Bollinger Bands: The Traps to Avoid and How to Use Them Right

Bollinger Bands are one of the most used indicators in the world, and one of the most misunderstood. The most common belief: "price touches the upper band, it's overbought, I sell; it touches the lower band, it's oversold, I buy." That interpretation turns an excellent volatility tool into a loss generator.

The problem is that Bollinger Bands don't measure what most traders think they do. They don't tell you whether a price is "too high" or "too low." They tell you how volatile the market is right now. Once you internalize that, you stop using them backwards, and you discover what they're actually good for.

What Bollinger Bands really measure

Let's go back to the construction, because it explains everything. Bollinger Bands are three lines:

• A moving average in the middle (by default a simple 20-period average).
• An upper band, placed two standard deviations above the average.
• A lower band, two standard deviations below.

Here's the key: the bands are built on standard deviation, a statistical measure of how spread out price is, which means volatility. When the market is choppy, standard deviation rises and the bands widen. When it's calm, standard deviation falls and the bands tighten.

In other words, Bollinger Bands are a volatility envelope around a moving average. Their width isn't decorative, it's the main information. The distance between the bands tells you whether you're in a wild phase or a quiet one. It's a natural complement to the ATR, which also measures volatility but differently.

By statistical construction, roughly 88 to 90% of price action happens inside the bands at two standard deviations. Hold onto that number, because it's the root of the biggest misunderstanding.

Trap 1: "touching the band means reversal"

This is the mistake that wrecks the most accounts with this indicator. The faulty logic: "price is at the upper band, so at a statistical extreme, so it'll revert to the mean, so I sell."

In a range, that can work. In a trend, it's a one-way ticket to a loss.

Why? Because in a strong trend, price does what's called "walking the band." It hugs the upper band and stays there, session after session, while the trend develops. Every time you sell "because it's at the band," you're shorting an uptrend in progress. You get shredded, again and again, exactly like the seller shorting an RSI above 70 in the middle of a trend.

Take a worked example. Picture gold in a strong uptrend. Price touches the upper band at $2,350. You sell, sure of the reversal. Price keeps going, hugging the band as it rises with it: $2,380, $2,410, $2,450. Over ten sessions, it "touched the band" ten times without ever reversing. Your stop got hit long ago. The band was never a ceiling: it was the wake of a powerful trend.

The rule: a band touch is not a signal by itself. It's a reversal signal only when the regime is a range, and confirmed by something else (structure, divergence). In a trend, it's more a sign of strength.

Trap 2: the bands aren't reliable support or resistance

Corollary of the first trap. Many traders treat the upper and lower bands as support and resistance, the same as a real horizontal level.

It's not the same nature. A support or resistance level comes from the real history of transactions: a price where the market has already reacted several times. The Bollinger band is purely mathematical: it's computed from the last 20 candles, it moves on every new close, and it has no memory of what happened before. It drifts with price.

Worse: the famous "90% of action inside the bands" rests on an assumption that returns are normally distributed. But markets have fat tails: extreme moves happen far more often than a normal distribution predicts. Band breaks are therefore more frequent than the theoretical statistic suggests, especially on news. Relying on the band as an airtight barrier ignores how markets actually behave.

Real use 1: the squeeze

Here's the most usable Bollinger Bands signal, and the one almost nobody talks about enough: the squeeze.

When the bands tighten sharply, it means volatility has collapsed: the market is calming down, contracting, hesitating. And volatility runs in cycles. A period of low volatility is almost always followed by an expansion. Concretely, an extreme squeeze of the bands often precedes a violent move.

But here's the crucial point: the squeeze tells you a big move is coming, not which direction. The bands never give you direction, only the imminence of the move. It's a classic error to think the squeeze is bullish or bearish in itself. It's neutral.

For direction, you have to pair it with something else: price structure, a level break, the trend context. The squeeze says "get ready"; structure says "this way." That's exactly why reading the market regime before acting is essential: the squeeze flags the transition from a range to a possible trend, but it's the breakout that decides.

In practice: spot a marked tightening, wait for the bands to expand alongside a clean close beyond a level, and enter in the direction of the break with a stop on the other side. You don't guess the direction, you follow it once it's confirmed.

A worked squeeze, with numbers

Take EUR/USD on an hourly chart. For two days, the pair stalls in a tiny range: band width drops to around 30 pips, against 80 to 100 in normal conditions. That's a clear squeeze: volatility has collapsed, the market is compressing a spring.

The trap would be to pick a direction now. Mistake: the squeeze doesn't say where, only that something is coming. You wait.

The break arrives. A candle closes cleanly above 1.0850, the top of the consolidation, and the bands snap apart. That's where you enter, long, in the direction of the break, with a stop below the bottom of the range at 1.0820. The move that follows a real volatility expansion often travels several times the width of the squeeze: here, 100 to 150 pips wouldn't be surprising. You guessed nothing. You waited for the market to reveal its direction, then followed it with defined risk.

Notice what the bands did and didn't do. They flagged the compression (get ready) and the expansion (it's happening). They never told you up or down. The level break did. That division of labor is the whole point: bands for volatility, structure for direction.

Real use 2: match your reading to the regime

The synthesis of everything above fits in one sentence: Bollinger Bands don't read the same way depending on the market regime.

In a range, the bands frame the oscillations. There, fading the extremes makes sense: sell toward the upper band, buy toward the lower band, target the return to the central average. It's the one context where the "overbought/oversold" reading roughly works, and even then, with confirmation.

In a trend, you flip the logic completely. The band in the direction of the trend becomes a continuation zone, not a reversal zone. The central average (the basis) often acts as dynamic support in an uptrend: pullbacks to the average become chances to enter with the move, not to bet against it.

That's also the philosophy behind the patterns John Bollinger himself described, like the "W-bottom" (a double bottom where the second low stays inside the lower band, a sign of selling exhaustion) or the "M-top" on the other side. These setups combine the position relative to the bands with price structure, never the band alone.

A last word on settings. The default pair (20 periods, 2 standard deviations) is a solid universal starting point. On shorter, jumpier timeframes, some shorten the period; to capture bigger moves, others push to 2.5 standard deviations. But don't fall into endless optimization: a setting that's perfect on the past guarantees nothing about the future, as the backtesting article reminds you.

It also helps to add a second indicator that reads %B or bandwidth as a number rather than eyeballing the bands. Bandwidth (the distance between the bands, normalized) lets you spot a squeeze objectively: when it drops to the low end of its recent range, you're in compression, no guessing required. %B tells you where price sits relative to the bands on a 0-to-1 scale, which is cleaner than judging "near the band" by eye. Neither changes the logic above, they just make the volatility read mechanical instead of subjective, which matters when you're trying to follow rules instead of moods.

Bollinger Bands are an excellent volatility radar once you stop asking them to do what they can't: predict a reversal because a price "is too high." Use them to read volatility and anticipate expansions, not to guess tops and bottoms.

And as always, the only way to know whether they improve your decisions is to measure. Tag the trades where the bands guided your entry and compare their expectancy to the rest by tracking your trades on TradesStack: you'll know whether this indicator gives you a real edge, or just clutters your chart.