What happens after a member of Congress buys a stock
By Daniel, The Philosopher Investor · published 2026-09-19 · prices through 2026-09-18 · free to read and to cite with a link
Congress trades get a lot of press. The question a trader cares about is simpler: if you bought what they bought, on the day the filing went public, what would you have made? This study answers it with the House filings we collect every day and the same daily prices the desk uses.
Ninety days after the filing, in one line
30 days vs SPY
-0.5%
median · 46% beat SPY · n=1,715
60 days vs SPY
-1.2%
median · 46% beat SPY · n=1,601
90 days vs SPY
-1.8%
median · 45% beat SPY · n=1,530
Sample
1,828
purchases · 38 members · 599 tickers
Average return against SPY, day by day after the filing
Average excess return of every purchase, in percent, by trading day after disclosure (day 0 = the first close a reader could buy at). Only days with at least 30 purchases are drawn. Table view.
What the numbers say
- The typical purchase did not beat the market. Ninety days after disclosure the median stock bought by a House member returned +2.1% outright and -1.8% against SPY, across 1,530 purchases with a finished window. Only 45% of them beat SPY.
- The picture is the same at every horizon: median excess return of -0.5% at 30 days (46% beat SPY), -1.2% at 60 days (46%), -1.8% at 90 days (45%).
- The average is a little above zero (+0.6% at 90 days) while the median is below it. A few big winners pull the mean up; most purchases sit slightly behind the index.
- Size of the purchase changed the result. Purchases filed in the $1K-$15K bracket, 1,595 of the 1,828 trades, ran a median -1.9% against SPY at 90 days and beat it 45% of the time. Purchases of $50K or more beat SPY 64% of the time, with a median excess of +4.6%. That group is small (55 finished windows), so treat it as a lead, not a law.
- Crowding did not help. Among the 15 names bought by the most distinct members, TSM did best (+13.4% median excess at 90 days, 6 members) and MSFT did worst (-10.1%, 11 members). The most bought names are mostly the largest US companies, and buying them after a filing was roughly a coin flip.
- Sample: 1,828 stock purchases by 38 House members across 599 tickers, disclosed between 2025-09-17 and 2026-09-17, prices through 2026-09-18. Senate filings are not in the data yet.
By size of the purchase
| Bracket filed | Purchases | 90d median vs SPY | 90d mean vs SPY | Beat SPY | Finished windows |
|---|---|---|---|---|---|
| $1K-$15K | 1,595 | -1.9% | +0.5% | 45% | 1336 |
| $15K-$50K | 174 | -3.3% | -1.8% | 45% | 139 |
| $50K-$100K | 34 | +4.9% | +8.2% | 61% | 33 |
| $100K+ | 25 | +4.1% | +7.1% | 68% | 22 |
The names bought by the most members
| Stock | Members | Purchases | 90d median | 90d median vs SPY | Beat SPY |
|---|---|---|---|---|---|
| MSFT | 11 | 29 | -4.5% | -10.1% | 13% (23) |
| AMZN | 11 | 25 | +5.8% | +2.8% | 71% (24) |
| AAPL | 10 | 33 | +4.8% | +3.8% | 62% (29) |
| NVDA | 9 | 18 | +2.0% | +1.2% | 54% (13) |
| META | 8 | 15 | -1.4% | -5.7% | 50% (12) |
| GOOGL | 7 | 19 | +4.8% | +4.9% | 80% (15) |
| AMD | 7 | 16 | +11.7% | +12.8% | 62% (13) |
| AVGO | 7 | 14 | -1.8% | -2.7% | 46% (11) |
| HD | 6 | 18 | +2.6% | -2.1% | 38% (16) |
| TSM | 6 | 15 | +21.0% | +13.4% | 93% (14) |
| UNH | 6 | 13 | -2.2% | -5.2% | 38% (13) |
| BRK.B | 6 | 10 | +0.1% | -2.4% | 17% (6) |
| JPM | 6 | 9 | -0.2% | -3.4% | 11% (9) |
| GS | 6 | 7 | +13.9% | +8.8% | 86% (7) |
| MCK | 6 | 7 | +6.7% | +5.8% | 100% (5) |
Ranked by distinct House members who filed a purchase in the window. Names with fewer than two buyers are not listed.
Method
- Universe: every stock purchase (asset type ST, side BUY) filed by a House member and disclosed in the 12 months to the latest filing in our table. Options, sales and exchanges are excluded. Trades whose filing date is before the trade date, and trades with no price bar within five days of the filing, are dropped.
- Entry: the close of the first trading day on or after the disclosure date. That is the first day a reader could have acted. Trade dates are not used: a filing can come 45 days after the trade, and nobody can buy on a date they do not know about yet.
- Exit: the close of the first trading day on or after disclosure plus 30, 60 and 90 calendar days. A window that has not finished by the last price date is not counted.
