BUILD: COURSE 2 | LESSON 2
Volatility-based position sizing and ATR stops
Learning objectives
Place stops using a multiple of ATR so that stop distance reflects current market volatility, not a fixed pip habit.
Compute position size from the formula: position = risk in money ÷ (stop distance × pip value), and adjust it when volatility changes.
Recognise the two classic sizing errors — same lot size everywhere, and same pip stop everywhere — and quantify what they cost.
In Foundations you learned the 1–2% rule: risk a fixed slice of your account per trade. The missing piece was how to keep that risk constant when markets don't cooperate. EUR/USD might move 60 pips on a quiet day and 140 on a CPI day; gold's daily range is a different animal from GBP/USD's. A fixed "30-pip stop, 0.5 lots, every trade" habit means your actual risk swings wildly from trade to trade while the number in your plan stays politely constant. This lesson replaces habit with a two-step process: volatility sets the stop, and the stop sets the size.
Step 1: Let ATR set the stop
You met ATR (Average True Range) in P1.4 as a volatility gauge: the average size of a bar's full range over, typically, the last 14 periods. Its job in risk management is to answer one question: how far can price wander purely through noise, without my trade idea being wrong?
A stop placed inside the noise gets hit by noise. The standard technique is to place the stop a multiple of ATR away from entry, beyond the structure that invalidates the trade:
Stop distance = k × ATR, with k commonly 1.5–3 depending on timeframe and style (swing traders often use 2× daily ATR; intraday traders 1.5× ATR of their trading timeframe).
Worked example. EUR/USD daily ATR(14) = 72 pips. You buy at 1.0850 with k = 2:
- Stop distance = 2 × 72 = 144 pips → stop at 1.0706.
Six weeks later volatility has halved: ATR = 36 pips. The same setup now gets a 72-pip stop. Same rule, different market, and in both cases the stop sits outside ordinary daily noise. Compare this with the trader who "always uses 50 pips": in the quiet market their stop is unnecessarily wide relative to the opportunity; in the volatile one it sits inside the noise band and gets clipped on trades where the idea was right. ATR stops don't predict anything — they simply scale your tolerance to the market's current breathing rate. One caveat: ATR is a minimum sanity distance, not a substitute for structure. If the invalidation level (below the swing low, beyond the range edge) is further than k × ATR, use the structure and accept the wider stop — or skip the trade.
Step 2: Let the stop set the size
Now the arithmetic that makes risk constant. With account risk fixed in money, the position size is whatever makes "stop distance × pip value" equal that money:
Position size (lots) = risk $ ÷ (stop distance in pips × pip value per lot)
Worked example, in full. Account: $10,000. Risk per trade: 1% = $100. EUR/USD, daily ATR = 72 pips, k = 2 → stop distance = 144 pips. Pip value for 1 standard lot of EUR/USD (USD account) = $10/pip.
- Position = 100 ÷ (144 × 10) = 100 ÷ 1,440 = 0.069 lots → round down to 0.06 lots (never round risk up).
- Check: 0.06 lots → pip value $0.60 → 144 pips × $0.60 = $86.40 risked. Slightly under $100 because of rounding. Correct.
Now the quiet market, ATR = 36 pips, stop = 72 pips:
- Position = 100 ÷ (72 × 10) = 0.138 → 0.13 lots, roughly double the size, same $risk.
This is the core insight: volatility-based sizing means trading smaller when markets are wild and larger when they are calm — automatically. Most losing streaks that "came out of nowhere" are really a volatility regime change hitting a fixed lot size: the trader's nominal risk never changed, but their real risk doubled. Two more worked variants so the formula generalises:
- JPY pair. USD/JPY at 155.00, 1 lot pip value ≈ ¥1,000 ÷ 155.00 ≈ $6.45/pip. Stop 90 pips, risk $100 → 100 ÷ (90 × 6.45) = 0.17 lots.
- Gold. XAU/USD, ATR = $28, k = 1.5 → stop $42. With 1 lot = 100 oz, a $1 move = $100. Risk $150 → 150 ÷ (42 × 100) = 0.035 → 0.03 lots. Traders who size gold like a forex pair discover this the expensive way.
What this does to your expectancy maths
Recall from P2.1 that your edge is measured in R. Volatility sizing is what makes "1R" mean the same thing on every trade, which is precisely what makes your journal statistics valid. If Monday's trade risks $60 and Thursday's risks $240 because the lot size never changed, your "+0.2R expectancy" is a fiction averaged over four different games.
There is also a subtle cost interaction. Halving the stop distance doubles the position size for the same $risk — which doubles the money paid in spread. A 1.2-pip round-trip spread on a 144-pip stop is 0.8% of R; on a 20-pip stop it is 6% of R. Tight-stop styles are not "lower risk"; they are the same risk paying higher tolls, and they need a correspondingly better raw edge to survive. Finally, cap total exposure: even perfectly sized individual trades stack up. A common practitioner rule is a maximum of 4–6% of the account at risk across all open positions (we'll sharpen this in P2.5, because correlated positions cheat this cap).
The pre-trade sizing checklist
Before every trade, in this order — it takes 30 seconds with the calculator (Tools → Position Size Calculator):
- Invalidation: where is the trade wrong? (structure level)
- Volatility check: is that level at least ~1.5 × ATR away? If structure demands a stop far beyond your ATR multiple, the trade may be too extended — reduce size further or pass.
- Risk $: account × your fixed % (1–2%).
- Size: risk $ ÷ (stop pips × pip value). Round down.
- Portfolio check: does this position push total open risk past your cap, or duplicate an existing exposure? (P2.5 preview.)
The order matters. Size is the last thing decided, and it is an output, never an input. The moment you catch yourself choosing the lot size first and hunting for a stop that "fits", you have inverted the process — that is how a 1% rule quietly becomes a 5% reality.
Key takeaways
Volatility sets the stop (stop = k × ATR, k ≈ 1.5–3, outside noise and beyond structure); the stop sets the size — never the other way round.
Position (lots) = risk $ ÷ (stop distance in pips × pip value per lot). Always round down.
Same $risk means smaller positions in volatile markets and larger in quiet ones — this is a feature, and it keeps 1R constant so your expectancy stats mean something.
Tighter stops mean bigger positions and proportionally higher spread costs per R — tight-stop trading needs a stronger edge, not less capital.
Cap total open risk (e.g. 4–6% across all positions) as a first defence before the correlation lesson refines it.