By the time you see a congressional stock trade, it is already old news, and that is the whole problem. Under the STOCK Act, members of Congress and their spouses have 45 days from the trade date to disclose a securities transaction, and plenty of filers use most or all of that window. A senator who buys $500,000 of pharmaceutical stock on January 5th does not have to file until roughly February 19th. For those six weeks the senator knows about the trade and probably knows what motivated it, and the rest of us do not.
The lag was never designed to be sneaky. The original legislation traded off transparency against the burden of fast reporting, and 45 days was the compromise. The practical effect is still a head start. By the time the disclosure lands, whatever information advantage drove the trade may already be sitting in the price.
How much edge survives the lag
People have actually studied whether congressional trades still make money after you account for the delay. The results are mixed but lean positive. Some find that a portfolio copying disclosed trades, bought on the disclosure date rather than the trade date, still beats the market by something like 2 to 5 percent a year. Others find no real outperformance once you subtract the lag, especially in recent years now that the whole thing gets more attention and more people are piling into the same filings.
The studies that do find post-disclosure edge tend to find it in specific corners of the data rather than across the board:
- Purchases in small-cap names, where a single informed buyer moves the price more.
- Trades by members who sit on committees that oversee the relevant sector.
- Trades that are unusually large compared to how that filer normally trades.
Those traits help separate the information-driven trades from routine portfolio housekeeping, which is most of the volume.
The timing pattern is its own signal
Once you start looking at when people file rather than just what they filed, some patterns show up. Some members consistently file right at the 45-day deadline, which quietly maximizes their head start. Others file within a few days, maybe out of conscientiousness, maybe to avoid looking bad. And a handful show a correlation where the longer delays line up with the more profitable trades, though I would be careful reading too much into that since it could easily be selection bias in the data.
The deadline also creates a clustering effect. Because so many members file near the 45-day limit, you get bursts of disclosure activity that trail bursts of actual trading by almost exactly 45 days. If several members of a committee bought pharma names after a private briefing, those trades tend to surface together, roughly six weeks after the briefing. The individual trades are stale, but the cluster itself tells you something.
Working with data you know is old
Given the lag, there are a few ways to still pull value out of these filings. At Blockcircle we lean on all three when we score the congressional feed.
The first is pattern recognition on the person. If a specific member has a track record of profitable trades inside their own committee's sector, a fresh disclosure from that member carries more weight than a random one, even at 45 days old. The advantage may have partly decayed, but if the trade was riding a policy change that takes months to actually roll out, a lot of it can still be there.
The second is to focus on the information instead of the timing. If three members of the Energy Committee all bought solar stocks inside the same two-week window, and it all showed up 45 days later, the clustering points at a shared catalyst. Rather than chasing the same stocks at a now-higher price, you go find the energy legislation or regulatory move that could have driven it and ask whether that impact is already priced in.
The third is to track the members who consistently file early. The ones who routinely disclose within 5 to 10 days give you a much faster signal, and leaning on them shrinks the lag problem a lot. The tradeoff is obvious, you are watching a much smaller slice of the trades.
Where this goes if the rules change
A few bills have floated cutting the window from 45 days down to as little as 48 hours with electronic filing. Others would just ban individual stock trading by members entirely and push them into blind trusts or index funds. Nothing has passed, but the pressure behind it keeps building.
If the window did drop to a couple of days, the value of this data for outside investors would jump. Near real-time disclosure would basically hand the public a live stream of trades from people with privileged access to policy. Until that happens, you are stuck filtering and analyzing around the 45-day lag and accepting that the delay shrinks the signal without killing it. That is the realistic bar for now, so build your process around what actually survives the wait rather than the headline that a senator bought something.