Mechanism: what Bollinger Bands measure
Bollinger Bands are a technical indicator built from a moving average and two “outer” lines. The middle band is usually a moving average of the selected price (for example, closing price). The upper and lower bands are placed a fixed number of standard deviations away from that middle band, using the same rolling window length.
Because the bands are based on standard deviation, they expand when price variability increases and contract when variability decreases. This makes Bollinger Bands primarily a volatility and distance-from-average tool.
Direct answer: what “signals” people look for
When people say “signals from Bollinger Bands,” they typically mean observable relationships between price and the bands, or changes in the bands’ behavior. Conventional, widely used interpretations include:
-
Price touching or riding the upper band A common interpretation is that price is strong relative to the moving average, or that volatility is elevated. It can occur during trending phases, but it can also happen during choppy conditions where the market keeps crossing bands.
-
Price touching or riding the lower band Similarly, frequent lower-band interactions are often read as weakness relative to the moving average, again with the caveat that repeated touches can occur without a sustained direction.
-
Band expansion (widening) When the distance between the upper and lower band increases, the usual takeaway is rising volatility. Expansion by itself does not specify whether subsequent movement will be upward or downward; it mainly signals that variability is changing.
-
Band contraction (squeezing) When bands narrow, volatility is decreasing. Some traders treat contraction as a precursor to larger moves, but the direction and timing remain uncertain, and “squeeze” periods can resolve in multiple ways.
-
Middle-band reversion and crossings Interactions with the middle band (the moving average) are often interpreted as shifts in balance between price and the average. However, crossings can be frequent in range-bound markets, creating many potential “entries” if treated as standalone signals.
Example scenario-impact: how a conventional interpretation can mislead
Imagine a market where recent volatility rises sharply. The rolling standard deviation increases, so the bands widen. In that environment, price may touch the upper band repeatedly even if the longer-term direction is unclear.
Material limitation / failure mode: the indicator reacts to what already happened within the rolling window. If the market quickly changes behavior—such as shifting from trend-like movement to a range—then the same visual band behavior can lead to opposite expectations. A “touch” that once coincided with continuation can later occur during mean-reversion, and the visual pattern alone does not tell you which regime you are in.
Assumptions for the example: the interpretation assumes typical Bollinger Band construction (rolling moving average plus standard-deviation bands) and a fixed window and deviation setting. If your chart uses different settings or a different price type, the observed “signal” may change.
Relevant limitations and risks (including false signals)
- Volatility ≠ direction: Band width and standard deviation relate to variability, not whether the next move is up or down.
- Parameter sensitivity: Window length and deviation multiplier change the bands’ responsiveness. The “same” market can produce different touches/crosses under different settings.
- Regime changes: Relationships that appear consistent in one period may fail in another when trend strength, range behavior, or volatility dynamics change.
- Data and calculation differences: Different data sources, price fields, and timeframes can alter the moving average and the standard deviation used for the bands.
- Costs and execution matter: Even when a pattern appears correct on a chart, real-world outcomes can be affected by spreads, slippage, and order handling.
Verification and next question
To use Bollinger Bands more independently (without treating them as guaranteed predictions), verify interpretations with methods such as:
- Backtesting with the same settings you plan to use, over multiple market periods.
- Consistency checks: confirm that the “signal” definition (for example, “touch,” “close above,” “cross”) matches your interpretation.
- Context filters: compare band behavior with broader context like volatility regime and whether price is generally trending or ranging.
A useful next question is: **How do different Bollinger Band settings (window length and standard-deviation multiplier) change what counts as a “touch” or “squeeze”?