Position sizing: how much to put in each trade
Why the size of each position decides how deep your losses go, the arithmetic of getting back, and the common ways systems size their trades.
By Baris Demirci · · Rev00
The idea
Most of the attention in trading goes to when to buy and when to sell. How much to buy gets far less, yet it decides something the entry rules cannot: how deep the falls in your account go, and whether you can keep following the system through them.
The same system can be calm or ruinous depending only on its position size. Sizing does not change which trades win. It changes what each win and each loss does to the account.
Losses are harder to undo than to take
A fall and the rise needed to undo it are not the same size, because the rise starts from a smaller account.
| Account falls from 100 to | To get back to 100 it must grow by |
|---|---|
| 90 | 10 on 90, about a ninth |
| 80 | 20 on 80, a quarter |
| 67 | 33 on 67, a half |
| 50 | 50 on 50, doubling |
| 25 | 75 on 25, four times over |
That is why sizing matters so much. Keeping each loss small keeps the account in the region where recovery is still realistic.
A run of losses is normal
Every system has losing streaks. For trend following and breakout systems, which are wrong more often than right, ten losses in a row is not unusual at all. What that run does to the account depends almost entirely on how much each trade risks.
Common ways to size a position
A fixed amount
Every trade gets the same amount of money, or the same number of shares. It is simple, but the risk varies from trade to trade: the same amount in a wild stock and a calm fund are very different bets.
Equal slots
The account is split into a set number of equal parts, one per position. Common in systems that hold a fixed number of positions at a time. Simple and easy to check, with the same weakness as a fixed amount.
A fixed share of the account at risk
Decide in advance how much of the account a single trade may lose if its stop is hit, then work backwards to the position size. The further away the stop, the smaller the position.
This is the method behind the chart above. It ties every position to a loss you chose in advance. Its weak point is the stop itself: a market that gaps past the stop loses more than planned.
Sizing by how much the market moves
Some instruments swing a lot each day, others very little. Sizing by volatility gives the calm one a larger position and the wild one a smaller one, so that each moves the account by about the same amount on a typical day. Trend following systems that trade many markets usually work this way, because otherwise the wildest market would dominate the account.
Formulas for the fastest growth
There are formulas, the best known being the Kelly criterion, that work out the position size that would grow an account fastest if you knew a system’s odds exactly. You never do, and the full Kelly size produces falls few people can live through. Where it is used at all, it is used as a ceiling, and traders take a fraction of it.
Limits that sit on top
Whatever the method, most systems also cap things:
- The largest single position, so one bad gap cannot do too much damage.
- Total exposure, so the account never holds more than it has, unless borrowing is part of the design.
- Positions that move together, so five holdings that are really one bet are not sized as five separate risks. The Holy Grail is uncorrelated systems explains why that matters.
What it asks of you
- Deciding before the trade. Size comes from the rule and the stop, not from how confident a trade feels.
- Accepting smaller positions than feel exciting. A size that keeps a losing streak survivable will feel too small on most days. That is the point.
- Sizing for the bad run, not the average one. Ask what ten losses in a row would do, and whether you would still follow the system afterwards.
The short version
Position sizing decides how deep the falls go, and falls are harder to undo than to take. Most systematic traders fix in advance how much of the account a trade may lose, adjust for how much a market moves, and cap single positions and total exposure. A good system with reckless sizing is a bad system.
Written for people building their own systematic portfolio management process. Nothing here is a recommendation to buy or sell anything. See the disclaimer.