What Is Prediction Market Arbitrage?
Prediction market arbitrage is the practice of simultaneously taking opposing positions across two or more markets or platforms to lock in a guaranteed profit when the combined implied probabilities of all outcomes sum to less than 100% — creating a mathematical edge that requires no forecasting skill to exploit.
In a perfectly efficient market, if you price every possible outcome of an event and add them up, you get exactly 100% (minus the house margin). When inefficiencies appear — because platforms have different liquidity, different user bases, or slow price discovery — the total drops below 100%, and that gap becomes extractable profit.
This isn't theoretical. During the 2024 U.S. presidential election, Kalshi's YES prices on Trump winning and Polymarket's NO prices on Harris winning were briefly misaligned for over 40 minutes, creating a window where a trader holding both sides would have collected approximately 4–6 cents on the dollar regardless of the outcome. At scale, that's meaningful alpha.
The Three Types of Prediction Market Arbitrage
Not all arbitrage is the same. Understanding the three primary structures helps you match opportunities to your capital, speed, and risk tolerance before committing to a trade.
1. Cross-Platform Arbitrage
The most common form: the same event is priced differently on Kalshi vs. Polymarket (or any two platforms). You buy YES on one and NO on the other, and if your combined cost is below $1.00, you profit at resolution regardless of outcome.
- Example: Fed holds rates — Kalshi YES at $0.58, Polymarket NO at $0.39. Combined cost: $0.97. Guaranteed profit: $0.03 per contract.
- Execution risk: Withdrawal delays, platform limits, and fee structures can eat the margin. Always account for trading fees on both legs before entering.
- Best for: High-liquidity political and macro events where both platforms are actively priced.
2. Cross-Market Correlation Arbitrage
This is subtler. Two logically linked markets are mispriced relative to each other — not perfectly opposing, but statistically correlated. For example: a market on "Fed cuts rates in Q3" and a separate market on "10-year Treasury yield below 4% by September" should move together. When they don't, one side is wrong.
- Example: If the rate-cut market prices a 70% chance of a cut, but the Treasury yield market implies only 45%, there's a divergence you can fade.
- Execution risk: These aren't riskless. You're expressing a view on the relationship between markets, not locking in a guaranteed payout. Size accordingly.
- Best for: Traders with macro knowledge who can identify when markets are genuinely inconsistent versus just pricing different scenarios.
3. Temporal Arbitrage (Pre-Event Mispricing)
Prices on the same platform fluctuate as new information arrives — but sometimes they overreact or underreact, creating short-lived mispricings against the objective probability. This is less "pure" arbitrage and more disciplined mean reversion, but the execution logic is similar: identify when the market has moved further than the information warrants, and fade it.
- Example: A March Madness upset causes the eventual-champion market to crater 15 points — but the actual impact on the favorite's path to the title is minimal. The overreaction is your entry.
- Best for: Fast-moving event markets during sports seasons or breaking news cycles.
How to Find Arbitrage Opportunities: A Practical Framework
The manual approach — refreshing both platforms and mentally adding probabilities — doesn't scale. Here's a systematic process that serious traders use:
- Build a market map: List every event traded on both Kalshi and Polymarket simultaneously. Focus on high-volume events (Fed meetings, major elections, key economic releases) where both platforms maintain active order books.
- Calculate implied probability totals: For YES/NO binary markets, add the best-ask YES price on Platform A to the best-ask NO price on Platform B. Any total below $1.00 (minus fees) is a candidate.
- Apply the fee filter: Kalshi charges roughly 2% on winnings; Polymarket fees vary by market. A $0.97 combined price becomes $0.992 after fees — not necessarily worth executing. Your gross margin needs to be at least 2–3 cents above all-in fees.
- Check liquidity depth: A great price on 10 contracts means nothing if the order book won't fill 100. Always verify you can execute both legs at the displayed price before committing.
- Execute simultaneously or in rapid sequence: The biggest execution risk is leg risk — one side fills, then the opportunity closes before the other side does. Prioritize the less-liquid leg first.
The Real Risks Arbitrage Guides Don't Tell You
True riskless arbitrage is rarer than most content suggests. Here are the genuine risks that kill otherwise sound setups:
- Counterparty and platform risk: Both Kalshi and Polymarket are relatively young platforms. CFTC-regulated exchanges like Kalshi offer more legal protection, but capital locked on any single platform carries non-zero platform risk.
