What is position sizing?
Position sizing is the rule that decides how much capital each trade gets. Given the same entry and exit signals, sizing alone determines whether a losing streak is a dip or a disaster, which makes it the most consequential and least glamorous part of any strategy.
Entries get the attention because they feel like the skill. Sizing is where the survival actually lives: it converts a strategy's edge into a growth rate while capping what any run of bad luck can take.
What are the main position sizing methods?
Fixed fractional sizing commits a set percentage of current equity to each position, so size shrinks in drawdowns and grows in good runs automatically. Fixed risk per trade goes a step further: choose the loss you accept if the stop is hit, often 1-2% of equity, and work backward from the stop distance to the position size.
Volatility-based sizing scales positions to current market movement, often using ATR, so a wild market gets a smaller position and a quiet one gets more. All three share the same principle: the market's behavior, not conviction, sets the size.
Why does sizing decide survival?
Losing streaks are a certainty, not a risk. A strategy with a 50% win rate will hit seven losses in a row regularly. At 1% risk per trade that streak costs about 7% of equity, an annoyance. At 10% per trade it costs half the account, and the mathematics of drawdown make half an account brutally expensive to rebuild.
Sizing also compounds in both directions. Oversizing does not just risk ruin; it forces the strategy to trade scared, and undersizing wastes an edge. The right size is the largest one whose worst realistic streak is boring.
What are the common sizing mistakes?
Sizing by feel, so conviction quietly replaces the rule. Doubling after losses to win it back, which turns a losing streak into a wipeout. And ignoring correlation: three full-size positions in assets that move together are one triple-size position wearing three names.
Algorithmic trading removes the first two by construction, the rule is code and applies every time, but correlation remains a portfolio-level decision the strategy builder has to make deliberately.
Frequently asked questions
- How much should I risk per trade?
- The widely used starting point is 1-2% of account equity per trade, meaning that is the loss if the stop is hit, not the position's full value. Smaller suits strategies with long losing streaks; going much larger requires evidence the streaks stay short, which backtests can estimate.
- What is the Kelly criterion?
- A formula for the sizing fraction that maximizes long-run growth given a strategy's win rate and payoff ratio. Full Kelly is aggressive and punishes estimation error brutally, so practitioners who use it at all usually trade a half or quarter of the Kelly size.
- Should every trade be the same size?
- Same rule, not necessarily same size. Fixed-risk and volatility-based methods produce different position sizes per trade from one consistent rule, sizing down when stops are wide or markets are wild. What should never vary is the rule itself.
- How does leverage relate to position sizing?
- Leverage lets a position exceed account equity, which makes disciplined sizing more important, not less. The risk-per-trade logic still governs: decide the acceptable loss first, and let leverage merely enable the computed size, never inflate it.
