How much should I risk per trade?

Start from the strategy's worst losing streak rather than from a rule of thumb. Multiply that streak by the risk percentage: if the product is under about 10% of equity the drawdown is a dent, and beyond about 35% an ordinary bad run becomes an account-ending one. For most EAs with a published record that puts the answer between 0.5% and 1%.

Should position size be based on balance or equity?

Equity. Balance ignores open positions, so during a drawdown a balance-based rule keeps sizing to money the account has already lost. Equity-based sizing shrinks as the account falls and grows again as it recovers, which is the behaviour the rule is supposed to have.

My EA only has a fixed lot input. What do I do?

Compute the lot yourself from the account you are running and set it there, then revisit it as equity changes. Take your risk per trade, divide by the EA's stop distance in pips and by the pip value per lot, and round down to the volume step. It is the same calculation the EA would do, performed once instead of per trade.

Is fixed percentage sizing always better than fixed lots?

It is more consistent, because it holds the risk constant as the account moves rather than letting it drift. Fixed lots have one advantage worth noting — the position size does not compound during a winning run, which caps how large the exposure can become while the edge is still unproven.

Category: Risk
Difficulty: Beginner
Used in: Risk managementEA evaluationMT5 operation

Position sizing

The rule that turns a trading signal into a lot size — most usefully by fixing the loss a single trade may cost and letting the stop distance decide the volume.

also: lot sizing, trade sizing, risk per trade

Updated 2026-08-14

In plain English

Deciding how much to trade is a separate decision from deciding what to trade, and it is the one that determines whether the account is still there when the strategy pays off. Fix what a loss costs, and the lot size follows from arithmetic.

Why it matters

Two accounts running the same EA on the same broker can end a year at plus thirty percent and at zero. The difference is not the strategy; it is this rule. Sizing is where an edge becomes an outcome.

It is the only input under the trader's control after deployment. The signal logic belongs to the EA; the size does not.
It scales returns linearly and ruin non-linearly, so there is a size beyond which more risk produces less money.
It is what makes a published backtest transferable. A result is a percentage; the lot size that produced it is not.
It is the correct response to uncertainty about an edge. An edge you are unsure of should be traded smaller, not analysed longer.

How it is calculated

Lots = (Equity × Risk%) ÷ (Stop distance in pips × Pip value per lot)

Equity: Current account equity, not balance — using balance ignores open losses and sizes up during a drawdown.
Risk%: The share of equity a single loss may cost. One percent is the conventional starting point.
Stop distance: Entry to stop-loss, in pips. A strategy without a stop has no denominator and therefore no sizing rule.
Pip value per lot: Account-currency value of one pip for one lot of the symbol — about $10 on a USD-quoted major, different on everything else.

The result must then be rounded down to the broker's volume step and checked against the symbol's minimum and maximum volume. Rounding up exceeds the risk that was chosen, on every trade, permanently.

What counts as high or low

0.25–0.5% per trade Conservative
1% per trade Conventional
2–3% per trade Aggressive
over 5% per trade Ruinous over a normal losing streak

Read it together with

No single number describes a strategy. These metrics change how this one should be read.

Maximum drawdown Worst losing streak Risk/reward ratio Expectancy

Example

A $10,000 account, risking 1% per trade, on two strategies whose stops differ. The money at risk is identical; the lot size is not.

Risk per trade: $100; 1% of $10,000 equity.
Scalper — 15-pip stop: 0.66 lots; Rounded down to 0.66 at a 0.01 volume step.
Swing EA — 120-pip stop: 0.08 lots; Eight times smaller, for exactly the same risk.
Same 0.10 lots on both: $15 and $120 at risk; What happens when a lot size is chosen instead of computed.

A fixed lot size across strategies with different stops is not a risk policy. It is eight different risk policies wearing the same number.

Calculation 100 ÷ (15 × 10) = 0.66 · 100 ÷ (120 × 10) = 0.08

Result One risk figure, two lot sizes, because the stops differ

How to interpret it

Choose the risk percentage from the strategy's worst losing streak, not from ambition. The question is what a normal bad run does to the account, and both terms of that are already in the backtest report.

