Combinatorial and Conditional Markets: Pricing Events That Depend on Other Events

Abstract

Combinatorial and conditional markets price bets on outcomes that logically or causally depend on other outcomes, using conjunction, disjunction, or conditional logic to create complex payoff structures. These markets emerge when correlated events require joint or sequential pricing, and their mechanics depend on order types that express dependencies and on pricing models that account for correlation. This paper examines how these structures are built, priced, and applied in modern prediction markets.

Core Concept

A combinatorial market is one in which a single contract's payoff depends on the joint outcome of two or more underlying events. The simplest forms are:

• An "AND" market: pays $1 if both event A and event B occur, $0 otherwise.
• An "OR" market: pays $1 if at least one of events A or B occurs, $0 otherwise.
• A conditional market: pays contingent on the realization of one outcome; for example, a market on the winner of a US presidential election given that a particular party wins the nomination.

These differ from standard wagering on single outcomes. A trader seeking exposure to two independent 50-50 events would expect, under fair pricing, to pay approximately $0.25 for an AND market (0.5 × 0.5) and $0.75 for an OR market (1 − (0.5 × 0.5)). However, when events are correlated, because they share a common cause, influence each other, or depend on a third variable, the prices diverge from simple multiplication or union rules.

Conditional markets formalize the idea of betting on one outcome, contingent on another. These are most naturally expressed as order types: "buy Contract Y at price Z, but only if Contract X resolves YES." This allows traders to build multi-leg positions that never execute if the condition fails, reducing basis risk and avoiding unintended exposure.

How Combinatorial and Conditional Markets Work Mechanically

Order-Based Construction

Prediction market platforms, particularly Kalshi, have formalized combinatorial markets through conditional order types ^[1]. When a trader places a "conditional order," the exchange holds it in reserve until the condition market resolves; only then does it execute at the specified price and quantity, if the condition met the stated outcome.

Example: Suppose Contract A (binary: "Will it rain on Monday?") and Contract B ("Will it rain on Tuesday?") are both live. A trader could place a conditional order: "If Contract A resolves YES, buy Contract B at $0.60 for 100 shares." The order remains inactive while A is undecided. Once A resolves YES, the order immediately executes; if A resolves NO, the order is cancelled.

Pricing Under Correlation

When a market explicitly prices joint outcomes, the fair price reflects the joint probability and the correlation structure. Denote P(A) as the implied probability of event A (derived from the market price), P(B) as the implied probability of B. Under independence, P(A AND B) = P(A) × P(B). However, markets often exhibit correlation:

• Positive correlation: both events more likely to co-occur (e.g., "GDP growth next quarter" and "Fed holds rates steady"). The AND market trades above independent pricing; the OR market trades below union pricing.
• Negative correlation: events trade off (e.g., "Large gain in equity markets" and "Large drop in bond yields on flight-to-safety"). The AND market trades below independent pricing; the OR market trades above union pricing.

Traders can exploit mispricing by constructing synthetic positions. If an explicit AND contract trades at $0.20 while the product of individual contract prices implies $0.22, a trader might sell the AND contract and simultaneously buy both underlying contracts at their market prices, locking in a $0.02 profit per share (before transaction costs). This arbitrage pressures prices toward consistency.

Conditional Order Mechanics

A conditional order on a prediction market is a wrapper around two constraints: a condition contract and a contingent contract. Once the condition resolves, the order becomes a simple limit order on the contingent contract, executed at the stored price if liquidity permits. If the condition resolves opposite to the stated requirement, the order is deleted.

The advantage for the trader is execution certainty for the contingent leg: no need to log back in and re-evaluate after the condition resolves. For the market, conditional orders reduce cancellation risk and create natural liquidity anchors on secondary outcomes.

Worked Example: 2024 Presidential Election Markets

Prediction markets on the 2024 U.S. Presidential election included both binary markets on individual outcomes (e.g., "Will Joe Biden win the presidency?") and conditional markets on subsequent events given each candidate's victory.

In March 2024, Kalshi offered conditional markets on which party would control the Senate after the election, contingent on each major-party presidential candidate winning ^[1]. The market on "Republicans control Senate given Trump wins" traded at a different price than a standalone "Republicans control Senate" market, reflecting the correlation between presidential and senatorial outcomes.

Suppose the standalone "Republicans control Senate" contract traded at $0.55 (implying a 55% probability), while "Trump wins presidency" traded at $0.40. If these events were independent, one might expect "Republicans control Senate given Trump wins" to also price at $0.55. However, Republican Senate strength is positively correlated with a Trump victory (shared voter sentiment, shared campaign effects). The conditional contract might trade at $0.62, reflecting the higher co-occurrence likelihood.

A trader believing the conditional probability was mispriceable could:

1. Place a conditional order: "If Trump wins, buy 100 shares of 'Republicans control Senate' at $0.60."
2. Simultaneously sell 100 shares of the standalone "Republicans control Senate" at $0.55.
3. If Trump wins and the order executes at $0.60, the trader is short the standalone contract at $0.55 and long the conditional at $0.60 in the post-Trump-win state, a loss of $5 per 100 shares.
4. If Trump loses, the conditional order is cancelled, and the short position in the standalone contract remains; if the Senate outcome is unresolved, the trader holds a naked short, incurring basis risk.

