The question has a popular answer and an academic answer. They are not the same.

"Do members of Congress beat the market" sounds like a clean empirical question. Collect the trades, compute the returns, compare to a benchmark, publish the answer. The problem is that every step of that sentence is contested.

Collect the trades: congressional disclosures report amounts in brackets, not exact values. A filing showing a trade between $100,001 and $250,000 could be $101,000 or $249,000. Analysts typically use the midpoint, which means every computed return is an estimate built on a 2.5x range assumption. Compute the returns: disclosures give transaction dates, not execution prices. The closing price on the disclosed date is a proxy, not ground truth. Compare to a benchmark: the S&P 500 is the popular choice, but a factor-matched benchmark that controls for size, book-to-market, and momentum gives very different results, which is part of why the academic literature has reached different conclusions from the same raw data.

Then there is the question of whose returns we are even measuring. Nancy Pelosi is the reporter. Paul Pelosi is the trader. Ritu Khanna's family trusts run the trading for the Khanna household, with Ro Khanna publicly stating he does not direct the trades. Linda Mays McCaul's accounts are managed by a third-party firm. Treating every disclosure as "Congressman X's return" erases the structural reality that the active traders are often spouses, trusts, and professional managers who are simply reportable because the member is in Congress.

The honest answer to "do members of Congress beat the market" depends on three narrower questions: which members, over what period, using what benchmark. Different answers to those three questions are what produces the gap between the popular narrative and the academic one.

The peer-reviewed literature is thin after 2019. Most of the public analysis since then has come from industry reports, most notably the annual congressional trading reviews published by Unusual Whales and by Quiver Quantitative, along with periodic investigations in the New York Times, the Washington Post, Business Insider, and Bloomberg.

These industry reports have produced consistent top-performer rankings. Nancy Pelosi and Paul Pelosi top the Democratic side in most years. Josh Gottheimer, Ro Khanna (through his family's trusts), and a rotating set of others appear in the upper quartile. On the Republican side, Mark Green, Kevin Hern, Tommy Tuberville, and Markwayne Mullin have each been flagged for high transaction activity and strong reported returns in individual years. The New York Times investigation published in September 2022 identified 97 current members or their immediate family who had traded stocks in industries their committees regulated during 2019 to 2021, which established a baseline for the kind of overlap that characterizes active congressional trading.

Two caveats are essential before accepting any specific annual return figure. First, every industry return calculation uses disclosure midpoints and approximated execution prices, so each "annual return" is an estimate with wide uncertainty bounds. Second, the popular framing that compares a top member's estimated return to the S&P 500 or to a hedge fund benchmark is selecting on the extreme end of a distribution: in any given year, with 535 members producing thousands of trades, statistical outliers will exist. The more meaningful question is whether the aggregate, or the median active trader, beats the market, which is exactly what Eggers and Hainmueller examined and found the evidence for to be weaker than headlines suggest.

The cleanest real-world test of the "Congress beats the market" thesis is not a research paper. It is a pair of exchange-traded funds.

Five structural reasons explain why the ETFs have failed to replicate the headline figures, and each of them applies to any attempt to systematically follow congressional trading at scale. First, diversification: the ETFs hold trades from roughly 150 members on each side. Top performers get washed out by average performers. Second, the 45-day disclosure lag: by the time a trade is disclosed, as much as a month and a half has passed. If the edge depends on timing relative to committee activity, the lag destroys most of it. Third, fee drag: 0.75 percent annually compounds against the strategy. Fourth, position sizing: if Paul Pelosi puts 20 percent of his portfolio into Alphabet call options, the ETF cannot match that concentration. Fifth, and most important, selection: the individual-return headlines describe the distribution's tail, not its center. Our analysis of NANC and the economics of copy-trading ETFs walks through these tradeoffs in full.

The ETF evidence does not say congressional trading has no edge. It says that the edge, to the extent it exists, lives in specific members, specific sectors, and specific trades, not in a broad pool that can be captured passively.

First, "does Congress beat the market" is the wrong question. The median member of Congress does not actively trade stocks. They hold index funds, target-date retirement funds, and mutual funds that track the broad market by definition. Asking whether the median member beats the market is roughly equivalent to asking whether the market beats itself.

Second, for the active-trader subset, the aggregate evidence is mixed. Ziobrowski 2004 and 2011 found meaningful outperformance in pre-STOCK Act data. Eggers and Hainmueller 2013 found mediocre to poor performance in a later sample using more conservative methods. Neither paper's verdict is definitively correct. The fairest reading is that the effect exists but is smaller and less consistent than the popular narrative claims, and that it depends heavily on which members, which years, and which benchmark.

Third, the narrower claim that is best supported by the evidence is the Karadaş result: members tend to do better on trades that align with their committee assignments than on trades outside that alignment. This is the claim that makes sector-by-sector tracking worthwhile. It is also the claim that the Pelosi household's Alphabet and Amazon options during active antitrust enforcement, defense contractor trades by members of the Armed Services Committees, and pharmaceutical trades by members of Health committees are all specific instances of. The information advantage does not show up uniformly across every trade every member makes. It shows up where the member sits close enough to the relevant decisions to know something the public does not yet know.

Kapitol.ai's editorial premise is built on exactly that observation. Most congressional trades are noise. A small minority are signal. The difference is not member-level performance over some arbitrary period. It is committee alignment, legislative timing, and position specificity on a trade-by-trade basis.