Range	What it means
Worst streak × risk% under 10% of equity	A normal bad run is a dent. The strategy gets to keep working.
10–20%	Recoverable, and uncomfortable enough that most people switch the EA off at the bottom.
20–35%	A 30% drawdown needs a 43% gain to recover. Reduce the risk percentage.
Over 35%	An ordinary losing sequence has become an account-ending event.

Multiply the strategy's worst losing streak by the risk percentage before deployment. That product, not the risk percentage alone, is the number to judge.
Compute from equity rather than balance, so size falls during a drawdown instead of holding steady while the account shrinks.
Round the computed volume down to the broker's volume step. On small accounts the rounding decides the risk more than the formula does.
Recheck pip value on anything that is not an FX pair — contract sizes differ, and carrying FX arithmetic across gives the wrong risk.
Restate any published result as a percentage of its account before comparing it with your own. The author's lot size describes the author's account.

Every EA listed here publishes its maximum drawdown, its worst losing streak and the account size its backtest ran on, which is exactly the input this calculation needs.

Common mistakes

Using a fixed lot across different strategies

The same 0.10 lots risks $15 behind a 15-pip stop and $120 behind a 120-pip stop. A fixed lot is only a risk policy when the stop distance is also fixed, and even then it stops matching the account as soon as the balance changes. Fixing the money at risk and computing the lot is the same amount of work and produces a consistent policy.

Sizing from balance instead of equity

Balance ignores open positions. During a drawdown with trades still running, balance-based sizing keeps opening full-size positions against an account that has already lost the money, which is precisely when the sizing should be shrinking. Equity-based sizing reduces automatically and recovers automatically.

Choosing risk percentage without the worst streak

Two percent per trade sounds moderate until it meets a twelve-trade losing run, which is an ordinary feature of a long record and takes roughly a quarter of the account. The worst streak is published in every tester report, and multiplying it by the risk percentage takes a few seconds and changes most people's answer.

Increasing size after a losing run

Raising risk to recover faster is martingale applied to the account rather than to the trade. It changes a recoverable drawdown into an unrecoverable one, and it does so at the exact moment the evidence for the edge is weakest.

In depth

What a 1% rule costs on the published records

The fourteen EAs listed here each publish the complete closed-trade record of our backtest for it, so every worst losing streak below is measured, not assumed.

Worst losing streak	EAs	That run at 1% risk
2 to 4 trades	7	2% to 4% of equity
8 to 9 trades	4	8% to 9%
10 to 14 trades	3	10% to 14%

At 1% per trade, three of the fourteen put a tenth of the account or more through an ordinary bad run — the band this page marks as recoverable but uncomfortable. At 0.5%, all fourteen stay inside the first band, because even the worst streak in the catalogue is 14 trades.

The two broker-side numbers this calculation depends on are in the terminal, not in the EA: right-click the symbol in Market Watch, choose Specification, and read the volume step and contract size before rounding anything down.

Frequently asked questions

How much should I risk per trade?: Start from the strategy's worst losing streak rather than from a rule of thumb. Multiply that streak by the risk percentage: if the product is under about 10% of equity the drawdown is a dent, and beyond about 35% an ordinary bad run becomes an account-ending one. For most EAs with a published record that puts the answer between 0.5% and 1%.
Should position size be based on balance or equity?: Equity. Balance ignores open positions, so during a drawdown a balance-based rule keeps sizing to money the account has already lost. Equity-based sizing shrinks as the account falls and grows again as it recovers, which is the behaviour the rule is supposed to have.
My EA only has a fixed lot input. What do I do?: Compute the lot yourself from the account you are running and set it there, then revisit it as equity changes. Take your risk per trade, divide by the EA's stop distance in pips and by the pip value per lot, and round down to the volume step. It is the same calculation the EA would do, performed once instead of per trade.
Is fixed percentage sizing always better than fixed lots?: It is more consistent, because it holds the risk constant as the account moves rather than letting it drift. Fixed lots have one advantage worth noting — the position size does not compound during a winning run, which caps how large the exposure can become while the edge is still unproven.

Related resources

Tool Position Size & Risk Run the calculation against your own equity, stop distance and symbol. Guide Testing Methodology The account size and lot size behind every published result here. EAs Browse EAs by historical drawdown Drawdown and worst streak are the two inputs that decide what risk percentage an EA can carry.