Multi Indicator Decision Tree Strategy IMACD EMA and Ichimoku

This strategy integrates multiple technical indicators like IMACD, EMA and Ichimoku to build a comprehensive decision tree model for generating trading signals.

Pine ScriptMachine learning Trend following Tools & templates MACD EMA SMA Ichimoku Pivots / S&R

Author: ChaoZhang
Complexity: Intermediate · 71 lines
Parameters: 8
Backtest bar: 30m · intraday
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Overview

This strategy integrates multiple technical indicators like IMACD, EMA and Ichimoku to build a comprehensive decision tree model for generating trading signals.

Strategy Logic

IMACD: Enhanced MACD to better capture market trends via ImpulseMACD and ImpulseHisto
Ichimoku: Plot conversion line, base line, leading span A, leading span B to identify support and resistance levels
EMA 40: Assist in determining trend direction
Long/short signals are triggered based on specific conditions between IMACD, cloud components and EMA 40

Long signal: When IMACD is a specific color and EMA 40 is above cloud top, go long

Short signal: When IMACD is red and EMA 40 is below cloud bottom, go short

Advantage Analysis

Combination of multiple indicators improves accuracy of trend judgment
Clear classification of decision tree model generates unambiguous trading signals
Flexible EMA length for better assistance in trend determination
Identify support and resistance levels better with cloud and trend indicators

Risk Analysis

Complex parameter tuning with multiple indicators
False signals may be triggered by improper EMA length
Difficulty in monitoring many indicators simultaneously

Risk Solutions: Optimize parameter settings, adjust EMA length, simplify workflow.

Optimization Directions

Enhance stability by optimizing parameters
Limit losses with stop loss strategies
Improve signal quality by backtesting with huge data
Build adaptive decision tree model with machine learning

Summary

This strategy identifies trends using multiple indicators to construct a decision tree model for generating trading signals. Pros are high-quality and accurate signals. Consists room for progressive optimization. Requires focus on parameter tuning and stop loss to control risks for long-term steady returns.

Parameters

Name	Default	What it controls
v_input_1	34	Length MA
v_input_2	9	Length Signal
v_input_int_1	9	Conversion Line Length
v_input_int_2	26	Base Line Length
v_input_int_3	52	Leading Span B Length
v_input_int_4	26	Lagging Span
v_input_3	40	EMA Length
v_input_4	false	Enable bar colors

Source Pine Script

/*backtest
start: 2024-01-14 00:00:00
end: 2024-01-21 00:00:00
period: 30m
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=5
strategy("Decision Tree Strategy: IMACD, EMA and Ichimoku [cryptoonchain]", overlay=true)

lengthMA = input(34, title="Length MA")
lengthSignal = input(9, title="Length Signal")
conversionPeriods = input.int(9, minval=1, title="Conversion Line Length")
basePeriods = input.int(26, minval=1, title="Base Line Length")
laggingSpan2Periods = input.int(52, minval=1, title="Leading Span B Length")
displacement = input.int(26, minval=1, title="Lagging Span")
emaLength = input(40, title="EMA Length")  // Added user-configurable EMA length

calc_smma(src, len) =>
    smma = float(na)
    smma := na(smma[1]) ? ta.sma(src, len) : (smma[1] * (len - 1) + src) / len
    smma

calc_zlema(src, length) =>
    ema1 = ta.ema(src, length)
    ema2 = ta.ema(ema1, length)
    d = ema1 - ema2
    ema1 + d

src = ohlc4
hi = calc_smma(high, lengthMA)
lo = calc_smma(low, lengthMA)
mi = calc_zlema(src, lengthMA)

md = (mi > hi) ? (mi - hi) : (mi < lo) ? (mi - lo) : 0
sb = ta.sma(md, lengthSignal)
sh = md - sb
mdc = src > mi ? (src > hi ? color.rgb(128, 255, 0, 26) : color.green) : (src < lo ? color.red : color.orange)

colorCondition = color.rgb(128, 255, 0, 26)

conversionLine = math.avg(ta.lowest(conversionPeriods), ta.highest(conversionPeriods))
baseLine = math.avg(ta.lowest(basePeriods), ta.highest(basePeriods))
leadLine1 = math.avg(conversionLine, baseLine)
leadLine2 = math.avg(ta.lowest(laggingSpan2Periods), ta.highest(laggingSpan2Periods))

// Use user-configurable length for EMA
ema40 = ta.ema(close, emaLength)

ebc = input(false, title="Enable bar colors")
barcolor(ebc ? mdc : na)

conversionLinePlot = plot(conversionLine, color=#2962FF, title="Conversion Line", display=display.none)
baseLinePlot = plot(baseLine, color=#B71C1C, title="Base Line", display=display.none)
laggingSpanPlot = plot(close, offset=-displacement + 1, color=#43A047, title="Lagging Span", display=display.none)
leadLine1Plot = plot(leadLine1, offset=displacement - 1, color=#A5D6A7, title="Leading Span A", display=display.none)
leadLine2Plot = plot(leadLine2, offset=displacement - 1, color=#EF9A9A, title="Leading Span B", display=display.none)
kumoCloudUpperLinePlot = plot(leadLine1 > leadLine2 ? leadLine1 : leadLine2, offset=displacement - 1, title="Kumo Cloud Upper Line", display=display.none)
kumoCloudLowerLinePlot = plot(leadLine1 < leadLine2 ? leadLine1 : leadLine2, offset=displacement - 1, title="Kumo Cloud Lower Line", display=display.none)
fill(kumoCloudUpperLinePlot, kumoCloudLowerLinePlot, color=leadLine1 > leadLine2 ? color.green : color.red)

a = (leadLine1 > leadLine2 ? leadLine1 : leadLine2) 
b = (leadLine1 < leadLine2 ? leadLine1 : leadLine2)  

if mdc == colorCondition and ema40 > a[displacement - 1]
    strategy.entry("Long", strategy.long)

if mdc == color.red and ema40 < b[displacement - 1]
    strategy.entry("Short", strategy.short)

Published by ChaoZhang on the FMZ strategy square · last modified 2024-01-22. 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-Indicator-Decision-Tree-Strategy-IMACD-EMA-and-Ichimoku.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.