1. Scope
This release stages the public-record files supplied in the JBR3 records binder dated August 23, 2026. It does not claim to be a complete historical corpus of every filing made by Dalton Investments or every connected entity. It is a point-in-time archive of the supplied collection.
The source binder remains unchanged. Public deployment uses a separate copy so that site organization, thumbnails, and metadata do not alter the retained source files.
2. Source collections
| Collection | Originating system | Preserved formats |
|---|---|---|
| Investment adviser filings | SEC IAPD and EDGAR materials | PDF, saved HTML, ZIP, JSON |
| Retirement-plan filings | U.S. Department of Labor EFAST2 | PDF, compiled CSV extract |
| Japan ownership filings | Japan FSA EDINET disclosure system | |
| Nonprofit tax returns | IRS Modernized e-File public corpus | XML |
| Fund materials | Public prospectus material | |
| Connected-entity filings | SEC EDGAR and IAPD materials | Saved HTML, JSON, text excerpt |
3. Source-status labels
Primary record
A preserved filing image, regulatory report, return, exhibit, or prospectus. “Primary” describes the file's documentary role in this archive; it is not a legal conclusion about admissibility, certification, completeness, or the truth of every statement made by a filer.
Source index
A saved regulator index or submissions file used to retain provenance and filing context. An index can point to records but is not a substitute for the underlying exhibit.
Bulk source data
A regulator-distributed data snapshot retained in its ZIP format. Snapshot dates and schemas vary, and a later agency release may supersede a copy in this archive.
Compiled extract
A locally generated CSV, JSON, or text excerpt made from source materials. Extracts are included for navigation and analysis, are labeled as derivatives, and should be checked against the underlying records before quotation or reliance.
4. File integrity
The deployment retains the binder's original directory structure and filenames under /records/. A SHA-256 digest was calculated from each deployed file. The digest, byte size, media type, archive path, source collection, and status appear in the machine-readable manifest and in the searchable document index.
A hash can show whether two byte streams match. It does not independently authenticate who created a document or whether a filing remains current.
5. Publication boundary
Raw public records and separately labeled data extracts were deployed at the archive's opening. Later on August 23, 2026, the compiled analytical memoranda and the binder index were also published — in a clearly separated Analysis section, labeled as compiled analysis rather than primary records, with their internal verification tags preserved and each page linking its source markdown and underlying records. Local extraction code and operating-system metadata remain unpublished. No source file was deleted or moved from the binder.
6. Limitations and corrections
Records may contain filer-supplied statements, scanned text, formatting errors, signatures, addresses, foreign-language text, or later-superseded information. Saved HTML may not reproduce every asset or link from the originating site. PDF thumbnails are visual aids only.
The current regulator or issuer record controls over this archive copy. Material corrections and collection changes are dated in the update log. A response or correction request does not become a finding merely because it is submitted; supporting documentation will be identified where available.