TradeStar Insider · What we tested
We sell filing-verified facts, not a predictive signal. Here is the honest record of every test we ran in-house to find a forward edge in insider buying — and why we do not claim one.
We ran four pre-registered, sealed tests. In each, the spec (universe, horizon, control, PASS thresholds) was frozen and sealed before any return was joined to any signal, then scored once with no re-slicing and no post-hoc tuning. All four KILL → NULL: no forward edge distinguishable from the same-week, same-universe base rate on our in-sample window.
A KILL is the pre-committed default and a clean outcome, not a disappointment. It means "no edge demonstrable here," not "no effect exists."
| Question | Do insider cluster buys — several insiders buying the same issuer in a window — beat the base rate? |
|---|---|
| Method | Entry on the filing's public filed_date; control = equal-weight mean forward return of all priced purchase events in the same ISO week; stationary block-bootstrap (block=5, 10,000 resamples), Benjamini–Hochberg FDR q=0.10. PASS requires all three: edge ≥ +1.5%, statistically significant, breadth ≥ 55%. |
| Sample | Study 1 — 151 U.S. trading days; 1,385 priced (ticker, filed_date) purchase events. |
| Number | +10d primary edge +0.283% (p=0.730), breadth 46.5%. |
| Verdict | KILL — fails all three legs (economic floor, significance, breadth). |
| Question | Do insider buys that increase holdings by ≥10% beat the base rate? |
|---|---|
| Method | Same frozen harness, same-week same-universe control, and same three-leg PASS gate as test 1 (block-bootstrap, BH-FDR q=0.10). |
| Sample | Study 1 — same 151-day window, 1,385 priced events. |
| Number | +10d primary edge +1.147% (p=0.332), breadth 47.4%. |
| Verdict | KILL — fails all three legs. |
| Question | Do large officer purchases in small-cap issuers beat the base rate? |
|---|---|
| Method | Study 2 (Cohen–Malloy–Pomorski officer/size conditioning). Small-cap = bottom cap tercile and absolute cap < $2.0B (dual gate; bottom-tercile cutoff $306M). +21d theory-matched primary horizon; same control, block-bootstrap, and BH-FDR q=0.10 machinery; n ≥ 100 power floor. |
| Sample | Study 2 — 157-day window; n=116 priced small-cap officer events at +21d. |
| Number | +21d edge +9.74% (p=0.243, 95% CI [−3.64%, +27.89%]), breadth 43.1%. |
| Verdict | KILL — genuinely powered (n≥100) but not significant and breadth < 55%; the large point estimate is carried by a minority of names, which is exactly what the frozen breadth gate exists to reject. |
| Question | Do opportunistic officer buys — stripping out routine-cadence buyers — beat the base rate? |
|---|---|
| Method | Study 2. Routine-proxy = a buyer with ≥3 distinct buy-months spanning ≥90 days; those are removed, the rest are opportunistic. +21d primary horizon; same frozen inference. |
| Sample | Study 2 — n=468 priced opportunistic officer events at +21d (only 18 routine-proxy buyers existed to remove). |
| Number | +21d edge +4.21% (p=0.157, 95% CI [−0.98%, +10.92%]), breadth 44.7%. |
| Verdict | KILL — not significant, breadth < 55%; removing routine buyers did not sharpen the signal to significance. |
All four tests will be re-run under the same sealed pre-registrations once our Form 4 backfill covers two years — and we will publish the result either way, KILL or not.