Managing Risk: The Kelly Criterion for Credit Put Spreads

In the world of options trading, particularly with defined-risk strategies like credit put spreads, the allure of consistent premium collection is strong. Yet, the difference between long-term success and a blown-up account often boils down to one critical, yet overlooked, skill: position sizing. You can have the perfect technical setup, but if you bet the farm on it, a single loss can be catastrophic. Today, we move beyond simple rules of thumb and explore a mathematical framework for position sizing: the Kelly Criterion. We'll adapt this powerful tool specifically for sizing credit put spreads in two distinct market environments—trending and range-bound—to help you manage risk and protect your trading capital.

What is the Kelly Criterion?

Developed by John L. Kelly Jr. in 1956, the Kelly Criterion is a formula used to determine the optimal size of a series of bets to maximize long-term growth of capital. In trading terms, it tells you what percentage of your portfolio to risk on a given trade based on your edge—the probability of winning and the payoff for a win versus a loss.

The classic formula is: f* = (bp - q) / b

f*: The fraction of your capital to bet.
b: The net odds received on the bet (profit/loss ratio).
p: The probability of winning.
q: The probability of losing (1 - p).

For traders, the core insight is that betting too small leaves money on the table, while betting too large increases the risk of ruin. The "Full Kelly" aims for maximum growth but can be volatile. Many practitioners use "Fractional Kelly" (e.g., half or quarter) to reduce volatility while still capturing most of the growth benefit.

Adapting Kelly for Credit Put Spreads

A credit put spread involves selling a put option at one strike price and buying a further out-of-the-money put to define your maximum risk. Your profit is the net credit received. Your maximum loss is the width of the strikes minus the credit.

To use Kelly, we need to define our variables for this specific strategy:

Win Probability (p): Your estimated chance that the spread expires worthless (or you buy it back for a fraction of the credit). This is NOT the probability of being in-the-money (delta). It's your trade's historical or modeled success rate.
Loss Probability (q): 1 - p.
Net Odds (b): (Credit Received) / (Max Loss). This is your reward-to-risk ratio on the trade.

The adapted formula becomes: f* = [ ( (Credit/Max Loss) * p ) - q ] / (Credit/Max Loss)

Practical Example: A Standard Trade

Let's say you sell a put spread on XYZ stock trading at $100.

Sell the $95 put for $2.00
Buy the $90 put for $0.60
Net Credit: $1.40 ($140 per spread)
Max Loss: ($5 spread width - $1.40 credit) = $3.60 ($360 per spread)
Reward-to-Risk (b): $140 / $360 = 0.389
Your Estimated Win Rate (p): Based on your backtesting, 70% (0.70). So, q = 0.30.

Plugging into the formula:
f* = [ (0.389 * 0.70) - 0.30 ] / 0.389
f* = [0.2723 - 0.30] / 0.389
f* = -0.0277 / 0.389 = -0.071

A negative result? This is the Kelly Criterion's critical warning. It tells you this bet, with these parameters, has a negative expected value and you should not take it. Your win rate is too low for the reward/risk profile. This immediately prevents you from risking capital on a mathematically losing proposition.

Applying Kelly in Different Market Regimes

The key variable you must adjust is your estimated win probability (p). This changes dramatically between market environments.

1. Sizing Spreads in a Range-Bound Market

In a choppy, sideways market, high-probability, low premium trades often thrive. You might sell spreads far out-of-the-money, aiming for a high win rate with smaller credits.

Example Trade: Market is range-bound between support and resistance.

Credit: $0.50 ($50)
Max Loss: $4.50 ($450)
b = 50/450 = 0.111
Estimated Win Rate (p): High, due to the range. Let's use 85% (0.85). q = 0.15.

f* = [ (0.111 * 0.85) - 0.15 ] / 0.111
f* = [0.09435 - 0.15] / 0.111 = -0.05565 / 0.111 = -0.501

Another negative! This reveals a common trap: the win rate feels high, but the terrible reward-to-risk ratio (< 1:8) destroys any edge. The Kelly Criterion steers you away from these "picking up pennies in front of a steamroller" setups. To get a positive Kelly, you would need an even higher win rate or a better credit-to-width ratio.

2. Sizing Spreads in a Trending Market

In a strong uptrend, you can confidently sell puts closer to the money for larger credits. The win probability might be lower than in a range, but the payoff is better.

Example Trade: Stock in a clear uptrend, pulling back to a moving average.

Credit: $3.00 ($300)
Max Loss: $2.00 ($200)
b = 300/200 = 1.5 (A favorable 1.5:1 reward-to-risk)
Estimated Win Rate (p): Good, but not extremely high due to closer strikes. Let's use 65% (0.65). q = 0.35.

f* = [ (1.5 * 0.65) - 0.35 ] / 1.5
f* = [0.975 - 0.35] / 1.5
f* = 0.625 / 1.5 = 0.417

A positive result! Full Kelly suggests risking ~41.7% of your capital on this single trade. This is extremely aggressive and highlights why Full Kelly is rarely used in trading. A single loss would be devastating. This is where Fractional Kelly is essential.

Using a more conservative Half-Kelly or Quarter-Kelly: Half-Kelly: 0.417 / 2 = 0.2085 (20.85% of capital)
Quarter-Kelly: 0.417 / 4 = 0.104 (10.4% of capital)

With a $25,000 portfolio, Quarter-Kelly suggests a max loss of $2,600 on this trade. Since your max loss per spread is $200, you could size into 13 contracts ($200 * 13 = $2,600). This provides a disciplined, mathematically sound position size.

Integrating Kelly with Overall Risk Management

The Kelly Criterion is a powerful input, not the entire system. It must be integrated with your overall rules:

Always Define Max Portfolio Risk: Never let your total risk across all trades exceed 5-10% of your portfolio, regardless of attractive Kelly numbers.
Use Conservative Win Rate Estimates: Be pessimistic. If you think your win rate is 70%, use 60% in the formula. This builds in a margin of safety.
Employ Fractional Kelly (Always): Treat Full Kelly as a theoretical maximum. Half or Quarter Kelly dramatically reduces drawdowns while preserving most of the growth benefit.
Have a Stop-Loss Plan: Kelly assumes you know your exact max loss. For a credit spread, this is defined, but you should still have a rule to exit if the short option reaches a certain delta (e.g., 0.70) before expiration to manage unrealized losses.

Conclusion: A Framework for Disciplined Sizing

The Kelly Criterion provides a rigorous, quantitative framework for answering the trader's eternal question: "How much?" By forcing you to quantify your edge (win rate and reward/risk) for every credit put spread, it brings discipline to your position sizing. As we've seen, it can save you from taking negative-expectancy trades in range-bound markets and help you capitalize on high-conviction setups in trending markets—all while controlling for maximum loss. Remember, the goal is not to maximize returns on a single trade, but to maximize the long-term growth of your capital while avoiding ruin. Start by incorporating Fractional Kelly calculations into your trade journal, use conservative estimates, and let this mathematical edge guide you toward more consistent and protected trading.