Cross-Platform Arbitrage: Polymarket vs Kalshi vs PredictIt
Cross-platform arbitrage exploits price differences for the same event across different prediction markets. When Polymarket prices an event at 60% YES and Kalshi prices it at 45% YES, you can buy YES on Kalshi and NO on Polymarket to lock in a spread. This guide covers how to find these opportunities, the risks involved, and why this strategy still works in 2026.
How Cross-Platform Arbitrage Works
Different prediction markets attract different traders with different information and biases. This causes the same event to be priced differently across platforms. When the price gap is large enough to cover fees on both platforms, you have an arbitrage opportunity.
You buy the underpriced side on one platform and the opposite side on another. No matter the outcome, one position wins and the other loses, but the combined profit is positive.
Platform Comparison
| Platform | Currency | Fees | Max Position | Availability |
|---|---|---|---|---|
| Polymarket | USDC (crypto) | ~1.5% taker | No limit | Global (not US) |
| Kalshi | USD | Variable (up to 3%) | $25,000-$100,000 | US only (CFTC regulated) |
| PredictIt | USD | 10% on profit + 5% withdrawal | $850 per market | US only (limited) |
Step-by-Step Example
Suppose 'Will the Fed raise rates in June?' is priced at YES = $0.40 on Polymarket and YES = $0.55 on Kalshi. This means NO on Kalshi is effectively $0.45.
- Buy YES on Polymarket at $0.40 (100 shares = $40)
- Buy NO on Kalshi at $0.45 (100 shares = $45)
- Total cost: $85 for one guaranteed $100 payout
- Gross profit: $15 (17.6% return)
- After fees (~$1.50 Polymarket + ~$1.35 Kalshi): ~$12.15 net profit
Where to Find Opportunities
The best cross-platform arbitrage opportunities appear in political markets (elections, policy decisions), where Polymarket and Kalshi have the most overlapping markets. Crypto-related markets and economic data releases are also good hunting grounds.
Tools like EventArb.com and ArbCalculator.cc scan multiple platforms in real-time and highlight price discrepancies. You can also build your own scanner using each platform's public API.
Risks Specific to Cross-Platform Arbitrage
- Resolution differences: Platforms may define the same event differently. 'Will X happen by December?' might use different time zones or definitions of 'happen'
- Settlement timing: One platform may resolve days before the other, leaving capital locked
- Currency risk: Polymarket uses USDC while Kalshi uses USD. USDC depegging (extremely rare) would affect one leg
- Regulatory risk: Kalshi markets can be suspended by the CFTC, leaving one leg stranded
- Capital requirements: You need funded accounts on multiple platforms simultaneously
Why This Strategy Still Works
Unlike within-platform arbitrage (which bots close in seconds), cross-platform arbitrage persists longer because no single bot can trade on all platforms simultaneously. Each platform has different onboarding, KYC, and capital requirements, creating natural friction that keeps spreads open.
Frequently Asked Questions
Which platforms are best for cross-platform arbitrage?
Polymarket and Kalshi have the most overlapping markets and are the primary pair for cross-platform arb. PredictIt has $850 position limits that make it less useful for scaling.
How much capital do I need?
You need funded accounts on at least two platforms. Minimum $1,000-$2,000 per platform to make meaningful returns. Serious arb traders keep $10,000+ on each platform.
How long do cross-platform arbitrage opportunities last?
Much longer than within-platform arb. Spreads of 3-10% can persist for hours or even days, especially in less liquid markets. This makes it more accessible for retail traders.
Is this legal?
Yes. Arbitraging across platforms is completely legal. However, you must comply with each platform's terms of service and applicable regulations in your jurisdiction.