Bollinger Bands and ATR: How to Measure Volatility Instead of Guessing

Every chart tool covered so far on Otrai answers the same underlying question: which way is price headed. Moving averages smooth out trend direction. RSI measures momentum. Candlesticks and chart patterns read price action for clues about the next move. None of them answer a different, equally important question: how much is this thing actually moving right now. That is what Bollinger Bands and Average True Range (ATR) are built for. Neither one predicts direction. Both quantify volatility, and once you can measure volatility instead of guessing at it, your stops, position sizes, and expectations all get more realistic.

What Bollinger Bands Actually Measure

A Bollinger Band is not a support or resistance line, even though it gets treated like one constantly. It is a statistical band wrapped around a moving average. The standard construction is three lines:

Middle Band = 20-period Simple Moving Average (SMA)

Upper Band = Middle Band + (2 × standard deviation of the last 20 closes)

Lower Band = Middle Band − (2 × standard deviation of the last 20 closes)

Standard deviation is a measure of dispersion — how far recent closing prices have scattered away from their average. When closes are clustered tightly together, the standard deviation is small and the bands sit close to the middle line. When closes are jumping around, the standard deviation is larger and the bands push farther apart. That means band width is a direct, numeric readout of recent volatility. Nothing about the bands' position tells you where price is going next — it tells you how much price has been swinging while getting there.

The Squeeze: What Narrowing Bands Actually Tell You

When a security trades in a tight range, standard deviation shrinks and the bands pull inward — traders call this a squeeze. A squeeze does not predict direction. It says volatility has compressed and, historically, compressed volatility tends to resolve into expanded volatility. A coiled spring is a fair analogy, minus the part where you know which way it's about to snap.

Here is a real 32-day sequence: a stock consolidates tightly around $50 for 20 days, then breaks into a sustained upward move. The chart below plots the closing price against its 20-period Bollinger Bands starting once the SMA has enough data to calculate, at day 20:

At day 20, the band width is $0.35 — about 0.7% of price, a textbook squeeze. By day 32, the width has widened to $11.00, over 21% of price. That is the squeeze resolving into expansion, exactly as the concept predicts. What the squeeze did not tell you in advance was that the breakout would be upward rather than downward — you only learn direction once price actually commits to it.

Band Touches Are Not Automatic Signals

The most common misreading of Bollinger Bands is treating a touch of the upper band as an automatic sell signal and a touch of the lower band as an automatic buy signal. That logic holds up reasonably well in a range-bound market, where price oscillates between statistical extremes and reverts toward the middle band. It falls apart in a trending market, where price can ride — or "walk" — the band for an extended stretch.

Look again at the data above. From day 26 through day 32, the close finishes above the upper band on every single day. A trader shorting each touch of the upper band during that stretch would have been stopped out or fighting the trend for a full week. The bands were doing exactly what they're supposed to do — flagging elevated volatility — but that is not the same as flagging a reversal. Band touches mean something different depending on whether the broader structure is range-bound or trending, which is why support and resistance levels and trend context matter more than the band touch itself.

ATR: A Volatility Number With No Direction At All

Average True Range strips out the directional question entirely. ATR does not care whether price went up or down — only how far it traveled. The calculation starts with True Range (TR) for each period, which is the largest of three values:

TR = max[(High − Low), |High − Previous Close|, |Low − Previous Close|]

That third comparison is what a simple high-low range misses: overnight or between-session gaps. If a stock closes at $100 and opens the next session at $105 before trading in a tight $104-$106 range, a naive high-low calculation reports a $2 range. True Range correctly captures the $6 move from the prior close, because the gap itself was real price movement. ATR then smooths TR over a lookback period — 14 periods is the standard, using Wilder's smoothing method — into a single rolling average.

Using the same 32-day sequence, here is the 14-period ATR value tracked from the point it first becomes available:

ATR sits flat around $0.33 through the squeeze, then climbs to $1.16 by day 32 — a 3.5x increase, echoing the same expansion the Bollinger Bands showed as widening width. What ATR will never tell you is which way that $1.16 of average daily movement is likely to go next. It is a magnitude, not a forecast.

The Real Use Case: Sizing Stops to Actual Volatility

ATR's main practical job in a trading plan is not prediction — it is calibration. A fixed-dollar or fixed-percentage stop ignores the fact that volatility changes constantly. A 2% stop is generous during a quiet consolidation and dangerously tight during an expansion, where normal noise can trigger it before the trade has a chance to work. An ATR-based stop scales automatically: stop distance = entry price ± (multiplier × ATR), commonly using a multiplier between 1.5 and 3.

That stop distance then plugs directly into position sizing math. If you're risking a fixed 1% of a $10,000 account ($100) per trade, a wider ATR-based stop means fewer shares to keep the dollar risk constant. Using the same price sequence and a 2x ATR stop multiplier:

The dollar risk never changes — it's $100 in every scenario. What changes is the number of shares and the total position value, shrinking from roughly $7,300 down to about $2,500 as volatility expands. That's the entire point: an ATR-based approach forces position size down automatically when a security starts moving harder, instead of leaving you with the same share count and a much bigger dollar swing riding on it.

How the Two Actually Pair Together

Bollinger Bands and ATR are measuring the same underlying phenomenon from two different angles. Bands give you a visual read on volatility state relative to the current price — are we compressed, expanding, or is price walking outside the normal range. ATR gives you a clean number you can plug into a stop-loss or position-size formula. Neither replaces a directional tool. Pair them with something that actually reads trend or structure, like moving averages or support and resistance, and use the volatility tools for what they're good at: sizing risk correctly and knowing when a quiet market is about to stop being quiet.

Key Takeaways

• Bollinger Bands = 20-period SMA plus/minus 2 standard deviations. Band width is a direct measure of volatility, not a support or resistance line.
• A squeeze (narrowing bands) signals that volatility has compressed and often precedes an expansion — it says nothing about direction.
• In a strong trend, price can walk the band for days. Band touches are more meaningful in range-bound conditions than trending ones.
• ATR measures the average size of price movement, including gaps a simple high-low range would miss. It has no directional component whatsoever.
• The primary practical use of ATR is scaling stop-loss distance and position size to current volatility, rather than using a fixed dollar or percentage stop.

Volatility tools tell you how much room a trade needs and how much size you can safely take. They will never tell you which way price is going next — that job belongs to a different set of tools entirely.

Disclaimer: This content is for educational purposes only and does not constitute financial advice. Trading involves substantial risk of loss. Past performance does not guarantee future results.