ATR-Based Position Sizing: The 10-Second Habit That Outlives Any Strategy

Strategies expire. Markets change, edges decay, the EA that printed money in 2023 goes flat in 2025. One habit survives all of it: sizing every position from volatility instead of from a fixed lot number. It takes about ten seconds per trade, and it is the closest thing to a permanent skill this business offers.

Lots = (Account × Risk%) / (ATR × Multiplier × Value per lot per point)

In words: decide how much money you are willing to lose if the stop is hit, place the stop at a multiple of the Average True Range, and divide. The ATR term is the whole trick — it makes the position automatically smaller when the market gets wilder and larger when it goes quiet.

Account: $10,000. Risk per trade: 1% = $100. Symbol: gold (XAUUSD), daily ATR around $22. Stop at 1.5 × ATR = $33 of price distance. On a standard contract, $1 of gold movement is worth $100 per lot, so a $33 stop costs $3,300 per lot. Position: $100 / $3,300 ≈ 0.03 lots.

Notice what did not enter the calculation: how confident you feel, how the last three trades went, how much you want to win back. The formula is boring on purpose.

A fixed 0.10 lots is not one risk — it is a different risk on every symbol and in every regime. On EURUSD in a quiet week it might be 0.5% per stop; on gold in a volatile week the same 0.10 lots can be 8%. Traders who blew up on gold in 2024-2025 mostly did not have bad entries. They had EURUSD position sizes on a symbol whose daily range had tripled. ATR sizing makes that mistake structurally impossible.

Because the risk is a percent of the account, the position grows as the account grows and shrinks after losses. A 1% risk per trade means ten straight losses cost about 9.6% of the account, not 10% — and, more importantly, a 30% drawdown needs a 43% gain to recover, while a 50% drawdown needs 100%. Volatility sizing does not prevent losing streaks; it keeps them in the survivable zone where compounding can still do its work.

ATR sizing does not create an edge, does not help against gaps through the stop, and 1% is not magic — a strategy with long streaks may need 0.5%. It also assumes you actually use a stop; without one the formula sizes nothing. This is exactly how our own EAs compute lots internally — every risk-mode preset is this formula with different multipliers, and our worst historical equity drawdowns per setting are published, not estimated.

The portfolio EA where every risk preset is this formula, with the worst equity drawdown per setting published: https://www.mql5.com/en/market/product/182751

Our products and free utilities: https://www.mql5.com/en/users/app.develop.sk/seller