Structured products for prediction markets

You don't bet on the event outcome. You bet on how bumpy the price path is between now and settlement. The realized volatility of an election contract from now to election day is priceable.

Buyers bet that significant new information will arrive and cause violent price swings. Sellers bet that current prices already reflect most uncertainty and the remaining path is smooth.

The pricing model diverges fundamentally from traditional volatility swaps because vol has a forced-to-zero boundary condition — at expiry, price must be 0 or 1, so volatility must collapse to zero regardless of path. Whoever solves this model first has a structural edge for as long as nobody else does.

Price jumps in prediction markets have a clear semantic meaning — new information arrived and changed the collective probability estimate. You can construct a contract: if price makes a single jump exceeding a threshold within a time window, it pays.

You're trading the pulse structure of information flow. You don't need to know what the information is. You're betting on whether significant news will arrive in a given period. This product hedges information asymmetry risk itself.

The closest traditional analog is event-driven funds buying straddles, but that's a bet on directional uncertainty, not on information arrival timing. No precise equivalent exists in traditional markets.

If contract A settles "yes", what is contract B's implied probability? This is a second-order contract trading the causal or correlative structure between two events.

Example: if the Fed cuts rates in June (event A), what does the market estimate for recession probability (event B)? You think the cut signals weakness; someone else thinks it prevents recession. You trade the causal hypothesis directly.

The pricing problem is hard — you need to infer joint probability distributions from marginal prices, and markets usually don't give you enough information to uniquely determine the joint. Whoever can estimate the optimal joint distribution from sparse data owns the pricing.

With enough liquid binary contracts, you can trade the implied correlation structure between them. A basket of political contracts, a basket of macro contracts — when correlation is underpriced, go long; when overpriced, go short.

This is the prediction market equivalent of CDO tranches. The credit crisis lesson applies directly: correlation is low and stable in normal times, then jumps to one in a crisis and everyone gets liquidated simultaneously.

Enormous profit potential. Enormous systemic risk. The product is straightforward to construct; the risk management is not.

You don't bet on the terminal value. You don't bet on volatility. You bet on the shape of the price path. "Price goes up then down" and "price goes down then up" are different contracts even if start and end points are identical.

Path shape encodes the evolution of market narrative — up-then-down means "initial optimism followed by disillusionment"; down-then-up means "initial panic followed by recovery." These different narrative trajectories have entirely different downstream consequences even when the terminal outcome is the same.

Trading path shape means you can hedge narrative risk, not just outcome risk. This is the most abstract layer and possibly the most valuable.

Portfolio-level margin on correlated positions. Credit facility for hedged books. Cross-venue settlement netting between Kalshi and Polymarket.

Derivative positions require margin infrastructure that doesn't exist yet in prediction markets. We're building it.