Cross-Platform Arbitrage: Sportsbook vs Prediction Market
Cross-platform arbitrage locks in 5-10% risk-free returns when sportsbooks and prediction markets disagree on price. 3 worked examples with fee-adjusted math.
What Cross-Platform Arbitrage Is and Why It Exists
Cross-platform arbitrage is the practice of betting both sides of an event across different platform types to lock in a guaranteed profit. You take one side at a sportsbook and the opposite side on a prediction market. If the prices disagree by enough, you profit regardless of the outcome.
This works because sportsbooks and prediction markets price events using fundamentally different mechanisms. Sportsbooks set odds based on internal models and liability management. Prediction markets discover prices through open order books. These two systems converge over time but frequently disagree in the short term, especially on political events, cultural outcomes, and anything that falls outside traditional sports.
The disagreements are not random. Sportsbooks are slow to adjust on events outside their core competency. Prediction markets can be thin on certain contract types. These structural differences create windows where the combined implied probability across platforms drops below 100%. That gap is your arb.
The Core Math Behind Arb Detection
An arb exists when the sum of implied probabilities across both sides is less than 100%.
Take a sportsbook offering a political candidate at +200 (implied 33.3%). On Kalshi, the same candidate's "No" contract trades at $0.58 (implied 58% the candidate loses). Add them up: 33.3% + 58% = 91.3%. That is below 100%, which means money left on the table.
The arb percentage is 100% - 91.3% = 8.7%. That is the theoretical profit margin before fees and execution costs.
Use the odds converter to get both sides into the same format. Sportsbooks display American odds. Prediction markets display contract prices. You need implied probabilities from both to compare.
Worked Example: Allocating a $1,000 Arb
Here is the full calculation on a real setup.
Setup:
- Sportsbook: Candidate Yes at +200 (decimal 3.00, implied 33.3%)
- Kalshi: Candidate No at $0.58 (implied 58.0%)
- Combined implied: 91.3%
- Total capital: $1,000
Optimal allocation formula:
The goal is to equalize profit across both outcomes. Let S = sportsbook stake and K = Kalshi stake. Total budget: S + K = $1,000.
If Yes wins: S x 3.00 must equal K / 0.58 If No wins: K / 0.58 must equal S x 3.00
Solving: S = $370, K = $630.
Outcome if Yes wins:
- Sportsbook pays: $370 x 3.00 = $1,110
- Kalshi position lost: -$630
- Net profit: $1,110 - $1,000 = $110 (11.0%)
Outcome if No wins:
- Kalshi pays: $630 / $0.58 = $1,086.21
- Sportsbook bet lost: -$370
- Net profit: $1,086.21 - $1,000 = $86.21 (8.6%)
Guaranteed minimum profit: $86.21 on $1,000. That is 8.6% risk-free. The arbitrage calculator handles this allocation math instantly for any odds combination.
How Fees Destroy Thin Arbs
Theory meets reality at the fee stage. Kalshi charges approximately 7% on net profits. Polymarket charges zero fees but has spread costs. Sportsbooks charge no explicit fee, but vig is built into the odds.
Take the example above. The $86.21 Kalshi profit gets hit with a 7% fee: $86.21 x 0.07 = $6.03. Your net Kalshi profit drops to $80.18. The sportsbook outcome is unaffected because you already accounted for the odds.
Now your guaranteed minimum is $80.18 instead of $86.21. Still profitable, but the margin shrank by 7%.
Here is where it gets dangerous. On a thinner arb:
Thin arb example:
- Sportsbook: +130 (implied 43.5%)
- Kalshi No: $0.60 (implied 60%)
- Combined implied: 103.5%
Wait. That sums to over 100%. No arb exists. You need the combined implied probability under 100%.
Marginal arb example:
- Sportsbook: +150 (implied 40%)
- Kalshi No: $0.57 (implied 57%)
- Combined implied: 97%
- Theoretical arb: 3%
After Kalshi's 7% fee on profits, this 3% raw arb shrinks to roughly 1.5%. Factor in slippage and execution timing, and it is breakeven or worse. Run the numbers through the fee calculator before committing capital.
The threshold: you generally need a raw arb of 4% or more for it to survive fees, slippage, and the opportunity cost of locked capital. Below 4%, the margin is too thin.
Slippage and Execution Risk
Sportsbooks fill instantly at the displayed price. Prediction markets do not. The order book can shift between the time you check the price and the time you execute.
A 5% arb can vanish in seconds on a fast-moving market. If you place the sportsbook leg first and the prediction market price moves before you fill, you are stuck with a one-sided position. That is not an arb. That is a naked bet.
Three rules for clean execution:
1. Execute the less liquid side first. That is almost always the prediction market. Place your limit order on Kalshi or Polymarket, wait for a fill, then immediately execute the sportsbook leg.
2. Only arb markets with sufficient depth. Check the order book. If there is not enough liquidity at your target price to fill your full size, the arb does not exist at your intended scale.
3. Account for settlement timing. Prediction market contracts can take days or weeks to settle. Your capital is locked until resolution. A 5% arb that locks capital for 3 months is equivalent to a 20% annualized return, which is strong. A 5% arb that locks capital for 12 months is a 5% return, which is mediocre for the effort involved. Read bankroll turnover and why it matters for the math on capital efficiency.
