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

• DUAL-CLASS DENOMINATOR. Workday has two classes: Class A common stock (WDAY, Nasdaq-listed, one vote/share) and Class B common stock (not publicly traded, ten votes/share, held almost entirely by co-founders Aneel Bhusri and David Duffield and their affiliated trusts/foundations). As of the 17 Apr 2026 proxy record date (2026 DEF 14A, filed 5 May 2026): 203,712,072 Class A shares outstanding and 45,973,479 Class B shares outstanding. company.shares_outstanding (249,685,551) is deliberately set to the SUM of both classes, per the brief's instruction that the register chart divides raw row 'shares' by this denominator and never reads pct_of_company: since founders' Class B holdings are put directly into the 'shares' field on several rows (Bhusri, Duffield, the officers/directors group), the denominator MUST include Class B too, or those rows would overstate their share of the company. Every row's pct_denominator repeats this choice explicitly.
• 13F filers (all 11 institutional rows in this register) report Class A shares only, since Class A is the only Workday class that is publicly traded and 13F-reportable. Their pct_of_company is computed against the all-class total (249,685,551), which understates their share of Class A alone; each row's pct_denominator also states the Class A-only percentage for reference (the brief's warning about conflating a stake's percentage of one class with its percentage of the whole company).
• VOTING POWER, the real story for this founder-controlled dual-class company. Per the 2026 proxy (as of the 17 Apr 2026 record date): Aneel Bhusri and David Duffield, bound by a mutual Stock Voting Agreement covering all Class B shares owned by either founder or their permitted transferees, together control approximately 69% of Workday's aggregate voting power (Class B carries 10 votes/share versus 1 for Class A). The 'Current Executive Officers and Directors as a Group (16 persons)' also shows 69% of voting power in the proxy table, because that figure is driven entirely by Bhusri's Voting-Agreement-inflated Class B attribution, not by any other group member. This 69% figure cannot be derived from the economic share-count register above (which nets out the reciprocal double-attribution between Bhusri and Duffield to avoid counting the same Class B block twice); it is recorded here as prose only, as the brief instructs, because the chart has no mechanism to display vote weighting.
• RECIPROCAL SHARED-VOTING DOUBLE COUNT, the defect this company was built to test for. Bhusri's and Duffield's officially-tabled proxy rows each show the SAME 45,660,277 Class B figure (99% of Class B, 69% of voting power), because under their mutual Stock Voting Agreement each is deemed under SEC beneficial-ownership rules to share voting power over the other's Class B stock. Bhusri's own economic Class B holding is only 8,131,443 (his 8,126,443 direct plus 5,000 held by his minor child); Duffield's own economic Class B holding is only 37,528,834 (via The David A. Duffield Trust). The two true economic blocks sum exactly to the disclosed combined figure (8,131,443 + 37,528,834 = 45,660,277), confirming there is no third, unaccounted-for block. Both the Bhusri row and the officers-and-directors group row (which is driven entirely by Bhusri's inclusion) are reduced in this file to exclude the Duffield-only portion, and Duffield's standalone row carries his own portion at full value, so the three rows together reconstruct the true total exactly once. Full arithmetic is in each row's method_note.
• Holdings are empty by design, a sourced finding rather than a gap. The most recent 10-Q (period 30 Apr 2026, filed 22 May 2026, Note 3 Investments) and the FY2026 10-K (period 31 Jan 2026, filed 6 Mar 2026) were both searched for 'equity method' (zero hits in either filing: Workday appears to have no equity-method investments at all), 'non-marketable', 'strategic investment', 'Workday Ventures' and 'unrealized gain'. Workday reports a single combined balance, 'Non-marketable equity investments measured using the measurement alternative' (privately held companies without readily determinable fair values, in which Workday does not own a controlling interest or exercise significant influence): carrying value $220 million at 30 Apr 2026 (initial cost $198 million, cumulative net unrealized gains $22 million) versus $230 million at 31 Jan 2026 (initial cost $205 million, cumulative net unrealized gains $25 million). No investee is named in either filing. The FY2026 10-K separately describes 'Workday Ventures,' the company's strategic investment arm, investing primarily in early-stage enterprise software companies, but again names no individual investee. Per the rule against assigning a combined balance to one investee, none of this figure is attributed to any named holding; the balance is recorded here and in unknowns instead of an invented holdings row.
• Exhibit 21.1 of the FY2026 10-K (subsidiaries list, checked per the brief's technique) is a bare name-and-jurisdiction list with no ownership-percentage column, and every entity on it appears to be a wholly owned operating subsidiary, including Workday's fiscal 2026 acquisitions Flowise (FlowiseAI, LLC), Paradox (Paradox, LLC and affiliates), Sana (Sana Labs AB/Inc/Ltd) and Pipedream (Pipedream, LLC), plus earlier acquisitions such as Adaptive Insights, HiredScore, Evisort, VNDLY and Peakon. A wholly owned acquisition is not a holding per the brief, so none of these appear as register rows; this is a genuine negative, not a place the researcher didn't look.
