Mean reversion bets that a price which has moved unusually far from its average will drift back towards it. Range trading is its simplest form: sell near the top of a sideways range and buy near the bottom. Both styles tend to win often and in small amounts, so a single loss left to run can undo weeks of careful trades.
What you will learn
- How to tell a range worth trading from a trend that is only pausing
- Four ways to measure a stretched price: ATR distance, Bollinger Bands, the z-score and RSI
- Where the entry, target and stop go in a range
- Why the stop decides whether the style works, shown in dollars on a Classic account
Find a range worth trading
A range is a stretch of sideways movement between a top and a bottom that price has turned from more than once. Look for at least two clear turns at the top and two at the bottom, and a 50-period moving average running flat through the middle. The support and resistance lesson shows how to mark the edges as zones.
Take a one-hour AUD/USD chart. Over three days, price has turned down near 0.6700 three times and turned up near 0.6600 twice. The range is 100 pips high, the midpoint is 0.6650, and the 50-period average is flat at about 0.6650. That is a range worth watching.
Be wary of two look-alikes. A short, tight range straight after a strong move is often a flag, a pause before the trend carries on, as the chart patterns lesson explains. And a range that forms ahead of a big data release can break the moment the number comes out.
Measure how far price has stretched
A stretch measure tells you how unusual the current price is compared with its recent past. Four are in common use:
- ATR distance: how many average true ranges price sits from its moving average.
- Bollinger Bands: a 20-period moving average with lines two standard deviations above and below it.
- The z-score: (price minus the average) ÷ the standard deviation. A z-score of +2 puts price on the upper Bollinger Band.
- RSI: readings above 70 or below 30 mark a stretched move, as the indicators lesson showed.
Back to AUD/USD. The 20-bar average is 0.6650 and the standard deviation is 18 pips, so the upper band is at 0.6650 plus 36 pips, which is 0.6686. Price rallies to 0.6695. The z-score is 45 pips ÷ 18 pips, which is +2.5. Price is 2.5 standard deviations above its average and at the top of the range, so the two signals agree.
None of these tools says when price will turn. In a strong trend, price can ride along the upper band for days, and a z-score of 2.5 can become 4. Use a stretch reading to decide which range edge to watch, and base the entry and the stop on the range itself.
Place the entry, target and stop
You sell at 0.6690. The stop goes outside the range, where the idea is proven wrong: 20 pips above the top of the range at 0.6700, which puts it at 0.6720. That is a 30-pip stop.
On a $10,000 Classic account at 0.5% risk, you can lose $50. $50 ÷ 30 pips is $1.67 a pip, so round down to 0.16 lots at $1.60 a pip, and 1R is $48. One plan is to close half at the midpoint and the rest just above the bottom of the range:
- Half, 0.08 lots, closes at 0.6650: 40 pips × $0.80 is $32.
- The other half closes at 0.6610: 80 pips × $0.80 is $64.
- The total is $96, or 2R. Move the stop to your entry once the first half is closed.
Keep the stop where it is
Ranges end, and when one breaks, a mean-reversion trader without a stop is short a market that may have started to trend. Run the same trade without the stop. You sell 0.16 lots at 0.6690, then, later the same day, sell another 0.16 at 0.6720 and again at 0.6750, because price looks even more stretched. A data release pushes AUD/USD up to 0.6830:
- The first sale is 140 pips down: $224.
- The second is 110 pips down: $176.
- The third is 80 pips down: $128.
The open loss is $528. The Classic daily loss limit on $10,000 is $500, measured on equity, so the account ends during the move, even if price later falls back into the range. With the stop at 0.6720, the same breakout cost $48, or a little more if the release made the stop fill past its level.
That one loss is 11R. Suppose the range method wins 60% of its trades, with an average win of 1.1R and an average loss of 1R. Expectancy is 0.66 minus 0.40, or +0.26R a trade, so an 11R loss wipes out the average result of about 42 trades. Set the stop before you enter, never add to a losing position, and stop trading a range once price closes outside it. The breakout lesson picks up from there.
Check your understanding
The 20-bar average is 1.0850 and the standard deviation is 15 pips. Price is at 1.0805. What is the z-score, and is price outside the lower Bollinger Band?
1.0805 is 45 pips below the average, and 45 ÷ 15 gives a z-score of -3. The lower band sits two standard deviations down, at 1.0820, so price is 15 pips outside it.
On a $25,000 Classic account at 0.5% risk, you buy EUR/USD at the bottom of a range with a 25-pip stop. What size do you trade, and what is a target at the midpoint, 50 pips away, worth?
0.5% of $25,000 is $125. $125 ÷ 25 pips is $5 a pip, or 0.50 lots. The 50-pip target is worth $250, which is 2R.
Price closes above the range you were selling. Why not add to the short at a better price?
The close outside the range is the evidence that the range idea has failed. Adding doubles your size just as price may start to trend, and the open loss counts against the daily limit straight away.
Key points
- Trade a range only after two turns at each edge, with a flat moving average through the middle.
- ATR distance, Bollinger Bands, the z-score and RSI measure how stretched price is, but none of them times the turn.
- Put the stop outside the range when you enter, and size the trade from that stop.
- Mean reversion wins often and in small amounts, so one large loss can undo dozens of trades.
- A close outside the range ends the range trade. Never add to a losing position.
Next lesson: Trade breakouts and filter out the false ones
All trading is simulated. Rewards are based on performance and are not guaranteed.