The idea is simple. Politicians disclose their stock trades. You can see what they bought. You buy the same thing. Profit.
In practice, it’s more nuanced than that. But it’s not complicated once you understand the mechanics.
Step 1: Understand the data source
Congressional stock trades are disclosed through two systems. House members file with the House Clerk. Senators file with the Senate Office of Public Records. Both are required by the STOCK Act.
The filings include the ticker (or asset description), the transaction type (buy or sell), the date of the trade, and an approximate dollar range. They don’t give you exact share counts. A filing might say “Purchase of NVDA, $250,001 – $500,000” rather than “Bought 1,200 shares at $412.”
That range-based reporting means you can’t perfectly replicate the exact position size. But you can replicate the direction and approximate weight.
Step 2: Account for the delay
This is where most people go wrong. They see a headline about a politician buying a stock and think they can still get in at the same price. They can’t.
The disclosure window is up to 45 days. Some politicians file within a week. Others take the full 45 days. In that gap, the stock can move significantly.
A concrete example: Nancy Pelosi disclosed AVGO (Broadcom) purchases on July 9, 2025, worth over $3M each. She actually traded on June 20–24. That’s a ~2.5 week delay. AVGO moved meaningfully in that window.
The strategy isn’t “buy as fast as possible after disclosure.” It’s “evaluate whether the position still makes sense at the current price.” Sometimes it does. Sometimes the stock has already moved 25% and the opportunity has passed.
Step 3: Pick your approach
There are roughly three ways people use congressional trading data.
Follow a single politician. Pick someone with strong historical returns. The top 15 ranking shows the best performers. The current leaders by 1-year CAGR:
- Ashley Moody: +76.51% (semiconductor-heavy portfolio)
- John Fetterman: +73.07% (single position in GOOG)
- David J. Taylor: +38.45% (diversified, 78 trades)
- Nancy Pelosi: +23.02% (tech-heavy, 117 trades)
This is the simplest approach but the most concentrated. Your returns are tied entirely to one person’s judgment.
Follow the consensus. Look for stocks that multiple politicians are buying at the same time. When three or four members of Congress from different committees and parties all buy the same ticker, that’s a stronger signal than one person acting alone.
Looking at our database: NVDA appears in portfolios of Pelosi (19.2%), Cleo Fields (36.0%), Ashley Moody (13.9%), and Terri Sewell (53.5%). When multiple politicians are independently buying NVIDIA, that’s the consensus signal.
Use it as a screener. Don’t copy the trade directly. Instead, use congressional disclosures as an alert to do your own research. If a senator on the Energy Committee is buying an oil company, maybe that’s worth looking into further.
What the backtests show
We’ve run simulations across our full dataset of 46 politicians and 3,619+ trades. Some patterns hold up.
Following top performers individually produces the highest potential returns but also the highest variance. The spread between the best performer (Ashley Moody at +76.51%) and the worst (Shri Thanedar at -67.69%) is 144 percentage points. That’s enormous.
The key across all approaches: accounting for the disclosure delay is non-negotiable. Backtests that ignore the delay overstate returns dramatically. Our simulations always use the disclosure date as the buy date.
What about 13F filings?
The same logic applies to hedge fund portfolios, with a slightly different filing system. Institutional investors file quarterly 13F reports with the SEC, also with up to a 45-day delay. The strategies for following Buffett or Ackman are similar in principle, though the frequency and concentration differ.
For the complete overview of both systems, the congressional trading guide covers everything in one place. And for a comparison of the tools available to track all of this, check out the best tracker apps roundup.