โ† Grizzly Bulls Research

SEC Ownership Evidence Funnel Across 20,834 Form 3/4/5 Candidates

Only 1,503 of 20,834 Extracted Position Candidates Were Simple Direct Non-Zero Positions
By Lee BaileyPublished September 12, 2026Measurement through September 10, 2026Version 1.020,834 extracted position candidates
Key findings

Only 1,503 of 20,834 extracted position candidates were simple direct non-zero positions

The reviewed Forms 3, 4 and 5 extraction parsed 7,359 of 7,375 targeted filings and produced 20,834 position candidates. Parsing was nearly complete, but only 7.2% of candidate rows landed in the simplest non-zero direct-position state. Footnote review alone accounted for more than half of all candidate assessments.

7,359Forms 3/4/5 filings parsed
99.8% of the 7,375 eligible filing rows; 16 failures stayed explicit.
1,503Simple direct non-zero candidates
7.2% of 20,834 extracted position candidates.
121Profiles with at least one simple direct candidate
61.1% of the 198 Forms-eligible profiles.
A simple-direct candidate is still not a published current holding. Security identity, current-position semantics and review remain separate requirements.

The selected-profile evidence funnel narrowed at several different boundaries

This is not a single conversion funnel from every billionaire to a current holding. Each step answers a different question: exact SEC identity, relevant filing metadata, Forms 3/4/5 eligibility, successful extraction, candidate production and then the narrowest direct-position assessment.

StageProfiles or rowsWhat the stage establishes
Current billionaire roster3,409 profilesThe population used to identify profiles needing additional SEC evidence work.
Profiles selected for SEC identity review2,881 profilesProfiles selected for exact SEC identity re-evaluation or new discovery work.
Exact SEC identities resolved212 profiles7.4% of the selected group reached an exact identity match. The remaining 2,669 stayed unresolved rather than being fuzzily matched.
Ownership filing metadata found211 profilesExact identity plus relevant ownership-filing metadata, not a share count or holding.
Forms 3/4/5 extraction scope198 profiles ยท 7,375 filing rowsFilings eligible for this Forms 3/4/5 extraction. Other ownership source families were analyzed separately.
Position-candidate profiles187 profiles94.4% of Forms-eligible profiles produced at least one conservative position candidate.
Simple direct candidate profiles121 profilesAt least one extracted row reached the simplest direct non-zero candidate state. Further review was still required.

More than half of extracted candidates required footnote review

The assessment states partition the 20,834-position-candidate population. Unlike the overlapping review reasons in the separate sale-footnote study, each candidate appears once in this table.

Candidate assessment states

Conservative review outcomes across the complete candidate population.
Footnote review10,537
50.6% of all position candidates.
Conflicting balances5,737
27.5% of all position candidates.
Indirect-holder review2,583
12.4% of all position candidates.
Numeric review399
1.9% of all position candidates.
Simple direct non-zero position1,503
7.2% of all position candidates.
Simple direct zero balance75
0.4% of all position candidates.
Footnote review represented 50.6% of all candidates. Only 7.2% were simple direct non-zero positions.

Schedule 13 evidence stayed in its own source-family review

The SEC evidence review found 211 profiles with relevant ownership filings across source families. The Forms extraction covered 198 of them. The remaining 13 profiles and 591 filing-metadata rows belonged outside Forms 3/4/5, principally Schedule 13D/G evidence. They were excluded by design rather than counted as extraction failures.

Methodology

  1. Start with the September 9 billionaire roster and select profiles needing additional exact SEC identity or ownership-filing review.
  2. Resolve reporting-owner identity only when an exact SEC identity match can be supported. Ambiguous or unmatched profiles remain unresolved.
  3. Keep ownership evidence separate by source family. Send only Forms 3, 3/A, 4, 4/A, 5 and 5/A through this filing extraction.
  4. Review the complete eligible Forms set. Preserve fetch or parse failures explicitly instead of dropping them from the denominator.
  5. Classify each extracted position candidate conservatively. A parsed balance can remain in footnote review, conflicting balances, indirect-holder review or numeric review.
  6. Publish aggregate funnel and assessment counts only. Do not redistribute filing bodies or person-level review records in the public research dataset.

SEC identity review measured September 9, 2026. Forms extraction measured September 10, 2026.

Primary SEC source semantics

Forms 3, 4 and 5 report beneficial-ownership information for covered insiders under Section 16. Grizzly Bulls reviews that filing evidence conservatively, but filing presence by itself does not prove a complete or current billionaire public-equity position.

Download the aggregate study data

JSON includes the funnel summary, aggregate assessment states, source notes and limitations. CSV contains the six aggregate candidate-assessment rows only. Neither download contains filing bodies or person- or issuer-level identifiers.

Citation and reuse

Lee Bailey. โ€œSEC Ownership Evidence Funnel Across 20,834 Form 3/4/5 Candidates: Only 1,503 of 20,834 Extracted Position Candidates Were Simple Direct Non-Zero Positions.โ€ Grizzly Bulls, September 12, 2026. Version 1.0. Data snapshot September 10, 2026. https://grizzlybulls.com/research/sec-ownership-evidence-funnel

When citing the 1,503-of-20,834 result, preserve that this is a selected-profile billionaire SEC review and a conservative Forms 3/4/5 extraction, not a representative sample of all SEC ownership filings. Simple-direct candidates are evidence requiring further review, not verified current holdings.