Multi timeframe Trading Strategy with Bollinger Bands

This strategy uses adaptive Bollinger Bands to design two types of trailing stop strategies and backtest them systematically across timeframes. It belongs to trend following strategies.

Pine ScriptMulti-timeframe Machine learning Breakout EMA SMA Bollinger Bands CCI

Author: ChaoZhang
Complexity: Starter · 89 lines
Parameters: 6
Backtest bar: 1d · swing
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Overview

This strategy uses adaptive Bollinger Bands to design two types of trailing stop strategies and backtest them systematically across timeframes. It belongs to trend following strategies.

Strategy Logic

Calculate the upper and lower bands of adaptive Bollinger Bands, with adjustable channel width.
Breakout tracking strategy to open positions on band breakouts and stop out when price reverts inside bands.
Reversion reversal strategy to open positions when price reaches bands and stop out when price reverts back inside bands.
Use CCI indicator to assist in determining long/short side.
Backtest across multiple timeframes to verify viability of both strategies.

Advantages

Bollinger Bands are intuitive in capturing price trends.
The two strategies fit different market conditions for robustness.
CCI helps determine long/short direction.
Multi-timeframe backtesting makes results more convincing.
Simple and clear strategy rules easy to implement.

Risks

Bollinger Bands can fail in certain situations.
Risks of premature or delayed stops in both strategies.
CCI may generate incorrect signals.
Handle backtest biases carefully.
Optimization risks overfitting.

Enhancement

Test parameters to find optimal combinations.
Evaluate adding filters with other indicators.
Optimize stops to reduce risks.
Research adaptive methods for channel width.
Verify with more symbols and timeframes.
Use machine learning to dynamically optimize parameters.

Conclusion

This strategy designs two trailing stop strategies based on Bollinger Bands and backtests them across multiple timeframes. Refining via parameter optimization, stop improvements etc can enhance robustness into a mature trend following system.

Parameters

Name	Default	What it controls
v_input_1	75	Length
v_input_2_close	0	Source: close \| high \| low \| open \| hl2 \| hlc3 \| hlcc4 \| ohlc4
v_input_3	1.9	Deviation
v_input_4	20	Period for CCI
v_input_5	false	Test Reverse to the Mean instead
v_input_6	true	Enable testing

Source Pine Script

/*backtest
start: 2022-09-13 00:00:00
end: 2023-09-19 00:00:00
period: 1d
basePeriod: 1h
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=4
strategy(title = "Underworld Hunter", overlay=true)

len = input(75, minval=1, title="Length")
src = input(close, title="Source")
basis = 0.0
basis := na(basis[1]) ? sma(src, len) : ema(ema(ema(src,len),len),len)

mult = input(1.9, minval=0.001, maxval=50, title="Deviation")
dev = mult * stdev(src, len)
upper = basis + dev
lower = basis - dev

//CCI calculation and inputs

lengthcci = input(20, minval=1, title="Period for CCI")
ma = sma(close, lengthcci)
ccivalue = (src - ma) / (0.015 * dev(src, lengthcci))

//CCI plotting

cciover0 = ccivalue >= 100 and ccivalue <= 120
cciover1 = ccivalue > 120 and ccivalue <= 140
cciover2 = ccivalue > 140 and ccivalue <= 160
cciover3 = ccivalue > 160 and ccivalue <= 180
cciover4 = ccivalue > 180

cciunder0 = ccivalue <= -100 and ccivalue >= -120
cciunder1 = ccivalue <= -120 and ccivalue > -140
cciunder2 = ccivalue <= -140 and ccivalue > -160
cciunder3 = ccivalue <= -160 and ccivalue > -180
cciunder4 = ccivalue <= -180

plotshape(cciover0, title="CCIO0", location=location.abovebar, color=#c6ff1a, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciunder0, title="CCIU0", location=location.belowbar, color=#c6ff1a, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciover1, title="CCIO1", location=location.abovebar, color=#ffff00, transp=0,style=shape.circle, size=size.tiny)
plotshape(cciunder1, title="CCIU1", location=location.belowbar, color=#ffff00, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciover2, title="CCIO2", location=location.abovebar, color=#ff9900, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciunder2, title="CCIU2", location=location.belowbar, color=#ff9900, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciover3, title="CCIO3", location=location.abovebar, color=#ff0000, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciunder3, title="CCIU3", location=location.belowbar, color=#ff0000, transp=0, style=shape.circle, size=size.tiny)
plotshape(cciover4, title="CCIO4", location=location.abovebar, color=#cc00cc, transp=0,style=shape.circle, size=size.tiny)
plotshape(cciunder4, title="CCIU4", location=location.belowbar, color=#cc00cc, transp=0,style=shape.circle, size=size.tiny)

//plotting

plot(upper, title="Upper shadow", color=color.black, transp = 30, linewidth = 4)
plot(upper, title="Upper line", color=#FF2E00, transp = 0, linewidth = 2)
plot(lower, title="Lower shadow", color=color.black, transp = 30, linewidth = 4)
plot(lower, title="Lower line", color=#FF2E00, transp = 0, linewidth = 2)
plot(basis, title="Basic line", color=color.red, transp = 50, linewidth = 2)

mean = input(title="Test Reverse to the Mean instead", type=input.bool, defval=false)
test = input(title="Enable testing", type=input.bool, defval=true)

ordersize=floor(50000/close)

if(close>upper and strategy.opentrades==0 and not mean and test)
    strategy.entry("Hunt Up", strategy.long, ordersize)
if (close<upper and close[1]<upper and close[2]<upper)
    strategy.close("Hunt Up", qty_percent = 100, comment = "Hunt End")

if(close<lower and strategy.opentrades==0 and not mean and test)
    strategy.entry("Hunt Down", strategy.short, ordersize)
if (close>lower and close[1]>lower and close[2]>lower)
    strategy.close("Hunt Down", qty_percent = 100, comment = "Hunt End")

//bounce of bands

if(close>upper and strategy.opentrades==0 and mean and test)
    strategy.entry("Sneak Down", strategy.short, ordersize)
if (close<upper and close[1]<upper and close[2]<upper and close>high[1])
    strategy.close("Sneak Down", qty_percent = 100, comment = "SneakEnd")

if(close<lower and strategy.opentrades==0 and mean and test)
    strategy.entry("Sneak Up", strategy.long, ordersize)
if (close>lower and close[1]>lower and close[2]>lower and close<low[1])
    strategy.close("Sneak Up", qty_percent = 100, comment = "Sneak End")

Published by ChaoZhang on the FMZ strategy square · last modified 2023-09-20. 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/Multi-timeframe-Trading-Strategy-with-Bollinger-Bands.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.