Price Hit the Bollinger Band. Why Are You Trading Against It?

Price touches the upper Bollinger Band. You sell.

It touches the lower band. You buy.

Simple. Elegant. Until the market keeps going.

Picture the baggage conveyor at an airport. A suitcase travels along the outer edge.

“It has reached the edge,” someone announces. “It must reverse now.”

You would probably look at the belt first. The suitcase can stay near that edge while continuing in the same direction. Where it sits and where it is going are different questions.

Bollinger Bands deserve the same treatment. They describe price relative to a changing statistical envelope. They do not build a wall that price must respect.

Zenith's stance: use the bands to understand the market's behaviour, then let price structure qualify the trade. Never buy or sell merely because a band was touched.

What the three lines actually tell you

The traditional setup uses a 20-period simple moving average in the middle. The upper and lower bands sit two standard deviations above and below it. A conventional TradingView setup is length 20, SMA basis, closing prices, standard-deviation multiplier 2 and offset 0. TradingView's indicator documentation.

“Period” means a bar on the selected calculation timeframe. Twenty hourly bars are not twenty trading days. Check the indicator's timeframe setting as well as the chart's.

Think of the middle line as the recent centre of price and the outer lines as a changing envelope around it. The envelope responds to the dispersion of the prices in its calculation window.

Suppose the average is $100 and the standard deviation is $2. The bands would be $104 and $96. If dispersion increases to $4 while the average stays at $100, the bands become $108 and $92. These are hypothetical calculations, not market quotes. Bollinger’s calculation method.

The price did not discover two new brick walls. The measuring tool changed its description of the market.

The upper band can mean strength, not exhaustion

John Bollinger's own rules explicitly distinguish a band touch from a trading signal. They also explain that price can repeatedly follow the upper band in an uptrend or the lower band in a downtrend. Bollinger's official rules.

That is often called “walking the band”.

Imagine a market forming higher highs and higher lows. The middle band is rising. Pullbacks are shallow, and buyers repeatedly push price back toward the upper band.

The repeated touches are consistent with strength. They do not prove a reversal is due because the indicator looks stretched.

Now compare a sideways market. Price repeatedly fails near the same high, the middle band is relatively flat, and moves through the range edge cannot hold.

An upper-band touch inside that context may help identify a rejection area. But the useful evidence is the failure to sustain higher prices, not the touch itself.

The same indicator can accompany two opposite trade ideas. If you decide “sell” before identifying which environment you are looking at, the indicator becomes an excuse rather than analysis.

Choose the playbook before the signal

Here is a practical three-way decision.

In a range, investigate rejection. Look for repeated boundaries and failed attempts to hold beyond them. A band touch near a prior wick zone is context. A subsequent failure and return into the range is additional evidence.

In a trend, investigate continuation. Respect the sequence of swing highs and lows. A pullback that holds an established support area may be more useful than trying to sell every upper-band touch. A lower-band touch during a persistent downtrend is not a bargain sticker.

In a squeeze, investigate expansion. The market has become quieter relative to its recent history. You still do not know which direction will produce a trade.

These are proposed decision principles, not a tested strategy with a claimed win rate. Borderline conditions are common. “I cannot classify this cleanly” is a perfectly good reason to do nothing.

A rising middle band alone also does not prove a trend will continue. It summarizes recent prices; it does not know the next buyer.

A squeeze tells you something is quiet, not which way it will go

When the bands narrow, traders often become excited: “A big move is coming!”

Perhaps. But narrowness must be judged relative to the instrument and timeframe. A quiet reading for one market may be ordinary for another. TradingView's BandWidth documentation makes that relative comparison important. Bollinger BandWidth.

BandWidth divides the distance between the upper and lower bands by the middle band, often displaying the result as a percentage. It helps compare the envelope's width through time rather than eyeballing a zoomed chart.

It does not provide an expiry time for the quiet period. Nor does a squeeze specify “up”.

Before acting, mark the actual price range that formed during the compression. Then observe which side breaks, whether price can remain outside that range, and where the nearest opposing structure sits.

A breakout followed immediately by a return into the old range is different evidence from a break that survives a retest. That does not make retests compulsory or successful. It gives a trader a specific event to evaluate rather than chasing the first exciting candle.

