Skip to content
Performance Analysis

Do Politicians Beat the Market? We Analyzed 99,546 Trades.

The short answer: the question is harder than the headlines make it. GovGreed tracks 99,546 deduplicated STOCK Act trades from 354 members of Congress. What the record shows plainly is timing: the median trade reaches the public 28 days after it happens, and 12,766 trades (12.8%) were filed after the 45-day legal deadline.

Source: STOCK Act filings + GovGreed signal scoring Published: 99,546 trades analyzed
Trades Analyzed
99,546
Deduplicated, since 2012
Filed Late
12,766
12.8% past 45 days
Members Trading
354
With disclosed trades
Median Gap
28 Days
Trade to disclosure
Avg Disclosure Gap
50 Days
45-day deadline

The Short Answer

The question “do politicians beat the market?” has been asked since the landmark 2004 Ziobrowski study found that U.S. senators' stock portfolios outperformed the market by 12% annually. Twenty-two years later, with the STOCK Act forcing public disclosure, the raw material is public. Turning it into one honest performance number is harder than it looks.

GovGreed's database contains 99,546 deduplicated STOCK Act filings from 354 members of Congress across 14 years (2012–2026). We cross-reference every trade against 7 independent intelligence layers — committee assignments, bill timing, campaign contributions, lobbying activity, herd behavior, sector momentum, and technical indicators — to produce a master signal score from 0 to 100.

Key finding: The signal score tells you how much independent context lines up behind a trade — the committee, the bills, the donors, the other members buying. It is not a return forecast. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate.

Overall Congressional Trading Performance

Across all 99,546 trades, the picture is more mundane than the headlines suggest. Many trades are routine portfolio management, spousal diversification, or small-dollar positions that have nothing to do with legislative inside knowledge.

The trades worth a closer look are the ones where the context stacks up — a committee member buys a stock in a company their committee oversees, around the time related legislation moves, while their campaign contributors include the same industry. GovGreed's signal engine surfaces those trades so you can read the record behind them.

Out of 354 members who have filed at least one trade, GovGreed's signal engine scores only 61 politicians. The filter is intentional: only trades with an amount midpoint of $50,000 or more are scored, because small-dollar trades (the $1,001–$15,000 range) are statistically meaningless for detecting insider-style signals. A $5,000 trade in Microsoft is portfolio noise. A $500,000 trade in a defense contractor by a member of the Armed Services Committee is a signal.

Signal Tier Performance Breakdown

GovGreed assigns every qualifying trade a master signal score (0–100) and a corresponding tier. The tier describes how much independent context converges on a trade:

Tier Score Range What it means
S 75+ Rare. Nearly every layer converging on one trade.
A+ 60–74 Strong convergence across several layers.
A 50–59 Multiple layers aligning.
B 40–49 Some context, mostly routine.
C 30–39 Little context behind the trade.
D/F <30 No convergence. Routine activity.

We do backtest the tiers, comparing each scored trade's price 30 days after the trade date with its price at the trade. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate. A few dozen trades per tier, measured without subtracting the market's own move, is not a result we are willing to put on a marketing page.

Convergence multipliers matter. A trade with 3 converging signals gets a 1.3x multiplier. 4 signals = 1.5x. 5 or more signals = 2.0x. This is why a trade must light up several independent layers to reach the top tiers.

Congress vs. the S&P 500

The most common benchmark question: do congressional trades beat a simple buy-and-hold of the S&P 500? The answer depends entirely on which trades you're measuring.

A fair test has to subtract what the S&P 500 did over the same window. A trade that returns +8% during a month when the S&P also returns +8% has earned nothing special. It also needs enough trades that a handful of big winners can't carry the average, and it has to treat a sale as a bet that the stock falls.

Our current tier backtest doesn't clear those bars yet. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate. Until it does, the honest answer to “does Congress beat the S&P 500?” is: some members have striking individual trades, and the timing record around those trades is public and worth reading. See it member by member on the Congress leaderboard.

Important caveat: The 50-day average disclosure gap means that by the time these trades are public, any move tied to them may already have happened. Whatever a politician earned from trade date forward is not what a copy-trader would earn from disclosure date forward.

Committee Members vs. Non-Committee

One of the most powerful predictors of congressional trading performance is whether the politician sits on a committee with oversight jurisdiction over the company they're trading. GovGreed's signal engine captures this through the bill correlation and contribution pattern layers.

