What Congress Sold Beat What Congress Bought. And One Member Is 38% of All the Trades.

Source: public STOCK Act disclosures, collected by our weekly job (data/politician-trades.json) from the open congress-trading-monitor dataset. 1,398 trades by 48 members, traded July 28, 2025 – October 16, 2026. "Return since trade" runs from the trade date to our latest refresh, compared with SPY over the same period. Holding periods differ from trade to trade, and none of this is risk-adjusted.
Sells beat buys
| Trades with return data | Median vs SPY since trade | Average vs SPY | Beat SPY | |
|---|---|---|---|---|
| Stocks they sold | 782 | +1.09% | +0.67% | 54.7% |
| Stocks they bought | 550 | −0.57% | +0.07% | 47.3% |
If Congress had an edge, you'd expect the stocks they bought to beat the ones they sold. In this window it was the other way round. The things they sold kept going up, which matters if you're thinking of copying their buys. We found the same thing in August from a different angle, when 355 purchases showed no edge.
One member is a third of the data
One member of Congress (listed in the source data as Rohit Khanna) accounts for 532 of 1,398 trades (38%), and 381 of 1,003 since August 1. Any "follow Congress" tracker built on this data is mostly tracking one portfolio.
Take that one account out and the picture gets worse for copying buys:
| Excluding the most active member | n | Median vs SPY | Beat SPY |
|---|---|---|---|
| Buys | 278 | −2.00% | 41.7% |
| Sells | 532 | +1.05% | 54.9% |
That one account's buys did better: a median of +0.93%, with 52.9% beating SPY. Without it, the other 47 members' purchases lagged SPY by a median of 2 points.
You find out a month later
- The median trade was disclosed 25 days after it happened.
- 189 trades (13.5%) were disclosed more than 45 days later. The source data flags these as late.
- 78% of trades fall in the smallest disclosure band, $1,001 to $15,000.
By the time a trade shows up in a feed, the move has usually happened. That's why copying it doesn't work, even if the original decision was well-informed.
What we do with this
Congressional trades show on our Smart Money page as information, not as a signal. They don't feed the score. Our own confluence test found that a Congress purchase landing on a Buy Zone score did worse, not better.
Scores in this piece come from the ZonedInOut model. How the score works →
Frequently asked
Do stocks bought by members of Congress outperform?
Which member of Congress trades the most?
How long does it take Congress to disclose stock trades?
Get the Sunday Brief
One email a week, before the open: what the board did, which calls worked, which didn't, and what changed. Free. No spam, unsubscribe anytime.