One Trade Scored 0.848 Out of 1.0
A single congressional stock trade, buried in a periodic transaction report and filed weeks after it happened, scored 0.848 on CapitolExposed's five-component conflict model. That number represents something unusual in the data: a transaction that simultaneously triggered committee-jurisdiction overlap at 100%, trade-before-vote timing proximity at 56%, and volume anomaly flags. The congressional average across all 15,948 trades in the database is 0.058. This trade scored more than 14 times that baseline.
The trade did not receive a press conference. No committee held a hearing about it. The member who made it disclosed it on the timeline required by the STOCK Act, and the filing sits in a public database alongside thousands of others. What makes it different is not any single characteristic. What makes it different is that every measurable dimension of potential conflict fired at once.
CapitolExposed built its conflict-scoring model to identify exactly these transactions. The system assigns a weighted score across five components: committee overlap (30% of the total weight), timing proximity to legislative activity (25%), lobbying connections between the member and the traded company (20%), campaign donation patterns from the company's industry (15%), and trade size relative to the member's typical activity (10%). Each component generates a subscore between 0 and 1, and the weighted combination produces the final conflict score.
Only 20 Trades Cross the Critical Threshold
Of the 15,948 trades analyzed, 20 scored at or above 0.7, the threshold CapitolExposed classifies as CRITICAL. Another 96 scored between 0.5 and 0.7, classified as HIGH. Combined, these 116 trades represent 0.7% of all congressional stock transactions, but they concentrate the most troubling patterns in the dataset.
The 20 critical trades share a profile. Every one involves a stock in a sector directly regulated by the trading member's committee. Every one occurred within a narrow window around significant legislative action. And every one involved a transaction size that deviated meaningfully from the member's typical trading pattern.
The lowest of the 20 critical trades scored 0.702. The highest, 0.848. The gap between them is smaller than the gap between any of them and the next tier down. These are not borderline cases elevated by a single outlier metric. They are transactions where multiple independent indicators converge.
How the Five Components Work
The committee-overlap component checks whether the traded company operates in a sector under the jurisdiction of any committee the member sits on. A member of the House Financial Services Committee trading shares of JPMorgan Chase would score high on this axis. A member of the Agriculture Committee trading the same stock would not. The weight is 30% because committee jurisdiction creates the most direct channel through which a member's legislative actions can affect a company's stock price.
Timing proximity, weighted at 25%, measures the distance in days between a trade and a committee vote, floor vote, or markup involving legislation relevant to the traded company's sector. CapitolExposed identified 1,159 trades made within a narrow window of a related committee vote. Of those, 61 involved stocks in the member's own committee jurisdiction, triggering the trade_before_vote flag simultaneously with the committee_trade flag.
Lobbying connections, at 20%, cross-reference the traded company against lobbying disclosure filings. If the company or its parent entity filed lobbying reports targeting the member's committee during the same session, the subscore increases. This does not mean the member was personally lobbied. It means the company had active lobbying operations directed at the committee where the member serves.
Donation patterns, at 15%, check whether the member received campaign contributions from the traded company's employees, its political action committee, or industry-aligned donors during the same election cycle. The presence of donations does not prove a quid pro quo. It establishes a financial relationship between the member and the company beyond the stock trade itself.
Trade size, the final 10%, compares the transaction amount against the member's own baseline. A member who typically trades in the $1,001 to $15,000 range and suddenly executes a $250,001 to $500,000 transaction scores high on this component. The logic is straightforward: outsized bets on specific companies warrant more scrutiny than routine portfolio management.
The Average Score Is 0.058
Context matters. The average conflict score across all 15,948 trades is 0.058. The median is even lower. Most congressional stock trades are small, routine transactions in broad-market ETFs or blue-chip stocks with no meaningful connection to the member's committee work. The system is not designed to flag every trade. Roughly 90% of all trades receive no flags at all.
But among the subset of trades on committee-jurisdiction stocks, the flag rate rises sharply. When a member trades a company that falls under their committee's regulatory authority, the probability of triggering at least one additional indicator increases by a factor of four. The committee-overlap flag, on its own, correlates with higher scores on timing, lobbying, and donation components. This is not because the scoring model double-counts. The components are independently sourced from different datasets: committee assignments from Congress.gov, trade filings from the House and Senate disclosure offices, lobbying data from LD-2 filings, and donation records from FEC reports.