The pattern is what you'd expect if committee knowledge matters:

  • Members of the Armed Services Committee who trade defense stocks score higher on the bill correlation layer than members who trade the same stocks without committee overlap.
  • Members of the Financial Services Committee trading bank stocks before regulatory announcements show elevated bill correlation scores.
  • Members of the Energy and Commerce Committee trading energy and healthcare stocks show significantly higher sector alignment in GovGreed's lobbying pattern layer.

GovGreed tracks bill-trade correlations — the timing relationship between when a politician trades and when related legislation moves through their committee. A high bill correlation score means the trade landed inside a window of related legislative activity.

Committee-Aligned Trades
Flagged
Committee, bill, and donor layers
all fire on the same trade
Non-Committee Trades
Routine
No committee overlap
behind the trade

This doesn't prove insider trading — GovGreed is a data platform, not a prosecutor. But it does show that the information environment inside congressional committees overlaps with what members trade. The signal engine surfaces these patterns so users can evaluate them.

Top Performers (Bipartisan)

Active trading is bipartisan. Both parties have members who rank highly on GovGreed's signal scores. The table below shows traders ranked by buy win rate (percentage of purchases that beat the S&P 500 over the following 30 days), loaded live from GovGreed's database.

Loading top performer data from Supabase...

Source: per-politician trade-return aggregates from GovGreed's live database. Win rate = percentage of stock purchases that beat S&P 500 over 30 days. Minimum 20 buy trades required for inclusion. Updated daily. Raw per-member figures on small samples; not a forecast.

Does Being Rich Make You a Better Trader?

You'd assume the wealthiest members — the ones with hundreds of millions in declared assets — would be the sharpest traders. We matched each member's declared net worth (from their Personal Financial Disclosure) against their trading record so you can compare the two.

First, the very richest barely trade at all. Senator Jim Justice (R-WV, ~$1.3 billion) and Rick Scott (R-FL, ~$480M) have almost no active trading record — their wealth sits in operating companies and blind trusts, not a brokerage account they're churning. Among members who do trade actively, the table below loads their record live from GovGreed's database.

# Member Declared net worth Trades Buy win rate Trader tier
Loading live data…
Wealth and trading record are separate questions. Large-portfolio members often hold operating companies, managed accounts, or blind trusts, so a big declared net worth says little about how actively a member trades individual stocks. Per-member win rates in this table are raw and often rest on a few dozen trades — read them as a description of the record, not a skill ranking.

Source: declared net worth from Personal Financial Disclosures (most recent parsed filing, assets−liabilities midpoint); buy win rate = share of purchases beating the S&P 500 over 30 days, minimum 25 buys. Large-portfolio members often trade through managed accounts and blind trusts, which dampens any single-name “skill” signal. Loaded live from GovGreed's database.

The Disclosure Delay Problem

The elephant in the room for anyone trying to copy congressional trades: the disclosure gap. Under the STOCK Act, members of Congress have 45 days to disclose a trade. The actual average delay is 50 days (median 28), and 12,766 deduplicated trades (12.8%) were filed after the 45-day deadline.

This creates a structural information asymmetry. By the time the public learns that a senator bought $500,000 of a defense contractor, the stock may have already moved. Any return measured from the trade date describes what the politician saw — not what a copy-trader would see from disclosure date forward.

That said, GovGreed's signal engine partially addresses this by scoring trades at the time of disclosure and using freshness labels (FRESH, NEW, RECENT, AGING, STALE) to indicate how much time has passed since the trade was executed. A high score on a fresh disclosure is the most timely thing the record can show you.

Average disclosure gap: 50 days. Median: 28 days. Worst single case: 997 days. 12,766 trades (12.8%) were filed late. The $200 fine for late filing has produced zero deterrent effect. See the full breakdown on GovGreed's Late Filing Tracker.

