How to Find Prediction Market Arbitrage Across Polymarket and Kalshi
Cross-platform price discrepancies are real, documented, and worth real money. Here's how they work, why they exist, and how to spot them before they close.
Academic researchers have documented over $40 million in arbitrage profits extracted from prediction markets between April 2024 and April 2025. Automated bots have turned hundreds of dollars into six-figure returns trading nothing but pricing inefficiencies. Wall Street quant firms like DRW, Susquehanna, and Jump Trading are building dedicated prediction market desks specifically to capture these opportunities.
And yet, the vast majority of retail prediction market traders are still trading on a single platform, completely blind to the price discrepancies sitting right in front of them.
This guide explains how cross-platform prediction market arbitrage works, why persistent pricing gaps exist between Polymarket and Kalshi, what it takes to find and evaluate these opportunities, and the tools that give traders a real edge.
What Is Prediction Market Arbitrage?
Prediction market arbitrage exploits the fact that the same real-world event is often priced differently across different platforms. When Polymarket and Kalshi both list a contract on the same outcome — say, whether the Fed will cut rates at its next meeting — their prices frequently diverge.
The core principle is straightforward. In any binary prediction market, a "Yes" contract and a "No" contract should sum to $1.00. If you can buy "Yes" on one platform and "No" on another for a combined cost of less than $1.00, you've locked in a guaranteed profit regardless of how the event resolves. One of your positions will always pay $1.00.
For example: if Kalshi prices "Yes" at $0.08 and Polymarket prices "No" at $0.087, your combined cost is $0.167 for a guaranteed $1.00 payout. That's 83.3 cents of profit per contract pair — a 500% ROI before fees.
Of course, opportunities this clean are rare. More commonly, cross-platform spreads are measured in single-digit percentage points. But even smaller discrepancies, traded consistently with the right position sizing, compound into meaningful returns over time.
Why Price Discrepancies Exist Between Platforms
If markets were perfectly efficient, the same event would trade at identical prices everywhere. But prediction markets in 2026 are far from perfectly efficient. Several structural factors create persistent pricing gaps between Polymarket and Kalshi.
Different Trader Populations
Polymarket's user base is predominantly crypto-native. These traders bring different information sources, risk tolerances, and behavioral patterns than Kalshi's base, which skews toward traditional finance users, sports bettors, and traders who discovered prediction markets through Robinhood. When two populations with different priors and information diets are pricing the same event independently, their consensus prices will frequently diverge.
Liquidity Asymmetries
Polymarket has significantly deeper liquidity in political and macroeconomic markets. Kalshi often has deeper liquidity in sports and in U.S.-specific economic indicators that require CFTC approval to list. When one platform has thin order books for a given market, prices are more susceptible to being pushed by individual trades and less reflective of true consensus probability.
Market Creation Timing
Polymarket can spin up new markets within hours of breaking news. Kalshi's listing process involves regulatory review and takes longer. When a market exists on one platform but not yet the other, early pricing can be highly inefficient. And when the same market finally appears on both platforms, initial prices often disagree as each platform's order book independently finds equilibrium.
Fee Structure Differences
Kalshi and Polymarket charge fees differently, which means the "fair value" price differs by platform. On Kalshi, fees are calculated using a formula tied to the contract price and peak around 40–60 cents. On Polymarket's global platform, fees have historically been near-zero for most markets. These fee differences mean that rational traders on each platform will price contracts slightly differently even if they have identical probability estimates.
Information Propagation Speed
When news breaks, prices on one platform often update faster than the other. Research has documented that Polymarket generally leads price discovery due to higher liquidity, but Kalshi sometimes lags by minutes during fast-moving events. These windows — where one platform has repriced but the other hasn't caught up — create some of the most exploitable short-term discrepancies.
Types of Prediction Market Arbitrage
Not all arbitrage opportunities are the same. Understanding the different types helps you evaluate which ones are worth pursuing.
Cross-Platform Arbitrage
This is the most accessible form for retail traders. You identify the same event listed on both Polymarket and Kalshi, buy "Yes" on whichever platform has the lower price, and buy "No" on whichever platform has the lower price for that side. If the combined cost is less than $1.00 after accounting for fees on both platforms, you have a positive-expectation trade.
