Strategy 07 · Event Contracts
Probability Pricing
The price is a probability. Is it the right one?
A contract that pays one unit if an outcome occurs and trades at a fraction of that unit is quoting a probability. Buying it is a statement that the true probability is higher; selling, that it is lower. The mandate’s edge, where it has one, is a better estimate than the market’s, built from the same public information, read with more discipline, and acted on only when the gap is wide enough to pay for being wrong.
From price to probability
The conversion is direct, with adjustments: the time value of money between now and resolution, the venue’s fees, and the bid-ask spread, which on thin contracts can be wide enough to hide any edge. A contract quoted as a range rather than a point implies a range of probabilities, and the mandate’s estimate must fall outside the whole range before there is a trade.
Forming an estimate
For scheduled data, from the distribution of forecasts, historical surprise and the current data flow. For regulatory outcomes, from the procedural record, the timeline and the stated positions of decision-makers. For protocol events, from the state of the code and the governance process. Every estimate is written down with its reasoning before a position is taken, so that it can be checked afterwards against what happened.
Why the gap must be wide
Event contracts are binary. A position that is right most of the time and wrong occasionally loses its entire stake when it is wrong. The required gap between market and estimate therefore scales with the confidence in the estimate and with the size of the loss on the wrong side. Small edges on binary outcomes are not traded; they are noise.
Calibration
The record of estimates against outcomes is the strategy’s primary quality control. If the mandate’s stated probabilities do not match observed frequencies over time, if events it called at a given confidence happen more or less often than that, the process is recalibrated or the category is dropped.
Practice
How Wesvel approaches it
- 01Every position has a written probability estimate and reasoning recorded before entry, and is scored against the outcome after resolution.
- 02A trade requires the estimate to lie outside the full quoted range, net of fees and time value, by a margin that scales with the size of the loss if wrong.
- 03Calibration is reviewed on a schedule by category; categories that do not calibrate are suspended.
- 04Estimates use only public information; the eligibility test excludes events where the fund could hold anything else.
More in Event Contracts
- What Is TradedThe categories of event contract the mandate trades, and the test every contract must pass before it is eligible.
- Venues and RegulationRegulated designated contract markets versus on-chain prediction markets, and how an unsettled regulatory position shapes what the fund can do.
- AI and Mispriced ProbabilitiesHow machine learning and language models screen hundreds of event contracts for prices that disagree with a calibrated estimate, and how that gap becomes a position with positive expected value.
- Cross-Venue Relative ValueBuying an outcome where it is cheap and selling it where it is rich, and why contract definitions decide whether the trade is real.
- Portfolio Event HedgingUsing event contracts to offset exposure elsewhere in the book to outcomes that price instruments cannot hedge.
- Resolution Risk and Position LimitsHow contracts resolve, what goes wrong when they do, and the caps that keep the mandate small.