Bollinger Band Squeeze Screener

Fixed width versus historical percentile

Choose the compression rule before examining outcomes. The thresholds below are illustrative research settings, not validated trading recommendations. Keep the band settings, timeframe, session and data source consistent.

A reproducible compression example

Assume a completed daily bar has a middle band of $100, an upper band of $102 and a lower band of $98. Width is 100 x (102 - 98) / 100 = 4%. With a $101 close, position is (101 - 98) / 4 = 0.75, displayed as 75%. This passes the inclusive 4% rule but is still inside the bands.

• For a custom percentile test, sort the previous 100 valid daily widths ascending and use the 10th value as a nearest-rank 10th-percentile cutoff. If that value is 4.2%, the current 4% width qualifies. Equality qualifies under this example rule.
• If another symbol has the same 4% width but its 10th value is 2.5%, it fails the historical test despite passing the fixed test. The thresholds answer different questions.
• Require the full history before evaluating the percentile. With 20-close bands and 100 prior widths, the first test needs 120 daily closes. Skip missing widths and invalid denominators; do not replace them with zero.

Define squeeze release before confirming direction

One testable workflow requires three consecutive completed bars to pass the chosen compression rule. At the third close, save its threshold and the high and low of those three bars. For the next five completed bars, call the squeeze released when width first rises above that saved threshold and above the previous width. If confirmation does not occur within five bars, expire the setup. These are example rules, not a claim that three or five bars is optimal.

• For an upside confirmation, require release on the same bar as a close above both that bar's upper band and the saved three-bar high. For a downside confirmation, require release and a close below both the current lower band and saved low.
• An intrabar wick outside the band does not meet this close-based rule. A release without the required close changes volatility status but does not confirm direction.
• Use only data available at each close. Freeze the setup range and threshold; never redraw them using later highs, lows or widths. Record a signal at the close and model any execution afterward, with costs and gaps.
• If confirmation needs an additional bar or volume test, specify it in advance and timestamp the signal only when that evidence exists. Extra confirmation delays a decision and can still fail.

A sideways false breakout that passes the rules

Hypothetical chart-review sample: three compressed bars define a $99 to $102 range and a saved 4.2% width threshold. The next completed bar has middle $100.50, upper $102.70, lower $98.30 and close $103. Width is 4.3781%, up from 4%, and the close exceeds both $102.70 and $102. It passes the upside confirmation. A following close of $101.20 returns inside the saved range: the breakout failed to hold. This is an arithmetic illustration, not observed market data or a backtest.

• A separate bar that trades to $103 but closes at $101.50 below its $102.70 upper band fails the confirmation immediately. Do not count the intrabar high as a closing breakout.
• Quiet sideways markets may remain compressed for many bars or repeatedly break and return. Low activity, wide spreads and event gaps can make a plausible chart signal difficult to trade.
• Keep failed candidates in the research log. Define invalidation, maximum holding time and transaction costs before testing; a squeeze label supplies none of these rules.

Screen available fields, then verify the sequence

This public page reports daily Bollinger position, width percentage and width rank for its covered collections. It is a candidate table, not a historical sequence scanner. The public endpoint caches results for up to four hours, which does not bound the age of the underlying market data. Inspect the reported update time and confirm the relevant daily bar has closed before using a reading.

• Select the intended market collection and check symbol coverage. Missing symbols or values are unknown, not negative signals. Use the Market Screener link for the product workspace and its available fields.
• Compare bb_width_pct and bb_width_rank, then open each retained symbol on a matching TradingView daily chart. Match exchange, session, source and band inputs before comparing values.
• Check the preceding compression bars, saved range, release and close confirmation in order. A low rank today does not prove a breakout occurred after a prior squeeze.
• For the custom previous-100-bar percentile, consecutive-bar count or expiry rule, ask Pine Script AI Coding Agent to draft a script for a finite symbol list. Include the exact rules above, inspect its source and validate its chart behavior. This does not add unsupported conditions to the public table or scan every exchange symbol.