// Pillar Four

THE EDGE

Prediction markets price the future. When two platforms disagree on the same outcome, you lock in the spread. No opinions. No forecasting. Just arbitrage.

Learn the Strategy See the Platforms
// The Concept

WHAT IS PREDICTION MARKET ARBITRAGE?

Prediction markets let you trade on real-world outcomes — elections, Fed rate decisions, sports results, weather events, even whether a CEO gets fired. Platforms like Polymarket, Kalshi, and PredictIt each price these outcomes independently.

Here's the play: when Platform A prices "Yes" at 42 cents and Platform B prices "No" at 53 cents, you've found an arb. Buy both sides for 95 cents total, and one of them pays out $1.00. That's 5 cents of guaranteed profit per dollar risked — regardless of what actually happens.

This is the same mathematical principle behind matched betting. You're not predicting anything. You're exploiting the gap between two prices. The outcome doesn't matter — you profit either way.

HOW PREDICTION ARB WORKS

The same math, different markets. Four steps to lock in the spread.

01
Scan the Markets
Monitor the same event across multiple prediction platforms. Look for prices that don't add up to 100% when combined.
02
Identify the Gap
When "Yes" on Platform A + "No" on Platform B totals less than $1.00, you've found an arb opportunity.
03
Place Both Sides
Buy "Yes" on the cheaper platform. Buy "No" on the other. You're now covered on every outcome.
04
Collect the Spread
When the event resolves, one position pays out $1.00. Subtract your total cost. The difference is your guaranteed profit.
// Show Me the Numbers

A REAL EXAMPLE

Let's say two platforms are pricing the question: "Will the Fed cut rates in June 2026?"

FED RATE CUT — JUNE 2026

Polymarket "Yes" price $0.42
Kalshi "No" price $0.53
Total cost for both sides $0.95
Guaranteed payout (one side wins) $1.00
Guaranteed profit per contract +$0.05
Arb Spread = $1.00 - (Yes Price + No Price)
Arb Spread = $1.00 - ($0.42 + $0.53)
Arb Spread = $1.00 - $0.95 = +$0.05 per contract (5.26% return)

At 100 contracts: $95 invested → $100 payout = +$5.00 guaranteed
At 1,000 contracts: $950 invested → $1,000 payout = +$50.00 guaranteed

The returns per trade are smaller than matched betting, but the volume scales differently. Prediction markets run 24/7, new events resolve constantly, and the capital requirements are flexible. Stack enough small edges and it compounds.

SUPPORTED PLATFORMS

You need accounts on at least two platforms to arb. Each has different markets, fee structures, and regulatory status. Here's the breakdown.

Polymarket
Crypto-based prediction market. Largest volume, deepest liquidity. Politics, crypto, world events. No fee on winning trades.
Best liquidity · Crypto deposit
Kalshi
CFTC-regulated US exchange. Real USD deposits. Economics, politics, weather, company events. 10% fee on profits (varies).
US-regulated · USD deposit
PredictIt
Political prediction market. $850 max per contract limit. 10% fee on profits, 5% fee on withdrawals. Limited markets.
Political markets · Position limits
Robinhood (Event Contracts)
Prediction contracts on select events via Robinhood's platform. Limited markets but zero-fee trading and easy USD funding.
Zero fees · Limited events
Betfair Exchange
World's largest betting exchange. Back and lay on sports, politics, specials. 2-5% commission. Not available in all US states.
Massive market range · Exchange model
Sportsbooks (Cross-arb)
DraftKings, FanDuel, and others sometimes offer prediction-style prop bets that can be arbed against dedicated prediction platforms.
Cross-platform arbs · Welcome bonus stacking
// Finding the Arbs

WHERE THE GAPS APPEAR

Arb opportunities show up most often when:

Breaking news hits. One platform adjusts prices faster than another. The lag creates a window.
Low-liquidity markets. Smaller markets on one platform may be mispriced relative to a deep market elsewhere.
Binary vs. multi-outcome events. When a multi-outcome market (like "Which party wins?") doesn't sum to 100% across platforms.
Different fee structures. Platform A's zero-fee structure can make a slim arb viable when Platform B charges 10%.
Cross-asset arbs. A prediction market question about a sports result may be priced differently than the same prop on a sportsbook.

BEFORE YOU START

Prediction arb is lower risk than traditional trading, but it's not zero risk. Understand these factors.

💰

FEES EAT INTO MARGINS

Kalshi charges ~10% on profits, PredictIt charges 10% on profits + 5% on withdrawals. Always factor fees into your arb calculation before placing trades.

CAPITAL LOCKUP

Your money is tied up until the event resolves. A market that resolves in 6 months locks your capital for 6 months, even if the per-contract return is solid.

📉

RESOLUTION RISK

Platforms occasionally resolve ambiguous outcomes differently. If Platform A says "Yes" and Platform B says "No" on the same event, both sides could lose. Stick to clear-cut binary questions.

🏛

REGULATORY DIFFERENCES

Kalshi is CFTC-regulated. Polymarket uses crypto and may not be available everywhere. PredictIt has position limits. Know the rules before you fund accounts.

// Go Deeper

PREDICTION ARB GUIDES

Step-by-step walkthroughs for every strategy. More guides published weekly.

FAQ

Yes. Prediction markets are legal in the US. Kalshi is regulated by the CFTC. Polymarket operates using cryptocurrency. PredictIt runs under a CFTC no-action letter. Trading on these platforms and exploiting price differences between them is not illegal — it's just trading.
You can start with as little as $50 on most platforms. Prediction market contracts typically trade between $0.01 and $0.99 per share. However, since arb margins are thin (2-8%), you'll need more capital to generate meaningful dollar amounts. $500-$2,000 across two platforms is a reasonable starting point.
Prediction market exchanges generally don't ban arbitrageurs — unlike sportsbooks. These are open markets where you're just another buyer. Arb activity actually provides liquidity, which platforms want. That said, PredictIt has position limits ($850 per contract) which naturally cap your exposure.
Raw arb spreads typically range from 2-8% before fees. After accounting for platform fees (Kalshi's ~10% profit fee, PredictIt's 10% + 5%), net returns are usually 1-5%. The trade-off is that this is risk-free capital allocation, not speculative. Compare it to a savings account, not to day trading.
Right now, manual scanning works — check the same event on Polymarket and Kalshi, add up the cheapest Yes and No prices. If the total is under $1.00, you have an arb. We're building automated scanning tools to surface these faster. Sign up for alerts to get notified when we launch.
The math is identical — you're covering both sides of a binary outcome. The difference is the venue. Matched betting uses sportsbook free bets; prediction arb uses price discrepancies between prediction markets. Matched betting has higher per-trade profit (because the free bet is the edge), while prediction arb has ongoing opportunity (because price discrepancies recur).
// Get Started

THE MARKET IS
MISPRICED.

Get alerts when we find prediction market arbs, new guides drop, and tools launch. Free forever.