Members of Congress disclose their trades, with a lag of up to forty-five days, and a popular claim says copying them beats the market. We tested it on the full record, more than eleven thousand disclosed purchases across over one hundred and fifty members since 2020. The pooled portfolio tracks the index and ends slightly behind it, and the median disclosed buy lags. The outperformance everyone cites belongs to a few standouts, not to Congress.
Members of the United States Congress must disclose their securities trades, with a lag of up to forty-five days. A popular claim says those disclosures are a structural edge: copy the filings and beat the market. We tested it properly, not on a handful of famous names but on the full disclosed record, eleven thousand purchases across more than one hundred and fifty members since 2020. The headline that survives a small, well-chosen sample does not survive the whole one.
The easy read: Congress beats the market, so copying members is a structural edge available to anyone willing to read the filings.
Our reading: on a small sample weighted toward a few celebrated names in a technology bull run, copying looks unbeatable. On the full record of more than one hundred and fifty members, it is close to a coin flip that slightly lags the index. The famous result is a selection effect, not a property of Congress as a whole.
This note is not about any individual or any party. It is about a measurable question: how much information, on average, is left in a disclosed trade by the time the public is allowed to see it. On the full sample, the answer is close to none.
The academic starting point is Ziobrowski and co-authors (2004, 2011), who found that the common-stock investments of U.S. senators and representatives earned abnormal returns in the years before the STOCK Act. The popular version of the claim is that members trade with an informational advantage tied to their committee work and legislative knowledge. The evidence, though, is contested and has shifted. Eggers and Hainmueller (2013) re-examined the data and found members' portfolios actually lagged a simple index fund; post-disclosure-law studies (Belmont and co-authors, 2020; Chen and Sacerdote, 2026) find little broad edge and conclude that Congressional trades look much like uninformed retail trading. The most striking recent exception, Wei and Zhou (2025), is that whatever outperformance exists is concentrated almost entirely in the handful of members who reach leadership positions.
The testable part is simpler than the motive. The trades become public, with a delay, under the STOCK Act. If those disclosures still predict returns after the delay, then public information is being absorbed slowly, which is the same family of effect as the drift after an earnings report. If they do not, then by the time the filing is public the information is already in the price.
The forty-five-day lag is the crux. By the time a disclosure is filed, the trade is old news in the literal sense. For copying to work, the move the member was positioned for has to keep going after the public learns of it. That is a strong condition, and it is exactly what the test checks on the full record.
If a disclosed purchase still earns excess return after the public sees it, the market is slow to price information that is, by then, fully public. On the full sample, it is not.
We took every disclosed purchase by every member in the public filings from 2020 onward, kept only ordinary equity purchases with a clean ticker, and entered each at the adjusted close on the date the trade became public, the notification or filing date, never the trade date. That is 11,376 purchases across 155 members and 1,750 tickers. Each position is held for 252 trading days. We read the result two ways: a single copy-all portfolio that holds every disclosed purchase equally weighted, and the distribution of each trade's twelve-month return measured against the S&P 500 over the identical window, which is the cleaner test of whether a disclosed buy carries information.
Prices are adjusted daily closes from a public source, so dividends are included. Delisted tickers cannot be priced and are dropped, which leaves 10,696 of 11,376 purchases priced, about 94 percent; this introduces a survivorship effect whose direction is ambiguous, since both failed companies and acquired ones disappear. Disclosed amounts are reported only as ranges, so positions are equally weighted rather than dollar-weighted. Sales are not shorted, options are copied as the underlying, trading costs and taxes are ignored, and the twelve-month windows overlap, so the naive significance test understates uncertainty rather than overstating it. This is a faithful study of the disclosed purchase record, not an investable replication of any person's portfolio.
Held as one equally weighted book, copying every disclosed purchase produced a portfolio that shadowed the S&P 500 for the whole period and finished a little behind it, about 2.30 times capital against 2.50 for the index. The drawdown was marginally deeper and the Sharpe ratio lower. There is no separation, no persistent lead. Whatever signal individual trades carry, pooling all of them together yields the market with a small drag, not an edge.
The distribution is blunter than the equity curve. Of 10,696 priced purchases, only 41 percent beat the S&P 500 over the year after disclosure, and the median trade lagged it by about 5.5 percentage points. The average is dragged back to roughly flat only by a thin right tail of large winners. In plain terms, picking a disclosed Congressional buy at random and holding it for a year was a slightly losing bet against simply owning the index.
The averages hide enormous variation between members, and this is where the popular story comes from. Ranking the most active filers by the average market-relative return of their disclosed buys, a small group stands clearly above the rest. Donald Beyer's purchases beat the market by roughly forty-seven points per trade on average; Mark Green's beat it on about eighty-five percent of trades; Nancy Pelosi, the household name, is genuinely strong but is not the leader. Against them sits a tail just as long in the other direction, members whose disclosed buys underperformed the index by twenty points a trade or more.
The party split is similarly undramatic. Democratic disclosed buys compounded at about 13.9 percent and Republican at about 12.1 percent, both near the index, with average per-trade alpha within a point of zero on each side. There is no party that systematically beat the market on this record.
A natural question is whether the same exercise applies to the President. It does not, for a simple data reason. The STOCK Act periodic transaction reports tested here are a Congressional instrument. The President and senior executive-branch officials instead file an annual public financial disclosure, the OGE Form 278e, which reports holdings and broad income ranges once a year rather than individual transactions within forty-five days. There is no timely, trade-by-trade stream to copy.
Beyond the filing format, a sitting President's disclosed wealth is concentrated in operating businesses and private assets, for example a media company and real estate, rather than an actively traded portfolio of listed equities. That is a register of ownership, not a sequence of timed market entries, so it cannot be turned into the kind of copy-on-disclosure test this note runs on Congress. We therefore leave the executive branch out rather than force a comparison the data does not support.
On the full record, copying Congress is not a market-beating strategy. The pooled book tracks the index and finishes slightly behind it, the median disclosed buy lags the market, and fewer than half of all purchases beat it over the following year. The famous outperformance is real for a few individuals and absent for the body, which is precisely what a careful reading of the academic literature predicts. The honest verdict is the opposite of the popular one: the disclosures are a fascinating dataset and a weak edge.
Pick the right member and you beat the market. Pick them all and you are the market, minus a little.
Congressional trading performance is a contested literature, and the broad-sample evidence increasingly points to little reliable edge. The work below spans the early supportive studies, the skeptical re-examinations, the most recent findings on concentration, and the statute itself. The disclosures, rules and calculations in this note are Iron Hall Capital's own.
This document has been prepared by Iron Hall Capital for informational and educational purposes. Its content does not constitute personalised investment advice, a recommendation to buy or sell financial instruments, a public offering, or a solicitation to subscribe to any financial product. The opinions and readings reflect Iron Hall Capital's judgement at the date of publication, are based on data considered reliable but not independently audited, and may be revised without notice.
The results shown are from historical simulations on past data. Backtested performance is hypothetical, is computed with the benefit of hindsight, does not reflect trading costs, financing, taxes, slippage or the market impact of real execution, and is not a reliable indicator of future results. Where a data series was not available, an equivalent real series has been substituted and labelled as such in the text. Where a method ignores costs or makes a simplifying assumption, this is stated. Markets can move sharply and without warning.
The author and Iron Hall Capital may hold, have held, or come to hold positions in the instruments referenced. Any reproduction, in whole or in part, requires written authorisation.
Iron Hall Capital · A private investment office · June 2026