This example illustrates the dual nature of conditional bets: they reduce execution risk but introduce selection or truncation risk if the condition does not materialize.

Limitations

Liquidity and Execution Risk

Combinatorial contracts often trade with wider spreads and lower volume than single-outcome markets. When a conditional order is placed on an illiquid secondary market, the stored execution price may not be competitive once the condition resolves and the order activates. A trader might intend to execute at $0.60, only to find that real liquidity appears at $0.58 or $0.65 by the time the condition triggers.

Correlation Estimation and Model Risk

Pricing combinatorial contracts requires estimating correlation between outcomes. Individual market prices for A and B imply marginal probabilities but not the joint distribution. A trader relying on a simple Gaussian copula or assuming independence will misprice asymmetric or fat-tailed correlations. Binary prediction markets have limited history; correlation estimates are often derived from external data (e.g., historical election results, economic data) that may not apply to the current betting environment.

Selection and Truncation Bias

If a conditional order is cancelled because the condition resolves unfavourably, the trader is left with a potentially large unhedged exposure in the underlying market. For example, the trader who sold the standalone "Republicans control Senate" contract and planned to hedge it with a conditional buy loses that hedge if Trump loses. The effective exposure becomes a directional bet on Senate outcomes in a non-Trump world, which was not the intended trade.

Basis Risk and Event Timing

Conditional markets assume clear, binary resolution of the condition. In reality, event space is multidimensional. "Does Trump win?" is binary, but "Will Senate Republicans control the chamber?" depends on definition, timing (does control mean a majority on day one, or after expected retirements?), and interpretation. Conditional contracts can be exposed to disputes or ambiguity that would not arise for the same contract as a standalone market.

Limited Historical Data on Pricing Accuracy

Prediction markets on combinatorial outcomes are relatively new. There is limited empirical evidence on whether these markets price correlations efficiently or whether traders successfully arbitrage combinatorial mispricings in practice. Market-wide studies on correlation accuracy in combinatorial betting do not yet appear in the public literature.

Summary

Combinatorial and conditional markets extend single-outcome betting to joint and sequential outcomes by pricing correlation and enabling conditional execution. Mechanically, they rely on either explicit contracts whose payoff depends on multiple events or on conditional order types that activate only if a prior event resolves as specified. Fair pricing requires estimating not just individual event probabilities but their correlation structure; mispricing can create arbitrage opportunities, though liquidity constraints and execution risk limit the ability to capture them.

Prediction markets such as Kalshi have formalized these structures, particularly in political and event-driven markets where outcomes are strongly correlated ^[1]. However, limitations in liquidity, correlation estimation, and the treatment of edge cases remain significant practical constraints. For traders, combinatorial markets offer a tool to express correlation views and reduce certain types of basis risk, but at the cost of accepting execution uncertainty and truncation risk if conditions fail to materialize.

Key Definitions

AND contract: A binary contract that pays $1 if all of two or more specified outcomes occur, $0 otherwise.

OR contract: A binary contract that pays $1 if at least one of two or more specified outcomes occurs, $0 otherwise.

Conditional market: A contract whose payoff is contingent on the resolution of another contract; often implemented as a conditional order type that activates only if the condition market meets a specified state.

Correlation: A measure of how the probabilities of two or more events move together; positive correlation means both events are more likely to co-occur than under independence.

Implied probability: The probability derived from a market price, calculated as the observed market price (assuming a $1 payoff at resolution).

Conditional order: A market order that remains inactive until a specified condition contract resolves; once the condition is met, the order activates as a limit order on a contingent contract.

Basis risk: The risk that a hedge does not perfectly offset the underlying position due to imperfect correlation, timing mismatches, or contract specification differences.

References

1. Kalshi Markets, "Conditional Markets," Kalshi (2024). Https://kalshi.com/

2. Manski, Charles F., "Interpreting the Predictions of Prediction Markets," Economics Letters, vol. 91, no. 3 (2006): 425-429. Doi:10.1016/j.econlet.2006.03.009

3. Wolfers, Justin and Zitzewitz, Eric, "Prediction Markets," Journal of Economic Perspectives, vol. 18, no. 2 (2004): 107-126. Doi:10.1257/0895330041371339

4. Chen, Yiling and Pennock, David M., "A Utility Framework for Bounded-Loss Market Makers," Proceedings of the 23rd Conference on Uncertainty in Artificial Intelligence (2007). ArXiv:0812.3950

5. Othman, Abraham, "Zero-Intelligence Prediction Markets for Binary Outcomes," Proceedings of the 2012 IEEE 51st Annual Conference on Decision and Control, IEEE.

Educational research on historical data only. Not investment advice, not a signal, and never a performance promise. Past results do not predict future performance. Every reference is link-verified before publication and every paper is re-audited weekly against the library's editorial standard.

Last reviewed by the PropLedger research pipeline: 2026-09-27. Educational research on historical data, not financial advice.