What is arbitrage in prediction markets, and when is it actually possible?

Detailed Explanation

1. Classic binary relationship: In an ideal frictionless market, Yes price + No price ≈ $1.
2. Arb opportunity: If Yes + No < $1, you can buy both and lock profit at settlement (in theory).
3. Why it’s hard: You must execute both legs at favorable prices and account for fees.

Common Scenarios

• Same event listed across two venues with different prices
• Mispriced complementary outcomes (Yes/No, candidate A/B)
• Range markets where the sum of ranges doesn’t equal 100%
• Temporary dislocations during breaking news

Exceptions & Edge Cases

• If you can’t short or there’s no “No” leg, then full arb may be impossible.
• If position limits cap your size, then the opportunity is small.
• If settlement rules differ across venues, then it’s not the same event (hidden risk).

Practical Examples

• You can buy Yes at $0.48 and buy No at $0.49 (combined $0.97)
  • Gross locked value at settlement = $1
  • Gross profit = $0.03/share
  • After fees/slippage maybe net $0.01 or negative

Actionable Takeaways

• ✅ Check combined pricing (Yes + No, or all ranges sum)
• ✅ Include realistic execution (spread + slippage + fees)
• ✅ Confirm identical settlement definitions across venues
• ✅ Move fast—true arbs close quickly

‍