The correlation exists because members who trade committee-jurisdiction stocks tend to trade them at sensitive times, in sectors where they receive donations, from companies that lobby their committees. Whether this reflects informed trading, structural coincidence, or some combination remains an open question that the data alone cannot answer.
449 Volume Anomalies
Beyond the conflict score, CapitolExposed flags trades that deviate from a member's typical trading pattern. Of the 15,948 trades analyzed, 449 triggered the volume_anomaly flag. These are transactions where the dollar amount, the frequency, or both exceeded the member's trailing 12-month baseline by two or more standard deviations.
Volume anomalies do not, on their own, indicate misconduct. A member liquidating a position after a divorce, receiving a stock grant, or rebalancing a portfolio could trigger the flag. But among the 20 critical-conflict trades, every single one also carried a volume anomaly. The combination of high conflict scores and unusual trading volume is the rarest pattern in the dataset: 20 trades out of 15,948, or 0.13%.
1,159 Trades Near Committee Votes
The trade_before_vote flag fires when a transaction occurs within a defined window of a committee vote on legislation relevant to the traded company's sector. CapitolExposed identified 1,159 trades that met this criterion across all members and all sessions in the database.
That number requires context. Members of Congress vote on legislation frequently. Committee markups happen on set schedules. For active traders, some overlap between trade dates and vote dates is statistically inevitable. CapitolExposed's timing model accounts for this by weighting the subscore based on the proximity in days: a trade one day before a vote scores higher than a trade 14 days before. The model also considers whether the legislation in question was publicly known at the time of the trade. Bills introduced and scheduled for markup weeks in advance create different informational conditions than emergency measures or classified briefings.
Among the 1,159 trades flagged for timing, the average proximity was 4.2 days. The tightest cluster involved trades executed the same day as, or the day before, a committee vote on sector-relevant legislation. These tight-window trades are overrepresented among the 20 critical-conflict transactions.
61 Committee-Jurisdiction Trades
Only 61 trades in the database triggered the committee_trade flag, meaning the member sits on a committee with direct regulatory authority over the company they traded. The small number reflects two realities: most members diversify their trades across sectors, and the committee_trade flag uses a strict definition of jurisdiction. A member of the Armed Services Committee trading Lockheed Martin triggers the flag. The same member trading Apple does not, even though defense spending affects the broader tech supply chain.
The 61 committee trades carry an average conflict score of 0.39, nearly seven times the overall average. Among those 61, 14 also carry the trade_before_vote flag, meaning the member traded a committee-jurisdiction stock within days of a committee vote on related legislation. All 14 of those trades score in the HIGH or CRITICAL range.
What the STOCK Act Does and Does Not Do
The Stop Trading on Congressional Knowledge Act of 2012 requires members to disclose stock trades within 45 days. It prohibits the use of material, nonpublic information gained through official duties for private financial gain. It does not prohibit members from trading stocks in sectors their committees regulate. It does not prohibit members from trading stocks in companies that lobby their committees. It does not prohibit members from trading stocks in companies whose employees donate to their campaigns.
The proposed H.R. 7008 would add financial penalties of $2,000 per late disclosure. That is a fine smaller than the minimum reporting threshold for most trades. For the 20 critical-conflict trades identified by CapitolExposed, the maximum reported values range from $50,001 to over $1,000,000 per transaction. A $2,000 fine on a million-dollar trade in a committee-jurisdiction stock, executed days before a committee vote, represents a cost of doing business.
Methodology and Limitations
CapitolExposed's conflict-scoring model is a quantitative screening tool, not a legal determination. A high score does not mean a member broke the law. A low score does not mean a trade was proper. The model identifies statistical patterns across publicly available datasets. It cannot access private communications, brokerage records, or the subjective intent of any member.
The 20 critical-conflict trades represent the statistical tail of a distribution. They are the transactions where every available public indicator of potential conflict aligns. Whether that alignment reflects informed trading, structural incentives, or coincidence is a question for investigators with subpoena power, not a database query.
Public disclosure records form the basis of this analysis. All trade data comes from Periodic Transaction Reports filed with the Senate Office of Public Records and the House Office of the Clerk. Committee assignments come from Congress.gov. Lobbying data comes from LD-2 filings with the Senate Office of Public Records. Donation data comes from FEC filings. The scoring model weights and thresholds are published in CapitolExposed's methodology documentation and are applied uniformly to all members regardless of party, chamber, or seniority.