Does copying Congressional stock trades actually work?

An honest look at whether following politicians' disclosed trades produces excess returns — what the academic research found, why the 45-day lag matters so much, and how to test it yourself.

2026-08-05 · 14 min read · By Insider Option

The pitch writes itself: lawmakers vote on legislation that moves industries, they sit in classified briefings, and the law forces them to publish their trades. Copy the trades, capture the edge. The pitch is popular enough that two ETFs exist to sell it.

The evidence is considerably more equivocal, and the gap between the evidence and the marketing is the most interesting thing in this whole subject. This post is our attempt to lay it out fairly, including the parts that argue against the product we sell.

The study everyone cites

The foundation of the entire "Congress beats the market" genre is a set of papers by Ziobrowski and co-authors in the mid-2000s. Examining Senate financial disclosures from 1993–1998, they reported that a portfolio mimicking senators' common-stock purchases substantially outperformed the market — the widely quoted figure is roughly 12 percentage points a year of abnormal return. A companion study of the House found a smaller but still positive effect.

Those results are real, published, and peer-reviewed. Four caveats are usually dropped when they are quoted:

• The sample predates the STOCK Act by well over a decade. Before 2012, disclosure was annual, paper-based, and effectively unmonitored. The behaviour being measured happened in a world with no meaningful scrutiny.
• It measures the senators' returns, not a copier's. The study computes performance from the transaction date. A copier does not have the transaction date; they have the disclosure date, weeks later.
• Small samples, concentrated effects. A modest number of very active traders can dominate an aggregate result of this kind.
• Later work using different methods and periods found much less. Studies covering the 2000s and the post-STOCK Act era generally report Congressional portfolios performing close to the market, with results that are sensitive to benchmark choice and frequently not statistically significant. Eggers and Hainmueller, among others, found no persistent outperformance in the periods they examined.

So the honest summary of the literature is: strong evidence of abnormal returns in 1990s Senate data, weak and contested evidence since. Anyone telling you "studies prove Congress beats the market by 12% a year" is quoting a thirty-year-old sample as if it were a current fact.

The four things that eat a copier's edge

Even granting that some members possess genuine informational advantage, four mechanical frictions sit between their edge and your return.

1. Information decay over the disclosure lag

This is the big one. Suppose a senator learns in committee that a defence appropriation is going to pass and buys a contractor. Over the following weeks, that information reaches the market through the ordinary process of legislation becoming public. By the time the Periodic Transaction Report publishes, the catalyst has frequently already occurred. You are buying after the news, not before.

2. The disclosure itself moves the price

For a handful of heavily followed members, the filing is a news event. Retail flow, financial media, and copy-trading products all arrive at once. That pop is not edge you capture — it is a cost you pay, because you are buying into demand the disclosure created. And it frequently fades.

3. You cannot actually replicate the position

Disclosures give ranges, not amounts. "$500,001–$1,000,000" could be either endpoint. You do not know the member's total portfolio, so you cannot compute position weight. You do not know their time horizon, their tax situation, whether the trade was a hedge against something you cannot see, or — critically — whether it was their spouse's decision, since spousal trades file under the member's name.

You are also blind to why they will exit. A sale disclosed 45 days late means you may hold something the member abandoned six weeks ago.

4. Sector tilt masquerading as skill

Congressional portfolios have historically skewed toward large-cap technology and other high-beta names. During long tech bull runs, a portfolio that copies Congress outperforms the S&P 500 for reasons that have nothing to do with political information — it is a factor bet.

This is why benchmark choice determines the answer. Against SPY, a tech-concentrated Congressional portfolio in the 2010s looks brilliant. Against QQQ, much of the apparent alpha evaporates. Any performance claim that does not specify its benchmark, and ideally control for factor exposure, is not telling you anything.

The counterargument, taken seriously

There are reasons not to dismiss the idea entirely.

• The average is the wrong statistic. "Congress" is 535 people, most of whom hold index funds and trade rarely. Aggregating them buries any effect from the small number who trade actively and unusually. The interesting question is not whether Congress outperforms but whether specific, identifiable members do, persistently, out of sample.
• Committee-relevant trades are a sharper hypothesis. A member of the Armed Services Committee buying a defence contractor is a more specific claim than a member buying Apple. Some research has found effects concentrated in exactly this kind of subset.
• Timing around known catalysts. The most damning Congressional trading episodes — the February 2020 sales ahead of the COVID market decline being the canonical example — are timing events near non-public briefings, not general stock-picking. That is a real phenomenon even if it is not a repeatable strategy.
• Disclosure momentum is tradeable on its own terms. If a filing reliably produces a short-term move, that is a strategy — just a completely different one from "Congress knows things". It is a flow/attention trade, and it requires being early and disciplined about exiting.

How to test this yourself, properly

You do not have to take anyone's word for it, including ours. A defensible backtest of "copy Congress" needs all of the following, and most published claims fail at least one:

1. Enter on the disclosure date, not the transaction date. This single choice explains most of the difference between impressive and unimpressive results.
2. Use the filing timestamp, not the filing day. If a PTR publishes at 3pm, you cannot have the open price.
3. Include every disclosed trade, not the ones that worked. Survivorship and selection bias are rampant in this space — screenshots of winning trades are not evidence.
4. Define an exit rule in advance. "Hold forever" and "sell when they sell, 45 days late" give very different results. Pick one, state it, apply it uniformly.
5. Subtract costs. Spread, commission where applicable, and the slippage from buying a name that just got attention.
6. Benchmark against a factor-matched portfolio, not just SPY. At minimum report results against a comparable market-cap and sector mix.
7. Split your sample. Fit any rule on one period, test on another. With a few hundred active traders and thousands of trades, you will find a member with a spectacular record by chance alone. Out-of-sample testing is the only defence.

Our strategy performance pages publish the price basis for every figure, and the methodology in our methodology page, including where the approach has not worked. If a tracker will not show you its losing periods, that is informative.

Where we actually land

Our view, stated plainly:

• "Copy Congress and beat the market" is not a well-supported claim as a general strategy in the post-2012 era.
• The 45-day lag is a genuine structural barrier, not a detail. It is the single strongest argument against the whole premise.
• The data is nevertheless valuable — as a research and accountability tool. Knowing that a member of a committee regulating an industry holds stock in that industry is a fact worth having whether or not you trade it.
• If you do trade it, treat it as one input among many, size positions as if you might be wrong, and never confuse a disclosure with a thesis.

If you want the exposure without doing any of this, the NANC and KRUZ ETFs package the idea for you — with the same underlying limitations, plus a fee.

This post is for informational purposes only and is not investment advice. Past performance does not predict future results. See our disclaimer.