Prediction Market Arbitrage: Strategies, Tools & Real Examples
April 12, 2026 · PolyMath Team · 11 min read
Arbitrage is the closest thing to free money that exists in financial markets. In prediction markets, arbitrage opportunities arise when the same event is priced inconsistently — either across different platforms, or within a single platform through correlated markets.
The math is simple. The opportunities are real. But they require speed, tools, and a clear understanding of the risks involved. This guide covers everything: how prediction market arbitrage works, the different types of opportunities you'll encounter, the mechanics of executing a trade, and how to use the PolyMath Arbitrage Scanner to find them automatically.
What Is Prediction Market Arbitrage?
Arbitrage, in any market, is the simultaneous purchase and sale of related assets to profit from a price discrepancy — with zero or near-zero risk.
In prediction markets, the core prices are probabilities. A market on "Will Candidate X win the election?" trades between 0¢ and $1. If you buy at 40¢ and the market resolves YES, you collect $1. If NO, you lose your 40¢.
Arbitrage occurs when the prices quoted by the market are inconsistent with each other.
The simplest case: a binary YES/NO market where YES is priced at 55¢ and NO is priced at 55¢. Both can't resolve true. If you buy both sides for $1.10, one will resolve to $1.00 — guaranteeing a 10¢ loss. That's negativearbitrage (known as "vig" or juice).
True arbitrage is the opposite: YES at 45¢ and NO at 45¢. Buy both for $0.90, collect $1.00 on resolution — that's a guaranteed 10¢ profit regardless of outcome. In Polymarket's order book model, these pure book-level arbitrages are rare and fleeting. The more common and practical opportunities fall into three categories.
Type 1: Cross-Exchange Arbitrage
The most textbook form of arbitrage. The same event is listed on two different prediction market platforms at different prices.
📊 Example
Kalshi: "Federal Reserve cuts rates in June" → YES at 52¢
Polymarket: "Fed rate cut June 2026" → YES at 61¢
Buy YES on Kalshi at 52¢ and sell YES on Polymarket at 61¢ (by buying NO at 39¢). The 9¢ spread creates the arbitrage opportunity — before accounting for friction.
Cross-exchange arbitrage is trickier than it looks because of execution friction:
Capital on both platforms simultaneously
You need funded accounts on each exchange at the same time — and capital tied up on one cannot be deployed on the other.
Withdrawal/deposit friction
Getting USDC between platforms takes time, during which prices move. On-chain delays can invalidate the opportunity entirely.
Identical resolution criteria required
The markets must be measuring the identical event with identical resolution criteria. Kalshi and Polymarket can resolve the same real-world event differently.
Spread threshold
When spreads exceed 8-10%, the math can work even accounting for friction. Below that, execution costs eat the profit.
Type 2: Within-Platform Correlated Market Arbitrage
This is the most accessible form of arbitrage on Polymarket, and it's what the PolyMath Arbitrage Scanner is built to find. Some markets are logically connected. If the probabilities implied by two related markets are inconsistent, a risk-free or near-risk-free trade exists.
The "Exclusive OR" Pattern
The clearest case: mutually exclusive, collectively exhaustive outcomes.
📊 US Presidential Election Example
"Trump wins 2024" → 52¢
"Harris wins 2024" → 47¢
"Other wins 2024" → 4¢
Sum: 52 + 47 + 4 = 103¢ (vig — no arb)
If "Other" dropped to 0¢ and the sum became 99¢, you could buy all three outcomes for 99¢ and collect $1.00 — a guaranteed 1% return.
The Nested Market Pattern
Some markets are logically nested — one outcome is a strict subset of another.
📊 Nested Example
"Inflation below 3% in Q2" → 55¢
"Inflation below 2.5% in Q2" → 30¢
The second is a subset of the first. If inflation is below 2.5%, it's also below 3%. Therefore P("below 2.5%") cannot exceed P("below 3%"). If it did, you have a pure arbitrage.
Type 3: Temporal Arbitrage
Markets update in real time, but not always simultaneously. Breaking news can move one market before another has repriced.
⚡ Example
A surprise jobs report drops at 8:30 AM. "Fed cuts rates in June" should immediately re-price. But "Inflation expectations 6 months out" might lag by minutes. Speed is everything here — this is the domain of algorithmic traders. But for retail traders, temporal arbitrage creates occasional "gift" opportunities when you notice a market that hasn't updated to obvious news.
How to Use the PolyMath Arbitrage Scanner
The PolyMath Arbitrage Scanner is purpose-built to find Type 2 arbitrage — correlated markets within Polymarket that are inconsistently priced.
Step 1
Enter related market prices
Input the YES probability for Market A and Market B from Polymarket.
Step 2
Define the relationship
Select: Exclusive OR (mutually exclusive outcomes), Conditional (one implies another), or Nested (one is a subset).
Step 3
Calculate implied spread
The scanner calculates the implied arbitrage spread based on your relationship type.
