Position sizing methods compared

2026-11-15

Every sizing method holds something constant. Which thing you hold constant is the actual decision, and most people make it by accident.

Short answer

You want	Method
The same risk on every trade	Fixed fractional — the sensible default
Simplicity while you learn	Fixed dollar risk
Comparable risk across instruments of different volatility	Volatility-adjusted
To be surprised by your losses	Fixed quantity

Fixed quantity — what people do by default

Always trade 1 BTC, or 100 shares.

Holds constant: the quantity. Therefore varies: the risk, by however much the stop distance and price vary.

This is the default because it is the easiest thing to type, and it is the worst of the four. A wide stop and a narrow stop produce losses that differ by multiples with no decision having been made. Your largest losses arrive in the trades that happened to need the widest stops, which is uncorrelated with anything about your edge.

It also makes your records uninterpretable: an R multiple computed across trades of varying risk is an average of incomparable things.

Fixed dollar risk

Risk exactly $200 on every trade, whatever the stop distance.

Holds constant: the loss per trade in currency. Therefore varies: the loss as a fraction of the account, as the account changes.

A real improvement, and fine while the balance is stable. Its weakness appears in a drawdown: as the account shrinks, a constant dollar risk becomes a growing percentage. Risking $200 on $10,000 is 2%; on $5,000 it is 4%. The method increases risk exactly when the account can least afford it — see risk of ruin.

Fixed fractional — the default worth using

Risk a constant percentage of current balance.

position size = (balance × risk %) ÷ |entry − stop|

Holds constant: risk as a fraction of the account. Therefore varies: the dollar loss, shrinking as the account does.

This is what the position size calculator computes, and its important property is automatic de-risking. After k consecutive losses the balance is (1 − f)^k — which decays more slowly than k × f would, because each loss is smaller than the last.

Risk	After 10 losses, fixed fractional	If losses were fixed-dollar
1%	−9.6%	−10%
5%	−40.1%	−50%
10%	−65.1%	−100%

At 10% risk, fixed-dollar sizing empties the account in ten losses. Fixed fractional leaves 35% of it. That gap is the method doing its job, and it widens precisely where it matters.

The trade-off: recovery is slower, because position sizes are smaller while you are down.

Volatility-adjusted

Set the stop distance from recent volatility rather than from a fixed percentage, then size from that.

stop distance = k × recent range (e.g. ATR)
position size = (balance × risk %) ÷ stop distance

Holds constant: risk as a fraction of the account, and the stop’s relationship to normal price movement. Therefore varies: the stop distance, and position size inversely with volatility.

This is fixed fractional with one extra correction, and the correction matters when you trade instruments of different character. A 2% stop is far outside normal movement on one instrument and inside the noise on another. A fixed percentage stop applied across both gives you two very different strategies wearing one name.

The cost is a parameter (k) and a volatility estimate that lags. Neither is free, and both add something to get wrong.

What none of them decide

How much to risk. Every method above takes the risk percentage as an input. Choosing it is a separate decision, made from your expected longest losing streak and the drawdown you can recover from — see losing streak probability and the drawdown recovery calculator.

Where the stop goes. Sizing starts after invalidation is located. Deriving the stop from the size you wanted breaks the chain — how to choose a stop-loss level.

Correlation. Three positions sized independently at 1% each are not three 1% risks if the instruments move together. They are closer to one 3% risk, and no per-trade method can see that.

The failure that undoes all of them

Increasing the risk fraction during a drawdown. The impulse to recover faster raises f exactly when the balance is smallest, converting fixed fractional’s protective decay into the opposite. It is the most common path to ruin for accounts that had a genuine edge.

Which is why the risk percentage belongs in code, alongside a daily loss limit, rather than in a decision you re-make while losing.

FAQ

What is fixed fractional position sizing?

Risking a constant percentage of your current account balance on each trade, with position size derived as (balance × risk %) ÷ stop distance. Because the percentage applies to the current balance, position sizes shrink automatically during a drawdown, which slows losses without requiring you to decide anything.

Is fixed fractional better than fixed dollar risk?

In a drawdown, clearly. Fixed dollar risk becomes a larger fraction of a shrinking account, so it increases risk exactly when the account can least afford it. Fixed fractional does the reverse. Over ten consecutive losses at a 10% risk level, fixed dollar sizing empties the account while fixed fractional leaves about a third of it.

Should I size positions by volatility?

It helps when you trade instruments of different character, because a fixed percentage stop means very different things on a quiet instrument and a volatile one. The cost is an extra parameter and a volatility estimate that lags. If you trade one instrument, the added complexity buys little.

How do I size correlated positions?

No per-trade method handles this, because each sizing decision is made in isolation. The practical approach is a cap on total exposure to a group of instruments that move together, set independently of per-trade sizing — otherwise three 1% positions in correlated instruments are one 3% position you did not authorise.