The Kelly Criterion in Crypto Futures: Why Full-Kelly Causes Ruin

The Kelly Criterion provides the mathematically optimal fraction of wealth to risk on an investment with positive expected value. However, applying Full-Kelly to high-volatility crypto futures leads to certain drawdown ruin due to estimation error.

1. Kelly Fraction & Half-Kelly Volatility Haircut

Full Kelly fraction: f* = (p * b - q) / b, where p = win rate, q = 1 - p, and b = payoff ratio (Avg Win / Avg Loss). Because crypto return distributions have non-Gaussian fat tails, quant desks apply a 50% haircut: f_safe = 0.5 * f*, capped at maximum 1.5% to 2.0% equity risk per trade.

Python: Half-Kelly Position Sizing Calculator with Leverage Cap

def calculate_safe_position_size(equity: float, win_rate: float, avg_win: float, avg_loss: float, current_price: float, stop_distance: float, max_risk_cap: float = 0.02):
    b = avg_win / avg_loss
    p = win_rate
    q = 1.0 - p
    
    full_kelly = (p * b - q) / b
    if full_kelly <= 0:
        return 0.0 # Negative expected value, do not trade
        
    half_kelly = full_kelly * 0.5
    allocated_risk = min(half_kelly, max_risk_cap) * equity
    
    position_qty = allocated_risk / stop_distance
    return round(position_qty, 4)

Frequently Asked Questions