How the machine reads disclosed congressional trading data.

Sources

The primary trading source is Capitol Trades public disclosure data. Returns Spectrum combines the disclosure bundle with end-of-day market price signals. Policy Proximity also uses Congress.gov public-law listings for the 118th and 119th Congresses.

Value ranges and volume

Congressional financial disclosures use reported value ranges, not exact order sizes. The project uses representative disclosed values for visual weight and ranking. Capital Flow separates scale from direction: gross disclosed volume sums buys and sells, while net flow treats buys as positive and sells as negative.

disclosed value; gross volume sums buys and sells; net flow uses buys positive and sells negative

Timing and sector heuristics

Committee exposure and Policy Proximity are triage heuristics. They are designed to help readers inspect patterns, not to make claims about motive or inside information.

• Committee-sector exposure compares a member's committee context with sectors represented in disclosed trades.
• Policy Proximity links public-law events to trades by date proximity and sector proximity.
• Returns Spectrum compares disclosed trades with trailing market movement for visualization context only.
• Reporting dates can lag transaction dates, and some disclosures include spouse, dependent, or joint transactions.

Caveats

Congressional financial disclosures are incomplete as a market dataset. Some transaction types are excluded from buy/sell flow, reported values are ranges, and disclosure timing can vary. The current bundle skipped exchange and received transaction types for flow math because they are not clean buy/sell signals.

Data

Current disclosure bundle

A crawlable summary of the public data currently bundled into the interactive scenes, generated on 2026-11-02.

Capitol Trades public 3-year volume total, read 2026-06-14

Dataset windows and sources

Data use notes

This dataset describes an editorial visualization bundle derived from public disclosure sources. Original source records remain subject to their publishers' terms. Reuse should cite Capitol Alpha Machine and the underlying public sources, preserve the caveats about disclosure ranges and timing heuristics, and avoid presenting representative values as exact holdings, profit, intent, or evidence of wrongdoing.

Top tickers and members in the returns bundle

These rankings use representative disclosed values inside the returns scene bundle. They are not a complete accounting of holdings, profit, or intent.

Committee Risk

Committee-risk proximity dataset

This aggregate links public-law events to disclosed trade rows using date proximity and sector proximity. Committee-sector overlap is a heuristic flag, not a claim of causation.

Top committees in the current aggregate

Rows are aggregate only. Individual committee pages are not generated.

Risk flags

Flags describe why a row became visually prominent inside Policy Proximity. Read them as triage labels for inspection, not conclusions.