How do I calculate expected value (EV) for a trade in a prediction market?

Detailed Explanation

1. Define your probability (p): your independent estimate the event happens.
2. Define the payoff: binary “Yes” pays $1 if yes, $0 if no.
3. Compute EV per share:
   • EV = p×1 + (1−p)×0 − price − fees
   • EV = p − price − fees
4. Why it matters: EV forces discipline—price might be “low,” but not low enough.

Common Scenarios

• You have a model forecast (macro, sports, politics)
• You’re reacting to breaking news and re-pricing outcomes
• You’re comparing multiple markets to allocate capital
• You’re checking whether a hedge is “worth it”

Exceptions & Edge Cases

• If payout isn’t binary, then EV needs the full payoff distribution.
• If settlement timing matters and capital is tied up, then you should consider opportunity cost.
• If there’s a chance of “void” or alternative resolution, then EV must include those branches.

Practical Examples

• You estimate p = 0.58. Market price is 0.52. Fees are 0.01 per share equivalent.
  • EV = 0.58 − 0.52 − 0.01 = 0.05 (5 cents per share)
  • For 1,000 shares → expected profit $50 (in expectation)

Actionable Takeaways

• ✅ Write down your p (and why)
• ✅ Subtract realistic fees + spread impact
• ✅ Stress-test p (what if you’re wrong by 5–10 pts?)
• ✅ Avoid trades where EV is smaller than execution costs

‍