When a corporate insider buys or sells company stock, the transaction is usually reported on SEC Form 4. Headlines often reduce the filing to “insider bought” or “insider sold,” but the form is more specific: it encodes transaction type, partial ownership context, and sometimes footnotes about tax withholding or option exercises. Learning the codes is one of the fastest ways to reduce false signals in insider-based research.
Form 4 is a statement of changes in beneficial ownership. It is filed by officers, directors, and large shareholders subject to Section 16 rules for most U.S. public issuers. It is not a real-time order feed, not a guaranteed expression of investment intent, and not limited to “bullish” open-market purchases. Many transactions are administrative: grants, exercises, gifts, or sales to cover taxes.
Insiders generally must report transactions within two business days. That is fast compared with 13F quarter-end lag, but it still is not the moment of execution. Price may have moved, and subsequent filings can amend or clarify earlier lines. For research, record the filing date separately from the transaction date shown on the form.
The “code” column is the heart of Form 4 literacy. A few codes appear repeatedly:
Footnotes can explain 10b5-1 plans, indirect ownership through trusts, or multiple transactions netted in one line. Form 4/A amendments can correct mistakes. If your data source hides footnotes, you are flying partially blind—good platforms preserve them for a reason.
Insider filings answer “what did specific individuals report trading?” 13F snapshots answer “what did large managers hold at quarter end?” Forecast track records answer “how did stated predictions score under fixed rules?” Use each for its own question—our workflow guide shows an order of operations.
Insider Transactions: Context Beats Headlines · Combine 13F, Insider, and Forecast Signals
Form 4 is worth reading in detail because the codes and footnotes carry the meaning. Start with transaction type and role, demand repetition before strong conclusions, and treat headlines as a lossy compression of public data.