The Kelly Criterion Explained: How Much to Risk Per Trade (and Why Full Kelly Is Too Much)
Most traders spend years obsessing over what to trade and almost no time on how much to trade. Yet position sizing — not entry signals — is what separates an edge that compounds from one that blows up. The Kelly Criterion is the cleanest mathematical answer to the sizing question, and understanding it (including its dangers) will sharpen how you think about risk forever.
What the Kelly Criterion actually is
Developed by John L. Kelly Jr. at Bell Labs in 1956, the Kelly Criterion gives the fraction of your capital to stake on a repeated bet in order to maximize the long-run growth rate of your account. The key phrase is "long-run growth rate" — Kelly does not maximize your expected profit on the next trade; it maximizes how fast your capital compounds over many trades.
The formula
For a simple bet, the Kelly fraction is:
where:
A worked example
Say your strategy wins 55% of the time (p = 0.55, q = 0.45) and your average winner is 1.5x your average loser (b = 1.5):
Full Kelly says risk 25% of your capital on this trade. And right there is the problem.
Why nobody trades full Kelly
Twenty-five percent per trade is wildly aggressive. Kelly's math assumes you know your win rate and payoff ratio exactly and that they never change. In real trading, both are noisy estimates from a limited sample, and markets shift. Two consequences follow:
Fractional Kelly: what practitioners actually use
The standard fix is to trade a fraction of full Kelly — typically half-Kelly or quarter-Kelly. Half-Kelly captures roughly three-quarters of the growth rate of full Kelly while cutting the volatility of returns dramatically. In the example above, half-Kelly means risking ~12.5%, and quarter-Kelly ~6.25% — still aggressive for most discretionary traders, which tells you how conservative typical "risk 1-2% per trade" rules really are relative to a strong edge.
How to use the idea even if you never plug in the formula
The bottom line
The Kelly Criterion is the theoretical ceiling on how aggressively you should size a known edge. In practice you trade a fraction of it, because your edge is uncertain and your psychology is human. Treat full Kelly as a speed limit you never reach, use a half or quarter of it as a sane operating range, and let the formula remind you of the deeper truth: sizing should be driven by the size of your edge, and overbetting a good system is one of the fastest ways to ruin it.
Educational content only, not investment advice. Estimate your own statistics and manage your risk.
Most traders spend years obsessing over what to trade and almost no time on how much to trade. Yet position sizing — not entry signals — is what separates an edge that compounds from one that blows up. The Kelly Criterion is the cleanest mathematical answer to the sizing question, and understanding it (including its dangers) will sharpen how you think about risk forever.
What the Kelly Criterion actually is
Developed by John L. Kelly Jr. at Bell Labs in 1956, the Kelly Criterion gives the fraction of your capital to stake on a repeated bet in order to maximize the long-run growth rate of your account. The key phrase is "long-run growth rate" — Kelly does not maximize your expected profit on the next trade; it maximizes how fast your capital compounds over many trades.
The formula
For a simple bet, the Kelly fraction is:
f* = (bp - q) / bwhere:
- f* = the fraction of capital to risk
- b = the payoff ratio (how many units you win per unit risked — your reward-to-risk)
- p = the probability of winning
- q = the probability of losing (1 - p)
A worked example
Say your strategy wins 55% of the time (p = 0.55, q = 0.45) and your average winner is 1.5x your average loser (b = 1.5):
f* = (1.5 × 0.55 − 0.45) / 1.5 = (0.825 − 0.45) / 1.5 = 0.375 / 1.5 = 0.25Full Kelly says risk 25% of your capital on this trade. And right there is the problem.
Why nobody trades full Kelly
Twenty-five percent per trade is wildly aggressive. Kelly's math assumes you know your win rate and payoff ratio exactly and that they never change. In real trading, both are noisy estimates from a limited sample, and markets shift. Two consequences follow:
- Brutal drawdowns. Full Kelly routinely produces 50%+ peak-to-trough drawdowns. Mathematically optimal, psychologically unsurvivable.
- Overestimation is fatal. If your true edge is smaller than you think — and it usually is — full Kelly overbets and can grind you toward ruin even with a real edge. Betting above the Kelly fraction lowers your growth rate while increasing risk: the worst of both worlds.
Fractional Kelly: what practitioners actually use
The standard fix is to trade a fraction of full Kelly — typically half-Kelly or quarter-Kelly. Half-Kelly captures roughly three-quarters of the growth rate of full Kelly while cutting the volatility of returns dramatically. In the example above, half-Kelly means risking ~12.5%, and quarter-Kelly ~6.25% — still aggressive for most discretionary traders, which tells you how conservative typical "risk 1-2% per trade" rules really are relative to a strong edge.
How to use the idea even if you never plug in the formula
- It proves sizing scales with edge. Bigger edge (higher p or higher b) → larger optimal size. A coin-flip edge deserves almost nothing.
- It punishes overconfidence. Because overbetting is so costly, the safe error is to underestimate your edge and size down.
- It demands an honest win rate and payoff ratio. You cannot apply Kelly — or size rationally at all — without real statistics from your own trade log.
The bottom line
The Kelly Criterion is the theoretical ceiling on how aggressively you should size a known edge. In practice you trade a fraction of it, because your edge is uncertain and your psychology is human. Treat full Kelly as a speed limit you never reach, use a half or quarter of it as a sane operating range, and let the formula remind you of the deeper truth: sizing should be driven by the size of your edge, and overbetting a good system is one of the fastest ways to ruin it.
Educational content only, not investment advice. Estimate your own statistics and manage your risk.