Bollinger Bands and Volatility

Bollinger Bands wrap a moving average in an envelope that widens when volatility rises and tightens when it falls, giving you a live picture of how stretched price is relative to its own recent behavior. They are a volatility tool first and a signal generator a distant second — and the most expensive mistake traders make with them is reading "price touched the band" as an automatic buy or sell.

The Math, Briefly

A standard Bollinger Band setup has three lines:

• Middle band: a 20-period simple moving average of closing prices.
• Upper band: the middle band plus 2 standard deviations of those same 20 closes.
• Lower band: the middle band minus 2 standard deviations.

Standard deviation measures how spread out recent closes are. When price chops violently, the deviation is large and the bands sit far from the average; when price grinds quietly, they hug it. That adaptiveness is the whole point: unlike fixed-percentage envelopes, the bands recalibrate to current conditions automatically.

A worked example: suppose BTC's last 20 daily closes average $100,000 with a standard deviation of $2,500. The middle band sits at $100,000, the upper at $105,000, the lower at $95,000. If volatility doubles over the next month, the same average would carry bands at $110,000 and $90,000 — the tool has stretched to match the market.

The defaults (20 periods, 2 standard deviations) are conventions, not laws, but they're the settings the crowd watches, which gives them the same self-reinforcing relevance as other popular tools. One statistical caveat worth knowing: the textbook claim that 2 standard deviations should contain about 95% of observations assumes normally distributed data. Price returns — crypto's especially — are not normal; they have fat tails. In practice price spends more time outside the bands than the normal-distribution intuition suggests, which is precisely why a band touch is not an automatic reversal signal.

What the Bands Tell You at a Glance

Band width = volatility regime. Wide bands mean an active, expanded market; narrow bands mean compression. Many platforms plot this directly as Bollinger Band Width. Volatility is cyclical — quiet periods breed explosive ones and vice versa — so the width itself is information before you read anything else.

Position within the bands = relative stretch. A close near the upper band says price is expensive relative to its own last 20 periods — nothing more. In a strong uptrend, price can remain glued to the upper band for weeks. This behavior, called walking the band, is one of the most reliable visual signatures of a powerful trend, and it's exactly where "sell the upper band" reflexes destroy accounts.

The middle band as a reference. In trending phases, the 20-period average often serves as the pullback zone — the same dynamic-support logic covered in the moving averages guide, with the bands adding volatility context around it.

The Squeeze: Compression Before Expansion

The most widely used Bollinger setup is the squeeze: bands contract to their narrowest range in months, signaling that volatility has drained out of the market. Since volatility mean-reverts, unusually quiet markets tend to precede unusually active ones. The squeeze doesn't tell you direction — only that the spring is loading.

Example sequence. A mid-cap token trades between $0.95 and $1.05 for five weeks. Daily band width shrinks to its lowest reading in six months, with both bands inside a 6% envelope. Then a daily candle closes at $1.09, outside the upper band, on volume well above its recent average. That combination — squeeze, directional close beyond a band, volume expansion — is the classic squeeze-breakout template. The measured expectation isn't a specific target; it's that a new volatility expansion phase has likely begun in the breakout's direction.

Two failure modes to respect:

• The head-fake. Price breaks one way out of a squeeze, reverses within a couple of candles, and runs the other way. Squeeze traders often wait for a retest or use the opposite side of the breakout candle as invalidation for exactly this reason.
• Serial squeezes. In dead markets, bands can compress, expand mildly, and compress again without a real move. The squeeze raises the odds of expansion; it doesn't schedule it.

Mean Reversion vs Breakout: Know Which Game You're Playing

Bollinger Bands support two opposite trading styles, and mixing them up is the core Bollinger mistake.

Mean-reversion mode fits ranging markets. When price is chopping sideways, a tag of the lower band with a reversal candle — ideally at horizontal support — sets up a trade back toward the middle band. The middle band is the natural target, and the range's edge provides invalidation. In this regime, band touches genuinely do mark short-term extremes fairly often.

Breakout/trend mode fits expanding markets. After a squeeze breaks, or in an established trend, closes outside a band are strength, not excess. Walking the band means the correct play was holding or adding on middle-band pullbacks, not fading each upper-band touch.

The same event — price closing at the upper band — is a fade candidate in one regime and a continuation cue in the other. So regime identification comes first, and the bands themselves help: recently squeezed and now expanding favors breakout logic; wide, flat, sideways bands around a flat middle band favor range logic. Structure helps too — a market inside a well-defined horizontal range is in mean-reversion territory almost by definition; a market making new swing highs after compression is not.

Trying to fade band touches during an expansion is how the classic account-wrecker happens: shorting the upper band repeatedly as a trending market walks it higher, adding each time because "it's even more overbought now."

Practical Settings and Honest Limits

Keep defaults unless you have a tested reason. 20/2 on your trading timeframe is the shared reference. Some traders add a 2.5 or 3 standard-deviation outer band to distinguish "stretched" from "extreme," which is more defensible than tightening bands until they generate constant signals.

Bands lag, like everything built on moving averages. The middle band is a 20-period average; after a violent reversal, the whole structure takes time to catch up. Bollinger Bands describe the recent past's volatility — they don't foresee news, and a headline can blow through any band.

They pair best with independent evidence. Band signals firm up when they coincide with horizontal levels, volume behavior, or momentum divergence (an RSI divergence at a lower-band tag in a range is a stronger mean-reversion case than the band touch alone). Pairing bands with another volatility-derived indicator, by contrast, mostly duplicates information.

They are context, not a system. Nothing in the band math knows about liquidity, funding, or the news calendar. Treat the bands as a lens for asking better questions — "is this stretch normal for the current regime?" — rather than a machine that emits entries.

Key Takeaways

• Bollinger Bands plot a 20-period average with ±2 standard-deviation envelopes that expand and contract with volatility — they measure stretch relative to recent behavior, not absolute cheapness or expensiveness.
• A band touch is not a signal by itself; in trends, price can walk a band for weeks, and fading that is a classic account-killer.
• The squeeze — historically narrow bands — flags volatility compression that often precedes expansion, but it's directionless until a confirmed breakout, and head-fakes are common.
• Decide the regime first: mean-revert band tags toward the middle band in ranges; treat band rides as strength in expansions.
• Keep standard settings, expect fat-tailed markets to pierce the bands more than textbook statistics imply, and combine bands with structure, volume, or momentum rather than using them alone.