joe digs

The Sixty-Nine Days · claim A4

ANALYSISOur labeled inference. Falsifiers listed.

Our reading: the money calendar was an exit. A $7 billion employee tender closed at an unchanged $852 billion mark with the full incident account still unpublished. Three days later a retail-facing listed fund put $150 million into OpenAI exposure described in one filing as Class A common, the class insiders hold, and in three others as Series C preferred; the fund has not said which is right. The people who sold into those two events got a price that did not yet know what OpenAI knew. We cannot see who sold. The filings that would show it do not exist yet, and the one fund that had to disclose did so with zero mention of the incident.

The document

Document
DXYZ 424B3 (Aug 28, 2026); Bloomberg tender reporting (Aug 10, 2026); DXYZ N-CSRS holdings
Accession
0001575872-26-000624
Dated
2026-08-28
Source
open the primary
Archived
primaries/dxyz-424b3-0001575872-26-000624-aug28.htm

Who read it

joe, EDGAR direct 2026-09-03; tender facts REPORTED (Bloomberg) not filed.

The innocent reading

Boards approve tenders with full information routinely, employees receive disclosure packages, and the July 21 post put buyers on notice that an incident existed. 'Class A common means insiders' is market convention, not a verified fact about OpenAI's cap table. Goanna and DXYZ may simply have bought without knowing more than the public did.

What would prove us wrong

  • Produce the tender information statement showing the incident was disclosed to participants in full before the close.
  • Show, from the audited DXYZ annual report or the 26E vehicle's documents, that the Aug 13 purchase was newly issued preferred stock rather than secondary common.
  • Identify the sellers and show none was an insider with knowledge beyond the July 21 post.