Right now on Kalshi, a single senator's Yea market is priced meaningfully higher than every market predicting how many senators will vote Yea. Catherine Cortez Masto, the Democrat from Nevada, trades at 44¢ on "will vote Yea on the crypto market structure bill." The threshold market — "will ≥58 senators vote Yea" — trades at 10¢. The other swing Democrats sit between 13¢ and 19¢.
Read carelessly, this looks like an arb. It isn't. It's a structural observation about how the market is pricing vote-clustering correlation, and the math is interesting enough to walk through.
The Setup
The Senate is voting on a crypto market structure bill in the next few weeks. Kalshi has decomposed it into two layers:
- Per-senator markets — will Senator X vote Yea? One contract for each Republican and Democrat genuinely in play.
- Aggregate threshold markets — will ≥58 vote Yea? Will ≥60? Will ≥50?
The bill needs 60 to break filibuster. If Republicans hold all 50 and vote uniformly, they need 10 Democrats to cross. The aggregate market is, in effect, pricing the Democratic crossover rate.
The Price Stack (2026-05-18 20:00 UTC)
Pulled from /api/agent/world:
| Senator | Yea Price |
|---|---|
| Catherine Cortez Masto (D-NV) | 44¢ |
| Angela Alsobrooks (D-MD) | 19¢ |
| Rand Paul (R-KY) | 18¢ |
| Mike Lee (R-UT) | 17¢ |
| Jerry Moran (R-KS) | 17¢ |
| Josh Hawley (R-MO) | 15¢ |
| Lisa Blunt Rochester (D-DE) | 13¢ |
| Aggregate Threshold | Price |
|---|---|
| ≥58 Yea | 10¢ |
| ≥60 Yea | 18¢ |
The Republicans in that table are the ones with meaningful Yea-No spread. Everyone else — McConnell, Thune, Cruz, etc. — implies near 100% Yea.
What the Structure Says
Cortez Masto stands out for two reasons. She represents Nevada — a state with material crypto industry presence and a Republican governor who's been vocal on the issue. And she voted Yes on FIT21 in 2024 and the GENIUS Act in 2025. Among Democrats in play, she's the most likely cross.
The other swing Democrats clustering at 13-19¢ tells you the market thinks they're meaningful possibilities but not anchor votes. The ≥58 aggregate at 10¢ tells you the market thinks the bill probably doesn't clear filibuster.
The Republican outliers — Hawley, Paul, Lee, Moran — at 15-18¢ Yea actually mean they're likely No. These are the libertarian / populist faction. If they defect, you can't make the math work even with full Democratic cooperation.
The Implied Vote Distribution
Back out what the individual prices imply about the aggregate.
Assume the 46 "safe" Republicans are 95% each. The 4 outliers are 17% each. The 6 most-likely Democrats are at the prices above. Compute expected Yea count under independence:
- Safe Republicans: 46 × 0.95 = 43.7
- Outlier Republicans: 4 × 0.17 = 0.68
- Top 6 Democrats: 0.44 + 0.19 + 0.13 + ... ≈ 0.89
- Remaining 44 Democrats (~5% each): 44 × 0.05 = 2.2
Expected total ≈ 47.5 Yea.
Under independence, P(≥58) requires the realized count to exceed expectation by 10+. The standard deviation of the sum is roughly √(Σ pᵢ(1-pᵢ)) ≈ √(46×0.0475 + 4×0.14 + 6×0.13 + 44×0.0475) ≈ √(2.18 + 0.56 + 0.78 + 2.09) ≈ 2.4.
So 10 above expectation is ~4σ under independence — essentially zero probability.
But the market prices P(≥58) at 10¢. The only way to get there is positive correlation in the Democratic crossover. Politically, this is realistic. Bills don't pass by accident. If 5 Democrats cross, it's because Schumer signed off — and once Schumer signs off, more cross. Correlation makes the tail fat.
The 10¢ aggregate is internally consistent with the individual prices only under a strong vote-clustering assumption.
Where the Read Could Break
Three ways the structure could be mispriced:
Sentiment-inflated individual markets. Retail traders who like the bill bid up specific senators without believing the bill clears. Individual Yea markets get inflated. Aggregate stays sober. If this is happening, short senators and long aggregate.
Republican defection underpriced. If the bill carries provisions that Hawley or Paul find unacceptable in the final text, the outlier 15-18¢ Yea prices should be even lower. A 5-cent move on each of 4 Republicans collapses the aggregate.
Information asymmetry on whip count. If a sophisticated trader knows leadership has killed the bill in cloakroom negotiations, they short the aggregate (one trade) instead of shorting every senator (10+ illiquid legs). Aggregate gets ahead of individuals.
The trade you make depends on which of these you believe.
The Trade
If you think individual markets are sentiment-inflated → short Cortez Masto, long ≥58.
If you think Republican defection is underpriced → short ≥58, long the Republican-Yea contracts (they cancel out if a Republican defects).
If you think leadership killed the bill → short Cortez Masto specifically. She's the outlier and would be the first repriced.
If you think the market is right — Cortez Masto is genuinely 44% likely and the bill is 10% likely to clear filibuster — you don't have a trade. You just understand the world a little better.
What SimpleFunctions Flagged
This whole structure is what /api/agent/world surfaces as a "divergence" signal. The output literally reads:
[divergence] How many Senate members will vote Yea… lagging Catherine Cortez Masto vote — 47¢ gap [Crypto]
lhs: Cortez Masto vote (44¢)
rhs: ≥58 Senate Yea (10¢)
gap: 47
The signal isn't a trade recommendation. It's a flag — these two markets have an interesting numerical relationship; go look. The interpretation is yours. The math above is one read. There are others.
That's the discipline. Look at the structure, walk the math, identify which assumptions the prices require, decide whether you believe them.