Why congressional trades appear 45 days late
Every trade on Hillscore — and on every other tracker — appears weeks after it happened. That is not a technology delay, a scraping limitation, or a choice any site makes. It is the law working exactly as written.
The rule
The STOCK Act requires a covered transaction to be disclosed within 30 days of the member becoming aware of it, and in no case later than 45 days after the transaction itself.
The two-part wording matters. The clock starts on awareness, not execution, which accommodates trades a member did not personally direct — a managed account, or a spouse's transaction filed under the member's name. In practice the 45-day outer limit is the one that binds.
How filings actually arrive
Every trade here carries both dates, so the lag is measurable rather than assumed. Two patterns show up consistently.
Filings arrive in batches, not per trade. A single Periodic Transaction Report often covers a month or more of activity, which is why several trades frequently share one filing date on a member's page while their transaction dates spread across weeks. A member does not file when they trade; they file when the deadline approaches.
And they cluster toward the end of the window. The statutory maximum functions less as a limit than as a scheduling target.
Filings are typed by hand
Periodic Transaction Reports are completed manually by congressional staff, and it shows. Impossible values reach the published filings: mistyped years, transaction dates falling after the report that discloses them, tickers that do not exist.
Those errors originate upstream, not in any tracker. We reject the impossible ones rather than republishing them — see data-quality exclusions — but the underlying point stands. This is hand-keyed administrative data and should be read as such.
Does the lag actually matter?
If you are asking whether you can act on a disclosure the day it appears: it is stale by definition. The more interesting question is whether that staleness costs anything measurable, and the data gives a clear answer.
Returns in this dataset build gradually rather than spiking on the trade date. Across 4,791 scored buys, the average disclosed purchase is up 0.29% after a week, 1.60% after a month and 5.15% after three months. The move spreads across months rather than concentrating in the days a follower would miss.
Read alone, that makes the lag look survivable. Read against the sector, it stops mattering for a different reason. Measured net of each stock's own benchmark ETF over the identical window, those same buys come to −0.16%, −0.58% and −0.32%. Very little edge is being lost to the delay, because there is very little edge to lose. The full comparison.
That is the honest answer to "am I too late?" — you are, but it matters less than you would expect, and for a reason nobody selling a copy-trading product wants to lead with.
Why the delay exists at all
The window balances two things.
Real-time disclosure would itself become market-moving. A well-known member's purchase, published as it executed, would be front-run within minutes — a different problem rather than a solution, and one that would arguably hand the member's followers an edge the member never had.
Against that, 45 days is long enough that disclosure functions as a historical record rather than as accountability in any live sense. By the time a questionable trade is public, the vote it might have related to has usually happened.
Whether that balance is right is a live legislative question. Notably, the Stop Insider Trading Act does not attempt to shorten the window. It restricts purchases outright, on the reasoning that a conflict disclosed late is still a conflict.
How Hillscore handles the two dates
Every trade records both the disclosed transaction date and the later filed date. All performance figures are calculated from the transaction date, never the filing date.
That is the conservative choice, and it is worth explaining why. Measuring from the filing date would credit a member with whatever the stock did during the intervening weeks — a gain already earned by the time anyone could see it, and one a reader could never have captured. Measuring from the transaction date asks the narrower, more honest question: what did this decision do, from the moment it was made?
Feeds and tables sort by filing date, because that is when the information became public and the order a reader encounters it in. Both dates ship in the published dataset so you can measure it either way yourself. Full method.