Licence, sources, and fair use

Three different things are stacked on every page of this catalog, and they do not carry the same rights. The filings underneath are public records of the United States government. The parsing, normalisation, and presentation on top of them are our own work. And the open licence over that second layer is a decision the owner has not yet taken, so this page states what is settled and marks what is not, rather than asserting a grant that does not exist.

The three layers

The filings

Form 4 filings are public records of the US Securities and Exchange Commission and are not subject to copyright. The transactional facts in them, who traded, how much, at what price, on what date, are not ours to license and we make no claim over them. The authoritative copy of any filing is the primary document on sec.gov, which every filing page links to directly.

Our layer

The XML parsing, the normalisation into a schema, the name corrections, the coverage statements, the search dictionary, this presentation, and these pages are ebomi's own work.

The grant over our layer

Not yet decided. The choice is between CC BY 4.0, which makes attribution enforceable, and CC0 with a citation request, which matches the norm the published datasets already live under. It is recorded as an open question in the information-architecture specification and it belongs to the project's owner, not to this page. Until it is taken, no grant is asserted here and none should be inferred.

What we ask

Cite the dataset. The citation is Balogh, A. Insider trading. Scientific Data 10, 237 (2023), doi:10.1038/s41597-023-02147-6. Every entity page carries its own cite-as block with the stable URL and the retrieval date, and every page has a machine-readable twin whose envelope repeats the same statement, so a citation made from the JSON and a citation made from the page agree.

Fair use of the JSON endpoints

Every page in this catalog has a .json twin at the same path. They are open and unauthenticated, and the policy is published here rather than discovered through a rate-limit response.

• Read them at a human pace. There is no published rate limit today; if one becomes necessary, its numbers will appear on this page before they appear in a response header.
• Identify your client in the User-Agent header, with a way to reach you. An anonymous crawler that causes trouble is a crawler we can only block.
• The sitemaps enumerate every canonical page, so a crawler never needs to guess a URL.

What this catalog does not do

• It adds no enrichment from any other source. Insider pages carry no biography, no photograph, and no scraped detail: they render public regulatory filings and nothing else.
• It never merges two CIKs into one person and never splits one CIK into two. Identity is the SEC's own identifier.
• It publishes no score, no rating, and no recommendation, and it carries no analytic, because the analytic does not exist yet and a plausible figure would be worse than none.
• It does not claim completeness, and it is specific about what it omits. Every page states the span this catalog holds, and absence inside that span is not evidence that no filing was made. Beyond the span, one class of filing is left out by design: a Form 4 reporting only derivative activity is not stored, because the separation of open-market trades from compensation is what this catalog is for.

Corrections and removal requests

A name in this catalog is a name as filed with the SEC, or a maintainer's correction of EDGAR's casing. If a page misreports what a filing says, that is a defect and we want to know. If a page reports correctly what a public filing says, the record is the SEC's and a removal request is answered by reference to the public source record, which every page links to.