Congressional trading disclosures

The STOCK Act requires members of the US House and Senate to disclose transactions in stocks and similar assets, for themselves and for their spouses and dependent children. Filings are due within 45 days of the transaction. Every row on this page comes from those disclosures.

The Act was passed in 2012 and applies to transactions above $1,000. Late filings carry a fixed fee that is small relative to the transactions being reported, which is one reason filings arrive close to the deadline.

Yes. Holding and trading individual shares is permitted, and the obligation the STOCK Act creates is to report a transaction rather than to seek permission for it. Proposals to restrict or ban the practice are raised regularly and none has been enacted.

Most of what appears here is routine: index funds, retirement accounts, and holdings managed on the member's behalf by an adviser under an arrangement the member does not direct.

Members disclose a range rather than a number. The lower brackets are narrow and the upper ones very wide, so the precision of a total built from this data falls sharply as the transactions get larger.

A transaction reported as $1,001 to $15,000 could be at either end of that span, and one reported in the top band has no upper bound at all. Any account of what a member made or lost that states a precise figure has supplied a number the filings do not contain.

A transaction can be up to 45 days old before disclosure is due, and filings often arrive close to the deadline. The price has usually moved in the interval, so a disclosure describes a decision taken weeks earlier under conditions that no longer hold.

The transaction date is shown alongside the disclosure date for that reason, and the gap between them is part of reading the row. A filing that arrives on day 44 for a trade that has already worked is a different object from one filed within a week.

A disclosure records that a transaction occurred. It does not record who decided on it. Many members hold assets in adviser-managed accounts or blind trusts, and those transactions are reported in the member's name whether or not the member knew of them in advance.

A well-timed trade is not evidence in itself. Across hundreds of members and thousands of transactions a year, some will fall shortly before relevant news through chance alone, and the ones that do are the ones that get written about.

Establishing that a particular trade was informed requires evidence these filings do not contain: what the member knew, when, and whether it was material and non-public. That is a question for an investigation rather than for a table.

Questions about exposure and pattern, which the filings answer directly. Which sectors legislators hold. Whether members of a committee hold shares in the industries it oversees. How concentrated a member's holdings are, and how that changes over a term.

Aggregate behaviour is also visible: how many members trade individual shares at all, how activity changes around major legislation, and which sectors see the most transactions in a period.

Summing brackets into a portfolio value, which produces a figure with an error range wider than most of the underlying numbers.

Reading a disclosure date as a transaction date, when they can be six weeks apart.

Attributing a trade to a member who may not have directed it.

Treating a single well-timed transaction as evidence, when the data set is large enough to produce many by chance.