• Vanguard case confirmed: parent CIK 0000102909 filed Form 13F-NT (notice, no holdings) for Q2 2026, so per the brief's rule the successor entities were summed rather than one row substituted. Entity roster (names and CIKs only, never figures) was taken from this project's own 31-Mar-2026 WDAY.json register snapshot, then each of the 8 listed successor entities' own fresh Q2 2026 13F-HR was individually fetched; 7 hold WDAY, 1 (Vanguard Advisers Inc) holds zero. Total: 26,550,649 shares, materially higher than the snapshot's 27,675,362-share Mar-2026 figure would suggest on its face because the entity mix shifted between quarters; the fresh per-entity figures are used, not the stale snapshot total.
• Capital Group family checked across all three CIKs named in the brief (Capital World Investors 0001422849, Capital International Investors 0001562230, Capital Research Global Investors 0001422848): none held any WDAY shares in their Q2 2026 13F-HR filings, confirmed by both a 'WORKDAY' and a shorter 'WORKD' name-fragment search finding zero rows in each entity's full information table. This is a genuine finding, not a gap: no register row is written for Capital Group.
• Market capitalization ($47.08 billion) and share price ($190.64) are from stockanalysis.com's live snapshot fetched 19 Aug 2026, dated by the source to the 18 Aug 2026 close. stockanalysis.com's own shares-outstanding figure (246.97 million) is close to but below the 249,685,551 all-class SEC figure used here (17 Apr 2026 record date), consistent with continued share repurchases in the interim (the 10-Q shows 12,025,000 shares repurchased in the quarter ended 30 Apr 2026 alone) and possibly a different as-of date or methodology; both figures are recorded for cross-check. An aggregator (WallStreetZen) separately reports approximately 79.97% institutional ownership of WDAY, a figure not reconcilable with any denominator used in this file (against Class A alone the 11 institutional rows here total 50.26%; against the combined all-class total, 40.95%); it is recorded here as an unresolved discrepancy rather than used, since aggregator methodology for such summary statistics is opaque and not independently checkable against a filing.
• SELF-GATE, computed exactly as specified, from the final file on disk: 14 register rows. Institutional 13F rows (Vanguard, BlackRock, Hotchkis & Wiley, Eagle Capital, State Street, First Eagle, Geode, Generation IM, Norges Bank, FMR, T. Rowe Price) sum to 102,384,664 shares, all counted at full value. Aneel Bhusri (a member, rolled up into the officers and directors group) counts at full value, 8,578,914 shares. David A. Duffield (not a member of anything) counts at full value, 39,979,883 shares. The officers and directors group (an aggregate with one listed member) is reduced to its residual: 9,022,283 (its own adjusted raw value) minus 8,578,914 (Bhusri's member value) = 443,369. Total: 102,384,664 + 8,578,914 + 39,979,883 + 443,369 = 151,386,830 shares against 249,685,551 shares outstanding, or 60.63%. This sits ABOVE the brief's typical 25-45% band and above even the 'up to 55%' founder-led allowance the brief describes for a company like Airbnb. It is not a class mismatch: every institutional row is confirmed Class A (13F-reportable), every insider row is confirmed combined-class and cross-checked for internal consistency (Bhusri's + Duffield's own economic Class B holdings sum exactly to the disclosed combined Class B total, with no third unexplained block), and no fractional shares or pre/post-split issues were found. The elevated total instead reflects two independent, well-documented facts about Workday specifically: unusually heavy institutional Class A ownership (the top 11 institutions alone hold 50.26% of Class A), layered on top of an unusually large single-founder economic stake (David Duffield alone holds 16.01% of the whole company, a rare figure for a co-founder who is no longer an officer or director). Both figures are individually sourced and cited above; the total is reported as computed rather than adjusted to fit the typical band. Corroborated first-hand by the coordinator against a source independent of any aggregator. This project's own 13F-derived register (data/registers/WDAY.json, period 31 Mar 2026, built by a separate pipeline from the bulk 13F data, single CUSIP 98138H101 which is Class A only) reports total institutional ownership of 153,670,920 shares, 75.44% of Class A, with its top eleven holders summing to 104,592,968, 51.34% of Class A. The eleven institutional rows in this file sum to 102,384,664, 50.26% of Class A, matching that independent figure closely and confirming no double count on the institutional side. The 60.63% is therefore arithmetically forced by the capital structure rather than being a defect: institutions hold roughly three quarters of Class A while the founders hold about 99% of Class B, and Class B is 18.41% of all common on its own. The charted total reconciles exactly as 102,384,664 institutional plus 8,578,914 for Bhusri counted at full value as a member, plus the 443,369 residual of the officers container, plus 39,979,883 for Duffield outside that container, equals 151,386,830. Duffield sits OUTSIDE the officers group because he is a co-founder and not a current executive officer or director, which is the documented Microsoft Ballmer and Gates precedent, not an inference from holder_type. The reciprocal-attribution correction described above is what keeps this figure honest: had both founders' proxy rows been taken at face value, the same 45,660,277 Class B shares would have been counted twice, adding about 18 percentage points and charting the register near 79%, which is the Marriott defect that passed every gate.