Keltner Channel

A volatility-based breakout-continuation strategy built on the Keltner Channel - an ATR-scaled band wrapped around an EMA. Buy when price breaks above the upper band (treated as bullish momentum continuing, not an overbought signal), exit when price reverts back to the middle band.

Objective

Catch genuine breakout continuations by requiring price to clear a volatility-adjusted band, then give the trade room to run until price falls back to its own average (the middle band), rather than exiting at a fixed distance.

Entry

Close > UPPERKELTNER(20)

Price has closed above the upper band - EMA(20) + ATR(20) * 2.

Exit

Close < MIDDLEKELTNER(20)

Price has reverted back down through the middle band - EMA(20).

When it works best

Best in a trending, moderately volatile market - the band needs real width (from ATR) to make a breakout meaningful. In a very quiet market the band sits close to price and breaks trigger on small moves; in a genuinely range-bound market price will repeatedly poke through the upper band and immediately revert, producing losing round trips (see Performance Expectations).

Timeframes

The same statement works on any timeframe, but shorter intraday timeframes will produce more, noisier breakout signals and may need a longer period or volume confirmation.

Suitable instruments

Stocks, index futures, FX - any instrument with genuine trending behavior and real intraday/day-to-day range. Very low-volatility instruments will produce a tight band that breaks on small, insignificant moves.

Advantages

• Self-adjusting band width - wider in volatile conditions, tighter in calm ones, unlike a fixed percentage band
• Breakout-continuation framing (per the catalog description) rather than treating an upper-band touch as overbought - lets a strong trend keep running
• All three outputs (Upper/Middle/Lower) are computed once per candle no matter how many a strategy references
• Reuses the existing, already-tested EMAIndicator/ATRIndicator internally rather than re-deriving either calculation

Limitations

• Whipsaws in a range-bound market - repeated shallow breaks above the upper band with no follow through produce a string of small losses (see Performance Expectations)
• No trend-strength or volume confirmation in the base rule - any close above the upper band triggers, regardless of whether the underlying move has real participation
• The exit (revert to the middle band) can give back a substantial part of a move before firing, same tradeoff as any mean-reversion-style exit paired with a momentum-style entry

Best practices

• Add a trend/strength filter (e.g. ADX, a longer EMA regime check) before taking entries live, the same pattern used for other bare breakout rules in this library
• Re-tune period/multiplier per timeframe rather than reusing daily defaults unchanged intraday
• Watch for the middle-band exit giving back gains in a sharp reversal - consider pairing with a trailing stop for live use

Common mistakes

• Running the base statement in a range-bound market expecting every upper-band touch to be a real breakout - see Performance Expectations for how that plays out on real data
• Treating a Keltner upper-band break as an overbought/reversal signal - the catalog interpretation (and this strategy) treats it as continuation, the opposite reading
• Judging performance from one ticker or one short window - results below are mixed even across a single trending-large-cap basket