Cluster buys: three insiders, one week
Three or more insiders buying in the same ten-day window is the most quoted insider signal. It works, but mainly in stress, and the bank version of it works only in a panic.
- Trades: every open-market insider purchase reported to the SEC on Form 4 from 2006 onward that passes our routine-versus-opportunistic filter; for this piece, three or more distinct insiders buying within the same 10-day window.
- Entry: first session after the filing became public. Held 63 trading days (about three months).
- Benchmark: IWM (small caps), because insider-buy candidates skew small and cheap; beating the S&P alone would just be that tilt.
- Market state: stress = SPY at least 10% below its trailing 1-year high on the entry day; calm otherwise.
- Statistics: clustered by company (one vote per company) — five insiders buying the same stock in the same week count once. Cells under 40 trades or 15 companies are shown but greyed. No trading costs deducted.
- What is excluded: companies without a ten-year price history (delisted names are the main gap; see the survivorship note on the evidence page).
Three-month excess return versus small caps for cluster buy (3+ insiders)s made during sell-offs, by sector. Sectors with too few trades omitted from the chart, shown in the table.
| Sector (SEC industry group) | In sell-offs | In calm markets | Read |
|---|---|---|---|
| Transport & logistics | +42.9 pts n 38 · t 2.6 | +14.0 pts n 112 · t 2.2 | Too few trades to say |
| Healthcare services | +33.9 pts n 16 · t 0.9 | +6.2 pts n 40 · t 0.8 | Too few trades to say |
| Tech hardware & semis | +28.8 pts n 89 · t 3.3 | −0.0 pts n 195 · t -1.4 | Sell-offs only; loses in calm |
| Medical devices | +20.5 pts n 101 · t 1.8 | −2.3 pts n 198 · t -0.1 | No reliable edge |
| Telecom & media | +15.4 pts n 20 · t 1.0 | −8.7 pts n 210 · t -0.5 | Too few trades to say |
| Mining & metals | +13.6 pts n 16 · t 0.3 | −7.7 pts n 106 · t -1.0 | Too few trades to say |
| Software & IT services | +12.4 pts n 78 · t 2.1 | −3.6 pts n 243 · t 0.2 | Sell-offs only; loses in calm |
| Construction | +9.9 pts n 27 · t 1.1 | −13.9 pts n 110 · t -1.5 | Too few trades to say |
| REITs | +9.3 pts n 465 · t 0.9 | +1.8 pts n 839 · t -0.1 | No reliable edge |
| Retail & consumer | +9.0 pts n 232 · t 1.9 | +2.0 pts n 768 · t 0.8 | No reliable edge |
| Energy | +7.6 pts n 277 · t 2.7 | −0.2 pts n 572 · t -1.4 | Sell-offs only; loses in calm |
| Industrials & manufacturing | +5.3 pts n 711 · t 2.1 | +4.3 pts n 1,618 · t 1.2 | Sell-offs only |
| Insurance | +4.5 pts n 262 · t 0.5 | +1.1 pts n 373 · t -0.7 | No reliable edge |
| Biotech & pharma | +2.5 pts n 202 · t 0.8 | +7.2 pts n 550 · t 1.4 | No reliable edge |
| Other | +2.2 pts n 304 · t 1.3 | +5.9 pts n 469 · t 0.7 | No reliable edge |
| Services | +2.0 pts n 453 · t 2.2 | −1.0 pts n 703 · t 1.7 | Sell-offs only; loses in calm |
| Utilities | +1.2 pts n 367 · t 1.1 | +0.5 pts n 1,039 · t -1.3 | Too few trades to say |
| Finance & real estate | +0.0 pts n 80 · t 0.7 | −0.5 pts n 240 · t -0.6 | Too few trades to say |
| Banks | −0.0 pts n 2,221 · t 0.1 | +1.4 pts n 7,513 · t 3.9 | Calm markets only |
| Food, drink & tobacco | −1.7 pts n 17 · t -0.3 | +0.8 pts n 89 · t -1.8 | Too few trades to say |
t is the clustered t-statistic: above 2 means the result is unlikely to be luck; above 3, very unlikely. "n" is trades.
Market state now: NORMAL — the S&P 500 is 3.1% below its one-year high (as of 2026-09-16). The calm column above is the one that applies today; the sell-off numbers are what to expect if conditions change.
Sectors under pressure right now (their own ETF is in a drawdown):
Buys in the last 45 days that match this definition (newest first, only names we have a page for):
A match is a candidate, not a recommendation: the table above says what this definition has done in this market state and sector on average.
SAVNG is an educational research publication, not a Registered Investment Adviser. Everything here is general-circulation information produced by an automated pipeline — the same for every reader — and is not personalized investment advice, an offer, or a recommendation to buy or sell any security. Any performance figure is hypothetical; past performance does not predict future results. Details.
Published by SAVNG Research. Every figure on this page is read from the study data at render time; if the study changes, the page changes with it.