Public record · House STOCK Act disclosures
Everyone “knows” Congress beats the market. We checked all 19,670 trades.
The headlines are real. What they leave out is that you are not Nancy Pelosi — and the politician you’d actually copy loses to an index fund.
How we measured it
Enter the day you could — not the day they did
Every tracker shows the trade date. But a member discloses up to 45 days later, so that price was never available to you. The only honest test buys on the disclosure date — the first moment the trade is public — holds twelve months, and scores against the S&P 500. Buys only.
Result 1 · Copy everyone
Do exactly what the pitch says, and you trail the market
Buy every disclosed purchase the day it goes public. Here is the median outcome — and how many members actually clear the bar the whole strategy is sold on.
The median member of Congress returns −2.6% against the market on the typical trade. Fewer than a third beat a fund you can buy for three basis points.
Result 2 · The surprise
It isn’t the 45-day delay
The obvious explanation is staleness — you’re a month-and-a-half late. But that’s not it. Entering on the disclosure date instead of the trade date costs almost nothing over a year:
Which means the problem isn’t timing. The trades simply don’t carry a signal. Most congressional buying is ordinary portfolio churn — index funds, bonds, broad tech — not secret conviction. Waiting 45 days barely changes a return that was never there.
Result 3 · “But Pelosi!”
She’s real. She’s also not you.
The counterargument has a name, and the numbers behind it are genuine — even measured our conservative way, Nancy Pelosi’s disclosed buys beat the market. The question is why, and whether it’s anything you could repeat.
+70.9% in 2024, against the S&P’s +24.9%. Her Nvidia call options returned +273%; Palo Alto calls +93%. Over the decade, +838% versus the index’s +256%.
That number is real — and it’s a leveraged number. Paul Pelosi buys long-dated call options on big tech after dips, not shares. Strip the leverage and score her actual buys the same way we scored everyone else, and here is what’s left — all four bars are the same thing: median 12-month return versus a benchmark.
Two things fall out. Her raw buys return +22% a year — but so did big tech; benchmark her against the tech index and a quarter of the “genius” is just being long Nvidia and friends through the biggest tech run in history. What survives — about +7% over even the Nasdaq, with a 62% win rate — is real residual skill, or the one bet in a hundred that had to land somewhere. Because here’s the frame the headline never shows you:
Pelosi is one of the roughly 31% who beat the market. The typical member is on the other side of that line — and that’s who “follow Congress” actually buys you.
The verdict
The averages lie; the median tells the truth
| Strategy | Median α | Win rate | What it really is |
|---|---|---|---|
| Copy every member | −3.6% | 44% | index churn |
| The median member | −2.6% | 31%* | underperforms |
| Pelosi vs S&P | +9.1% | 62% | leverage + tech + selection |
| Pelosi vs tech index | +6.8% | 57% | thin residual edge |
* share of members beating the S&P, not a per-trade win rate.
The Pelosi headline and “Congress can’t beat an index fund” are not in conflict. They’re the same fact seen from two ends: a distribution with a heavy losing middle and a few leveraged, tech-heavy winners the press writes about precisely because they won. Copy the feed and you get the middle. You don’t get Pelosi — Pelosi got Pelosi.
What this is and isn’t
- Survivorship. Roughly 15–18% of tickers are delisted or acquired and drop out of the price history. That distorts absolute levels, but the direction — sub-50% win rates, negative medians — holds across every window.
- Not risk-adjusted. Raw returns vs a benchmark, equal-weight per disclosed buy, no costs, no sizing. A concentrated leveraged book like Pelosi’s carries far more risk than the index it’s measured against.
- Selection is the point. We looked at Pelosi because she’s famous for winning. Across 207 members, someone tops the table by chance; her ~7% edge over the tech index is suggestive, not proof of skill.
- Not advice. Analysis of public disclosures, reproducible from the House Clerk’s filings — not a recommendation to buy, sell, copy, or avoid anything.
