Dual ATR Channel Trend Following Strategy

The Dual ATR Channel trend following strategy is a trend tracking strategy that combines moving averages, ATR channels and multiple technical indicators to follow the trend after it has been established.

Pine ScriptTrend following RSI MACD EMA SMA ATR Ichimoku

Author: ChaoZhang
Complexity: Intermediate · 153 lines
Parameters: 22
Backtest bar: 15m · intraday
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Overview

How It Works

The strategy uses the Kijun line as the main moving average indicator to determine the trend direction. It also incorporates ATR channels to limit the price activity range - not going long when price is near the upper band and not going short when price is near the lower band to avoid chasing new highs and selling lows.

When the Kijun line has an upward crossover, a buy signal is generated. When a downward crossover happens, a sell signal is triggered. To filter out false signals, the strategy also employs multiple technical indicators for confirmation, including Aroon, RSI, MACD and PSAR. A buy or sell signal is only triggered when all the confirmation conditions are met.

Once in a trade, the strategy uses stop loss and take profit to manage positions. The stop loss is set at 0.5 ATR and take profit at 0.5%. When the price crosses the Kijun line in the opposite direction again, the position will be closed immediately.

Advantages

Using Kijun line to determine trend avoids being whipsawed by range-bound markets
ATR channels limit price activity for better risk control
Multiple confirmations greatly reduce false signals
Incorporating stop loss and take profit locks in profits while managing risks

Risks

Delayed signals from multiple confirmations, possibly missing early trend moves
Small stop loss may get stopped out frequently
Poor Kijun and ATR parameters may generate many wrong signals
Reliance on parameter optimization and curve fitting, may not work well in live trading

Improvement Opportunities

Test more advanced trend indicators like Ichimoku clouds
Optimize stop loss and take profit points for better risk reward ratio
Find optimal parameters for different markets
Add dynamic adjustment of parameters based on live market conditions
Test different combinations of confirmation indicators
Continuously optimize to ensure robustness of strategy

Conclusion

The dual ATR channel trend following strategy combines moving averages, ATR channels and multiple technical indicators to trade in the direction of the trend once established. Compared to single indicator strategies, it can greatly improve signal quality and win rate. The stop loss and take profit mechanisms also control risk. Through parameter optimization and combinatorial testing, this strategy has the potential to achieve steady profits. But its reliance on historical data is a concern and live performance requires further verification. Continuous optimization is key to ensuring robustness.

Parameters

Name	Default	What it controls
v_input_1	0	Baseline Type: Kijun \| EMA \| DEMA \| TEMA \| WMA \| VWMA \| SMA \| SMMA \| HMA \| LSMA \| ALMA \| McGinley
v_input_2	20	Baseline Length
v_input_3	12	Baseline Length Fast
v_input_4	false	* Least Squares (LSMA) Only - Offset Value
v_input_5	0.85	* Arnaud Legoux (ALMA) Only - Offset Value
v_input_6	6	* Arnaud Legoux (ALMA) Only - Sigma Value
v_input_7	14	ATR Length
v_input_8	true	donttradeoutside_atrcave
v_input_9	8	Length Aroon
v_input_10	false	dont_use_macd
v_input_11	13	Fast Length
v_input_12	26	Slow Length
v_input_13	9	Signal Smoothing
v_input_14	false	dont_use_rsi
v_input_15	14	RSI Length
v_input_16	false	dont_use_psar
v_input_17	0.03	psar_start
v_input_18	0.018	psar_increment
v_input_19	0.11	psar_maximum
v_input_20	false	use_exit
v_input_21	0.5	SL
v_input_22	0.005	TP

Source Pine Script

/*backtest
start: 2023-10-24 00:00:00
end: 2023-10-27 00:00:00
period: 15m
basePeriod: 5m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=4
// strategy(title="NoNonsense Forex", overlay=true, default_qty_value=100000, initial_capital=100)

