Aggregate Form 4 SEC filings over a trailing 90-day window; compare insider buy vs. sell dollar value and flag clustered same-side activity; classify posture
Answers whether company insiders are net buyers or sellers of their own stock. It aggregates SEC Form 4 filings into net buy and sell dollar values with buy/sell transaction counts and any clustered activity. Each ticker is classified accumulating when open-market purchases exceed open-market sales both by dollar value and by count, distributing when sales exceed purchases both ways, and neutral otherwise. Clustered insider buying is one of the more closely watched patterns, while selling is noisier because insiders sell for reasons unrelated to a view, such as diversification, taxes, and scheduled plans; filings also arrive with a reporting lag and some buys are routine plans.
Insider transactions, one of the few places in equities where a real information asymmetry is disclosed by law.
THE FILING. US securities law requires SECTION 16 persons - directors, officers, and beneficial owners of more than 10% of a registered class of equity - to report transactions on FORM 4, generally within two business days. That mandatory disclosure is why the dataset exists.
WHAT COUNTS. Form 4 carries far more than open-market trades: grants, option exercises, gifts, and dispositions withheld for tax all appear on it. Aggregating those as informational buying and selling would be meaningless, so a measure of this kind concerns NON-DERIVATIVE OPEN-MARKET purchases and sales specifically. This analysis counts only the transactions the data marks as open-market buys and sells, and leaves grants, exercises, gifts and tax withholding out.
THE EMPIRICAL ASYMMETRY. Lakonishok and Lee (2001) find that insider PURCHASES are more informative than sales. That is not the same as 'selling does not inform' - the finding is comparative, not a dismissal. The usual explanation is that an insider has fewer reasons to buy than to sell: diversification, liquidity needs, tax, and scheduled 10b5-1 plans can all produce selling unrelated to a current view. Be careful with that argument too - it explains why selling is NOISIER, and without plan-level data you cannot claim most selling was predetermined. Nor are 10b5-1 plans automatically uninformed: their adoption, modification and cancellation can themselves be strategic, and cooling-off requirements differ by person under the current rules.
CLUSTERING is the refinement: several reporting persons buying within a short window is harder to explain by one individual's circumstances. What the filings establish is distinct reporting persons and temporal proximity - not that their decisions were independent.
REPORTING LAG means this is a record of what was done, not of what is being done. And it covers US issuers subject to Section 16; equivalent disclosure elsewhere differs in timing and scope.
AGGREGATION CONVENTION. Net buy and sell values are compared over 30- and 90-day windows. The window is part of the measurement: a cluster that looks decisive over 30 days can be unremarkable over 90, and a longer window dilutes a recent change while a shorter one is dominated by any single large filing.
THE POSTURE RULE. A ticker is accumulating when its open-market purchases exceed its open-market sales both by dollar value and by count, distributing when sales exceed purchases both ways, and neutral in every other case, including when value and count point different ways. No dollar threshold is applied: the same sum means very different things at different company sizes and to holders of different wealth, which is why the rule asks value and count to agree instead of a size.
The accumulating / distributing / neutral classification is our own convention.
Published reports are readable without an account. The newest is Ether (ETH) and XRP extend a month-long slide as Bitcoin falls below $75,000: trend, momentum and news (Nov 9, 2026). Create an account to run research on the stocks and crypto you choose.