Congress Diff: What Legislators Bought and Sold This Quarter

In February 2024 I ran the quarterly Congress Diff view over the Q4 2023 disclosures and noticed something that struck me: net legislator flow in a specific set of rare-earth-mining and defense-supply-chain names had flipped from roughly neutral in Q3 to net-long by a statistically meaningful margin in Q4, not a landslide, but eight to twelve members net-long per name across a sector of about 15 names. No single trade was remarkable. The aggregate was. Three weeks later the White House announced new critical-minerals policy with specific contracts lined up, and the sector rallied 20% over the subsequent six weeks. The individual disclosures didn't tell me anything; the diff view showed a coherent positioning pattern across a coherent set of names that pointed to the policy shift before it was public. This is exactly what aggregated Congressional flow is supposed to show, and it's what any single trade can't.

This post is about the Congress Diff card, why quarterly aggregates are the unit of analysis that produces signal, and the three patterns that separate informative flow from random dispersion.

TL;DR

• Individual trades are noise; aggregates are where signal lives. Single-trade interpretations are almost always wrong.
• Sector rotation in Congress flow often leads policy narrative by 1–2 quarters.
• Committee-concentrated flow is higher-signal than generic flow. 80% of net flow from relevant committee members > same total from unrelated legislators.
• Combining with lobbying spend tightens the signal. Rising lobbying + rising congressional flow = policy tailwind base case.
• Heatmap view is the scanner; drill in on rotations. Don't read single-name diffs in isolation.

Why the aggregate is more informative than any single trade

One senator's $50k buy isn't a trading thesis. One member's trade in a single name is statistically noise, members trade for idiosyncratic reasons (diversifying a concentrated equity grant, financial-advisor rebalancing, estate planning) that carry no market-relevant information. Thirty members of Congress netting positive across Big Tech over a quarter, while netting negative across regional banks, is information, because it means something in the policy environment has shifted and enough people with varying information sets positioned for it that the shift is probably real.

The Congress Diff view aggregates thousands of individual filings into a quarter-over-quarter net flow per ticker and per sector, which is the unit of analysis that actually generates signal. At the aggregate level, the idiosyncratic noise averages out and what's left is correlated positioning, which, if clustered by committee relevance, is where the academically-verified alpha concentrates.

What the Congress Diff card shows

The Congress Diff card shows quarterly net legislator flow per ticker with:

• Net buy/sell count: legislators net-long vs. net-short this quarter, split by chamber and party
• Estimated net notional: midpoint of disclosed ranges summed across members
• Change vs. prior quarter: ΔQoQ in net-long count, net notional, and committee-weighted flow
• Sector aggregates: quarterly net flow rolled up by GICS sector
• Committee overlap ratio: fraction of the net flow sourced from legislators on committees with jurisdictional overlap to the sector
• Sector heatmap: quarterly flow by sector over trailing 8 quarters, colored by direction and intensity, for fast rotation detection
• Drill-down to the per-ticker Congress Summary: one click into the name-level view

Exports include CSV + a weekly email digest of the largest net flow moves.

Reading the flow

Sector rotation leads policy narrative. Often by 1–2 quarters. Legislators positioning net-long in, say, defense or rare-earths two quarters before the narrative turns bullish is a pattern that recurs, because legislators have advance knowledge of budget authorizations, committee markups, and regulatory direction. The heatmap view makes sector rotations visually obvious; when a sector flips from red (net sells) to green (net buys) over two quarters, it's usually worth a drill-down into what's driving the change.

Committee-concentrated flow is higher-signal than generic flow. If the net-long flow in a healthcare name is 80% sourced from Health Committee members, the implied information ratio is higher than a broad name with equal flow spread across 30 unrelated legislators. The committee overlap ratio column highlights this, names where the flow is committee-driven are where the academic alpha concentrates, not names with diffuse flow from members trading on news like the rest of us.

Combine with lobbying spend. When net-long congressional flow AND quarterly lobbying spend both jump for the same name, a legislative/regulatory tailwind is the base case. The card's sector aggregates cross-reference lobbying data so you can see the two together. Flow-up, lobbying-up is a different signal than either alone.

Flow shifts matter more than flow levels. A name that's been consistently net-bought for years, still net-bought this quarter, is an equilibrium. A name that flips from net-sold to net-bought is a change. Focus drill-down attention on the delta column, not the absolute level.

Example: a 2023 sector rotation

The heatmap over Q1–Q4 2023 for sectors I tracked:

Defense went from neutral in Q1 to heavily net-bought in Q3–Q4. Technology went the opposite direction. At the time, defense narrative wasn't yet clear, budget negotiations were ongoing and the FY2024 Defense Authorization Act wasn't finalized until December. The Congress Diff pattern was visible by end of Q3 2023. Defense stocks outperformed S&P 500 by 18 percentage points from Q3 2023 through Q2 2024. The flow pattern was a leading indicator; fundamental analysts caught up two quarters later.

What the card won't catch

• Intra-quarter timing. Diff is quarterly aggregated; timing within a quarter isn't visible at this resolution. For intraday decisions, use the Live Activity card.
• Spouse and dependent trades. Included in the aggregate but often carry less signal than direct member trades.
• Trades by members not in the aggregated universe. Some members trade very little; their absence from flow doesn't mean absence of information.
• Structural rotations vs. event-driven flow. A pre-budget-authorization flow shift and a post-earnings-rotation flow shift look similar in the aggregate; the context has to come from elsewhere.

Common mistakes

• Reading single-name diffs without sector context. A 3-net-long-buy name in a sector where every name is net-long is consistent with the sector flow; the name-specific read is weak.
• Ignoring committee overlap. Flow concentrated in non-relevant committees is much weaker signal than flow from relevant committees.
• Treating diff changes as tradeable without confirmation. Combine with lobbying, earnings, and news before sizing.
• Overweighting the largest notional movers. A few members with very large ranges can dominate notional-weighted flow; count-weighted flow is often more informative.
• Reading one quarter's flow as trend. Look at 2–4-quarter trajectories to distinguish trend from noise.

Where it fits

Pair with Congress Live Activity for the raw feed, the per-ticker Congress Summary for single-name drill-down, and Corporate Lobbying Spend to cross-reference the regulatory-tailwind read. For macro context, the Universe sector rotation view overlays sector-level flow with broader market flow.

FAQ

How fresh is the diff data?

Lags the underlying disclosure cadence. Since STOCK Act filings have up to 45-day lag, the quarterly diff is most reliable 1–2 months after quarter-end when most filings have posted.

What's the right aggregation window?

Quarterly is the natural cadence (matches 13F and committee cycles). Shorter windows (monthly) are noisy; longer windows (annual) smooth out useful rotations.

How do I use this without being political?

The card is agnostic to party, it shows net flow across all filers. Party breakdown is available in the drill-down but the aggregate is the signal.

Does this predict elections or policy?

Indirectly, sometimes. It predicts sectors where policy is likely to shift (which members are positioning for) before that shift is widely known. Not a direct election-outcome tool.

Can I export to a spreadsheet?

Yes, CSV export with per-ticker, per-sector, per-chamber breakdowns. Weekly email digest option in settings.

Related reading

• Congress House Trades
• Congress Live Activity
• Congress Senate Trades
• Congress Trades per Ticker

Open the Congress Diff card → /app/congress