//////////////////////
////// BASELINE //////
//////////////////////
ma_slow_type = input(title="Baseline Type", type=input.string, defval="Kijun", options=["ALMA", "EMA", "DEMA", "TEMA", "WMA", "VWMA", "SMA", "SMMA", "HMA", "LSMA", "Kijun", "McGinley"])
ma_slow_src = close //input(title="MA Source", type=input.source, defval=close)
ma_slow_len = input(title="Baseline Length", type=input.integer, defval=20)
ma_slow_len_fast = input(title="Baseline Length Fast", type=input.integer, defval=12)

lsma_offset  = input(defval=0, title="* Least Squares (LSMA) Only - Offset Value", minval=0)
alma_offset  = input(defval=0.85, title="* Arnaud Legoux (ALMA) Only - Offset Value", minval=0, step=0.01)
alma_sigma   = input(defval=6, title="* Arnaud Legoux (ALMA) Only - Sigma Value", minval=0)

ma(type, src, len) =>
    float result = 0
    if type=="SMA" // Simple
        result := sma(src, len)
    if type=="EMA" // Exponential
        result := ema(src, len)
    if type=="DEMA" // Double Exponential
        e = ema(src, len)
        result := 2 * e - ema(e, len)
    if type=="TEMA" // Triple Exponential
        e = ema(src, len)
        result := 3 * (e - ema(e, len)) + ema(ema(e, len), len)
    if type=="WMA" // Weighted
        result := wma(src, len)
    if type=="VWMA" // Volume Weighted
        result := vwma(src, len) 
    if type=="SMMA" // Smoothed
        w = wma(src, len)
        result := na(w[1]) ? sma(src, len) : (w[1] * (len - 1) + src) / len
    if type=="HMA" // Hull
        result := wma(2 * wma(src, len / 2) - wma(src, len), round(sqrt(len)))
    if type=="LSMA" // Least Squares
        result := linreg(src, len, lsma_offset)
    if type=="ALMA" // Arnaud Legoux
        result := alma(src, len, alma_offset, alma_sigma)
    if type=="Kijun" //Kijun-sen
        kijun = avg(lowest(len), highest(len))
        result :=kijun
    if type=="McGinley"
        mg = 0.0
        mg := na(mg[1]) ? ema(src, len) : mg[1] + (src - mg[1]) / (len * pow(src/mg[1], 4))
        result :=mg
    result

baseline = ma(ma_slow_type, ma_slow_src, ma_slow_len)
plot(baseline, title='Baseline', color=rising(baseline,1) ? color.green : falling(baseline,1) ? color.maroon : na, linewidth=3)

//////////////////
////// ATR ///////
//////////////////
atrlength=input(14, title="ATR Length")
one_atr=rma(tr(true), atrlength)
upper_atr_band=baseline+one_atr
lower_atr_band=baseline-one_atr
plot(upper_atr_band, color=color.gray, style=plot.style_areabr, transp=95, histbase=50000, title='ATR Cave')
plot(lower_atr_band, color=color.gray, style=plot.style_areabr, transp=95, histbase=0, title='ATR Cave')
plot(upper_atr_band, color=close>upper_atr_band ? color.fuchsia : na, style=plot.style_line, linewidth=5, transp=50, title='Close above ATR cave')
plot(lower_atr_band, color=close<lower_atr_band ? color.fuchsia : na, style=plot.style_line, linewidth=5, transp=50, title='Close below ATR cave')
donttradeoutside_atrcave=input(true)
too_high = close>upper_atr_band and donttradeoutside_atrcave
too_low = close<lower_atr_band and donttradeoutside_atrcave

////////////////////////////
////// CONFIRMATION 1 ////// the trigger actually
////////////////////////////
lenaroon = input(8, minval=1, title="Length Aroon")
c1upper = 100 * (highestbars(high, lenaroon+1) + lenaroon)/lenaroon
c1lower = 100 * (lowestbars(low, lenaroon+1) + lenaroon)/lenaroon
c1CrossUp=crossover(c1upper,c1lower)
c1CrossDown=crossunder(c1upper,c1lower)


