Advanced Guide to Event Arbitrage

What Event Arbitrage Actually Means on Kalshi and Polymarket

Event arbitrage is the practice of exploiting price discrepancies between correlated or identical contracts across prediction markets so that your combined position profits regardless of outcome, before fees and slippage eat the spread. On Kalshi and Polymarket, this usually shows up as the same event priced differently on each platform, or as a chain of related contracts (a market and its complement, or a market and a correlated proxy) that don't sum to the values they should. This guide assumes you already understand the basics of how these markets settle; if you don't, read How Kalshi Works first, since arbitrage strategies collapse fast if you misunderstand settlement mechanics, fee schedules, or contract expiration rules.

The opportunities are real but shrinking. Retail traders discovered cross-platform arb in 2024 and 2025, and spreads that used to sit at 4-6 cents now close in minutes. What's left requires speed, correct pricing math, and a filter for execution risk that most traders skip.

Cross-Platform Arbitrage Between Kalshi and Polymarket

The cleanest form of event arbitrage happens when Kalshi and Polymarket both list contracts on the same underlying event — a Fed rate decision, an election outcome, a Super Bowl winner — and the implied probabilities diverge enough to cover both platforms' fees. Kalshi charges a per-contract trading fee that scales with price and quantity; Polymarket's cost structure runs through gas (minimal on Polygon) and the bid-ask spread on its order book. Because the fee structures differ, a 3-cent gap that looks profitable on paper can be a 1-cent gap after you account for Kalshi's fee curve, which is steepest near 50 cents.

Before you commit capital, map the exact contract language on both sides. "Will the Fed cut rates by 25bps or more at the July meeting" on one platform is not always equivalent to "Will the target rate be reduced in July" on another — resolution sources, rounding rules, and cutoff timestamps can differ just enough to break the arbitrage thesis. If you're deciding where to route size in the first place, Kalshi vs Polymarket 2026 covers the structural differences in liquidity depth and settlement speed that determine which platform absorbs your order without moving the price against you.

Building a Correlated-Markets Arbitrage Guide for Related Contracts

Not every arbitrage opportunity requires two platforms. Within a single market, you'll find families of contracts that are mathematically linked — a candidate's win probability, their vote-share bracket contracts, and a head-to-head matchup contract should all be internally consistent. When they're not, you have a same-platform arbitrage.

The standard example: a multi-outcome election market where the sum of all "yes" prices across mutually exclusive outcomes should equal roughly 100 cents (adjusted for the platform's vig). If the sum prices out to 94 cents, you can buy all outcomes and collect the difference at settlement, minus fees. If it prices to 108 cents, market makers or informed traders are paying a premium for optionality you can sell into.

This only works if the outcomes are genuinely mutually exclusive and exhaustive. Traders lose money treating "will Team A win" and "will Team A make the playoffs" as complementary when they're correlated but not exclusive — a subtle distinction that a structured pillar-by-pillar review catches and a quick glance at the order book does not.

Timing Windows: When Arbitrage Spreads Actually Open

Spreads open at predictable moments: right after major news breaks and one platform's market makers reprice faster than the other's, during low-liquidity overnight hours when one side of a cross-platform pair goes stale, and immediately after a market first lists when initial pricing is thin. The spread you're chasing at 2pm on a Tuesday during normal volume is almost never worth the fees; the spread that opens for 90 seconds after a surprise data release is where the volume actually sits.

This is also where most manual arbitrage attempts fail — not on the math, but on latency. By the time you've confirmed a discrepancy by eye, checked both order books, and sized your position, informed flow has already closed the gap. Systematic scanning across both venues, refreshed on a short interval, is the only way to consistently catch these windows rather than the stale ones that show up on delayed dashboards.

