A moving-average centerline wrapped in bands set by a multiple of ATR — steadier than its better-known cousin, genuinely useful for judging trend strength, genuinely quiet exactly when a market goes nowhere.
American grain trader Chester Keltner described a simple channel built from a 10-day average of typical price and the 10-day average trading range, in his book "How to Make Money in Commodities."
Keltner never claimed the idea as original and didn't call it a "channel" himself — others attached his name to it after hearing about the rule.
Floor trader and later "Market Wizard" Linda Bradford Raschke replaced the simple range with Wilder's ATR, and the SMA with an EMA, creating the version traders use today.
Valued for smoother, more constant width and for showing genuine trend strength rather than raw statistical spread.
The centerline is typically a 20-period EMA; the upper and lower bands sit a multiple of ATR (commonly 2×) above and below it.
Because ATR is smoother than standard deviation, Keltner's width changes more gradually — it doesn't spike and contract as sharply around a single big bar.
Price staying inside the channel says nothing much is happening; a genuine close beyond either band is comparatively rare and often marks real strength or weakness.
Through that run, price rode along the upper band for weeks straight — the channel didn't flag it as overbought, it simply kept containing a genuine, strong trend.
After weeks of narrow, quiet trading inside the channel, a decisive close beyond the upper band marked the start of a genuine new trend leg.
Price rides along the upper band for two straight weeks in a strong, genuine uptrend. A trader treats each new touch as an automatic sell signal. Sound?
A quiet, narrow-channel consolidation is suddenly broken by a decisive close beyond the upper band. A trader dismisses it as noise. Sound?
A trader assumes Keltner Channels and Bollinger Bands will always look identical, since both are three-line price envelopes. Accurate?
Price and the channel, watched tick by tick on the left — and the mark it leaves in the ledger on the right. A confirmed band-ride, a quiet consolidation — and a genuine breakout.
Read the run leading into the band touch, then call it: a genuine trend worth riding, or an isolated spike likely to reverse?
The classic error is reading every band touch the same way. The discipline is mechanical: read the centerline's slope for trend context first, and require a genuine close, not a wick, before calling a breakout.
Wrapping an EMA in Wilder's ATR gives traders a steadier, genuinely calmer read on trend and volatility — but that same calm means it only speaks loudly when the market genuinely has something to say.
You have to be willing to take what the market gives you.