Kelly Criterion for Options Trading: Optimal Position Sizing Math

The Kelly Criterion is a formula from information theory that calculates the optimal fraction of your bankroll to risk on a bet with a known edge. Applied to options trading, it provides a mathematical framework for position sizing that maximizes long-term growth while avoiding ruin.

The Kelly Formula

Kelly % = W - (L / R)

Where:

  • W = Probability of winning (expressed as decimal)
  • L = Probability of losing (1 - W)
  • R = Ratio of average win to average loss

    • Example: Your iron condor strategy has a 72% win rate. Average winner: $120. Average loser: $280.

  • W = 0.72
  • L = 0.28
  • R = 120/280 = 0.4286

    • Kelly % = 0.72 - (0.28 / 0.4286) = 0.72 - 0.653 = 0.067 or 6.7%

    The Kelly formula says the mathematically optimal risk per trade is 6.7% of your account.

    Why Full Kelly Is Dangerous

    The Kelly Criterion assumes:

  • You know your exact win rate and average win/loss with certainty
  • You'll place an infinite number of trades
  • You have no emotional reaction to drawdowns
  • Your edge is constant over time

    • None of these are true in practice. Your actual win rate might be 68% instead of 72%. Market conditions change. And full Kelly sizing produces stomach-churning drawdowns — often 40-50% at some point during the journey.

    This is why virtually every professional who uses Kelly applies a fractional Kelly approach:

    Fractional Kelly: The Practical Approach

    Half Kelly (0.5×) is the most common recommendation. It captures about 75% of the growth rate of full Kelly while reducing drawdowns by roughly half.

    | Metric | Full Kelly | Half Kelly | Quarter Kelly | Growth rate100% of optimal~75% of optimal~50% of optimal Max drawdown40-60%20-30%10-15% Emotional stressExtremeModerateLow | Recovery time from worst drawdown | 1-2 years | 6-12 months | 3-6 months |

    For most options traders, half Kelly is the sweet spot. It provides strong compounding while keeping drawdowns within psychologically survivable territory.

    Calculating Kelly for Common Options Strategies

    Credit Spreads (30 Delta Short Strike)

    Typical parameters:

  • Win rate: 68%
  • Average win: $80 (closing at 50% of $160 credit)
  • Average loss: $250 (stopping at 2× credit)

    • Kelly % = 0.68 - (0.32 / 0.32) = 0.68 - 1.0 = -0.32

    Negative Kelly! This means the strategy as described has no edge. The high win rate is offset by disproportionately large losses. You'd need to improve the win rate, increase the average win, or decrease the average loss for this to have positive expected value.

    This is a critical insight: high win rates do not guarantee positive Kelly. The win/loss ratio matters equally.

    Credit Spreads (Optimized Management)

    Adjusted parameters:

  • Win rate: 70%
  • Average win: $90 (closing at 50%)
  • Average loss: $180 (tighter stop at 1.5× credit)

    • Kelly % = 0.70 - (0.30 / 0.50) = 0.70 - 0.60 = 0.10 or 10%

    Half Kelly: 5% risk per trade. Now we have a viable sizing framework.

    Long Options / Debit Spreads

    Typical parameters:

  • Win rate: 40%
  • Average win: $250
  • Average loss: $100

    • Kelly % = 0.40 - (0.60 / 2.50) = 0.40 - 0.24 = 0.16 or 16%

    Half Kelly: 8% risk per trade. The higher Kelly fraction reflects the favorable win/loss ratio despite the lower win rate.

    Step-by-Step: Applying Kelly to Your Trading

    Step 1: Track your results. You need at least 50 trades (ideally 100+) to have meaningful statistics. Record every trade's outcome.

    Step 2: Calculate your actual metrics:

  • Count your wins and losses → Win rate
  • Average your winning trades → Average win
  • Average your losing trades → Average loss

    • Step 3: Apply the Kelly formula.

    Step 4: Use half Kelly as your risk per trade.

    Step 5: Recalculate quarterly. As you accumulate more data and market conditions change, your Kelly number will shift. Update your sizing accordingly.

    Kelly Limitations for Options Traders

    Changing market conditions. A strategy that works beautifully in low-volatility bull markets may have very different statistics in bear markets. Kelly calculated from the last 6 months of a bull run may overestimate your edge during the next correction.

    Non-binary outcomes. Kelly assumes you either win a fixed amount or lose a fixed amount. Options trades have a range of outcomes — you might close at 30% profit, 50% profit, or 10% loss. Averaging these into single numbers loses information.

    Correlated positions. Kelly is for independent bets. If your options positions are correlated (and they usually are to some degree), the effective risk of multiple positions is higher than the sum of individual Kelly fractions suggests.

    Estimation error. If your win rate estimate is off by just 5% (68% instead of 73%), the Kelly fraction changes dramatically. Given the relatively small sample sizes in personal trading (a few hundred trades), your estimates carry significant uncertainty.

    The Practical Takeaway

    Kelly Criterion gives you a ceiling, not a target. Calculate it, take half, and use that as your maximum risk per trade. If half Kelly says 5%, and your standard risk rule says 2%, use 2%. If half Kelly says 1%, reduce below your standard rule.

    The real value of Kelly isn't the exact number — it's the framework of connecting your position size to your measurable edge. No edge, no trade. Small edge, small size. Large edge, larger (but still fractional) size.