Where to Find Cross-Platform Arbs
The most common arb opportunities appear when:
Sportsbooks underprice political events. Traditional books added political betting recently and their models are less refined than their sports models. Prediction markets, where political junkies concentrate, often have sharper political pricing.
Prediction markets overprice long shots. The favorite-longshot bias exists in prediction markets too. A candidate trading at $0.05 on Polymarket might be worth $0.02. If a sportsbook has the same candidate at +4000 (implied 2.4%), you can arb the difference.
Breaking news creates temporary dislocations. When news drops, one platform reacts faster than the other. Sportsbooks are slow to adjust lines on non-sports events. Prediction markets reprice within minutes. The window is short but real.
Multileg markets create structural arbs. When the same multi-outcome market exists on both Polymarket and Kalshi with different overrounds, buying cheap contracts on one and selling expensive ones on the other can lock in profit. The arbitrage calculator handles the allocation for these setups.
For a full comparison of how sportsbooks and prediction markets differ in pricing mechanics, read sportsbooks vs prediction markets. Understanding why prices disagree helps you find arbs faster.
Middling: Arbitrage's Higher-Variance Cousin
Pure arbs guarantee profit but they are rare and close fast. Middle bets offer a related approach with a different risk profile. Instead of locking in guaranteed profit, you bet both sides at different lines to create a window where both bets win.
Example cross-platform middle:
- Sportsbook: Candidate wins by more than 5 points at +130
- Kalshi: Candidate wins by 4-7 points at $0.28
If the candidate wins by 6 or 7, both positions cash. Every other result, you lose one side but win the other for a small net loss. The EV depends on the probability of landing in the window.
Why consider middling alongside pure arbs? Three reasons:
- More frequent. Line disagreements large enough for middles appear 3-5x more often than pure arbs.
- Less account heat. Sportsbooks flag arb patterns aggressively. Middle bets look like normal spread bets individually.
- Cross-platform gaps are wider. Sportsbooks and prediction markets reprice at different speeds, creating bigger windows than sportsbook-to-sportsbook comparisons.
The tradeoff is variance. Middles lose small most of the time and win big occasionally. Over a season of 100+ plays, the math works if you are disciplined about only taking +EV setups. Run the numbers through the EV calculator before committing.
Risk Factors That Kill Arb Profits
Cross-platform arbitrage is not free money. Several risks can turn a theoretical profit into an actual loss.
Account limits. Sportsbooks limit or ban profitable bettors. If you are consistently arbing, your sportsbook account will get restricted. This is not a question of if, but when.
Settlement disputes. Different platforms may settle the same event differently based on their rules. A contract that settles "Yes" on one platform and "No" on another due to rule interpretation differences is a worst-case scenario.
Counterparty risk. Prediction market platforms can freeze withdrawals, change fee structures, or face regulatory action. Kalshi's CFTC regulation provides some protection. Crypto-based platforms provide less. For a deeper look at platform differences, see the Kalshi vs Polymarket comparison.
Opportunity cost. Capital locked in a long-dated arb cannot be deployed elsewhere. A 6% arb on an event that settles in 6 months ties up capital that could earn more through active trading. Always calculate the annualized return, not just the raw percentage.
The math works. The execution is where most traders fail. Start small, track every trade, and only scale up when your process is reliable.
Frequently asked questions
- What is cross-platform arbitrage in betting?
- Cross-platform arbitrage means betting opposite sides of the same event on different platform types, such as a sportsbook and a prediction market, to guarantee profit regardless of the outcome. It works when the combined implied probabilities across platforms sum to less than 100%.
- How much profit can you make from cross-platform arbs?
- Typical raw arb margins range from 3-10% before fees. After platform fees and slippage, realistic guaranteed returns are 2-7% per trade. Annualized returns depend heavily on how quickly contracts settle and how fast you can recycle capital.
- Do prediction market fees eliminate arbitrage profits?
- They can. Kalshi's 7% winner fee eats directly into arb margins. A raw arb below 4% is usually breakeven or negative after fees. Always calculate fee-adjusted returns before committing capital. Polymarket's zero-fee structure preserves more of the arb margin.
- Is cross-platform arbitrage legal?
- Arbitrage itself is legal. However, sportsbooks can limit or close accounts of bettors who consistently arb. Prediction market access depends on your jurisdiction and the specific platform's regulatory status. Kalshi is CFTC-regulated and legal for US residents. Polymarket's US access is restricted.
- What is the biggest risk in cross-platform arbitrage?
- Execution risk. If you place one leg and the other side's price moves before you can execute, you are left with a one-sided bet instead of a hedged position. Always execute the less liquid leg first and have the sportsbook ready to fire immediately after.
- What is the difference between a cross-platform arb and a middle bet?
- An arb guarantees profit regardless of outcome. A middle bet creates a window where both sides win but accepts a small loss when the result falls outside that window. Middles are more common and draw less sportsbook attention, but have higher variance. Read the full breakdown in our middle bet guide.