Keltner Breakout
BullishContinuationIntermediateA close that jumps above EMA20 + 2.5 ATR. Exit on the first close back under EMA20.
Verdict: CautionBitcoin 4H: 271 trades, won 47%, avg trade +1.5% vs random day +0.8%. Backtest
Take the 20-candle exponential average of the closes. Then measure how far a normal candle travels, the average true range over the same 20 candles. The upper Keltner band sits two and a half of those ranges above the average.
step 1/5
Trade rules
| Direction | Long (buy) |
|---|---|
| Entry | A candle CLOSES above EMA20 + 2.5 x ATR20 (the 20-candle exponential average of the closes plus 2.5 times the 20-candle average true range, Wilder smoothing) after the candle before closed at or below its own band. Buy at that close. Exit at the first close below EMA20. Paper traded on 1-hour and 4-hour candles with that exit and a protective stop inside the candle at EMA20 or 2 x ATR14 below the entry, whichever is lower (for this rule it is always EMA20); the encyclopedia backtest uses that stop, a target of two times the risk and a 100-candle limit. Tested in docs/research/2026-10-indicator-tests.md. |
| Cancelled if | Nothing to wait for: the setup and the entry are the same candle. A wick above the band that closes back inside is not a signal, and neither is a candle that was already above the band the candle before. |
| Target | Two times the risk, projected from the breakout |
5 common mistakes
- Using the simple 20-candle average and a simple ATR. The tested rule uses the exponential average and Wilder's ATR; other versions give other signals.
- Exiting at the first red candle. The tested exit waits for a close below the 20-candle average.
- Expecting it to trade often. It fired about 5,000 times in 12 coins over 6-9 years of 1-hour candles and sits in a trade 13% of the time.
- Treating it as better than a Donchian breakout. It earns more per trade on 1-hour candles, but a 55-candle Donchian breakout earned more per year in the same test.
- Adding a volume filter and expecting a different strategy. Volume above 1.5 times its average raised the return per trade, but cut the number of trades and changed no verdict.
Backtest
same rules every time · equal size · no fees · method| Market | Trades | Won | Avg trade | Random day |
|---|---|---|---|---|
| Bitcoin · 4H · 2017-2026 | 271 | 47% | +1.46% | +0.78% |
| US large caps (50) · 1D · 2016-2026 | 2384 | 44% | +1.96% | +2.44% |
Verdict: CautionBeat a random entry in some markets, not in others.
pattern +395%random +210%
pattern +4678%random +5813%