forex-scalping.com

Volatility-Adjusted Sizing: Same Risk, Different Lot Sizes

Risking 1 percent on every trade does not mean using the same lot size or stop distance. It means the stop distance adapts to the market and the lot size adapts to the stop.

Different pairs and different times carry different volatility. GBP/JPY might move 150 pips a day while EUR/CHF moves 50. A fixed 30-pip stop is generous on one and tight on the other.

Volatility-adjusted sizing starts with ATR. Decide that your stop will be, for instance, 1.5 times the daily ATR. On a pair with an ATR of 100 pips, the stop is 150 pips. On a pair with an ATR of 50, it is 75 pips. Then calculate the lot size so that the stop distance equals your fixed risk in dollars.

The result is that a wild pair gets a smaller position and a calm pair a larger one. Each trade carries the same dollar risk, and each stop has about the same probability of being noise.

This also adapts over time. When volatility spikes, positions shrink automatically. When markets quiet down, positions grow.

The approach helps avoid a hidden bias: trading the same lots on volatile and quiet pairs implies very different actual risks.

Practical step: calculate the ATR-based stop and lot size for three different pairs with the same dollar risk, and compare the lot sizes.

Worked example. EURUSD with an ATR of 10 pips on the 15 minute chart, stop at 1.5 ATR, so 15 pips. GBPJPY with an ATR of 25 pips, stop at 1.5 ATR, so 37.5 pips. To risk the same 1 percent, the GBPJPY lot size has to be much smaller. Same rule, very different position. With 50 dollars at risk, so EURUSD would be about 0.33 lots at roughly 10 dollars per pip per lot. On GBPJPY the value of a pip moves with the yen, so use a calculator there. The point is that the money at risk is the same on both, even though the stops and lot sizes look completely different.

The same idea applies across sessions on the same pair. EURUSD in quiet Asia might have half the ATR of the London open. If you use a fixed stop for both, you're either too wide in Asia or too tight in London. On a 15 minute chart, EURUSD might average 5 or 6 pips per candle between 2am and 6am Lagos and 12 to 15 pips per candle after 8am. A 10 pip stop is generous in the first window and gets clipped constantly in the second. Use the ATR reading at the time you enter, not an average across the whole day.

Common mistake I see: using ATR from the daily chart to size a 15 minute trade. The daily ATR is much bigger than the moves that matter for an intraday stop. Use ATR from the timeframe you're actually placing the stop on. Say the daily ATR on EURUSD is 70 pips and you set a 1.5 ATR stop from that, you'd have a 105 pip stop on a scalp, which makes no sense. The reverse is also a problem: a 15 minute ATR stop on a swing trade will be far too tight for daily swings. Match the ATR timeframe to the trade.

Volatility-adjusted sizing is central to risk management. A note on the example: with a 150-pip stop on a 100-pip ATR pair, the position must be small enough that 150 pips equals 1% of the account. On a small account this can run into the minimum lot problem, as you covered in another post.

In that case, a shorter timeframe with a smaller ATR, or skipping the pair, may be the only way to keep risk correct.

Fairman's session point has a timing trap at the London open that caught me for a long time. A 14 period ATR on M15 covers three and a half hours, so just after 08:00 London most of the bars it averages are still quiet Asian ones.
The chart shows a typical morning with round numbers: after the 08:15 bar the ATR reads about 7 pips, while the two London bars just printed ranged 15 and 14 pips. A 1.5 ATR stop of under 11 pips sits inside one ordinary bar, and with 50 dollars at risk the size comes out near 0.47 lots, against 0.25 once the reading catches up.

The ATR does not climb back above 12 pips until 10:30, so the first two hours are exactly when the sizing is most wrong. That matches Fairman's 12 to 15 pip London candles, just with a lag nobody mentions.

Until 10:30 I size from yesterday's ATR at the same hour instead of the live reading. After that the live number is close enough to use again.

Does anyone use a shorter ATR period around the opens instead of borrowing the previous day? I tried a 5 period version for a while, but it overreacts to the very first bar.