How long until a House member's stock trade becomes public

Key findings

• 22 daysthe median time from a stock trade to the public record, across 1,819 filings, 2021-2026. Each filing counts once, at its own median gap.
• 12 to 31.5 daysthe middle half of filings. One filing in four took 31.5 days or more to become public.
• Filings, not tradesone filing can carry hundreds of trades from a single broker notice, so every figure gives each filing equal weight.
• 5 → 7 daysthe notification delay in filings with a single stock trade, 2021 to 2026, against 6 → 13 across all filings. The rise is much smaller where batching cannot apply.

This study covers every Periodic Transaction Report in the House Clerk's public disclosure index filed from 2021 to 9 September 2026: 3,114 filings, of which 1,819 contain stock trades that could be read and dated. In the typical one of those filings, a stock trade reached the public record 22 days after it was made.

Each filing counts once. For every filing, the gap is the median across the stock trades it reports, and the figures in this study are medians of those per-filing values, so a report listing hundreds of trades weighs the same as a report listing one. Gaps are in calendar days. A report covers trades by the member, a spouse or a dependent child, and all of them are counted. The public record means the filing date in the Clerk's index.

The wait has two stages, each with a median of about a week: from the trade to the member being notified of it, and from that notification to the report being filed. The two stage medians describe separate distributions, so their sum need not equal the median of the whole wait. The 22 days is measured directly; adding the stages would understate it by about a third.

Why the lag matters

A trade becomes visible to the public only once its report is filed. This study takes the filing date in the House Clerk's index as that moment. It does not measure when a report is posted online or when another service picks it up, both of which come after it. On that basis, the middle half of filings became public between 12 and 31.5 days after their trades, and one filing in four took 31.5 days or more.

For traders who read these disclosures, that age is part of the data. By the time a trade is public, the stock has had weeks to move, and the report records a transaction, not a position: it does not say whether the holding was kept. Each report lists the transaction date alongside the filing date. For anyone testing a strategy on disclosures, the date that matters is when the information became available, not when the trade was made.

This study measures disclosure timing only. It does not test what prices did afterwards or whether following disclosed trades would have paid.

For anyone reporting on or studying these filings, two findings change how the usual numbers read. The two stages each have a median of about a week, but they do not add up to the total. And the rise in notification delay since 2021 is much smaller among filings with a single stock trade, which is consistent with larger filings driving much of it.

What the rule requires

Under the STOCK Act, a House member must report a purchase, sale or exchange of stocks, bonds, commodities futures or other securities worth more than $1,000, including trades by a spouse or dependent child. The report is due no later than 30 days after the member is notified of the trade, and in no case later than 45 days after the trade itself, whichever comes first.

Three things follow from that wording, and all three matter for reading the numbers below. There are two limits: one counted from notification, and an outer one counted from the trade. The 45-day figure is that outer limit on the whole process, not the reporting window. And the notification date is entered by the filer — it is what a member reports being told, not an independently verified fact.

The two stages of the clock

Each figure counts filings, not trades, for the reason set out in the next section.

In 78 of the 1,819 filings, about 4%, at least one trade was reported as notified more than 45 days after it took place.¹

Why this counts filings, not trades

A Periodic Transaction Report is one document that can carry one trade or several hundred. Each filing gets equal weight so that reports containing hundreds of trades do not dominate the result; a trade-weighted analysis answers a different question. The trades inside a single filing are also not independent observations — they typically arrive on one broker notification. Among the 690 filings that report five or more trades, the median filing carries one distinct notification date for every eight trades it contains.

The effect is not subtle. Measured across trades rather than filings, 7.9% of the 21,496 stock rows were notified more than 45 days after the trade. But those 1,703 rows are concentrated: a single filing accounts for 30% of them, three filings account for 51%, and ten filings account for 84%. Remove the ten largest filings and the same trade-level figure falls to 4.7%.

So a trade-level headline about this data is, arithmetically, a claim about roughly ten documents. The filing-level figure of about 4% is the one that describes the population, and it is stable across every year in the census.

Has notification delay changed?

The per-filing median notification delay, from trade to notification, rises across the census: from 6 days in 2021 to 13 days in 2026. Restricting every year to filings dated 1 January to 31 August, so that the partial 2026 year is compared like for like, leaves the rise intact: 6, 5, 7, 8, 11.5 and 13 days.

Much of that rise tracks filing size. The two years with the longest median delay are also the two with the largest filings: on the same January-to-August basis the median filing carried 3 trades in 2021 and 2 in 2024, but 5 in 2025 and 4 in 2026. Larger filings carry longer notification gaps, because one broker letter covering a month of activity reports its oldest trade with a long gap. Among filings with a single stock trade, where batching cannot apply, the median rises much less: from 5 days in 2021 to 7 days in 2026, against 6 to 13 across all filings.

That pattern is consistent with batching — more trades grouped into each filing — accounting for much of the increase. It does not rule out some slowing of individual trades, because the single-trade series rose too. We are publishing the check alongside the trend.

How this was measured

Every figure above comes from one run over a frozen list of filings: the House Clerk's index as it stood on 11 September 2026, pinned so that the same run always returns the same numbers. Transaction and notification dates come from each report; the filing date comes from the index.

Of the 3,114 filings in the frame, all 3,114 were retrieved. 314 carry no text layer — they are scans — and are therefore absent from every figure. 1,833 yielded at least one parseable stock row, for 21,598 rows in total.

Dates are cleaned by one rule applied before any figure is computed: a row is kept only if 3 July 2012 ≤ transaction date ≤ notification date ≤ filing date. The floor is the STOCK Act's own commencement. A row that fails any clause is removed from every figure and no date is ever corrected, because a plausible repair is a fabricated observation — most of the impossible rows admit no unambiguous single-character fix, and the majority of those that do would have manufactured a longer gap than the filing itself reported. That rule removes 102 rows: 30 notified after filing, 7 transacted after filing, 65 notified before the transaction, and none before the STOCK Act. 21,496 rows in 1,819 filings remain — the rule empties 14 filings entirely — and those are the figures above.

What this does not show

This is House-only and stock-only. Senate electronic filings report a single date rather than separate notification and transaction dates, so including them would contribute a pile of zero-day gaps and halve the median; they are excluded entirely. The results describe the 1,819 readable filings with stock trades; scanned filings and other asset classes are outside the analysis, and amended reports are counted as the index lists them. The notification date is filer-entered, so every figure here describes what members reported being told, not an independently established fact. The study classifies no filing and no member as timely or untimely, and deliberately reports no such measure: the rule for deadlines falling at a weekend changed inside the 2021-2026 window, so any such classification would depend on which year's guidance applied. 2026 is a partial year, ending with filings dated 9 September 2026. And figures describe filings already made — they say nothing about any filing not yet submitted.

Questions and answers

QuoTrend is an information tool, not investment advice. Figures describe historical data only.