Educational content only, not financial advice. Read the full disclaimer.
By Options AI Screener · Last updated 2026-10-12 · 8 min read
Congressional stock disclosures are the most widely shared and most widely misread dataset in retail investing. The filings are real, public and worth reading. What they are not is a feed of trades you can follow, and the gap between those two descriptions is where the money goes.
The Stop Trading on Congressional Knowledge Act of 2012 — the STOCK Act — requires members of Congress, their spouses and dependent children to disclose securities transactions in a periodic transaction report, or PTR.
Three features of that requirement shape everything you can do with the data.
The deadline is long. A filing is due within 30 days of the member becoming aware of a transaction, and in no case later than 45 days after it happened. The two clocks mean the practical reporting window is six weeks, and filings cluster near the end of it.
Amounts are brackets, not numbers. A transaction is reported as a range. The lowest is $1,001 to $15,000; the brackets widen as they rise, and the top ones span millions. Nothing in the form tells you where inside the bracket the trade sat.
The filer is not always the decision-maker. A spouse's trade is reportable. So is activity in a managed account where the member gave no instruction. The form does not separate a conviction purchase from an index rebalance inside a portfolio someone else runs.
Everything else about this data is a detail. The lag is structural.
Consider the sequence. A member buys on 3 March. The filing appears on 14 April. You read it on 15 April. Six weeks of earnings reports, guidance changes and macro news have happened in between, and the stock is wherever that news left it.
| Trade date | 3 March |
| Filing visible | 14 April |
| Elapsed | 42 days |
| Price you pay | whatever 42 days of news produced |
This is not a small friction to be managed with a limit order. On a volatile mid-cap, six weeks routinely covers a 20% range in either direction. You are not copying a trade; you are opening a new one, at a new price, on information the market has already absorbed.
The practical consequence: the disclosure cannot be the reason you enter. If a name only looks interesting because a representative bought it in early March, there is no version of that trade where you get what they got.
The brackets are easy to underrate because they look like numbers.
A filing in the $1,001 to $15,000 bracket could be a rounding error in a multi-million-dollar portfolio, or a meaningful commitment from someone with modest assets. The form gives you no way to tell, because it discloses the transaction and not the portfolio.
That matters most for the question people actually want answered, which is how convinced were they. Conviction is a position size relative to net worth, and neither of those two quantities appears in the filing. What you have is a bracket, and a bracket supports exactly one kind of comparison: this filing against that filing.
None of the above makes the dataset useless. It narrows what the dataset is good for, and three things survive.
Repetition by one member. A single purchase is noise. The same member buying the same name across three separate filings over several months is a different observation, because it is harder to explain as portfolio mechanics and it is visible without knowing the amounts.
Clustering across members. Several unconnected filers buying the same name inside a short window is the one pattern here that is genuinely hard to produce by accident. It is still not a forecast, but it is a reason to look.
Unusual size against that filer's own history. Someone whose filings sit consistently in the lowest bracket appearing two brackets up is a change in behaviour you can see despite the ranges, precisely because you are comparing like with like.
What all three have in common is that they survive the lag. They are reasons a name is worth examining on its own merits weeks later, rather than claims about a price.
The intuition that a member of a committee overseeing an industry may know more about that industry is reasonable, and it is also where this analysis most often stops being evidence.
A representative on a defence appropriations committee buying a defence contractor is a story. It is not a disclosure of non-public information, and the filing contains nothing that distinguishes it from the same person buying a sector they simply understand. Committee membership is worth recording as context. Treating it as a causal explanation is reading a motive into a form that reports none.
The honest comparison is with corporate insider filings. Officers and directors file Form 4 within two business days and report the exact number of shares at the exact price. On both dimensions that matter here — speed and precision — it is a strictly better dataset, and it is worth reading first. What Form 4 filings do and do not tell you covers the ways that data misleads in turn, which are different ones.
Used together, the sensible order is: insider filings for the timely, precise signal; congressional filings for breadth, and for the clustering pattern that only shows up across many filers.
The Congress Trades tracker shows filings by member and by ticker with the disclosure date beside the transaction date, which is the one comparison that keeps the lag visible rather than buried. The Insider Buying tracker is the faster-reporting half of the same question.
Neither is advice about a specific position. The discipline this data rewards is reading it as a record of what was disclosed, and then deciding whether the name is worth owning on grounds that have nothing to do with who filed it.
The STOCK Act requires a periodic transaction report within 30 days of becoming aware of a transaction, and no later than 45 days after it occurred. In practice most filings land near the end of that window, and late filings carry a standard fee small enough to work as a cost of doing business rather than a deterrent.
No. Amounts are reported in brackets, starting at 1,001 to 15,000 dollars and widening from there. A filing showing the 1,000,001 to 5,000,000 bracket covers a five-fold range, so you cannot compute a position size or a portfolio weight from it. You can only compare brackets with other brackets.
Not as a direct copy. The disclosure lag alone means you are entering at a different price, often weeks of news later, and the ranges prevent you from matching size. The data works better as one input into why a name is worth examining than as a trade signal in itself.
Speed and precision. Corporate insiders file Form 4 with the SEC within two business days and state the exact share count and price. Congressional reports allow up to 45 days and report a dollar bracket. The insider filing is a far better dataset on both counts, which the companion guide to insider buying covers.