If you've worked with charts at any point in the last ten years, you know Bollinger Bands. John Bollinger developed them in 1980 — standard deviation × 2 around an SMA. They're built into every charting tool and mentioned in every trading guide.
What many don't know: Bollinger's concept is based on an older idea. Chester Keltner described his own volatility-band system in 1960 — twenty years before Bollinger. Linda Bradford Raschke modified it in 1980 into the form we know today as "Keltner Channels". Instead of standard deviation, Keltner uses the Average True Range (ATR) as the volatility measure.
Today Keltner Channel Breakout goes live as the tenth strategy in the Backtesting Arena. Here's an honest introduction — what it is, when it works, and when it doesn't.
The Mechanic in One Table
| Component | Calculation | Default |
|---|---|---|
| Mid-Line | EMA over closing prices | 20 periods |
| Upper Band | Mid + ATR × Multiplier | ATR(10), 2.0× |
| Lower Band | Mid − ATR × Multiplier | ATR(10), 2.0× |
The bands are dynamic: they widen in volatile market phases and narrow in calm ones. That's the central difference to fixed indicators like SMA crosses.
Trade Logic
The standard breakout rule:
- BUY: Close crosses above the upper band → the market breaks through the upper volatility threshold, we take this as a trend signal.
- SELL: Close crosses below the mid-line (default in our implementation) or alternatively below the lower band (classic variant).
The two exit modes are a tradeoff:
- Mid-Line exit: softer, fewer whipsaws. You often get a second chance because the band itself doesn't have to be broken.
- Lower-Band exit: classic, fewer trades. When the trend really flips, you sit deeper in drawdown before the signal kicks in.
In our backtest defaults we use mid-line exit because it performs better in most market phases in our tests.
What Keltner Is Good For
Clear trending markets. When an asset has a clean up- or downtrend, the breakout above the upper band gives a clear signal. Standard-deviation-based Bollinger Bands have a problem here: in strong trends, standard deviation shrinks relative to price (because there's less mean reversion), and Bollinger gives reversion signals too early. ATR doesn't shrink as much in this configuration because ATR measures the absolute range value.
High volatility. ATR scaling automatically gives more room when the market is choppy. On BTCUSDT 2021, a fixed 5% stop setup would have been way too tight — a 2× ATR setup adapts to real volatility.
Multi-asset consistency. ATR is asset-class-agnostic — a 2× ATR move means structurally the same on forex as on crypto, just in different absolute magnitude. Standard deviation scales differently across asset classes.
What Keltner Is NOT Good For
Sideways markets. When price oscillates, it also fluctuates through the upper band up and down — the strategy produces whipsaws. That's a general problem of all breakout systems. On 1W charts the sample is small enough to absorb, on daily it can hurt.
Short timeframes (under 1d). On 1h or 15min, Keltner produces too many false signals because ATR moves at those timescales are more noise than signal. We take 1d as minimum, 1W as default.
Mean-reversion traders. Keltner is a trend-following strategy, not a reversal tool. If your thesis is "the market reverts to the middle", use rsi_ob_os (RSI Overbought/Oversold) or bb_rsi (Bollinger + RSI, coming later).
Concrete Defaults and Why
| Parameter | Default | Reasoning |
|---|---|---|
| EMA period | 20 | Linda Raschke's standard setup |
| ATR period | 10 | Welles Wilder's original |
| ATR multiplier | 2.0 | LeBeau's recommendation — gives enough room without being too late |
| Exit mode | mid-line | Better risk/reward in our tests |
You can adjust all four via the backtest form. Higher EMA period = slower signal, fewer trades. Higher ATR multiplier = wider bands, later breakout signal but fewer whipsaws.
Where the Strategy Is Available
Free+ as the 10th strategy in the Backtesting Arena, available now:
- Backtest page — all 5 asset classes (Crypto, Stocks, ETF, Commodities, Forex)
- Bulk Backtest (Admin) — comparison runs across multiple assets
- Custom Report Wizard — self-service strategy reports
- Live Alerts — email alerts on BUY/SELL triggers (Pro+)
- Signal traffic light — live status per pair (Pro+)
Free has access to 3 strategies (RSI/SMA, Golden Cross, RSI OB/OS). Pro+ unlocks all 10, plus filter combinations, trailing-stop modifiers, and the other advanced features.
A Transparent Closing Note
Keltner Channel Breakout is a robust, simple trend-following system. It's not the best system for every market — no system is. But it's methodologically clean, has clear inputs and outputs, and its behaviour can be backtested well.
What you should test yourself: try Keltner with mid-line exit on your favourite pairs across different market phases — bull (2020-2021), bear (2022), sideways (mid 2023), recovery (2024). Performance varies noticeably.
Try the strategy:
→ tradingstrategies.work/dashboard/crypto
Backtesting Arena · Your own backtests, your own parameters — across 2,000+ assets in every major market.