- Benchmark: SPY over the same two dates. Excess return is the stock return minus the SPY return. Hit rate is the share of purchases with a positive excess return.
- Prices: adjusted close, so splits and dividends do not create fake moves.
- Brackets: the legal disclosure ranges as filed ($1,001-$15,000 and up). Each trade is grouped by the floor of its range. No dollar figure is a fill; they are all ranges.
- The chart draws the average excess return of all purchases by trading day after the filing, day 0 to day 63 (about 90 calendar days), on every day with at least 30 purchases in the sample.
The chart as a table
| Trading day | Average vs SPY | Purchases |
|---|---|---|
| 0 | +0.00% | 1,828 |
| 1 | +0.02% | 1,828 |
| 2 | +0.12% | 1,822 |
| 3 | -0.09% | 1,822 |
| 4 | +0.09% | 1,821 |
| 5 | +0.07% | 1,817 |
| 6 | +0.05% | 1,817 |
| 7 | -0.01% | 1,787 |
| 8 | +0.02% | 1,765 |
| 9 | +0.11% | 1,762 |
| 10 | +0.24% | 1,724 |
| 11 | +0.06% | 1,724 |
| 12 | +0.14% | 1,721 |
| 13 | +0.33% | 1,721 |
| 14 | +0.30% | 1,721 |
| 15 | +0.30% | 1,721 |
| 16 | +0.32% | 1,720 |
| 17 | +0.36% | 1,720 |
| 18 | +0.30% | 1,720 |
| 19 | +0.32% | 1,720 |
| 20 | +0.19% | 1,717 |
| 21 | +0.06% | 1,715 |
| 22 | +0.05% | 1,715 |
| 23 | -0.01% | 1,707 |
| 24 | -0.20% | 1,707 |
| 25 | -0.18% | 1,706 |
| 26 | -0.22% | 1,706 |
| 27 | -0.14% | 1,706 |
| 28 | -0.27% | 1,706 |
| 29 | -0.28% | 1,700 |
| 30 | -0.41% | 1,683 |
| 31 | -0.20% | 1,679 |
| 32 | -0.09% | 1,641 |
| 33 | +0.04% | 1,641 |
| 34 | +0.22% | 1,610 |
| 35 | +0.18% | 1,610 |
| 36 | +0.39% | 1,609 |
| 37 | +0.29% | 1,609 |
| 38 | +0.31% | 1,609 |
| 39 | +0.56% | 1,603 |
| 40 | +0.36% | 1,601 |
| 41 | +0.42% | 1,601 |
| 42 | +0.31% | 1,601 |
| 43 | +0.34% | 1,601 |
| 44 | +0.39% | 1,601 |
| 45 | +0.32% | 1,601 |
| 46 | +0.16% | 1,600 |
| 47 | +0.35% | 1,592 |
| 48 | +0.35% | 1,592 |
| 49 | +0.29% | 1,590 |
| 50 | +0.48% | 1,590 |
| 51 | +0.49% | 1,587 |
| 52 | +0.39% | 1,586 |
| 53 | +0.49% | 1,573 |
| 54 | +0.75% | 1,569 |
| 55 | +1.05% | 1,534 |
| 56 | +1.14% | 1,532 |
| 57 | +0.83% | 1,532 |
| 58 | +0.70% | 1,532 |
| 59 | +0.73% | 1,530 |
| 60 | +0.72% | 1,530 |
| 61 | +0.57% | 1,530 |
| 62 | +0.63% | 1,530 |
| 63 | +0.48% | 1,529 |
Common questions
- Do stocks go up after Congress buys them?
- Not on average, once you buy at the filing date and compare with SPY. Over the last 12 months the median House purchase trailed SPY by about two points at 90 days and beat it less than half the time.
- Why measure from the disclosure date and not the trade date?
- Because the trade date is private until the filing. A member has 45 days to report under the STOCK Act, and many file later. The disclosure date is the first day anyone outside the office could copy the trade.
- Which Congress purchases did best?
- The larger ones. Purchases filed at $50,000 or more beat SPY more often and by more than the small ones, but there are few of them, so the result is a lead to watch, not a rule to trade.
- Does the study include the Senate?
- Not yet. The table holds House filings from the Clerk's public periodic transaction reports. Senate filings will be added when the fetch is reliable.
- Can I reproduce these numbers?
- Yes. The filings are public on the House Clerk site, the prices are daily adjusted closes, and the rules are listed under Method. The script and the JSON it writes are named on this page.
- Where can I see the latest Congress trades?
- The free page Congress stock trades this week lists the latest House filings and the names bought by the most members, updated after each US close.
Sources
- US House Clerk, Financial Disclosure Reports (periodic transaction reports)
- STOCK Act of 2012 (45-day reporting rule)
- Daily adjusted prices: our market data store (Polygon and Yahoo Finance)
- Script: scripts/research_congress_90d.py, output data/research/congress-buys-90-days.json
Quote any number on this page with a link to it. Informational only, not investment advice.