Short answer: a bollinger bands strategy uses three lines, a moving average with an upper and lower band placed a set number of standard deviations away. The bands widen when volatility rises and narrow when it falls. The most reliable use is reading volatility conditions and setting expectations for range, not treating a touch of a band as a buy or sell signal.
The three components
The middle line is a simple moving average, conventionally twenty periods. The upper and lower bands sit two standard deviations above and below it, calculated from the same twenty periods of closing prices.
Because standard deviation measures dispersion, the bands respond directly to recent volatility. When price has been moving in large steps, the bands are far apart. When it has been quiet, they contract toward the average.
| Component | Default | What it represents |
|---|---|---|
| Middle band | 20 period simple moving average | The recent mean price |
| Upper band | Middle plus 2 standard deviations | Statistically stretched upside |
| Lower band | Middle minus 2 standard deviations | Statistically stretched downside |
| Bandwidth | Distance between the bands | Current volatility level |
| Percent B | Where price sits within the bands | Position relative to the range |
What a band touch actually means
This is the point where most traders go wrong. A touch of the upper band does not mean overbought, and a touch of the lower band does not mean oversold.
By construction, roughly ninety five percent of observations fall within two standard deviations of the mean if the data is normally distributed. Price data is not normally distributed, but the general point holds: a band touch is a statistically notable event, not a rare one. In a strong trend, price can ride the upper band for many consecutive periods, and traders selling every touch lose steadily.
The more accurate reading is that a band touch tells you price is extended relative to its own recent behaviour. Whether that means reversal or continuation depends on the trend, which the bands themselves cannot tell you.
Strategies that use the bands well
The squeeze. When bandwidth falls to a multi-month low, volatility has compressed. Compression is followed by expansion eventually, so the squeeze identifies conditions ripe for a large move without saying which way. Traders use it to prepare rather than to enter, often placing orders on both sides of the range.
The walk. In a strong trend, price rides one band while the middle line acts as support or resistance. Recognising a walk stops you from fading a trend repeatedly. The practical rule is that if price has closed outside the band more than twice in recent periods, treat the market as trending rather than ranging.
Mean reversion in confirmed ranges. When a market is genuinely range-bound, fading band touches back to the middle line is a workable approach. The critical word is confirmed: the strategy requires evidence of a range before it is applied, not the assumption of one.
Band width as a position sizing input. Wide bands mean larger typical moves and therefore wider stops and smaller positions. This is arguably the single most useful application and the least discussed.
Combining the bands with other tools
The bands measure volatility and location. They say nothing about direction or momentum, so pairing them with a directional input is sensible.
A trend filter, such as the slope of the middle line or a longer moving average, answers whether to fade band touches or respect them. A momentum reading can flag when a walk is losing force. Volume tells you whether an expansion out of a squeeze has participation behind it.
What does not help is adding a second volatility measure. Bands, envelopes, and volatility channels all describe the same property, and stacking them produces agreement rather than information.
Risk management with volatility bands
- Set stops outside the band, not on it. A stop resting exactly at a band sits inside the range price routinely covers.
- Scale position size to bandwidth. When bands are twice their usual width, halve the position for the same risk.
- Define invalidation from structure. The bands move as price moves, so a level derived from them is not fixed. Use swing points for the actual stop.
- Limit mean-reversion attempts. Fading a trending market is the classic failure with this tool, and capping attempts per session contains the damage.
- Account for expansion around events. Scheduled releases produce sudden band widening, and calendars such as the Federal Reserve FOMC schedule tell you which sessions carry that risk.
- Include transaction cost. Mean reversion strategies trade frequently, and spread plus commission charges are a larger share of the expected profit than in trend strategies.
Leverage amplifies each of these decisions. FINRA's explanation of margin accounts sets out the mechanics, and under a prop firm evaluation with a fixed daily loss limit, a run of failed mean reversion trades in a trending market can end an attempt within a single session. Pivex sets those limits out in advance, so it is practical to convert bandwidth into a maximum position size in advance.
Settings and when to change them
The twenty period, two standard deviation default is a convention rather than an optimum. It has the advantage of being widely watched, which gives the levels a mild self-reinforcing quality.
Shortening the lookback makes the bands more responsive and produces more touches. Lengthening it produces a smoother envelope with fewer, more significant excursions. Raising the deviation multiple to two and a half or three reduces the number of touches and makes each one more notable.
The mistake is to adjust settings after a losing trade. That guarantees you are always using the configuration that just failed and prevents any meaningful assessment of whether the approach works.
A worked read of a full cycle
Following one complete cycle makes the tool much easier to use than any list of rules.
Start with a quiet market. Bandwidth falls steadily as daily ranges shrink, and price oscillates in a narrow area around the middle line. Nothing here is tradeable, and the correct action is preparation rather than entry.
Bandwidth then reaches a multi-month low. This is the squeeze. Mark the high and low of the compressed area, because those two prices are where the eventual expansion will be measured from.
Price breaks out of the compressed area and closes outside the upper band. Bandwidth jumps as the standard deviation calculation absorbs the larger move. A common error at this moment is to fade the band touch, since the reading looks stretched. In practice a breakout from compression frequently produces a walk rather than a reversal.
The walk develops. Price closes outside or near the upper band for several consecutive periods while the middle line rises beneath it and holds on pullbacks. During the walk, band touches are continuation evidence and the middle line is the level worth watching.
Eventually price fails to reach the band on a new high, or closes back below the middle line. That is the first genuine sign the move is losing force. Bandwidth begins to flatten, then contract, and the cycle starts again.
Recognising which phase you are in matters far more than any single signal, because the same band touch means opposite things in the compression phase and in the walk.
Common mistakes to avoid
- Treating band touches as signals. They are context, and in a trend they are continuation evidence rather than reversal evidence.
- Trading the squeeze directionally. Compression tells you a move is likely, not which way.
- Ignoring the middle line. The moving average is a genuine reference level and often more useful than either band.
- Using bands in illiquid instruments. Erratic prices produce erratic standard deviations, and the bands become meaningless.
- Assuming a fixed statistical property. The two standard deviation rule assumes a distribution markets do not follow, particularly in the tails.
- Adding more indicators to fix the ambiguity. The ambiguity is inherent, and the answer is a directional filter, not another oscillator.
Next step: chart bandwidth as its own line beneath price for the instrument you trade, mark the ten narrowest readings over the past year, and note what happened in the following ten sessions. That single exercise separates the squeeze concept from the folklore around it.
Before participating, review Pivex's current program rules and decide whether a simulated proprietary trading evaluation fits your experience and risk limits. Participation does not guarantee a reward, income, or any particular trading result.