Do stock trades by members of Congress beat the market?

By Daniel, The Philosopher Investor · updated October 2, 2026

What does the research actually say?

Studies going back two decades have looked at this and landed in different places. Early work on Senate filings found a meaningful edge. Later studies covering the House found smaller effects, and some found none at all once the results were adjusted for the size and style of the stocks held. Sample periods differ, methods differ, and the answers differ with them.

Where the findings agree is on the narrow cases. Trades by members who sit on committees overseeing an industry have shown more predictive power than the average trade. So has buying concentrated around legislation or federal contracts. The broad claim that Congress as a group beats the market does not hold up as cleanly.

Why does the disclosure lag change the answer?

Members have up to 45 days to report a trade. By the time a purchase appears on a public site, the stock has had weeks to move, and any catalyst the member saw coming may already sit in the price.

That gap is the difference between a member's return and a follower's return. Even where the filings show an edge, tests that buy on the disclosure date rather than the transaction date find most of it gone. The data is a record of what happened, never a live feed.

Which trades carry the most information?

Purchases carry more than sales. Members sell for taxes, for a house, or because an adviser rebalanced. They buy when they expect the stock to rise, which is a much narrower set of reasons.

Clusters carry more than single trades. Several members buying the same sector within a few weeks says something about where a well-connected group is putting money. One member buying one bank says almost nothing at all. A cluster also survives the lag better, because a shift in positioning stays interesting for months.

What do the trackers get wrong?

Amounts are ranges, never figures, so a tracker showing a dollar total has estimated it from bracket floors or midpoints. Two sites can show very different numbers for the same filing and both be defensible.

Coverage is uneven. The Senate system is harder to read automatically than the House, some members still file on paper that no scraper reads, and managed-account trades sit in the data next to personal decisions. Any performance table built on that record inherits those gaps.

How should you use the data instead?

Treat it as a theme generator. When filings cluster around defense, utilities or one policy area, that is a prompt to go and understand why, rather than a list to copy line by line. The filings are a free public record of where a connected group put money, and that is worth reading even when it arrives late.

Then check the idea on its own terms. The chart, the business, and the catalyst still have to make sense. A filing is a starting point that costs nothing to read, and a poor reason on its own to own a stock.

Common questions