Position Sizing 101: Fixed Fractional, ATR-Based, and Kelly

Even with identical logic, how much you put on per trade dramatically changes long-run return and max drawdown. Before chasing win rates or Sharpe ratios, it’s worth mastering position sizing.

Three Classical Methods

Method	Formula	Character
Fixed fractional	Size = Capital × fixed %	Simple; DD scales with %
ATR-based	Size = Risk $ ÷ (ATR × k)	Volatile names automatically sized smaller
Kelly criterion	Size = p − (1−p)/R	Theoretically optimal, practically aggressive

Fixed Fractional

“Always put 10% of capital in each trade,” or “equal weight across N names.” Intuitive and a good starting point, but it ignores volatility — you end up taking more risk in more volatile names.

ATR-Based

Measure volatility via ATR (Average True Range) and size so that dollar risk per trade is constant. If you allow 1% of capital at risk, high-ATR names get smaller sizes automatically.

Kelly Criterion

Calculates the growth-optimal fraction from win rate and payoff ratio. The catch: it assumes those parameters are known and stable, which they never are. In practice, people run half-Kelly or smaller to dampen estimation error.

Backtest Lenses

• Absolute max drawdown. If fixed fractional gives 50% DD, it’s a sizing signal
• Return / DD ratio. Often improves substantially with better sizing
• Balance during losing streaks. Can you still recover after 10 consecutive losses?

Pitfalls

Full Kelly assumes your estimated win rate and payoff ratio are correct. They never are — so full Kelly produces violent equity swings from estimation error. Conservative sizing often produces better long-run outcomes on the same strategy.

Combining sizing experiments with Monte Carlo simulation gives a statistical read on how sizing changes affect DD.

Try It in QuanTest

In QuanTest you can switch sizing methods on the same strategy and compare outcomes. Moving from fixed fractional to ATR-based alone often reshapes the DD profile significantly.

---

This article is for educational purposes. It does not guarantee the profitability of any strategy or future performance. Investment decisions are your own responsibility.