Verified-edge traders — timing that survives statistical correction

Most members of Congress show no abnormal trade timing — on average, it really is a coin flip. But a small, bipartisan subset beats the market at a rate that survives multiple-comparison correction. These are the members whose buy timing (beating the S&P) or sell timing (avoiding losses) is statistically distinguishable from luck. We show the win rate, the sample size, and the corrected significance — you draw the conclusions.

Method: per-trade · 30-day S&P-500-excess return · execution-date anchored · option exercises excluded · Fisher-exact + Benjamini-Hochberg FDR · "Proven edge ✓" = q < 0.05

Proven edge members significant after FDR correction (q < 0.05)

Buy timing (beat the market)
Sell timing (avoided losses)
Loading FDR-corrected results…

All members tested every member with ≥20 trades of that side · ranked by corrected significance

Loading…
See every member we tested — and how close they came
The full ranked table (132 member-sides, win rate + corrected q-value + sample size for each) is a Founders feature. The proven-edge names above are always free.
Unlock the full table — Membership $29.99/mo

Recent smart exits latest sells by members whose exit timing is FDR-significant — loss-avoidance, not a bullish call

Loading recent exits…
Methodology. Unit of analysis is the individual trade. Outcome is the S&P-500-excess return over 30 days, anchored to the execution (trade) date. A buy "wins" if the stock beat the market; a sell "wins" if the stock subsequently underperformed the market (good exit timing — loss avoidance). Option exercises are excluded. For each member with at least 20 trades of a side we run a one-sided binomial test against the pooled base rate and a one-sided Fisher-exact test against all other members, then apply a Benjamini-Hochberg false-discovery-rate correction across the tested set (separately for buys and sells). A member earns "Proven edge ✓" only if their corrected q-value is below 0.05 — i.e. their record is unlikely to be one of the high win-rates you'd expect by chance from testing this many members. This is the skilled/informed subset that an aggregate test averages away; it is evidence of abnormal return, not proof of trading on nonpublic information. Not financial advice. All data from public STOCK Act disclosures, prices from market data.