Correlated Markets

#Definition

Correlated markets are prediction markets whose outcomes are logically connected. Two markets are correlated when knowing the resolution of one provides information about the resolution of the other—or, in strong cases, completely determines it.

The two primary types are:

• Same-outcome markets: Resolve identically. If one resolves Yes, the other must also resolve Yes.
• Different-outcome markets: Resolve oppositely. If one resolves Yes, the other must resolve No.

Understanding market correlation is essential for arbitrage, hedging, and cross-market trading strategies.

#Why It Matters in Prediction Markets

Correlated markets create opportunities and risks that single-market analysis misses.

Price consistency: If two same-outcome markets should resolve identically, their prices should be equal. Divergence indicates either an arbitrage opportunity or a misunderstanding about whether the markets are truly equivalent.

Information propagation: When a correlated market moves, logically connected markets should adjust. Delays in this adjustment create trading opportunities through leader-follower strategies.

Portfolio construction: Traders building positions across multiple markets need to understand correlations to avoid unintended concentration or to construct effective hedges.

Risk management: Holding positions in correlated markets without awareness can multiply risk. A trader long on "Candidate X wins primary" and "Candidate X wins general" has correlated exposure—both positions lose if the candidate underperforms.

#How It Works

#Types of Market Correlation

Same-Outcome Correlation

Two markets have same-outcome correlation when they must resolve to the same value.

Examples:

• "Will the Lakers win the NBA Finals?" vs. "Lakers NBA Champions 2025?"
• "Will US GDP growth exceed 2%?" on Platform A vs. Platform B

#Logic Table: Same-Outcome

Logically: Resolution(Market A) = Resolution(Market B)

If both markets are truly same-outcome but trade at different prices, arbitrage exists:

Market A: Yes at $0.65
Market B: Yes at $0.60

Strategy: Buy Yes on B ($0.60), Sell Yes on A ($0.65)
Net cost: -$0.05 (you receive $0.05)
Outcome: If Yes, you pay $1 and receive $1 (net zero)
         If No, you pay $0 and receive $0 (net zero)
Profit: $0.05 guaranteed

Different-Outcome Correlation

Two markets have different-outcome correlation when one resolving Yes means the other resolves No.

Examples:

• "Will Team A win?" vs. "Will Team B win?" (in a two-team matchup)
• "Will inflation exceed 3%?" vs. "Will inflation stay below 3%?"

Logically: Resolution(Market A) = NOT Resolution(Market B)

For perfectly different-outcome markets:

Market A: Yes at $0.55
Market B: Yes at $0.50

If truly opposite: Price(A Yes) + Price(B Yes) should ≈ $1.00
Actual sum: $1.05

Strategy: Sell Yes on both (or buy No on both)
This exploits the overpricing.

Conditional Correlation

Some markets are correlated only under certain conditions.

Example:

• "Will Candidate X win the primary?" trades at 70%
• "Will Candidate X win the general election?" trades at 45%

These are conditionally correlated: winning the general requires (in most cases) winning the primary. The general election probability can't logically exceed the primary probability (assuming the primary is prerequisite).

If the general election market priced higher than the primary, a logical inconsistency exists.

#Correlation Strength Spectrum

Markets exist on a correlation spectrum:

#Identifying Correlated Markets

Correlation can be identified through:

1. Logical analysis: Reading market questions and resolution criteria to identify dependencies

2. Semantic analysis: Using AI to identify textually similar or thematically linked markets

3. Historical price correlation: For recurring market types, analyzing how prices move together

4. Platform metadata: Some platforms explicitly link conditional markets or use shared conditions (NegRisk)

#Examples

#Example 1: Cross-Platform Same-Outcome

An interest rate decision market exists on two platforms:

If resolution criteria are equivalent (same meeting, same definition of "raise"), these are same-outcome markets. A trader can:

• Buy Yes on Kalshi at $0.68
• Sell Yes on Polymarket at $0.72
• Lock in $0.04 profit regardless of outcome

#Example 2: Different-Outcome Pair

A binary market on a championship game:

• "Team Alpha wins championship?" Yes at $0.58
• "Team Beta wins championship?" Yes at $0.45

If only these two teams can win, the prices should sum to $1.00. The actual sum is$1.03.

A trader can sell Yes on both markets (or buy No on both), collecting $1.03 while knowing they'll pay out exactly$1.00 on one. Net profit: $0.03 per share.

#Example 3: Conditional Dependency

Political markets:

• "Will Senator Y run for President?" at 60%
• "Will Senator Y win the Democratic nomination?" at 25%
• "Will Senator Y win the Presidency?" at 12%

These form a conditional chain. Nomination requires running; winning requires nomination. If "win Presidency" priced higher than "win nomination," an opportunity exists.

#Example 4: Imperfect Correlation

Economic indicators:

• "Will unemployment exceed 4.5% in Q1?" at 40%
• "Will the Fed cut rates in Q2?" at 55%

These are correlated (high unemployment often precedes rate cuts) but not deterministically linked. The Fed might not cut despite high unemployment, or might cut for other reasons. This correlation informs position sizing but doesn't guarantee arbitrage.

