Bollinger Bands and ATR Explained: Volatility, the Squeeze and ATR Stop Placement
Lesson 7 of 10 · Intermediate
How Bollinger Bands are built
Bollinger Bands were introduced by John Bollinger in the 1980s. The standard setting is 20 and 2:
- Middle band = SMA(20) of the close.
- Upper band = SMA(20) + 2 × the 20-period standard deviation.
- Lower band = SMA(20) − 2 × the 20-period standard deviation.
Because the distance between the bands comes from standard deviation, the bands widen when volatility rises and contract when it falls. Two derived measures are also useful: %B = (price − lower band) / (upper band − lower band), which shows relative position (above 1 means outside the upper band, below 0 means under the lower band), and bandwidth = (upper − lower) / middle, a relative measure of volatility.
A common belief is that “95 percent of prices stay inside the bands.” That figure comes from the normal distribution, but market returns are not normal and extreme moves occur more often than expected. The real proportion varies by market and period, so do not treat the bands as hard boundaries.
The squeeze and band walks
The squeeze. When bandwidth falls to its lowest level in a relatively long window, say several months on the daily chart or several days intraday, volatility has compressed. Market volatility tends to alternate between quiet and active phases, so compression often precedes expansion. But a squeeze does not tell you the direction. The first move out of a squeeze is sometimes a head fake: price breaks one side, then reverses and makes the main move the other way.
Band walks. In strong trends, price can touch the upper band or close above it for many consecutive candles. A touch of the upper band is not by itself a sell signal; in such a phase, a return to the middle band is often the first serious test of the trend. The classic beginner error is selling every touch of the upper band; that might work in a range but is damaging in a trending market.
ATR: a measure of true volatility
ATR (Average True Range) was introduced by Wilder in the same 1978 book. First the true range (TR) of each candle is computed, the largest of three values: high minus low, the absolute value of high minus the previous close, and the absolute value of low minus the previous close. This definition accounts for price gaps. Then ATR(14) is obtained with Wilder's smoothing: new ATR = (previous ATR × 13 + current TR) / 14.
ATR shows no direction; it simply tells you how much the market typically moves per candle. Its value is in price units, so on gold it is expressed in dollars. High ATR means a volatile market and low ATR a quiet one; neither is bullish or bearish on its own.
Using ATR for stop distance and position size
A stop placed much closer than the market's normal noise is often triggered by ordinary fluctuation. ATR helps you match stop distance to real volatility. A common method is to set the stop at a multiple of ATR (for example 1.5 to 3 times) from the entry, or beyond a structural level plus a fraction of ATR. The right multiple should be tested for your style and timeframe.
Hypothetical calculation: the 1-hour ATR(14) of XAUUSD is 8 dollars. With a 1.5 multiple, the stop distance is 12 dollars. Suppose the account is 10,000 dollars and risk per trade is set at 1 percent, which is 100 dollars. At many brokers one standard lot of gold is 100 ounces, so each one-dollar move is worth 100 dollars per lot. Approximate size = 100 ÷ (12 × 100) ≈ 0.08 lots. Always check the contract specifications with your own broker.
ATR on gold varies a lot through the day. An ATR computed in quiet Asian hours is far too small for the moment of the US jobs report. After a big news event, ATR also stays elevated for days and can make stops wider than necessary.
Hypothetical combined example
The numbers are hypothetical. On the 1-hour chart gold has ranged between 3860 and 3885 for two days, Bollinger bandwidth is at a multi-day low, and ATR has fallen from 11 to 6 dollars. This is the classic picture of compression. The analyst does not know the next direction, but knows volatility expansion is likely, especially if an important release is on the economic calendar.
Scenario: a break above 3885 with a close outside the band and follow-through on the next candle activates the bullish scenario; a quick return inside the range raises the odds of a fakeout and a move the other way. In both cases the low ATR of the compression period is not appropriate for setting a stop after the breakout, because volatility is rising; a stop based on the range structure is better.
Common mistakes
- Selling every touch of the upper band in a strong uptrend.
- Trying to guess the breakout direction from the squeeze alone.
- Placing a stop closer than the market's normal noise.
- Using quiet-hours ATR around major news releases.
- Believing that 95 percent of prices must stay inside the bands.
What does volume say here?
In volume-profile language, a Bollinger squeeze is a balance area: price has traded for a while in a narrow range and the profile has built a bell shape with a clear POC. The key question at the breakout is whether the market accepts new value. A breakout with rising volume, aligned delta and volume building outside the balance area is usually more credible; a breakout on low volume with price quickly returning into the value area is the head fake that Bollinger alone cannot identify. A common profile rule of thumb is that a return into the value area raises the odds of a move to the other side of it; like everything else, that needs testing.
For stops too, combining ATR with volume structure is more logical than either alone: rather than a fixed 1.5 ATR distance, you can place the stop beyond a high volume node (HVN) or on the far side of a low volume node (LVN), using ATR only to judge whether that distance is reasonable. For gold, profile and delta data come from COMEX GC futures converted with the basis; CFD volume is tick volume.
📊 See real gold futures volume behind these ideas: free order flow course · today's gold volume map
Frequently asked questions
What is the difference between Bollinger Bands and Keltner Channels?
Bollinger Bands take their width from standard deviation around an SMA, while the common version of the Keltner Channel uses ATR around an EMA. Some traders treat the Bollinger Bands moving inside the Keltner Channel as a sign of extreme volatility compression.
What ATR multiple should I use for a gold stop?
There is no fixed number. Multiples of 1.5 to 3 are common, but the right one depends on timeframe, style and time of day, and should be tested on enough data.
Are Bollinger Bands better for ranging markets?
In a range, the band edges are often reaction zones; in a trend, price can walk along the band. First identify the market regime, then interpret the bands accordingly.
Education only, not a recommendation to buy or sell; we never give signals. Price examples are illustrative. Test any method on a demo account first, with strict risk management.