Does touching the lower band mean I should buy?

No. A touch identifies price relative to the bands; it is not a complete trading signal. Define the reversal or continuation hypothesis and test it with explicit exits and costs.

Do two standard deviations imply 95% future price coverage?

No. The band construction does not establish a normal distribution or guaranteed future containment. John Bollinger cautions against making that statistical assumption.

Do wider bands guarantee fewer false signals?

Wider bands change the levels for the same calculation window. Whether resulting entries are more useful depends on the strategy and data; reduced activity does not prove improved accuracy.

Will adding RSI improve the strategy?

It adds a condition and additional research choices. Compare a defined baseline against the filtered version with consistent costs and untouched evaluation data rather than assuming improvement.

How should I define a squeeze?

Choose a specific BandWidth measure, lookback and low-width criterion. Record the rule and test the subsequent entry separately; a squeeze does not specify an assured direction or return.

Can grid search settle the best exit rule?

It can compare tested historical results under fixed assumptions. It cannot establish a universally best exit or guarantee that the selected rule will remain preferable.

Bollinger Bands Strategy Optimization for TradingView

Define Bollinger Bands strategy rules before searching length, width and entry thresholds. Compare reversal and breakout assumptions without treating band touches as proof.

View Chrome Extension

Building a Strategy with Bollinger Bands

Traditional Bollinger Bands place outer bands around a simple moving average using a standard-deviation multiplier. The default length and multiplier are starting conventions, not demonstrated best settings for every market. John Bollinger explicitly distinguishes a band touch from a standalone buy or sell signal; price can continue along a band in a trend.

Choose the hypothesis before searching: a reversal rule after price returns inside a band differs from a continuation rule after an outside close. Specify the exact bar, comparison operator, entry timing, exit and position size. Define any percentage threshold using an explicit denominator. A “1% beyond the lower band” rule is not the same as moving one percent of the band width.

If you compare exits, name both behaviors. For a long entered near the lower band, a middle-band target is generally closer than an upper-band target at that moment, but bands move and fills depend on the order rules. It is incorrect to universally describe opposite-band exits as earlier than middle-band exits. A squeeze filter also needs a measurable width criterion and lookback; narrow bands do not determine the direction of a later move.

Optimizable Parameters

Bollinger Bands strategy inputs and illustrative test values
Parameter	Description	Example Test Values
Band length	Length used consistently by the chosen middle-band and dispersion calculation	Illustrative: 10–40, step 2
Width multiplier	Multiplier applied to standard deviation, not a guaranteed probability interval	Illustrative: 1.5–3.0, step 0.5
Entry threshold (%)	Define whether the denominator is band price, middle-band price or band width	Illustrative: 0.5–2.0, step 0.25
Exit rule	Compare explicitly implemented middle-band and opposite-band exits	Hold one rule fixed, or test both separately

Optimization Example and Test Design

Hypothetical design: lengths 10–40 by 2 provide 16 values, width multipliers 1.5–3.0 by 0.5 provide 4, and thresholds 0.5–2.0 by 0.25 provide 7. That is 16 × 4 × 7 = 448 combinations for one fixed exit rule. Testing two exit rules raises the nominal total to 896. This is a proposed grid, not an EURUSD experiment or a measured runtime. Make the threshold reproducible: if a lower band is 100 and the rule means one percent below that band price, the level is 99. If the bands are 100 and 110 and the rule instead means one percent of their width, the level is 99.9. These illustrative values show why the unit definition matters. Keep the source, feed, timeframe, test dates, costs and sizing with every result. Review trade count and drawdown as well as profit, and inspect nearby parameter settings. Freeze the chosen rule before testing a later interval. A drop in profit factor can reflect selection bias, sampling variation or changed conditions; the two numbers alone do not establish overfitting.

Sources

Frequently Asked Questions

Related Strategy Optimizations

rsi moving average crossover stochastic macd

Start Optimizing Your Bollinger Bands Strategy

Install the extension, sign in and check your plan access. Automated parameter optimization requires an Advanced plan or above.

View Chrome Extension