#Empirical Evidence: Correlated Market Trading

Research on Polymarket provides data on how traders exploit correlated markets:

#IMDEA Networks Institute Study (2024-2025)

A comprehensive analysis of 86 million bets placed between April 2024 and April 2025 found:

The "multi-condition" profits came specifically from trading across correlated markets—where outcomes in one market logically constrain outcomes in related markets.

#Key Findings

• Sports markets had the most frequent correlation opportunities
• Political markets (especially during the 2024 U.S. election) offered higher-value opportunities
• Most trades captured 1-5% profit margins
• The top individual trader extracted over $2 million from these strategies

This research confirms that correlated market relationships create real, exploitable value—though competition among sophisticated traders captures much of it.

#Risks and Common Mistakes

Assuming perfect correlation when it doesn't exist

Two markets that seem equivalent may have different resolution criteria—different sources, timing, or edge case handling. "Will X happen by December 31?" on one platform might use UTC; another might use local time. This isn't true same-outcome correlation.

Missing conditional dependencies

Traders sometimes hold positions in markets where one is prerequisite to another without recognizing the implied correlation. Diversification requires truly independent markets.

Correlation breakdown

Markets that historically correlate may diverge under unusual conditions. Economic relationships that held for years can break during crises.

Execution timing

Exploiting correlated markets requires executing both sides quickly. If Market A moves while you're trading Market B, the correlation trade may become unprofitable.

Resolution timing differences

Same-outcome markets that resolve at different times create period of uncertainty. The first market resolves, but the second isn't tradeable yet, leaving one leg exposed.

Overestimating correlation strength

"Moderate positive correlation" doesn't mean markets will resolve the same way. Traders often overweight correlation when sizing positions.

#Practical Tips

• Map resolution criteria carefully: Before treating markets as correlated, compare exact resolution criteria including source, timing, and edge cases.

• Calculate correlation-adjusted position sizes: If you hold positions in correlated markets, your effective exposure is larger than it appears. Reduce sizes accordingly.

• Use correlation for hedging: Positions in different-outcome correlated markets can hedge each other. A long Yes on "Team A wins" and long Yes on "Team B wins" provides partial hedging in an uncertain matchup.

• Monitor leader markets: In correlated pairs, identify which market typically moves first (more liquid, more attention) and use it as a signal for the follower.

• Document your correlation assumptions: Write down why you believe markets are correlated and what could cause that correlation to break. Review when market conditions change.

• Check for NegRisk structures: On platforms like Polymarket, formally linked markets through NegRisk are guaranteed to have specific correlation structures. These are safer than assumed correlations.

• Consider transaction costs: Exploiting small price differences between correlated markets requires accounting for fees on both sides. A 2% discrepancy with 1% fees per trade yields no profit.

#Related Terms

• Arbitrage
• Semantic Trading
• Leader-Follower Strategy
• Conditional Markets
• NegRisk
• Categorical Market
• Hedging
• Market Fragmentation

#FAQ

#How do I know if two markets are truly same-outcome?

Read both resolution criteria carefully, including the resolution source, exact timing (date and timezone), and edge case handling. Markets with identical questions can still resolve differently if they use different data sources or interpret ambiguous situations differently. When in doubt, assume they're not perfectly correlated.

#Can correlated markets have different prices?

Yes, for several reasons: imperfect correlation (markets are related but not identical), different liquidity (thin markets have more noise), timing (one market updated more recently), or fees and frictions that prevent perfect arbitrage. Persistent price differences in truly same-outcome markets indicate either genuine arbitrage opportunity or a hidden difference in resolution criteria.

#What's the difference between correlated markets and conditional markets?

Conditional markets are formally structured so one market only activates or has meaning if another condition is met (e.g., "If Team A reaches finals, will they win?"). Correlated markets have related outcomes but are independently structured—both can be traded regardless of other outcomes. Conditional markets are explicitly linked by platform mechanics; correlated markets are linked by logical or causal relationships.

#How can AI help identify correlated markets?

Semantic trading systems use natural language processing to analyze market questions and identify textually similar or thematically related markets. AI can cluster markets by topic, detect near-duplicate questions, and flag logical relationships (e.g., identifying that "primary" and "general election" markets for the same candidate should be conditionally correlated). This automated detection scales better than manual analysis.

#Is trading correlated markets riskier than trading independent markets?

It depends on the strategy. Holding same-direction positions in positively correlated markets concentrates risk—if your thesis is wrong, multiple positions lose simultaneously. However, trading price differences between correlated markets (arbitrage-style) can be lower risk than directional trading because profits don't depend on predicting the outcome. Understanding and intentionally managing correlation is key.

---

Meta Description (150–160 characters): Learn about correlated markets in prediction markets: same-outcome and different-outcome relationships that enable arbitrage and cross-market trading strategies.

Secondary Keywords Used:

• same-outcome markets
• different-outcome markets
• market correlation
• cross-market trading
• linked markets