Multi-Scale Range Detection with Rotation and Failed-Break Recovery
Summary
The document presents a price range detector that tests three nested lookback windows and selects the longest window satisfying a set of structural filters. These include volatility compression ranked against the instrument’s own history, repeated closes across the range midpoint, repeated boundary touches, limited slope, close containment, and a full observation window. The rotation test is intended to distinguish a trading range from a V-shaped reversal, while percentile, extreme, or candle-body boundaries provide alternative ways to define the box.
After detection, the method can extend the range’s start to earlier contained bars and absorb small overshoots. A confirmed break can remain provisional for a set period; if price returns inside, the range is revived and the excursion is marked as a deviation. The document explains display states and trade interpretation, and reports synthetic-series testing, but gives no broad market backtest or evidence of profitability. It cautions that deviation markers are retrospective, and that percentile boundaries increase computation cost.
Key ideas
- Testing nested windows lets the longest window passing all filters define the range.
- Ranking volatility compression against its own recent history adapts thresholds to the instrument and interval.
- Midline crossing frequency helps separate repeated rotation from a one-pass reversal.
- Boundary touches, limited drift, close containment, and sufficient history add structural checks.
- A failed confirmed break can revive the range and mark the excursion extreme as a deviation.
- Deviation markers appear after a failed break and do not make the preceding breakout signal reliable.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.