Insider Buying Stocks: A Source-First Research Workflow
How to screen reported insider purchases without turning every Form 4 row into a false conviction signal.
An insider-buying screen finds open-market or private purchases reported by officers, directors, and qualifying owners. The screen should filter Form 4 transaction codes, security type, ownership, price, quantity, post-transaction holdings, filing time, and footnotes. It must exclude or separately label grants, option exercises, tax withholding, gifts, and amendments.
Define buying before ranking it
A transaction described loosely as acquisition may come from compensation, exercise, conversion, or another non-open-market event. Start with the SEC code and table, then inspect footnotes and derivative details.
Dollar value alone can be misleading. Compare the purchase with the insider’s existing holdings, role, compensation context, transaction price, and other transactions in the same filing.
Make the screen historically valid
Use the SEC accepted time, not the transaction date, as the first public observation. Apply a next-tradable-session rule for after-hours filings and keep amendments as later versions.
A point-in-time universe must include companies that later delisted or changed tickers. Rebuilding a 2018 screen from today’s active symbols removes many of the outcomes the strategy would actually have faced.
- Filter codes before calculating purchase value.
- Group related rows without losing them.
- Enter signals only after public filing time.
- Reject issuers with incomplete coverage windows.
Use the result as research input
Insider purchases can identify events worth investigating, but they do not prove undervaluation or predict returns. Join them to filings, earnings, fundamentals, news, and market data available at the same cutoff.
Record the exact screen definition with every result. Changing minimum value, role filters, clustering windows, or universe rules creates a different strategy and should be versioned accordingly.
How DataCedar preserves the evidence
DataCedar separates acquisition from serving. Permitted source responses are retained with retrieval time and identifiers, normalized into DataCedar-owned tables, checked against expected coverage, and exposed through a stable versioned API. A collector can be replaced without changing the customer contract or making an upstream provider a runtime dependency.
Every research stream carries effective and known-at time where the distinction matters. Rights-restricted, unavailable, partial, stale, and genuinely empty states remain visible, so a backtest can fail closed and a buyer can see the product boundary before committing engineering time.
Key takeaways
- 01Filter true purchase codes from grants and exercises.
- 02Use filing accepted time for public availability.
- 03Normalize by holdings and context, not value alone.
- 04Treat the screen as evidence, not advice.
Query public company evidence through one stable API. Free Explorer tier, no card.
Start freeView pricing