Kelly Criterion Calculator
The mathematically optimal fraction of your account to risk per trade.
How it works
Kelly answers one question exactly: what fraction of capital maximises long-term growth given an edge. Below it you grow more slowly than you could, above it you grow more slowly than you could and eventually go broke — the curve falls off a cliff on the right.
Almost nobody trades full Kelly. It assumes your win rate and payoff are known precisely, which they never are, and it produces drawdowns most people cannot sit through. Half Kelly gives roughly three quarters of the growth with far less pain, which is why it is the practical default.
Related calculators
How the Kelly Criterion Calculator works
The Kelly criterion answers one question precisely: given an edge, what fraction of your capital maximises long-term growth? This calculator applies it to trading, taking a win rate and a reward-to-risk ratio and returning the optimal risk per trade — plus the safer fractional sizings most traders should actually use.
The shape of the growth curve
Plot long-term growth against position size and you get a curve that rises, peaks at the Kelly fraction, and then falls — steeply. Betting less than Kelly means slower growth than you could have had. Betting more means slower growth as well, and past roughly twice Kelly, expected growth turns negative even with a genuine edge.
That asymmetry is the whole practical lesson. Undersizing costs you some upside; oversizing eventually costs you the account. When you are uncertain — and you always are — err low.
Why nobody trades full Kelly
Kelly assumes your win rate and payoff are known exactly. In trading they are estimates from a limited sample, and they drift as market conditions change. Overestimate your win rate by a few points and full Kelly quietly becomes an over-bet, which is the one error the curve punishes hardest.
Full Kelly also produces drawdowns that are correct in theory and unbearable in practice — 50% peak-to-trough is entirely normal. Half Kelly captures around three quarters of the growth with roughly half the volatility, which is why it is the practical default. Quarter Kelly is common among traders whose edge estimates are rough. Check what your chosen size implies with the risk of ruin simulator.
Reading the expectancy figure
Expectancy is quoted in R — units of risk. An expectancy of +0.35R means that over many trades you make 0.35 times your risk amount per trade on average. Risk 1% per trade with that expectancy and you gain about 0.35% of the account per trade before costs.
If expectancy is zero or negative, Kelly returns zero and the correct position size is no position. No sizing scheme rescues a strategy without an edge — it only changes how quickly the account erodes.
FAQ
What is a reward-to-risk ratio?expand_more
How far your target sits from entry compared with your stop. A ratio of 2 means the target is twice the stop distance, so a winner earns twice what a loser costs. Combined with win rate it fully determines expectancy: you can be profitable at a 35% win rate with a ratio of 3, or unprofitable at 60% with a ratio of 0.5.
Why is my Kelly fraction so large?expand_more
Usually because the win rate is optimistic. Kelly is very sensitive to that input — a few points can double the recommended size. If the tool suggests risking more than a quarter of the account, treat it as a signal that the inputs need auditing rather than as a position size.
Does Kelly work with a fixed stop-loss strategy?expand_more
Yes, and that is the cleanest way to apply it. The Kelly fraction becomes your risk per trade, which the position size calculator turns into an actual quantity given your stop distance.