How GovGreed's Signal Engine Works

GovGreed's signal scoring isn't a single model — it's an intelligence fusion engine that combines independent data sources into a composite score. Each input contributes a weighted percentage to the master score, and convergence multipliers amplify trades where multiple inputs agree:

  • 20% Politician Quality — Historical trading track record. Win rates, trading styles, options usage, portfolio concentration. 247 politicians scored.
  • 20% Herd Signals — 3+ politicians converging on the same stock within a detection window. Currently tracking 31 active herd signals. Multiple members buying the same stock is the clearest convergence pattern.
  • 16% Bill Correlation — Timing analysis between trades and legislative activity across every scored trade. Detects trades that occur suspiciously close to favorable committee activity.
  • 12% Technical Context — RSI, SMA crossovers, volume surge analysis. Confirms whether the stock's technical picture supports the politician's trade direction.
  • 12% Sector Momentum — Congressional buy/sell ratio by sector, combined with sector price momentum. Identifies sectors where insider buying is concentrated.
  • 10% Campaign Contributions — 565 contribution patterns mapping campaign donations to subsequent trades in the donor's industry. Follow the money.
  • 10% Lobbying Alignment — 2,101 lobbying patterns connecting lobbyist activity to politician trading in the same tickers. Detects influence-to-trade pipelines.

The master score is computed as a weighted sum of all 7 layers, then multiplied by a convergence factor: 1.3x for 3 converging signals, 1.5x for 4, and 2.0x for 5 or more. This multiplicative design means that a trade must light up multiple independent indicators to reach A+ or S tier. A single hot signal alone isn't enough.

The engine runs on 2,790 scored signals across 61 politicians. Only trades with an amount midpoint of $50,000 or more qualify, because small-dollar trades don't carry enough conviction to be statistically meaningful as insider signals.

Explore the signals yourself

Browse every trade, every signal score, every politician. Free access.

View Signal Dashboard →

Frequently Asked Questions

Do politicians beat the stock market? ▼
There is no honest single number. GovGreed tracks 99,546 deduplicated congressional trades from 354 members, and most of them are routine portfolio activity. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate. What the record does show clearly is timing: the median trade is disclosed 28 days after it happens, the average 50 days, and 12,766 trades (12.8%) were filed after the 45-day deadline.
What is GovGreed's signal scoring? ▼
GovGreed scores congressional trades using 7 independent intelligence layers: politician quality, herd signals, bill correlation, technical indicators, sector momentum, campaign contributions, and lobbying patterns. Each layer contributes to a master score (0–100), which determines a tier from S (75+) down to F (<20). Convergence multipliers amplify trades where 3+ layers agree. The system identifies which trades carry the most insider-style conviction.
How accurate are congressional trading signals? ▼
GovGreed backtests each signal tier by comparing a scored trade’s price 30 days after the trade date with its price at the trade. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate. A tier describes how many independent layers line up behind a trade — committee, bills, donors, lobbying, herd activity — not a forecast of its return.
Which politicians have the best stock returns? ▼
The top-performing politicians span both parties. Performance is measured by excess return (trade return minus S&P 500 return) and win rate (percentage of trades that beat the benchmark). GovGreed's Congress leaderboard ranks all 354 trading politicians by these metrics, updated daily from live data.
Can you copy congressional stock trades? ▼
Copy-trading Congress is theoretically possible but practically challenging due to the 50-day average disclosure delay (median 28 days). By the time a trade is public, any move tied to it may already have happened. GovGreed's signal engine helps by scoring disclosed trades and identifying which ones still have forward-looking conviction, but the disclosure gap remains the biggest structural obstacle. Read our copy-trading guide for a full breakdown.
Why do most congressional trades score low? ▼
Not all congressional trades are informed by insider knowledge. Many are routine portfolio management, spousal trades, or small diversified allocations. GovGreed's D/F tier signals (scores below 30) are trades with no convergence behind them — no committee alignment, no bill correlation, no contribution pattern, and no herd activity.
Does GovGreed publish a tier win rate? ▼
No. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate. A+ tier signals (master score 60–74) represent trades with strong convergence across multiple intelligence layers.
About this data. All statistics on this page are derived from GovGreed's database of 99,546 deduplicated STOCK Act filings, sourced from QuiverQuant and FMP APIs. Signal scores are calculated using GovGreed's intelligence engine with convergence multipliers. Tier backtests are small-sample and raw (not excess vs the S&P 500), so we don’t publish a tier win rate. Only trades with an amount midpoint of $50,000+ are scored. Data coverage: 2012–2026, 354 members with trades, 61 members with scored signals. Updated daily.
Not financial advice. All data sourced from public federal disclosures: STOCK Act filings via QuiverQuant and FMP APIs, cross-referenced with Congress.gov, FEC, SEC EDGAR, and Senate LDA data. GovGreed provides data and transparency — it does not make investment recommendations or legal accusations. Past performance does not guarantee future results. Trade counts current as of September 2026.