Prediction Market Arbitrage: How It Works, After Fees

Arbitrage means buying something in one place and selling the same thing for more somewhere else, so you profit whatever happens. On prediction markets it usually means buying every outcome of an event for less than $1 in total, since exactly one outcome will pay $1.

On a single exchange, and between a regulated U.S. exchange and a crypto-based one, these gaps do appear. Most are too small to survive fees, and the ones that look big often aren't the same bet. Here's how it works and what to check.

Three kinds of arbitrage

1. YES and NO on one market. If you can buy YES and NO on the same contract for less than $1 combined, you lock in the difference. On a single market this almost never happens at prices you can actually trade, because the two sides come from the same order book.
2. Every outcome of one event. Some events have several mutually exclusive outcomes: which candidate wins, which range a number lands in. If the YES prices of all outcomes add up to less than $1, buying one of each pays $1 whatever happens. If they add up to more than $1, the opposite trade can work.
3. The same question on two exchanges. Different exchanges often list similar questions. If YES is 45¢ on one and NO is 52¢ on the other, buying both costs 97¢ for a pair that pays $1.

A worked example, after fees

Take the third case with 100 contracts: YES at 45¢ on a regulated U.S. exchange and NO at 52¢ on a crypto-based one.

The second exchange charges its own taker fee on many markets, so the real profit is smaller again, and that's before the cost of moving money between a dollar account and a crypto wallet. A 3¢ gap became roughly 1¢ or less. That's typical: gaps of a few cents are mostly eaten by fees. Run your own numbers with the fee calculator.

The risks that aren't on the price

Settlement rules. This is the big one. Two markets with the same headline can settle on different sources, different times or different definitions. If one exchange says YES and the other says NO for the same real-world event, you lose on both legs. Always read both rule pages word for word.

Execution. Prices move. If one side fills and the other doesn't, you're no longer hedged; you simply hold a position. Thin order books make this worse, and the size available at the quoted price is often small.

Money tied up. Your money sits in both legs until the market settles, which can be months. A 1% return locked up for six months is a poor use of capital.

Moving money. A regulated U.S. exchange holds dollars; a crypto-based exchange often holds a dollar stablecoin. Transfers between them take time and can cost fees, and which exchanges you can use depends on where you live.

Account limits. Large or frequent cross-exchange trading can hit position limits or trigger reviews. Read each exchange's terms.

When it's worth doing

• The gap is wide enough to pay both exchanges' fees with room to spare.
• You've read both settlement rules and they truly match.
• There's enough size at your price on both sides, and you can enter both quickly.
• The market settles soon, so your money isn't locked up for long.

That combination is rare, which is why the traders who do this well tend to use software to scan for it and act in seconds.

Arbitrage and prop firm accounts

A prop firm evaluation is a different game. At Lenium you trade one simulated account on live market prices, and positions in the same event must stay in one account, so cross-exchange arbitrage isn't part of it. What an evaluation tests is the skill behind most real profits: pricing events better than the market. Read how to make money on prediction markets or see Lenium's rules.