Arithmetic Detection of Bollinger Band Double-Bottom Patterns

Summary

This example turns a visual Bollinger Band double-bottom, or W-shaped pattern, into arithmetic rules. It builds 20-period bands from a rolling mean and standard deviation, then scans a 75-bar window for a sequence of candidate nodes: an initial lower-band touch, a middle point near the moving average, and a second low near but above the lower band and below the first low. A later price move above the upper band confirms a long signal. The script records node positions for plotting and exits when band width contracts below a threshold.

The example uses GBP/USD bid-ask averages and illustrates signals with a chart, but supplies no numerical backtest statistics or evidence of profitability. Its fixed thresholds and window may depend heavily on price scale and instrument; the source does not describe parameter calibration, transaction costs, out-of-sample validation, or position sizing. The author favors explicit arithmetic pattern rules over machine learning and notes that a related top pattern would reverse the bottom pattern’s logic.

Key ideas

• The method calculates 20-period Bollinger Bands using two standard deviations around a moving average.
• A 75-bar scan identifies candidate points for a double-bottom shape using band and price relationships.
• A move above the upper band confirms a long signal after the candidate pattern forms.
• The script clears the position when band standard deviation falls below a threshold.
• The example plots GBP/USD signals but supplies no performance statistics or validation.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.