SwealthEdge

Explainers / INSDR

Insider clusters

Several distinct insiders filing open-market purchases in the same issuer inside a rolling window. The detection is about the filings, not about the company.

Engine family slug insider_cluster · methodology insider_cluster/1.0.0

What has to be true

  • Three or more distinct insiders of the same issuer file Form 4 transactions coded P — an open-market purchase — with transaction dates inside a rolling ten-trading-day window.
  • Grants, option exercises, gifts, and derivative transactions are excluded. The exclusion is made from the transaction codes on the filing itself.
  • Transactions disclosed as made under a Rule 10b5-1 plan are excluded where the filing says so, because a planned purchase carries no timing information.
  • The aggregate disclosed consideration is measured against the issuer's average daily dollar volume, so a cluster in a name that trades nothing is described as such rather than ranked alongside one that does not.

Thresholds

The settings this family applies. They are published because they are choices: a different window or a different count produces a different set of signals from the same data.

Distinct insiders
at least 3

Counted by reporting person, not by filing: one person filing three times is one insider.

Window
10 trading days, rolling
Aggregate disclosed consideration
reported, not gated

Published whatever the size, and shown so a reader can discount a small one.

Average daily dollar volume
reported, not gated
Sample
US issuers with Section 16 reporting obligations

There is no equivalent filing in most other jurisdictions, so this family is US-only.

The published work it draws on

Named, not reproduced and not linked. Each entry says what that work reported, in its own terms. None of it was conducted by us, and none of it is a statement about the current market.

  • Lakonishok, J. and Lee, I. (2001), ‘Are Insider Trades Informative?’, Review of Financial Studies 14(1).

    Examined aggregate insider activity and reported that purchases carried more information than sales, with the effect concentrated in smaller companies.

  • Jeng, L., Metrick, A. and Zeckhauser, R. (2003), ‘Estimating the Returns to Insider Trading: A Performance-Evaluation Perspective’, Review of Economics and Statistics 85(2).

    Measured insider purchases as a portfolio and found abnormal performance that was materially smaller than earlier event-study estimates.

  • Cohen, L., Malloy, C. and Pomorski, L. (2012), ‘Decoding Inside Information’, Journal of Finance 67(3).

    Separated routine filers from opportunistic ones and found that the informative component sat almost entirely with the latter — an argument that who files matters as much as how many.

Where this does not work

Every detection method has conditions it was measured under and conditions it fails in. These are this one’s, written down rather than left out.

  1. A Form 4 is public the moment it is filed, and filings are due within two business days of the transaction. Nothing detected here is private information, and any effect is measured on data every other reader already has.
  2. The cluster definition is a choice, not a fact. Moving the window from ten trading days to fifteen, or the count from three insiders to four, changes which issuers appear and how many. We publish the settings so the choice is visible; we do not claim it is the right one.
  3. The exclusions depend on how a filing was coded. Rule 10b5-1 disclosure has been inconsistent across filers and across time, so some planned purchases will be counted as opportunistic ones and will dilute the signal.
  4. The published literature above concentrates in small-capitalisation issuers, where bid-ask spreads are wide. An effect measured before costs is not the same as an effect that survives them, and we do not model costs.
  5. Those studies cover sample periods that ended years before this engine ran. We have not re-run them on current data, and an anomaly that is widely published is an anomaly that other people are also acting on.
  6. A cluster is a statement about a set of filings. Insiders buy for reasons that have nothing to do with their view of the business, including to meet an ownership requirement or because a plan matured.

What this page cannot tell you

  • Whether this issuer is a suitable holding for anybody, including you.
  • What the insiders knew, believed or intended. We observe filings.
  • What happens next. The detection describes a condition that has occurred.