What Is Prediction Market Arbitrage Between Kalshi and Polymarket?

Prediction market arbitrage between Kalshi and Polymarket is the practice of buying one side of an event contract on one platform while simultaneously taking the opposite side on the other, when the combined cost of YES on platform A and NO on platform B is meaningfully below $1.00 net of fees. Because YES and NO must sum to $1.00 at resolution, any combined entry price below $1.00 (after fees and slippage) is a locked-in profit regardless of how the event resolves.

The opportunity exists because Kalshi and Polymarket have structurally different user bases, liquidity profiles, and pricing models — so the same event routinely trades at different probabilities on each platform.

The Mechanics

A simplified worked example, using realistic price patterns:

• Polymarket lists "Will Event X happen?" with YES trading at $0.42 and NO at $0.58
• Kalshi lists the same (or functionally identical) event with YES at $0.51 and NO at $0.49
• An arbitrageur buys YES on Polymarket at $0.42 and NO on Kalshi at $0.49
• Combined entry cost: $0.91 per contract pair
• At resolution, exactly one side pays $1.00 and the other pays $0
• Gross profit: $0.09 per contract, or roughly 9.9% on capital deployed

That gross has to absorb taker fees, withdrawal fees, capital costs, and any slippage from sweeping the order book. Net spreads of 2–5% on liquid major events have been widely reported in 2026 — smaller than the gross looks but still attractive relative to most market-neutral strategies.

Why the Spreads Exist

Three structural reasons, none of them going away quickly:

1. Different User Bases

Kalshi is CFTC-regulated and serves primarily US users with conventional brokerage onboarding. Polymarket is offshore, crypto-funded, and serves a global audience with a heavier crypto-native skew. The two pools have different beliefs, different information access, and different reaction speeds to news — so prices on the same question diverge.

2. Different Pricing Models

Polymarket runs a central limit order book with a maker-taker fee model. Kalshi also operates an order book but with different fee structures, contract sizing, and tick increments. The mechanical differences alone produce small structural spreads even before any informational divergence.

3. Capital Frictions

Moving capital between the two platforms is non-trivial. Kalshi requires US bank rails; Polymarket requires USDC on a supported chain. Most retail participants don't bother bridging, so capital doesn't equalise prices the way it would in a frictionless market. Those friction costs are exactly what create the persistent spreads.

What It Looks Like in Practice

The arbitrage is real, but it's not free money. Practitioners deal with:

• Resolution risk. "Functionally identical" markets sometimes resolve differently because of subtle wording differences in contract specs. A market that resolves YES on one platform and NO on the other turns the trade into a directional bet at full risk.
• Capital tied up until resolution. Most prediction market contracts don't allow early exit at the entry price — capital is locked until the event resolves, which can be days, weeks, or months.
• Liquidity-adjusted spreads. Quoted top-of-book prices look attractive; sweeping size against the actual depth often closes most of the gap.
• Fees. Maker rebates help; taker fees on both sides eat into thin spreads.
• Regulatory and platform risk. Geo-restriction changes, account closures, or platform-level disputes (Polymarket has been blocked in several jurisdictions; Kalshi has faced state-level actions) can strand positions.

Why It Matters for the Industry

For B2B operators and infrastructure providers, the existence of cross-platform arbitrage matters for two reasons.

First, arbitrage activity is a leading indicator of market health. Healthy spreads and active arbitrage mean the prediction market layer is functioning as a price discovery mechanism rather than a closed gambling product. That's the case Kalshi and Polymarket need to make to regulators.

Second, the same data that powers arbitrage is the data operators want. Cross-platform price feeds, depth-of-book data, and resolution metadata are exactly what a sportsbook or hybrid operator needs to integrate prediction market contracts into their own product surface — whether as a direct feed, a hedge venue, or a competitive benchmark for their own pricing.

The arbitrageurs are doing the work of stitching the prediction market layer together. Operators who build on top of that layer benefit from a more efficient, more liquid, and more credible market.

---

Last verified: April 2026