Position Sizing Methods Compared: Which Is Best?
Five position sizing methods — fixed fractional, fixed lot, Kelly, volatility-based, and martingale — compared on safety, growth, and complexity.
Position Sizing Methods Compared: Which Is Best?
The same strategy with different sizing rules produces wildly different outcomes. Position sizing is where edge meets survival.
Position sizing is how you decide how much to trade. The method you choose changes your returns, your drawdowns, and your odds of blowing the account. Here are five methods compared.
1. Fixed fractional (risk a constant %)
Risk the same percentage of your current account on every trade. As the account grows, position size grows; as it shrinks, risk shrinks.
| Pros | Cons |
|---|---|
| Drives risk of ruin toward zero | Slower compounding than aggressive methods |
| Auto-adjusts to account size | Tiny on small accounts |
| Survives long losing streaks | Requires recalculating each trade |
Best for: Almost everyone, especially beginners. The default recommendation.
2. Fixed lot (constant share/contract count)
Trade 100 shares every time, regardless of account size or stop distance.
| Pros | Cons |
|---|---|
| Simplest to execute | Risk varies wildly per trade |
| Easy to backtest | Same-size bet on a tight stop and a wide stop means very different risk |
| No calculation needed | Can blow the account in a bad streak |
Best for: Pure execution practice on paper. Not recommended for live accounts.
3. Kelly Criterion
Risk the mathematically optimal fraction to maximize long-term growth: Kelly% = W − (1−W)/R.
| Pros | Cons |
|---|---|
| Maximizes theoretical growth | Full Kelly has brutal drawdowns |
| Mathematically grounded | Inputs (win rate, R) must be accurate |
| Scales with edge strength | Edge estimates are unreliable in practice |
Best for: Use fractional Kelly (quarter-Kelly) as a ceiling on your risk, never as a target. See the Kelly Criterion guide.
4. Volatility-based (ATR scaling)
Risk a constant percentage, but adjust position size so the distance to stop equals a fixed multiple of ATR. Effectively: risk the same %, with stops that adapt to volatility.
| Pros | Cons |
|---|---|
| Adapts to market conditions | Requires computing ATR |
| Stops sit outside normal noise | More complex than fixed fractional |
| Works across instruments | Stops can be wide on volatile assets |
Best for: Discretionary traders using ATR-based stops. A refinement of fixed fractional.
5. Martingale (double after losses)
Double position size after each loss to "recover" the previous loss when you eventually win.
| Pros | Cons |
|---|---|
| None worth the risk | Mathematically guaranteed to eventually wipe the account |
| Requires infinite capital to be safe | |
| Loses to table limits / margin calls |
Best for: Nobody. This is a known ruin strategy dressed up as a "system." Avoid completely.
Comparison table
| Method | Safety | Growth | Complexity | Recommendation |
|---|---|---|---|---|
| Fixed fractional | High | Moderate | Low | Default for most traders |
| Fixed lot | Low | Low | Lowest | Paper trading only |
| Kelly (fractional) | Medium | High | High | Use as a ceiling |
| Volatility-based | High | Moderate-high | Medium | Best for discretionary traders |
| Martingale | None | — | Low | Never |
Which is best?
For 90% of traders, the answer is fixed fractional at 1%–2% risk per trade, optionally refined with volatility-based stops. It's the simplest method that's also mathematically sound — and the only one that survives both your worst losing streak and your worst psychological moments.
The remaining 10% — typically systematic trend followers with large samples — can layer fractional Kelly on top as a sanity check on their sizing. Even they cap at 1%–2% in practice.
How to start
- Pick fixed fractional
- Set risk at 1% (0.5% if you're new)
- Use the position size calculator every trade
- Log your actual risk in a journal
- After 200 trades, re-evaluate — only then consider a more sophisticated method
The best position sizing method is the one you'll actually follow for 1,000 trades. Fixed fractional wins that test.