Kelly Criterion

What Is the Kelly Criterion?

The Kelly Criterion calculates the optimal fraction of capital to wager on a bet or trade, given a known edge and win probability. Developed by John Kelly at Bell Labs in 1956, the formula maximizes long-term growth rate—not any single trade’s expected value, but compounded return over many trades.

The core formula: fraction to bet = (probability of win × payoff ratio - probability of loss) / payoff ratio. A trade with a 60% win rate and 1:1 payoff suggests betting 20% of capital. A 40% win rate with 3:1 payoff suggests approximately 6.7%.

The Kelly Criterion produces sizing that maximizes the geometric mean of returns. Bet less than Kelly and you grow more slowly than optimal. Bet more than Kelly and you also grow more slowly—and crucially, you risk much larger drawdowns. Overbetting can lead to ruin even with a genuine edge.

Why It Matters for Options Traders

Options trading involves explicit probability estimates and defined payoff ratios in many common structures, making the Kelly Criterion more directly applicable than in most other trading contexts. A credit spread with a 70 percent probability of profit and a maximum profit of 1.00 dollar against a maximum loss of 4.00 dollars has a calculable Kelly fraction. The math is available to anyone willing to do it.

In practice, most sophisticated options traders use a fraction of full Kelly — commonly one-quarter to one-half — rather than the full theoretical optimum. Full Kelly maximizes growth but also produces very large swings in account value. Half Kelly reduces the volatility of outcomes significantly while still capturing most of the long-term edge. This reflects a real-world insight: the inputs to the Kelly formula (win rate, average win, average loss) are estimates, not certainties, and overconfidence in those estimates should be penalized with conservative sizing.

The Kelly Criterion also makes explicit something that position sizing rules often obscure: the importance of the win rate and payoff ratio together. A high win rate strategy with small wins may deserve less capital per trade than a low win rate strategy with large wins, if the math works out that way. Intuition about “feels like a high probability trade” without quantifying the payoff distorts these calculations.

Key Characteristics

• Full Kelly is rarely used in practice: The volatility of full Kelly sizing is extreme enough that even skilled practitioners cap exposure at half or quarter Kelly to manage drawdown risk
• Estimate error penalizes Kelly sizing: If your edge estimate is wrong — and it usually has some error — full Kelly will overbet; conservative fractions provide a margin of safety for estimation error
• Kelly applies to repeated, similar bets: The formula is most useful when a trader is making many similar trades over time, not for one-off position sizing in unique situations
• Probability of profit is a Kelly input: Options pricing models provide POP estimates, which can be combined with the potential profit and maximum loss to calculate a Kelly fraction — making it more tractable for options than for discretionary equity trades
• Negative Kelly means no bet: When the formula produces a negative number, the math is telling you the edge is actually against you — the bet should not be made at any size
• Kelly does not account for correlation: When multiple positions are held simultaneously, correlated positions reduce the effective edge; Kelly calculated on each position independently will overestimate how much total capital can be safely deployed