The key variables are the current spread between platforms, the fee structure on each side, the available liquidity at the quoted prices, and the time until the market resolves (which determines how long your capital is tied up).
Intra-Market Rebalancing
Sometimes the "Yes" and "No" prices within a single platform's market don't sum to exactly $1.00. In low-liquidity markets, you can occasionally buy both sides for less than $1.00 on the same platform. This is simpler to execute since everything happens on one venue, but opportunities tend to be smaller and shorter-lived.
Combinatorial Arbitrage
This is the most sophisticated form. Prediction markets frequently list multiple related markets about the same underlying event. For example, a platform might have separate markets for "Will the Democrats win the presidential election?" and "Who will be the Democratic nominee?" Logical dependencies between these markets can create pricing inconsistencies that are harder to spot but potentially more profitable.
Academic research analyzing Polymarket data identified both rebalancing and combinatorial arbitrage as significant sources of extracted profit, with the most successful wallets earning millions through systematic combinatorial strategies.
The Math Behind Cross-Platform Spreads
Before chasing any spread, you need to understand the numbers. A visible price difference between platforms is not the same as a profitable opportunity once fees and execution realities are factored in.
Break-Even Calculation
For a cross-platform arbitrage to be profitable, the combined cost of both positions must be less than $1.00 minus total fees on both platforms.
Start with the "Yes" price on the cheaper platform and the "No" price on the other. Add the fees you'll pay on each side. If the total — both contract costs plus both sets of fees — is still under $1.00, the difference is your profit per contract pair.
Kalshi's fee formula produces fees that peak around 1.5 to 2 cents per contract at mid-range prices (40–60 cents) and decline toward zero at extreme prices. Polymarket's fee structure varies by market type and the specific contract, but has historically been minimal on the global platform.
As a practical rule of thumb, cross-platform spreads below about 4 to 5 percentage points often don't justify the execution complexity once fees, slippage, and capital lockup are considered. Spreads above that threshold start becoming interesting, and spreads above 8 to 10 points are uncommon but can represent substantial opportunities when they appear.
Capital Lockup Considerations
Unlike arbitrage in equities or crypto — where positions can be entered and exited in seconds — prediction market arbitrage ties up your capital until the market resolves. A contract that doesn't resolve for three months means your capital is locked for three months. A 5% gross return on a 90-day lockup annualizes to roughly 20%, which is attractive. The same 5% on a 365-day lockup is just 5%.
Smart traders factor resolution date into every opportunity assessment. Short-duration markets (2 to 8 weeks) with meaningful spreads are the sweet spot for capital-efficient arbitrage.
Liquidity and Slippage
The quoted price is only valid at the quoted size. If one side of the trade has thin liquidity, you may not be able to fill at the price you're evaluating. Partial fills are common, and attempting to fill a large order on a thin book can move the price against you, erasing the spread entirely.
Always check order book depth on both platforms before committing. An opportunity that looks like an 8-point spread at the top of the book might only be a 3-point spread at the size you want to trade.
How to Find Cross-Platform Arbitrage Opportunities
Finding these opportunities requires solving two problems simultaneously: identifying which markets exist on both platforms, and comparing their prices in real time. Neither is trivial when done manually.
The Manual Approach (and Why It Doesn't Scale)
In theory, you could open both Polymarket and Kalshi side by side, search for the same event on each, and manually compare prices. In practice, this falls apart quickly. Market titles and descriptions differ between platforms. A market called "Will the Fed cut rates in March?" on Kalshi might be phrased as "Fed decision in March?" on Polymarket. Manually matching equivalent markets across thousands of listings is tedious and error-prone.
Even once you've identified a matched pair, you need to check prices on both platforms simultaneously, calculate the spread net of fees, assess order book depth on both sides, and decide whether the opportunity justifies the capital lockup. By the time you've done this manually for one market, the spread may have already closed.
As one arbitrage guide noted, manual identification of profitable opportunities has become essentially impossible in 2026's competitive environment. The windows are too short and the calculation overhead too high.
