Trading Glossary
The Kelly Criterion is a formula that calculates the position size that maximizes an account's long-run geometric growth rate, given a known edge. In trading terms, that edge is expressed as two numbers: your win rate and your payoff ratio (the average winning trade divided by the average losing trade). Feed those two numbers in, and the formula returns a single percentage of capital to risk on the next trade.
The standard version is Kelly % = W − [(1 − W) / R], where W is the win rate as a decimal and R is the payoff ratio (average win divided by average loss). A trader who wins 55% of the time with an average win 1.5 times the average loss gets a Kelly figure of about 25% of capital per trade. That output isn't a dollar amount, it's a percentage of the account the formula says is mathematically optimal to risk, assuming the win rate and payoff ratio hold going forward exactly as measured.
The formula is only as good as its two inputs, and both are estimates pulled from a limited trade history, not fixed constants. A win rate measured from 40 trades can easily drift once the sample grows, and a single outsized winner or loser can swing the payoff ratio significantly. Because full Kelly sizing is also mathematically aggressive by design, meaning it accepts large, realistic swings in account value in exchange for the fastest theoretical growth rate, most traders who use it at all size at a fraction: half Kelly or quarter Kelly. Cutting the recommended size in half gives up some long-run growth but cuts volatility by roughly the same proportion, a trade most traders consider worth it given how uncertain the underlying win-rate and payoff-ratio inputs really are.
Traders with enough completed trades to measure a real win rate and payoff ratio use the Kelly formula as one input into position sizing, usually cross-checked against a flat risk-per-trade rule rather than followed at full size. It's most useful as a ceiling: a way to see whether a current risk-per-trade habit is already conservative relative to what the math would allow, or already pushing past it.