Skip to main content
Trump Conflict Ledger
  • Ledger
  • Timeline
  • Network
  • Topics
  • Actors
  • Sources
  • Methodology

Methodology

Every record on this site is built to the same contract: it must name its sources, state what those sources establish, state what they do not establish, carry the subject’s response or an explicit note that none was located, and show when it was last verified. This page explains the rules.

Evidence classes

  • Documented fact — the asserted fact appears in a primary record: an OGE disclosure, a court or clemency document, a government press release, a congressional record, or an equivalent inspectable filing.
  • Reported correlation — reputable, named reporting establishes a timing overlap or relationship, but not causation. Most of the ledger’s timing-based entries are in this class, deliberately.
  • Formal allegation — an official complaint, referral, letter, staff report, or sworn filing alleges the conduct. An allegation is recorded as an allegation. It is never restated as a finding.
  • Active investigation — an official source, or independent named reporting of a referral, confirms an open inquiry. An investigation is not an adjudication.

Confidence grades

Confidence measures evidence completeness, not moral certainty:

  • High — a primary record plus independent corroboration, or two independent primary records.
  • Medium — one primary record without adequate corroboration, or multiple high-quality secondary sources.
  • Low — excluded from publication entirely. Candidate items that fail the bar are held in an internal rejection log and never rendered on this site.

Prohibited language

The words insider trading, bribery, corruption, pay-to-play, and quid pro quo are legal and quasi-legal conclusions. This site uses them in exactly three situations: quoting an attributed source, describing a precise formal legal status backed by a formal record (for example, an adjudicated federal bribery conviction described by the Justice Department), and this explanation. They never appear in the site’s own editorial voice, headlines, or interface labels, because the cited evidence does not establish them.

Two structural legal facts frame everything here:

  1. The President is statutorily exempt from the principal federal conflict-of-interest statute. 18 U.S.C. § 202(a) excludes the President and Vice President from the definition of covered officials under 18 U.S.C. §§ 202–209. A presidential holding or policy overlap is therefore never described as a § 208 violation, because it cannot be one.
  2. The President remains a public financial-disclosure filer under the Ethics in Government Act, and the STOCK Act states that executive-branch officials are not exempt from generally applicable securities-fraud prohibitions. Disclosure obligations and reporting deadlines documented on this site are real legal obligations.

Timing is not causation

A timeline can establish that A occurred before B. It cannot establish that A caused B, that any trader knew B was coming, that a trader was connected to an official, or that a crime occurred. Every timeline cluster and every timing-based claim on this site carries a specific statement of what the sequence does not establish. Same-day events without timestamps are treated as unordered: the disclosure forms at the center of this ledger record dates, not times.

Source hierarchy

  1. Primary public records — OGE, SEC, DOJ, CFTC, FEC filings; court and clemency documents; pardon warrants; congressional records; government and company statements; inspectable blockchain transactions.
  2. Independent wire and public-service reporting — Reuters, AP, NPR, PBS.
  3. Major outlets with named sourcing — NYT, WSJ, ABC, NBC, CBS, Bloomberg, FT, ProPublica, and peers.
  4. Specialist watchdogs that link primary records — used with explicit labeling of their methodology, never as government statistics.
  5. Advocacy and party statements — usable only for attributed responses.
  6. Anonymous social-media content — never a source. At most a lead requiring independent verification, and no such lead is published here without it.

Amount precision is explicit

OGE disclosures report values in broad statutory bands (for example, “$1,000,001–$5,000,000”). This site never converts a band into a point estimate. Amounts are separately labeled as an exact primary-record figure, reported point estimate, range, minimum, or maximum. A point estimate reported by an outlet can keep equal numeric bounds for machine processing without being mislabeled exact. Every amount cites its source, and the validator reserves exactness for a primary record.

Responses and denials

A response or denial is displayed beside the claim it answers, in full-size text, attributed and dated — never in a footnote. Where no public response was located, the record says so explicitly, with the dated search note recording when sources were reviewed. That search note is not a record of direct contact. Each claim separately carries a direct right-of-reply status; because no audited outreach channel is active in v0, every record is marked not_contacted, and an existing public quote is never read as evidence that this project reached out. Silence is never construed as admission.

Verification and corrections

Every source record carries an access date and a last-verified date. The versioned integrity manifest records the current result for every source URL, including explicit 401/403 paywall or anti-bot blocks and named timeout exceptions; a blocked automated fetch is not treated as a missing source. The correction log is append-only and includes prepublication audit fixes as well as future post-launch corrections: corrected text is struck through and preserved, never deleted. See Corrections. A public direct submission channel is not active in v0; the ledger states that limitation rather than implying that direct outreach occurred.

What we will not publish

  • Identification of pseudonymous wallets or unidentified traders with named people, absent a primary filing or two strong independent reports.
  • Aggregate “conflict scores” or any number that dresses editorial judgment as measurement.
  • Claims whose only source is an anonymous post, an unnamed aggregation, or a single outlet where the contract requires corroboration.
  • Any assertion that a documented timing overlap proves unlawful conduct.

    Temporal adjacency does not establish causation. Every entry on this site pairs what the evidence establishes with what it does not establish; classification and confidence describe evidence completeness, not guilt. See the methodology and corrections pages.

    • Evidence ledger
    • Data downloads
    • Methodology
    • Corrections
    • About

    © 2026 Hermanity. Independent public-interest research project. Records last verified 2026-07-21.