Using a Terminal for Cross-Platform Scanning
This is where a cross-platform prediction market terminal provides transformational value. Instead of manually hunting for matched markets across separate browser tabs, a terminal aggregates both platforms into a single interface and does the matching, price comparison, and spread calculation for you.
skreenr, for example, automatically matches equivalent markets across Polymarket and Kalshi and presents over 1,000 matched pairs in a unified Compare view. For each matched market, you see both platforms' current prices, the spread, and which platform offers the cheaper "Yes" and "No."
The Arbitrage Scanner goes further, specifically identifying opportunities where the combined cost of both sides falls below $1.00. Each opportunity is displayed with the ROI, profit per contract, total cost, volume on both platforms, and the spread — sortable by whichever metric matters to your strategy. Clicking into any opportunity shows the full strategy: which platform to buy each side on, the position calculator for modeling different stake sizes, a probability-over-time chart with both platforms overlaid, and side-by-side order books so you can verify liquidity before acting.
This kind of systematic scanning across hundreds of matched markets in real time is what separates traders who consistently capture cross-platform value from those who stumble into one-off opportunities by accident.
Risk Factors That Every Arbitrageur Must Understand
Cross-platform prediction market arbitrage is often described as "risk-free." It isn't. Understanding the real risks is essential to trading profitably over time.
Settlement Risk
Different platforms may resolve the same event differently if their resolution criteria diverge. One platform might use AP's election call, another might use official certification. Edge cases — recounts, disputed results, ambiguous outcomes — can cause one platform to settle differently than the other, turning a "risk-free" position into a loss on one leg.
Before entering any cross-platform position, verify that both markets use equivalent or compatible resolution criteria.
Execution Risk
Non-atomic arbitrage — where you fill one side before the other — introduces the risk that the second leg moves against you during execution. You buy "Yes" on Kalshi at 8 cents, then switch to Polymarket to buy "No," and the price has already moved from 8.7 cents to 15 cents. Your "arbitrage" is now a directional bet.
Professional arbitrageurs mitigate this by pre-funding both platforms and using limit orders on both sides simultaneously. Even so, fast-moving markets can slip.
Platform Risk
Your capital is custodied on the platform where you hold each position. If a platform experiences an outage, regulatory action, or operational issue before settlement, your ability to realize the arbitrage is impaired. Spreading capital across two platforms inherently doubles your platform exposure.
Capital Opportunity Cost
Money locked in a low-spread arbitrage for months could potentially earn more in a higher-conviction directional trade. The "risk-free" label can lure traders into deploying too much capital in low-return positions, reducing their overall portfolio performance.
Building a Sustainable Cross-Platform Strategy
The traders who profit consistently from cross-platform prediction market trading aren't chasing individual arbitrage opportunities. They're building a systematic process.
Start with a tool that gives you continuous visibility into cross-platform pricing. Use a terminal like skreenr to monitor matched markets, track spreads over time, and receive alerts when spreads widen beyond your threshold. Develop a personal framework for evaluating opportunities: minimum spread after fees, maximum resolution time, minimum order book depth on both sides, and maximum capital allocation per position.
Track every trade. Record entry prices, fees paid, resolution dates, and actual P&L. Over time, you'll develop an understanding of which market categories (politics, sports, crypto, macro) produce the most reliable cross-platform spreads, which time windows offer the widest divergences (hint: immediately after major news events), and which resolution timeframes give you the best risk-adjusted returns.
The prediction market industry is growing at an extraordinary pace — $63.5 billion in volume in 2025, weekly volumes exceeding $5 billion in early 2026, and institutional firms now building dedicated desks around these exact strategies. The opportunity set is expanding, not contracting. But the traders who capture it will be the ones with the best tools, the most systematic processes, and the clearest view across platforms.
Start Scanning for Cross-Platform Opportunities
skreenr matches markets across Polymarket and Kalshi in real time, calculates spreads automatically, and ranks arbitrage opportunities by ROI — so you can focus on evaluating and executing, not searching and calculating.
skreenr is an information and analytics platform for prediction market traders. We do not execute trades, hold funds, or provide financial advice. Cross-platform trading involves risk, including settlement differences, execution slippage, and capital lockup. Trade only with capital you can afford to lose. Please review our Terms of Service and Privacy Policy before using the platform.