Cross-market arbitrage in prediction markets is the practice of simultaneously buying and selling the same event contract on different platforms — such as Kalshi and Polymarket — to lock in a guaranteed profit when those platforms price the same outcome at different probabilities, typically exploiting discrepancies of 2–10 percentage points before transaction costs.
TL;DR — Key Takeaways
- Arbitrage profit exists when the same binary outcome trades at different prices across two or more platforms simultaneously.
- A viable arb requires the combined cost of YES + NO across two platforms to be less than $1.00 (or 100 cents).
- Kalshi and Polymarket frequently show 3–8% price gaps on major events like Fed rate decisions and election outcomes.
- Execution speed matters: arb windows on high-volume markets can close within minutes of opening.
- Fees, withdrawal timing, and counterparty risk are the three most common hidden costs that erode arb profits.
- Low-volume markets offer larger price gaps but carry meaningful liquidity risk that can prevent full position execution.
What Is Cross-Market Arbitrage in Prediction Markets?
Arbitrage is one of the oldest concepts in finance — buy low in one place, sell high in another, pocket the difference. In prediction markets, the "asset" is a probability: you're trading on whether a specific event will occur. When two platforms disagree about that probability, a profit opportunity emerges.
Here's the simplest version: Suppose Kalshi prices a Fed rate cut in June 2026 at 42 cents (42% probability), while Polymarket prices the same event at 51 cents (51% probability). You can buy NO on Polymarket for 49 cents and YES on Kalshi for 42 cents. Your total cost is 91 cents. If the event resolves either way — rate cut or no cut — one of your contracts pays $1.00. That's a 9-cent guaranteed profit, or roughly a 9.9% return on your outlay, assuming frictionless execution.
This is the core mechanic. The rest of this guide is about finding these gaps, executing them profitably, and managing the real-world friction that erodes the returns.
Why Do Price Discrepancies Exist Between Platforms?
If markets were perfectly efficient, gaps like the one above would be arbitraged away instantly. But prediction markets are still maturing, and several structural factors keep inefficiencies alive:
- Different liquidity pools: Kalshi and Polymarket draw different user bases with different information sets and risk tolerances. A retail-heavy Polymarket market and an institutional-leaning Kalshi market can diverge meaningfully.
- Regulatory fragmentation: CFTC-regulated platforms like Kalshi operate under different frameworks than offshore or crypto-native platforms, limiting capital flow between them and slowing arbitrage correction.
- Delayed information propagation: Breaking news updates one platform's order book before the other has time to react. The window is short but real.
- Low market volume: Thin markets have wider bid-ask spreads and slower price discovery. As of May 2026, some markets are seeing significantly below-average daily event volume, which paradoxically creates more arb opportunities — and more liquidity risk simultaneously.
The Arbitrage Viability Formula
Before placing any trade, you need to quickly calculate whether a gap is actually profitable after fees. The core check is simple:
Arb Profit = $1.00 − (Cost of YES on Platform A + Cost of NO on Platform B)
If this number is positive before fees, you have a candidate. Then apply the fee haircut:
Net Profit = Arb Profit − (Fee on Platform A × Position Size) − (Fee on Platform B × Position Size)
Kalshi charges a fee on winnings (typically 2–7% depending on market type). Polymarket's fee structure varies by market. For a $100 position with a 5-cent gross arb and 2% fees on each side, you're looking at roughly $5 gross − $4 fees = $1 net. Barely worth it — which is why only gaps above 5–8 cents typically survive fees and remain worth trading.
For a deeper look at how position sizing interacts with these returns, see our guide on Kelly Criterion position sizing in prediction markets — the same framework applies to scaling arb trades optimally.
Step-by-Step Execution: How to Run a Cross-Market Arb
Step 1: Identify the Candidate Market
Focus on markets that exist on both Kalshi and Polymarket with identical or near-identical resolution criteria. Federal Reserve rate decisions, major election outcomes, and economic indicator releases (CPI, unemployment) are the most common overlap. Verify the resolution rules match — a seemingly identical market can resolve differently if the contract wording differs.
Step 2: Check the Spread in Real Time
Price gaps are dynamic. Build a simple spreadsheet or use a monitoring tool that tracks the mid-price on both platforms for your target markets. A gap that looked profitable 10 minutes ago may already be closed. You're looking for: Platform A YES price + Platform B NO price < $0.93 (leaving room for fees).
Step 3: Calculate Maximum Position Size
Your position is capped by the available liquidity on both platforms. If Kalshi only has $200 of YES available at your target price, you can't arb more than $200 — and trying to fill a larger order will move the price against you. Check the order book depth, not just the last traded price.
Step 4: Execute Both Legs Simultaneously
This is the hardest part. The longer the gap between your first and second trade, the more price risk you carry. In fast-moving markets — think: a major political announcement, a surprise Fed statement — both legs need to fill within seconds or the arb collapses. For high-stakes executions, have both platform tabs open and execute in rapid sequence.
Step 5: Account for Withdrawal Timing
Profits aren't real until they're in your bank account. Kalshi and Polymarket have different withdrawal timelines and potential holds on new accounts. Factor this into your capital allocation — money locked in a winning position you can't withdraw is an opportunity cost.
