I've been tracking what Congress buys for 6 months. Here's what I learned and the free tool I built.
about 6 months ago i started manually tracking congressional stock trades after reading about how consistently some members outperform the S&P. the STOCK Act requires disclosure within 45 days but the filings are buried in PDFs on senate.gov and the house clerk site. pain to parse.
after a few weeks of doing this manually i started noticing patterns. committee members trading in their sectors. clustering where 3+ members buy the same industry within a month. trades that preceded major contract announcements or regulatory changes by weeks.
so i started cross-referencing with two other data sources: unusual options flow (sweep orders, ascending fills, large premium bets) and dark pool block trades. when all three signals align on the same ticker, the hit rate goes way up compared to any single signal alone.
i ended up building a free tool that pulls all three into one place and scores the signals. its called MarketSignals (marketsignals.pages.dev). no paywall, no upsell. i just wanted something that saved me the manual work and figured other people might find it useful.
what it tracks:
- congressional trades as they're disclosed (with committee context)
- unusual options activity scored by sweep type, fill pattern, premium size
- dark pool block trades above threshold with clustering detection
a few things i learned along the way:
1. the 45 day disclosure window means congressional data is useless for day trading but excellent for swing/position trades
2. options flow is 99% noise. the scoring system matters more than the raw data
3. dark pool prints are only meaningful when they cluster. one big block is a data point, five blocks in a week is a signal
4. the real edge is when 2-3 of these signals converge on the same name
happy to answer questions about the methodology or how i built the scoring system. wrote up detailed guides on the site too if anyone wants to dig deeper into any of the three signal types.