Risk Factors That Break the Arbitrage Thesis

Four failure modes account for most losses that traders mistakenly attribute to "bad luck":

• Resolution mismatch — contracts that look identical resolve on different sources, different cutoff times, or different rounding conventions, so your "arbitrage" is actually a directional bet on a technicality.
• Execution slippage — one leg fills at your quoted price, the other moves before you can complete the position, turning a locked spread into open directional exposure.
• Platform-specific settlement risk — withdrawal delays, disputed resolutions, or platform-level freezes can strand capital on one leg while the other settles normally.
• Fee erosion — thin spreads that clear on a spreadsheet often don't clear once you model Kalshi's fee curve and Polymarket's effective spread cost at your actual order size.

Position sizing should scale down, not up, as apparent spread size increases — a suspiciously wide gap is more often a data or resolution error than free money.

How PillarLab AI Fits Into This

Manually scanning Kalshi and Polymarket for pricing discrepancies across hundreds of active markets isn't a workflow a human can sustain during the windows that matter. PillarLab AI runs continuous real-time data ingestion from both platforms and applies a structured 9-pillar analysis to every market it evaluates — covering liquidity depth, resolution-source verification, historical base rates, cross-platform price correlation, news-catalyst timing, order-book imbalance, settlement risk, contract-language matching, and position-sizing guidance. That combination is what turns raw price differences into vetted opportunities instead of false positives caused by mismatched contract terms.

Because the engine checks resolution language and settlement sources as part of the pillar framework, it filters out the mismatched-contract trap that catches manual traders — the single biggest reason apparent arbitrage spreads turn into directional losses. The edge-detection layer flags cross-platform divergences as they open rather than after informed flow has already closed them, and it does so across the full market list simultaneously rather than the handful of contracts a trader can watch by hand. For anyone treating event arbitrage as a repeatable process rather than an occasional lucky find, that continuous, structured scan is the difference between a strategy and a hobby.

Position Sizing and Execution: A Practical Arbitrage Guide

Once you've confirmed a genuine spread — matching resolution language, adequate liquidity on both legs, fees modeled at your actual size — execution order matters. Enter the less liquid leg first. If you're pairing a deep Polymarket order book against a thinner Kalshi contract, fill the Kalshi side first so you're not left holding one leg while the other repriced during your delay.

Cap position size to what both order books can absorb without moving the spread you're trying to capture; a 6-cent gap on paper that requires walking through three price levels to fill your full size might net out closer to 2 cents in practice. Reading order-book depth correctly, not just the top-of-book quote, is a skill worth building before you scale size — see How to Read Prediction Market Odds for the mechanics of converting displayed prices into true fill-weighted cost.

Track your realized spread against your modeled spread after every trade. Consistent underperformance versus your model usually means your fee assumptions or slippage estimates are wrong, not that the market is inefficient in your favor.

Frequently Asked Questions

Is event arbitrage between Kalshi and Polymarket legal?

Yes. Trading on both regulated platforms is legal in permitted jurisdictions; arbitrage itself is a standard trading strategy, not a rules violation, provided you follow each platform's terms of service.

How much capital do you need to start event arbitrage?

Enough to clear both platforms' minimum order sizes and absorb fees on both legs — typically a few hundred dollars minimum, though thin spreads need larger size to generate meaningful returns.

Why do arbitrage spreads disappear so quickly?

Increased retail awareness and automated scanning tools compress spreads within minutes of opening, especially on high-volume markets like elections or major sports events.

Can PillarLab AI execute trades automatically?

PillarLab AI focuses on structured analysis and edge detection across Kalshi and Polymarket; you review flagged opportunities and execute manually on each platform.

What's the biggest mistake new arbitrage traders make?

Assuming contracts with similar wording resolve identically. Mismatched resolution sources or cutoff times turn a supposed arbitrage into an unhedged directional bet.

Start free with Start free with 10 credits and let the 9-pillar engine flag mismatched pricing across Kalshi and Polymarket before the spread closes. For platform selection guidance beyond arbitrage, see Best Prediction Market 2026, and if your focus extends to sports-specific markets, review Best AI for Sports Betting for a broader comparison of analytical tools.