In this guide
The primary reason skilled forecasters struggle in prediction markets is not inaccurate forecasts — it's inadequate capital preservation. An astute probability assessment becomes worthless if a prolonged losing run depletes your entire stake. This guide outlines the approach that safeguards against this outcome.
The Kelly Criterion: The Mathematical Foundation
Kelly Criterion determines the theoretically ideal proportion of your stake to allocate to each trade: f = (bp - q) / b
- b = net odds received (e.g., if YES costs 0.40, b = 1.5)
- p = your probability estimate
- q = 1 - p
- Result: optimal fraction of bankroll for this position
In practice: use half-Kelly. Whilst Kelly delivers mathematical optimality under certainty, our probability assessments carry inherent uncertainty, making half-Kelly the superior choice for risk-adjusted performance.
Hard Rules: Never Break These
- Maximum 5% of bankroll per single position — no exceptions regardless of conviction
- Maximum 25% of bankroll in any single correlated cluster — e.g., all US election markets
- Stop-loss: if you lose 25% of your starting bankroll in a month, stop trading for the rest of the month
- Never add to a losing position to "average down" — reevaluate the fundamental thesis first
Drawdown Recovery
Inevitable downturns occur even when you possess genuine edge. Following a 20% drawdown, cut your position sizes in half until you return to your previous peak. This approach prevents a difficult run from spiralling into ruin.
FAQ
- How much starting capital do I need for serious prediction market trading?
- $500-1,000 furnishes sufficient capital to build adequate diversification across 10-20 positions using half-Kelly sizing. Below $100, position constraints undermine your capacity to execute systematic approaches effectively.
- What should I do after a winning streak?
- Adopt greater caution, not complacency. Winning streaks breed excessive confidence. Maintain your systematic sizing discipline irrespective of recent results.