Dataset
sec-form4-insider-trading
AlphaAI's near-real-time parse of SEC EDGAR Form 4 filings, providing 20,832 individual insider transaction tranches and 8,538 grouped economic events for US public-company officers, directors, and 10% owners.
Data preview
Dataset card
The SEC Form 4 Insider Trading dataset, published by mikhailmakeev and sourced from AlphaAI, aggregates parsed insider transactions filed on SEC EDGAR Form 4 by officers, directors, and 10% owners of US public companies. It is distributed under a CC BY 4.0 license with a reported size between 10K and 100K records.
What this dataset contains
The dataset captures insider transactions filed on SEC EDGAR Form 4, parsed roughly six minutes after publication, with prices validated by an anchor-based plausibility check at ingest. Two granularities are provided: per-transaction rows and event-level rows grouped by accession_number, transaction_code, and ownership_form, so multi-tranche events such as 10b5-1 ladder sales collapse into a single record. Each event is annotated with a deterministic 1-10 relevance score derived from the event's summed total value.
Structure and fields
Although the full schema is listed as unknown, the documentation describes insider_transactions.csv grouped by (accession_number, transaction_code, ownership_form). Fields include `insiders` (distinct insider names), boolean flags `any_officer`, `any_director`, `any_ten_percent_owner`, and `any_10b5_1`, summed `shares` and `total_value`, and `avg_price_usd` computed as a value-weighted average over priced tranches. The dataset flags caveats such as non-USD filers (for example Taiwanese issuers reporting in NT$), 10b5-1 plan rows, and items that look like open-market buys but are actually private placements or financing rounds.
How analysts use it
Equity analysts and event-driven funds use the dataset to monitor insider buying and selling patterns, with the 1-10 score providing a starting point for ranking clusters by dollar magnitude. The grouping logic and explicit flags (any_officer, any_director, any_10b5_1) help separate discretionary trades from routine plan-based selling, supporting signals that feed into event-driven or sentiment-overlay strategies. Before aggregating, users are expected to filter non-USD tickers, 10b5-1 sales, and apparent 'buy' rows that are actually private placements.
Documentation & references 1
Related datasets
Browse moreParsed eligibility criteria from 13,229 ClinicalTrials.gov studies representing the candidate pool gathered by two first-stage retrievers during TREC Clinical Trials 2021–2023, formatted as typed entity-relation graphs for reranking models.
The clinical-trials-patient-graphs dataset is a small tabular collection of patient-level clinical records distributed in Parquet format by the 2001jdev publisher. It organizes entities and relations extracted from trial data into structured rows spanning 2021 through 2023.
This dataset is a parsed snapshot of ClinicalTrials.gov records, distributed in Parquet format on the Hugging Face Hub under an unspecified license. It contains roughly 52,000 trial entries drawn from public registry filings.
clinical-trials-trec-qrels is a tabular relevance-judgment file distributed by the 2001jdev user on the Hugging Face Hub. It maps clinical-trial topic identifiers to NCT registry IDs with graded relevance scores.
The clinical-trials-trec-topics dataset is a small tabular collection of clinical-trial topic records distributed in Parquet format on the Hugging Face Hub. It appears to be derived from TREC clinical-trial retrieval topics, with rows partitioned by year.
SEC EDGAR 10-K & 10-Q financial statements for 2,600+ NASDAQ companies.
Discussion
Questions or observations about this dataset? Join the conversation on X.
No comments yet. Start the conversation on X.