Step 4
Get exact trade sizing
If a spread exists, the scanner shows the exact trade: how many shares of each to buy to lock in the guaranteed profit.
For two mutually exclusive outcomes A and B, if P(A) + P(B) < 1.0, you can buy both and profit by 1 - P(A) - P(B) per dollar invested.
🔍 Get arbitrage opportunities emailed to you weekly
Free newsletter — no spam, unsubscribe anytime.
The Math: Calculating Your Arbitrage Return
Let's work a clean example.
📊 Georgia Senate Seat Example
"Democrats win Georgia Senate seat" → YES at 44¢
"Republicans win Georgia Senate seat" → YES at 52¢
Total cost: 44 + 52 = 96¢ (4¢ opportunity)
Sizing for equal payouts:
Democratic YES: cost = X × 0.44
Republican YES: cost = X × 0.52
Total: X × 0.96 → 4.17% guaranteed return
If market resolves in 30 days: 50%+ annualized return
Risk Management in Arbitrage
Arbitrage sounds risk-free, but real-world execution has genuine risks:
⚠ Resolution Risk
Risk: Polymarket can resolve markets in unexpected ways. If your "arbitrage" sides are in different markets that resolve using different criteria, you could lose on both legs.
Mitigation: Read resolution rules for both markets carefully before trading. Ensure they are genuinely mutually exclusive and exhaustive.
⚠ Liquidity Risk
Risk: You may be able to buy one side at the quoted price but not the other. By the time you execute the second leg, the price has moved.
Mitigation: Use limit orders when possible. Only trade markets with sufficient volume (>$10K in liquidity is a rough minimum). The PolyMath scanner shows estimated slippage.
⚠ Timing Risk
Risk: Markets can stay mispriced longer than expected, tying up capital. Or they can reprice before you can execute.
Mitigation: Only pursue arbitrage in markets with clear, near-term resolution dates. Capital locked in a 6-month arbitrage at 2% isn't worth it.
⚠ Platform Risk
Risk: Polymarket is crypto-native. Smart contract bugs, oracle failures, or platform downtime are non-zero risks.
Mitigation: Size positions appropriately. Arbitrage positions should be part of a diversified approach, not your entire bankroll.
Why Not Everyone Does This
If arbitrage is risk-free profit, why doesn't everyone do it until the opportunities disappear? They do — but slowly. Prediction market arbitrage has natural friction:
Capital requirements: Meaningful returns require meaningful capital. A 3% arb on $100 is $3. On $10,000, it's $300.
Monitoring cost: Opportunities are fleeting. You need to actively scan for them. Manual scanning is time-consuming.
Execution speed: By the time you notice, calculate, and execute, the price may have moved. Professional arbitrageurs use automated systems.
Opportunity cost: Your capital is locked until resolution. You can't redeploy it for other trades.
Most "arbitrages" aren't: What looks like a free lunch often isn't, once you've read the fine print on resolution criteria.
The PolyMath Arbitrage Scanner handles the monitoring and calculation. But the capital, execution speed, and judgment calls on resolution criteria are still on you.
Practical Tips for Getting Started
✓ Start with large, well-known markets
Election markets, Fed rate decision markets, and major sporting events have clear resolution criteria and high liquidity. Avoid niche markets where resolution rules are ambiguous.
✓ Paper trade first
Identify what you think are arbitrage opportunities and track what happens without real money. This teaches you where the hidden risks are.
✓ Use the scanner during news cycles
The best opportunities emerge when news hits and markets lag in repricing. Check the scanner during high-volume news cycles.
✓ Keep position sizes moderate
Don't bet your entire bankroll on a single "guaranteed" arbitrage. Guaranteed is a strong word.
✓ Factor in bid-ask spreads
Polymarket has no explicit trading fees, but bid-ask spreads are a real cost. For thin markets, the spread can eat your entire arbitrage profit.
Getting Started on Polymarket
If you don't have a Polymarket account yet, sign up here. The platform accepts USDC deposits via most major crypto on-ramps.
Once you're set up, bookmark the PolyMath Arbitrage Scanner and the Kelly Criterion Calculator — you'll use both constantly. The arbitrage scanner finds the opportunity; Kelly tells you exactly how much to bet given your confidence in the risk-free nature of the trade.
Summary
Prediction market arbitrage is real, recurring, and profitable — but not free. The best opportunities are:
1. Cross-Platform
Same event priced differently on Polymarket vs Kalshi.
2. Correlated Markets
Logically linked markets on Polymarket priced inconsistently.
3. Temporal
Fast-moving news creates lag between related markets.
The PolyMath Arbitrage Scanner automates the detection and sizing calculation for Type 2 opportunities within Polymarket. It won't make you rich overnight, but consistent, systematic execution of well-structured arbitrage trades is one of the most defensible edges available to retail traders.
🔍 Explore More PolyMath Tools
Do the math. Read the resolution criteria. Execute fast. That's the arbitrage playbook. Subscribe to the PolyMath newsletter for new strategy guides and high-EV market opportunities delivered to your inbox.