Kelly criterion calculator
Enter your win probability and your payoff ratio (average win ÷ average loss) to get the Kelly fraction — the share of capital that maximises long-run growth — plus the half-Kelly most traders actually use, and a plain-English read on whether you even have an edge.
What is the Kelly criterion?
The Kelly criterion is the bet size that maximises the long-run geometric growth of your capital, given a known edge. Risk more than Kelly and volatility drags your compound growth down; risk less and you grow more slowly than you could. The full-Kelly fraction is:
f* = W − (1 − W) ÷ b
where W is your win probability (as a fraction) and b is the payoff ratio — your average win divided by your average loss. If f* comes out at or below zero, you have no edge, and the growth-maximising bet is nothing at all.
Why most people bet FRACTIONAL Kelly
Full Kelly is mathematically optimal but wildly volatile — it courts drawdowns most traders can't stomach, and it assumes your edge estimate is exact. It never is. A small overestimate of your win rate or payoff pushes full Kelly into over-betting, and over-betting compounds toward ruin. That's why half-Kelly or quarter-Kelly is standard: half-Kelly keeps roughly three-quarters of the growth for a fraction of the swings, and forgives an optimistic edge estimate.
How to read the result
- f* ≤ 0 — no edge; the growth-maximising bet is zero.
- 0–10% — a thin edge; size cautiously and prefer fractional Kelly.
- > 10% — a meaningful edge; still, half-Kelly is the practical choice.
Kelly, risk of ruin, and backtests
Here's the trap: your win rate and payoff ratio come from a backtest. If that backtest is overfit, your real edge is smaller than measured — so a Kelly fraction sized on the inflated numbers is really over-betting, and full Kelly on an overfit edge is a fast road to ruin. This is why sizing and honesty about your edge are the same problem.
QUANTHEON Lab attacks both ends. Its 1,000-path Monte-Carlo simulation estimates your risk of ruin directly by reshuffling the trade sequence, and the Deflated-Sharpe haircut checks whether the edge is even real before you stake capital on it. Pair this with a hard cap: size each trade so a single loss can't do real damage.
FAQ
What is the Kelly criterion?
Kelly gives the fraction of capital to risk on a bet with a known edge to maximise long-run compound growth: f* = W − (1 − W) ÷ b, where W is win probability as a fraction and b is the payoff ratio (average win ÷ average loss).
Why do traders use fractional Kelly?
Full Kelly is extremely volatile and assumes your edge estimate is exact. A small overestimate can push you toward ruin, so most traders use half- or quarter-Kelly for far smoother equity with only a little less growth.
Can Kelly cause ruin?
Yes — if your win rate and payoff come from an overfit backtest, your real edge is smaller than measured, and full Kelly on an inflated edge over-bets toward ruin. Estimate risk of ruin directly and haircut the edge before sizing.
Related: Position size calculator · Monte-Carlo simulation · Adaptive position sizing