Real Market Example: 2026 Fed Rate Decision Arbitrage
In early May 2026, ahead of the Federal Reserve's May meeting, Kalshi's "Fed cuts rates in May" market briefly showed a YES price of 18 cents while Polymarket showed 26 cents. The corresponding NO prices were 82 cents and 74 cents. Running the formula: buying YES on Kalshi (18¢) + NO on Polymarket (74¢) = 92¢ total cost against a $1.00 guaranteed payout — an 8-cent gross margin. After estimated fees of ~4 cents combined, net profit was approximately 4 cents per dollar of position, or 4.3% return in under 30 days. Annualized, that's substantial — but the window was open for less than 90 minutes before both platforms converged.
The Hidden Risks Most Beginners Miss
Arbitrage sounds risk-free by definition, but prediction market arb carries real risks that pure financial arbitrage doesn't:
- Resolution ambiguity: If the two platforms resolve the same event differently due to contract wording differences, you can lose on both legs.
- Platform insolvency risk: Prediction markets are still relatively young. Capital held on a platform is subject to counterparty risk.
- Partial fill risk: You get filled on one leg but not the other, leaving you with a directional position you didn't intend.
- Correlated market drift: If the underlying event probability shifts dramatically between your two fills, one leg's market price moves against you before you complete the trade.
Managing these risks is part of any serious prediction market portfolio strategy — arb positions should be sized conservatively as a percentage of total capital precisely because execution risk is non-trivial.
When Arbitrage Conditions Are Best
Arb opportunities concentrate around specific conditions. Based on platform data patterns, the most productive windows are:
- Immediately after major news breaks — one platform's users react faster than the other's
- Low-volume periods — fewer active traders means slower price correction, though liquidity is also thinner
- New markets opening — initial pricing is often inconsistent across platforms before the market finds equilibrium
- Evening trading windows — reduced professional trading activity can leave gaps open longer
Academic research on prediction market efficiency, including work published via the National Bureau of Economic Research on information aggregation in prediction markets, consistently finds that price discovery is fastest in high-volume markets and slowest in low-liquidity or newly-opened contracts — exactly where arb hunters should focus attention.
Building an Arbitrage Monitoring System
Manual scanning is time-consuming and error-prone. A basic monitoring setup involves:
- A list of 10–20 markets that exist on both Kalshi and Polymarket simultaneously
- A spreadsheet that auto-refreshes prices from each platform's API or public market data
- A threshold alert when the combined cost of YES + NO falls below your fee-adjusted target (e.g., 92 cents)
- A log of every attempted trade with outcome — this data is invaluable for identifying which market categories generate the most reliable gaps
For traders who want to go beyond manual systems, dynamic position sizing frameworks can be adapted to automatically scale arb positions based on gap size, liquidity depth, and remaining time to resolution.
Frequently Asked Questions
Is prediction market arbitrage actually risk-free?
No — while the mathematical structure of arbitrage eliminates directional price risk, prediction market arb carries execution risk (partial fills), resolution risk (differing contract terms), and counterparty risk (platform solvency). True risk-free arbitrage requires perfect simultaneous execution and identical contract terms across both platforms.
What is the minimum viable price gap for cross-market arbitrage?
Most experienced traders require a gross spread of at least 6–8 cents ($0.06–$0.08 per dollar of position) before fees to ensure a net-positive trade after Kalshi and Polymarket fee structures are applied. Gaps below 5 cents rarely survive execution costs.
How do I find arbitrage opportunities on Kalshi and Polymarket?
Start by identifying markets with identical or near-identical resolution criteria on both platforms. Then compare mid-prices manually or via a simple spreadsheet connected to each platform's public API. Focus on Fed rate decisions, major elections, and economic indicator releases — these have the highest market overlap.
How much capital do I need to start arbitrage trading in prediction markets?
You need sufficient capital on both platforms simultaneously. A practical starting point is $200–$500 per platform ($400–$1,000 total) to capture meaningful position sizes while staying within individual market liquidity constraints. Spreading across multiple arb opportunities also requires more capital than running a single position.
Can I automate prediction market arbitrage?
Kalshi and Polymarket both offer API access that advanced traders use to monitor prices and execute trades programmatically. Full automation requires API credentials, coding ability, and careful handling of rate limits and order execution logic. Most retail traders start with semi-automated monitoring (automated alerts, manual execution) before building fully automated systems.
What types of events produce the most arbitrage opportunities?
Federal Reserve rate decisions, presidential and congressional elections, and major economic releases (CPI, jobs reports) generate the most consistent overlap between Kalshi and Polymarket. Sports markets occasionally show gaps, particularly around major events like March Madness or the Super Bowl, though resolution criteria differences are more common in sports contracts.
The Bottom Line on Prediction Market Arbitrage
Cross-market arbitrage is one of the most intellectually honest strategies in prediction markets — you're not trying to predict the future, you're exploiting the market's failure to agree on the present. The gaps are real, the profits are achievable, and the framework is learnable.
But it rewards preparation over impulse. The traders who do this consistently aren't clicking around hoping to get lucky — they have monitoring systems, pre-calculated fee thresholds, and execution discipline built in advance of the opportunity. The gap waits for no one.
If you want to track cross-platform price discrepancies, analyze market timing patterns, and size your arbitrage positions optimally, tools like Prevayo are built exactly for this — giving prediction market traders the analytics infrastructure to find and act on opportunities before they close.