////////////////////////////////
////// CONFIRMATION: MACD //////
////////////////////////////////
dont_use_macd=input(false)
macd_fast_length = input(title="Fast Length", type=input.integer, defval=13)
macd_slow_length = input(title="Slow Length", type=input.integer, defval=26)
macd_signal_length = input(title="Signal Smoothing", type=input.integer, minval = 1, maxval = 50, defval = 9)
macd_fast_ma = ema(close, macd_fast_length)
macd_slow_ma = ema(close, macd_slow_length)
macd = macd_fast_ma - macd_slow_ma
macd_signal = ema(macd, macd_signal_length)
macd_hist = macd - macd_signal

macdLong=macd_hist>0 or dont_use_macd
macdShort=macd_hist<0 or dont_use_macd

/////////////////////////////
///// CONFIRMATION: RSI /////
/////////////////////////////
dont_use_rsi=input(false)
lenrsi = input(14, minval=1, title="RSI Length") //14
up = rma(max(change(close), 0), lenrsi)
down = rma(-min(change(close), 0), lenrsi)
rsi = down == 0 ? 100 : up == 0 ? 0 : 100 - (100 / (1 + up / down))
rsiLong=rsi>50 or dont_use_rsi
rsiShort=rsi<50 or dont_use_rsi

//////////////////////////////
///// CONFIRMATION: PSAR /////
//////////////////////////////
dont_use_psar=input(false)
psar_start = input(0.03, step=0.01)
psar_increment = input(0.018, step=0.001)
psar_maximum = input(0.11, step=0.01) //default 0.08
psar = sar(psar_start, psar_increment, psar_maximum)

plot(psar, style=plot.style_cross, color=color.blue, title='PSAR')
psarLong=close>psar or dont_use_psar
psarShort=close<psar or dont_use_psar

/////////////////////////
///// CONFIRMATIONS /////
/////////////////////////
Long_Confirmations=psarLong and rsiLong and macdLong
Short_Confirmations=psarShort and rsiShort and macdShort

GoLong=c1CrossUp and Long_Confirmations and not too_high
GoShort=c1CrossDown and Short_Confirmations and not too_low

////////////////////
///// STRATEGY /////
////////////////////

use_exit=input(false)
KillLong=c1CrossDown and use_exit
KillShort=c1CrossUp and use_exit

SL=input(0.5, step=0.1)/syminfo.mintick
TP=input(0.005, step=0.001)/syminfo.mintick

strategy.entry("nnL", strategy.long, when = GoLong)
strategy.entry("nnS", strategy.short, when = GoShort)
strategy.exit("XL-nn", from_entry = "nnL", loss = SL, profit=TP)
strategy.exit("XS-nn", from_entry = "nnS", loss = SL, profit=TP)
strategy.close("nnL", when = KillLong)
strategy.close("nnS", when = KillShort)

Published by ChaoZhang on the FMZ strategy square · last modified 2023-11-01. Collected in the Quant Trading Vault. Unvalidated research: re-test with your own data and costs before trusting it.

Build this with AgenKit

A spec is not a trading system. Paste this into your coding agent and AgenKit runs a specialist team — architecture, test-first build, risk gates, review — with paper trading on by default.

in Claude Code, Cursor, Codex & 4 more

/agenkit build a production-grade trading bot with a backtest engine, risk controls and paper trading by default from strategies/Dual-ATR-Channel-Trend-Following-Strategy.md

npx agenkit install engineering-kit in your project
Paste the command into your agent and approve each phase
Get tests, backtest, risk controls and a paper-trading switch

Get AgenKit How /agenkit works

Understand it first

Before you run this, read the lesson behind the idea and how to backtest it honestly. Specs like this one are the survivors of an unknown number of trials; deflate their results accordingly.