- Resolution disputes: If two platforms resolve the same event differently (it happens more than you'd expect with ambiguously worded contracts), your "hedged" position becomes an unhedged loss on the wrong side.
- Capital efficiency: Arbitrage ties up capital on both legs simultaneously. A 3% edge on a position requiring $200 of locked capital earns $6 — often less attractive than a well-sized directional trade with good Kelly sizing.
- Withdrawal timing: Winning your Polymarket position doesn't help if your Kalshi withdrawal takes 3–5 business days and you need the capital for the next opportunity.
This is why arbitrage works best as part of a broader portfolio approach rather than a standalone strategy. Pairing it with solid prediction market risk management principles — including position limits and bankroll discipline — is non-negotiable.
Sizing Arbitrage Positions: Modified Kelly for Near-Certain Bets
Even "guaranteed" arbs should be sized thoughtfully. The standard Kelly Criterion framework produces very large recommended sizes for high-probability bets — which is mathematically correct but practically dangerous when residual risks (platform, resolution, execution) remain.
A common adjustment: use 25–50% Kelly on arbitrage positions to account for unmodeled tail risks. If full Kelly says bet 40% of bankroll, size to 10–20% instead. You give up some expected value but dramatically reduce the risk of a catastrophic single-event loss from a resolution dispute or platform issue.
When Arbitrage Opportunities Actually Appear
Timing matters. Mispricings cluster around specific market conditions:
- Breaking news windows: The 10–30 minutes after a major announcement (jobs report, Fed statement, election result) before prices equilibrate across platforms.
- Low-volume periods: Weekend and late-evening sessions have thinner order books and slower price discovery — more inefficiency, but also harder to fill large positions.
- New market listings: Freshly listed markets on one platform before the other platform lists the same event often show significant divergence.
- Sports event resolution: During live sporting events, in-game markets move faster than cross-platform equilibration can keep pace with.
Frequently Asked Questions
Is prediction market arbitrage legal?
Yes, prediction market arbitrage is entirely legal on regulated platforms. Kalshi operates under CFTC oversight as a designated contract market, making it fully legal for U.S. residents. Polymarket has geographic restrictions for U.S. users, so check your eligibility before trading on that platform.
How much money do I need to start arbitrage trading in prediction markets?
You need enough capital to fund both legs of the trade simultaneously — typically $200–$500 minimum to make the per-trade profit meaningful after fees. Most serious arbitrage traders operate with $2,000–$10,000 to run multiple positions at once and compensate for the inherent capital inefficiency of locking funds on two platforms.
How often do real arbitrage opportunities appear in prediction markets?
Genuine cross-platform arbitrage opportunities that survive fee calculation appear several times per week on major event markets, but they close quickly — often within minutes. During high-activity periods like elections, Fed meetings, or March Madness, the frequency increases significantly but so does competition from automated systems scanning for the same gaps.
Is Kalshi or Polymarket better for arbitrage?
Kalshi is the better choice for U.S. traders who need regulatory safety and reliable fund access, because its CFTC-regulated status means clearer contract resolution rules and withdrawal protections. Polymarket offers deeper liquidity on some markets, but U.S. geographic restrictions limit its utility as an arbitrage leg for most American traders. See our Kalshi vs. Polymarket comparison for a full breakdown.
The Bottom Line: Arbitrage Is Real, But It Requires Infrastructure
Prediction market arbitrage is one of the most intellectually honest edges available to retail traders — it doesn't require you to know more than the market, just to act faster and more systematically than the average participant. The math is real. The profits are real. But so are the execution friction, the capital requirements, and the tail risks that most beginner guides gloss over.
The traders who extract consistent value from arbitrage aren't doing it manually. They're using tools that monitor multiple markets simultaneously, flag divergences in real time, and track position-level performance across platforms. Analytics platforms like Prevayo are built specifically for this kind of systematic edge-finding — giving prediction market traders the infrastructure that was previously only available to professional quant desks.
If you're serious about building a repeatable arbitrage process, start with a clearly defined target: one market, two platforms, documented entry and exit criteria. Master the basics before scaling. The edge is there — the question is whether your execution is disciplined enough to capture it.