Practice focus · Predictive Markets
Thousands of contracts, one missing price: correlation.
Event contracts have scaled into a genuine asset class — tens of billions a month, a Supreme Court question pending, and ETF wrappers already filed. What they still lack is the structure every mature derivatives market eventually grows: a benchmark tier where liquidity concentrates, a basis layer where idiosyncratic risk gets sold back, and a quoted price for the correlation that combo flow is already trading blind.
This section works that gap from both ends — the contract architecture that would close it, and the arithmetic that reveals it in prices already printing. It is also where the practice keeps the most tooling, because most of these findings only appear when you run the numbers across venues.
1 · Architecture — where liquidity should concentrate
The structural question that organizes everything else: event markets fragment into thousands of thin, idiosyncratic contracts, and mature markets solve that by concentrating liquidity in a parametric benchmark and selling the residual back as basis. That is the ILW and CDS-index pattern, applied here.
CME's NHL futures rebuilt from six seasons of official play-by-play — 15,410 team-games. A 55-row constituent table collapses to goal differential plus a milestone layer carrying 41% of per-game variance; 69% of the risk is outcomes Kalshi already trades, 31% has no outcome book, and the largest discretionary term is scored by the home arena's own off-ice crew
Three announcements in thirty-six hours put event contracts into the institutional risk stack — but the flow crossing in is weather and commodity risk, and its offsets sit in collateral pools the layer cannot reach. One exposure in four wrappers, three pools with no bridge, and the three routes by which the span eventually gets built
Cantor is the pipe, Kalshi the venue, Susquehanna the balance sheet — a six-avenue offset menu, two worked blocks (one that back-to-backs in fed funds futures, one that cannot be hedged at any price worth paying), and the argument that warehousing capacity rather than distribution is the moat
Half of Gen Z investors moved money meant for investing into sportsbooks — a capped, guardrailed event-contract allocation for the speculative dollars already committed, built on live venue books, with ten guardrails and the registration question of who could run it
Kalshi's 15-minute metals grid read through Derman's intrinsic time — 96 daily binaries as a market-quoted activity clock, and why the heat belongs to the trader rather than the metal
The CFTC's July 30 conflicts proposal for the vertically integrated exchange — mapped across thirteen corporate families, with the filled-last market-maker rule as the place it stops being principles and starts being money
Six event families tiered into benchmark and basis layers under the CFTC event-contract regime, with tranche and correlation models and market-scoring-rule sizing
A four-phase listing plan for litigation-outcome derivatives, sixteen candidate cases, and a perpetual-on-index design for a market with no natural expiry
2 · Correlation — the price nobody quotes
The partition rule is the sharpest tool in this section: the prices of a mutually exclusive, collectively exhaustive set must sum to one, or a riskless box exists. Applied to what already trades, it turns combo pricing into an implied-correlation surface — and exposes where venues violate their own arithmetic.
Every combo price minus the product of its legs is an implied correlation, printed thousands of times a day and quoted by nobody — with a Fed × BTC ladder join, a World Cup lookback, and $100M a week of flow read as a correlation tier with no benchmark to hedge it
A settlement-hardness spectrum across intraday event contracts and perp liquidation marks: distortion flows one way, and a visible on-chain liquidation cluster makes the push materially cheaper
Kalshi and Polymarket settling on the same URL — index basis verified at zero, a six-bracket exhaustive partition quoting to 180%, and a spread still unreadable
Also in Energy & Compute3 · Wrappers and adjacent structure
What happens when event-contract exposure gets packaged for buyers who cannot hold the contracts directly — and where the wrapper works against the product.
The filed cohort, the total-return-swap plumbing that makes a 1940 Act fund possible on a CFTC event contract, why the wrapper fights the binary payoff, and two structural concepts not yet in market
Kalshi's compute ladders as a strip of digitals — a discretised probability distribution, and therefore a term structure, extracted from event-contract order flow
Also in Energy & Compute4 · The tooling — the deepest cluster on the site
Five interactive engines, because this is the section where the analysis is most computational: margin under joint stress, cross-venue mispricing, position sizing, and two tournament-structure arbitrage models built on live venue data.
BTC/SPX perpetuals against event-contract binaries in one portfolio-margin framework — 8-cluster Monte Carlo over 5,000 paths, an implied-vol solver and surface, and venue-aware margin comparison
Also in Digital & DerivativesKalshi against Polymarket against Manifold — where the same question prints at different probabilities, and what survives fees
Kelly and fractional-Kelly sizing run against the margin framework, so position size and margin consumption are solved together rather than sequentially
Outright versus match-path dispersion — the tournament-structure arbitrage where the partition rule has real teeth, built on live venue quotes
Best-of-seven series math into implied per-game probabilities, with a dutching detector and LP-optimal portfolios across Polymarket and Kalshi
Where this connects
The threads that run out of this one.
The first continuously published compute forward curve was extracted from event ladders — the clearest case of this section's arithmetic doing work elsewhere.
Digital & derivativesEvent contracts and perpetuals now share a margin framework and, on several venues, a book — which is why the settlement-hardness comparison exists at all.
Market structure radarListings, CFTC actions and the state-preemption litigation tracked continuously — the fastest-moving regulatory surface the practice covers.