"Congress trades on inside information, so copy Congress" is the pitch behind a growing pile of apps and ETFs. We built the honest version of that test and ran it on every member we track.
The rules
The STOCK Act requires members to disclose trades within 45 days. A real copier can't trade earlier than the disclosure, so:
- Execution: each disclosed trade executes at the first close after its disclosure date — not the trade date the member actually got.
- Sizing: buys invest the midpoint of the disclosed amount range (disclosures give ranges like $15,001–$50,000, not exact figures); partial sales trim by the midpoint; full sales liquidate the position.
- Scope: stocks only — options and unpriced assets excluded. NAV is time-weighted, so the index measures decision quality, not cash-flow timing.
- Bar to qualify: ≥5 copyable buys and ≥1 year of history. 54 members qualify.
The spread is the story
Top and bottom of the leaderboard, by annualized alpha vs SPY over each member's own copy window:
| Member | Copied buys | Since | Copy CAGR | SPY CAGR | Alpha |
|---|---|---|---|---|---|
| John Fetterman (Sen.) | 6 | 2025-06 | +116.4% | +24.0% | +92.4pt |
| Tommy Tuberville (Sen.) | 20 | 2024-04 | +78.1% | +20.6% | +57.6pt |
| Dwight Evans (Rep.) | 8 | 2024-01 | +64.8% | +21.3% | +43.5pt |
| Tim Moore (Rep.) | 91 | 2025-02 | +58.1% | +17.3% | +40.8pt |
| Morgan McGarvey (Rep.) | 5 | 2024-07 | +49.5% | +17.7% | +31.8pt |
| … | |||||
| John McGuire (Rep.) | 2025 | +4.1% | −20.0pt | ||
| April McClain Delaney (Rep.) | 2025 | −1.0% | −22.4pt | ||
| John Hickenlooper (Sen.) | 11 | 2024-06 | −12.1% | +18.7% | −30.8pt |
Copying Tuberville turned $100 into $259 in about two years. Copying Hickenlooper turned it into $77 — while SPY made +42% over the same window.
What to make of it
The average doesn't matter — dispersion does. "Copy Congress" as a blanket strategy buys you the whole distribution, including the bottom of it. The useful question is never whether to copy Congress; it's who, and whether their edge is real or luck.
Most of these track records are short. The disclosure dataset only gets dense from 2024, so even the leaders have 1–2 years of history — a small enough sample that a single concentrated bet (Fetterman's six buys) can dominate the number. Treat the leaderboard as a watchlist, not a ranking of skill.
Sizing is fuzzy by construction. Disclosures report ranges; our midpoint convention can misstate any single trade's size by ~3x in either direction. Portfolio-level results are more robust than any individual line item.
Your signal source can retire on you. Tuberville — the most copyable senator by this measure — isn't seeking Senate re-election; he won Alabama's Republican gubernatorial nomination in May 2026. Governors don't file STOCK Act disclosures. A copy-trading strategy carries key-person risk that no backtest captures.
Every member's full simulation — NAV curve, drawdown, buy count — is on their page. Start with the copy-trade leaderboard and click through.
Methodology and code: scripts/build-congress-copytrade.ts. Disclosure data from STOCK Act periodic transaction reports; 53 option/derivative trades and 2,333 unpriced trades excluded across the dataset. Member status (Tuberville's gubernatorial run) verified against public reporting as of July 2026. Nothing here is investment advice.