Pattern Rotation

What It Does

Pattern Rotation reads two signals on each bar close: a Donchian-channel break (the entry trigger) and an ATR ratio (the regime classifier).

upper = max(highs over the prior N bars)
lower = min(lows  over the prior N bars)

ratio = ATR(recent) / ATR(baseline)

The regime classifier compares a fast ATR window to a slower baseline ATR window. When the recent ATR is at least the baseline ATR (ratio >= threshold, default 1.0), the strategy considers the market to be in an expanding-volatility regime and trades the break in its direction:

• expansion regime + close above upper band → long (continuation)
• expansion regime + close below lower band → short (continuation)

When recent ATR has shrunk below the baseline (ratio < threshold), the market is in a contracting-volatility regime and the strategy stays flat. No fade, no entry. The contraction filter exists because backtests showed that low-volatility breakouts mostly whipsaw, and skipping them improves both profit factor and trade economy.

Bracket exits handle the trade once entered: a fixed take-profit and stop-loss in instrument points.

Why This Beats Pure Breakout (when it does)

A plain Donchian-channel breakout strategy has to take every break, including the small ones during quiet, range-bound periods. Those false breaks erode the edge. Pattern Rotation’s volatility filter skips breaks that fire when the market is already calmer than its recent baseline, which is structurally where most chop happens. Same direction, fewer trades, more profit per trade.

The improvement is most visible on equity-index futures at intraday timeframes, where volatility regimes alternate cleanly. On daily bars, volatility regimes change too slowly for the filter to add much; the strategy effectively reduces to its breakout backbone with fewer trades.

Key Characteristics

• Entry signal: Donchian-channel break of the prior N bars (current bar excluded)
• Regime classifier: ratio of recent ATR to baseline ATR
• Trades only in expansion regime; stays flat in contraction
• Bidirectional within the expansion regime: long upside breaks, short downside breaks
• Fixed-points bracket exits
• One position at a time

Known Weaknesses

• The regime classifier is a single-bar reactive measurement. It can flip back and forth across the threshold from one bar to the next. A confirmation rule (require N consecutive bars in expansion) is a natural extension that this version skips
• Daily timeframes barely benefit. The volatility regime moves too slowly for the filter to add value when each bar is a full day
• Contraction-regime opportunities are forfeited entirely. If a real mean-reversion edge exists in low-vol periods, this strategy will not capture it
• All the weaknesses of the underlying breakout strategy apply: false breaks during chop are still possible inside expansion regimes; fixed-points sizing needs re-tuning per instrument

Best In

• Trending equity-index futures at intraday timeframes (1H has been the strongest in backtests)
• Markets where volatility regimes alternate clearly between active and quiet phases
• Use cases where you want a regime-aware version of breakout-continuation rather than always-on participation