If the breakout has already travelled most of the distance to the next obstacle, the opportunity may be gone even though the bands are still expanding.

Look for a failed continuation, not just a band touch

One interesting Bollinger pattern compares two lows, not just one dramatic candle.

In a classic W-bottom example, price first falls near or below the lower band, rebounds toward the middle band, then makes another low. The second price low can be lower than the first while remaining inside its lower band. A subsequent break of the intervening resistance adds evidence of a possible reversal. TradingView describes this sequence in its official guide. W-bottom explanation.

Why is that useful? The second decline can be less extreme relative to its own changing envelope, even though its raw price is lower.

But the envelope has also changed. This is comparative evidence, not proof that sellers have disappeared.

The practical lesson is to demand a sequence: attempted continuation, failure, and an observable change in structure. A single candle below the band cannot tell that whole story.

You do not need to predict the exact bottom to participate in a recovery. You do need a reason to believe the previous behaviour is changing.

A worked trade that can still be rejected

Consider an entirely hypothetical stock trading in a repeatedly tested $94–102 range. Its middle band is around $98. The upper band is near the range high.

Price briefly trades to $102.20, then returns into the range. Suppose subsequent price action supports a short entry at $101. A structure-based stop-market trigger is placed at $102.50, beyond the rejection high. The first objective is $98, provided no nearer support has appeared.

Planned gross risk is $1.50 per share. Gross potential reward is $3.

Allow an illustrative $0.10 per share in total round-trip costs. Net potential reward becomes $2.90 and the risk allowance $1.60: approximately 1.81:1.

That passes an illustrative 1.5:1 screen, but it does not establish a profitable strategy. The rejection can fail, costs can be higher, and the stop can fill worse during a gap. Shorting also requires availability and may involve borrowing costs; losses can theoretically be unlimited. SEC short-selling guidance.

Now suppose you hesitate and enter at $100, leaving the same stop and target.

Net potential reward is only $1.90 against $2.60 of risk, approximately 0.73:1. The chart may still look attractive. The trade no longer does.

Do not push the target to the lower band simply to make the ratio look better. The middle of the range may be the first real obstacle. Nor should you squeeze the stop inside the failed high to rescue the arithmetic.

The bands helped describe the setting. Entry quality, invalidation and available space decided whether the trade deserved capital.

Your stop should not move because the band moved

A dangerous habit is placing a stop “just outside the band”, then moving it again when volatility expands.

The indicator has recalculated. That does not automatically justify taking more risk.

For the hypothetical short, the planned stop remains $102.50 beyond the rejection high. If triggered, a stop-market order seeks to exit at the available price, even if price later falls back. SEC stop-order guidance. The band can shift without justifying a wider stop.

If you intentionally design a dynamic trailing rule, define it before entering and test it. Do not convert an unplanned loss into a new strategy halfway through the trade.

Targets deserve the same discipline. The opposite band is a moving reference, not a promised destination. A prior swing, range midpoint or nearby support may matter sooner.

Two statistical traps and one hindsight trap

First, two standard deviations do not give you a reliable 95% probability that price stays inside the bands. Bollinger explicitly warns against applying normal-distribution assumptions to their construction. Official rules.

Second, an extreme reading does not guarantee an immediate return to average. A trend can change the average while you wait.

Third, the current candle and its indicator values can change before the candle closes. When reviewing a chart, distinguish a temporary intrabar excursion from a completed close. If your proposed method requires a close, a screenshot taken halfway through the bar is not the confirmation.

This is also why endlessly adjusting the length and deviation until yesterday's chart looks perfect proves very little. Test a fixed method across different periods, include costs, and judge it on observations not used to select the settings.

More attractive historical arrows are not the same as a stronger trading process.

Four answers before the next trade

Before the next trade, write four short answers:

• Is this a range, a trend, or compression with direction still unresolved?
• What price action, beyond touching a band, would confirm my idea?
• What structure would show the idea has failed?
• Is there enough room to the first credible obstacle after costs?

Back at the airport, the suitcase can keep following the outside edge of the conveyor. Its location does not overrule the belt’s movement.

Watch what is carrying price before deciding that an edge must turn it around.