Kelly Criterion Calculator — Quantitative Position Sizing
Optimize trade bet sizing, prevent trader risk of ruin, and balance risk-reward using Full Kelly, Half Kelly, and Fractional Kelly algorithms.
Full Kelly maximizes theoretical mathematical growth but causes brutal 50%+ portfolio drawdowns. Professional quant desks use Half Kelly (0.5x) to retain 75% of growth while cutting drawdowns by half.
How It Works in 4 Steps
Input Historical Win Rate
Enter the percentage of winning trades from your verified trading journal.
Set Payoff Ratio
Define your average profit per winning trade divided by average loss per loser.
Define Account Capital
Input total available trading capital deployed across your brokerage account.
Apply Fractional Kelly
View Half Kelly and Quarter Kelly position sizes to optimize capital growth.
Why Quantitative Desks Rely on the Kelly Criterion Formula
Developed by Bell Labs scientist John L. Kelly Jr. in 1956, the Kelly Criterion is a mathematical formula that determines the optimal fraction of capital to allocate to an investment with positive expectancy: f* = (b*p - q) / b.
While Full Kelly mathematically maximizes long-term compounded growth, it induces violent 50%+ equity drawdowns. Consequently, top quantitative hedge funds and proprietary desks deploy Half Kelly (0.5x) to harvest 75% of peak growth while dramatically slashing volatility.
📋 Regulatory References & Data Sources
- Black-Scholes-Merton (1973) option pricing model
- SEBI circular SEBI/HO/MRD/DP/CIR/P/2019 — F&O margin framework
- NSE circular NSE/CMPT/39170 — SPAN margin methodology
- Securities Contracts (Regulation) Act 1956, as amended
Disclaimer: This calculator is for educational and planning purposes only. It does not constitute financial advice. Consult a SEBI-registered investment advisor for personalised guidance. Tax rules are updated as per the latest Finance Act — verify with a qualified CA before filing.
Explore Related Analysis Engines
Frequently Asked Questions — Kelly Criterion Calculator — Quantitative Position Sizing
f* = (bp - q) / b, where b is the payoff ratio (win/loss), p is the probability of winning, and q is the probability of losing (1 – p).