Skip to content
Insider tapeFilings through Fri, Oct 23
Watchlist

Signal Lab: test any insider buying setup

Pick the traits of an insider buy and see how every matching case since 2020 did over the next 90 days against the S&P 500. Each result shows its sample size, the full spread, and whether it held up in both the early and late years.

Build a setup

Every filter uses data the market knew on the filing date. The result updates as you change them.

Loading 16,899 cases…

How Signal Lab works, and its limits
  • Cases: 16,899 open-market insider purchases (Form 4 code P, $10K+) in U.S. stocks since 2020. Buys in one stock within 14 days count as one case.
  • Result: the stock's price change over the 90 days after the last filing became public, minus the S&P 500's over the same days. "Beat" means that difference was above zero.
  • Out-of-sample check: we split the cases at 2023-04-01. A setup that only worked in one half is more likely luck. Results with fewer than 25 cases aren't shown.
  • Limits: survivorship bias (delisted stocks often drop out, which flatters results), price-only returns, no trading costs, one 90-day horizon. Testing many combinations will turn up some by chance.
  • History, not a prediction. Not investment advice. Full methodology

Prefer a plain answer? Ask the filings. Or test your instincts with today's Insider Call.