Candlestick Patterns Based Strategy

This strategy is based on candlestick patterns to identify different candlestick signals and trade along the trend. With proper risk management methods like stop loss, take profit, and trailing stop, it aims to…

Pine ScriptTrend following SMA

Author: ChaoZhang
Complexity: Intermediate · 150 lines
Parameters: 19
Backtest bar: 1d · swing
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Overview

Strategy Logic

The strategy mainly identifies the following candlestick patterns for trading signals:

Engulfing: bullish engulfing and bearish engulfing patterns
Piercing Line/Dark Cloud Cover: bullish piercing line and bearish dark cloud cover patterns
Morning Star/Evening Star: bullish morning star and bearish evening star patterns
Belt Hold: bullish belt hold and bearish belt hold patterns
Three White Soldiers/Three Black Crows: three white soldiers and three black crows patterns
Three Stars in the South: three stars in the south pattern
Stick Sandwich: stick sandwich pattern
Meeting Line: bullish meeting line and bearish meeting line patterns
Kicking: bullish kicking and bearish kicking patterns
Ladder Bottom: ladder bottom pattern

When detecting these candlestick signals, it will place pending orders near the open price of the next bar, with pre-defined stop loss and take profit for trend following. It also uses trailing stop and breakeven stop to manage risks.

It also adds a moving average filter to avoid taking signals when price is on the wrong side of MA.

Advantages

Based on classical candlestick patterns, it has some universal applicability.
Trading mechanically based on pattern rules, not affected by subjective factors.
Reasonable stop loss and take profit settings to control single trade risks.
Trailing stop mechanism adjusts stop loss dynamically along with the market.
MA filter adds more logic to avoid being trapped in wrong trades.

Risks and Solutions

Candlestick patterns have some misidentification issues, which can lead to false signals. Can optimize parameters or filter out invalid patterns.
Static stop loss cannot fully avoid risks from market events. Can use wider stop or trailing stop.
Sensitive to trading sessions, cannot run 24x7. Can adjust trading times or add auction filters.
MA filter may miss some opportunities. Can lower MA period or remove the filter.
Hard to profit from both long and short sides together due to conflicts. Can separate strategies for long and short.

Optimization Directions

Optimize parameters of candlestick patterns to improve identification.
Test different trailing stop methods to find the optimal one.
Try more advanced risk management techniques like money management or volatility stop loss.
Add more filters to improve the filtering logic.
Build candlestick pattern recognition models using machine learning.
Develop strategy logic that can identify both long and short signals.

Summary

This strategy uses classical candlestick patterns for trend detection and trades mechanically based on the signals. It manages risks by strict stop loss and trailing stop, and improves logic by adding MA filter. The strategy is easy to understand and implement, but also has some issues like misidentification and difficulties in parameter tuning. Future optimizations can be done by introducing more technical indicators and machine learning models to achieve better performance.

Parameters

Name	Default	What it controls
v_input_1	true	Engulfing
v_input_2	true	Harami
v_input_3	true	Piercing Line / Dark Cloud Cover
v_input_4	true	Morning Star / Evening Star
v_input_5	true	Belt Hold
v_input_6	true	Three White Soldiers / Three Black Crows
v_input_7	true	Three Stars in the South
v_input_8	true	Stick Sandwich
v_input_9	true	Meeting Line
v_input_10	true	Kicking
v_input_11	true	Ladder Bottom
v_input_12	0.01	Tick Size
v_input_13	10	Stop Loss
v_input_14	100	Take Profit
v_input_15	10	Breakeven Margin
v_input_16	5	Price Movement Confirmation
v_input_17	false	MA Filter
v_input_18	50	MA Period
v_input_19	0000-0000	Trading Session

Source Pine Script

/*backtest
start: 2022-09-15 00:00:00
end: 2023-02-17 00:00:00
period: 1d
basePeriod: 1h
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=2
strategy("Candle Patterns Strategy - 2", shorttitle="CPS - 2", overlay=true)
// New risk management system: order entry, moving stop loss to breakeven + moving average filter (SMA)

//--- Patterns Input ---

OnEngulfing = input(defval=true, title="Engulfing", type=bool)
OnHarami = input(defval=true, title="Harami", type=bool)
OnPiercingLine = input(defval=true, title="Piercing Line / Dark Cloud Cover", type=bool)
OnMorningStar = input(defval=true, title="Morning Star / Evening Star ", type=bool)
OnBeltHold = input(defval=true, title="Belt Hold", type=bool)
OnThreeWhiteSoldiers = input(defval=true, title="Three White Soldiers / Three Black Crows", type=bool)
OnThreeStarsInTheSouth = input(defval=true, title="Three Stars in the South", type=bool)
OnStickSandwich = input(defval=true, title="Stick Sandwich", type=bool)
OnMeetingLine = input(defval=true, title="Meeting Line", type=bool)
OnKicking = input(defval=true, title="Kicking", type=bool)
OnLadderBottom = input(defval=true, title="Ladder Bottom", type=bool)

//--- Risk Management Input ---
tick = input (defval = 0.01, title="Tick Size", minval = 0.001)
inpsl = input(defval = 10, title="Stop Loss", minval = 1)
inptp = input(defval = 100, title="Take Profit", minval = 1)
inpbm = input (defval=10, title="Breakeven Margin", minval = 1)
inpindent = input(defval = 5, title="Price Movement Confirmation", minval = 0)
InpSmaFilter = input(defval=false, title="MA Filter", type=bool)
maPer=input(defval = 50, title="MA Period", minval = 1)
//inptrail = input(defval = 0, title="Trailing Stop", minval = 0)
// If the zero value is set for stop loss, take profit or trailing stop, then the function is disabled
//sl = inpsl >= 1 ? inpsl : na
sl = inpsl * tick
bm = inpbm * tick
tp = inptp //* tick
indent = inpindent * tick
//trail = inptrail >= 1 ? inptrail : na

//--- Session Input ---

sess = input(defval = "0000-0000", title="Trading Session")
t = time(timeframe.period, sess)
session_open = na(t) ? false : true

// --- Candlestick Patterns ---

//Engulfing 
bullish_engulfing = high[0]>high[1] and low[0]<low[1] and open[0]<open[1] and close[0]>close[1] and close[0]>open[0] and close[1]<close[2] and close[0]>open[1] ? OnEngulfing : na
bearish_engulfing = high[0]>high[1] and low[0]<low[1] and open[0]>open[1] and close[0]<close[1] and close[0]<open[0] and close[1]>close[2] and close[0]<open[1] ? OnEngulfing : na

//Harami
bullish_harami =  open[1]>close[1] and close[1]<close[2] and open[0]>close[1] and open[0]<open[1] and close[0]>close[1] and close[0]<open[1] and high[0]<high[1] and low[0]>low[1] and close[0]>=open[0] ? OnHarami : na
bearish_harami =   open[1]<close[1] and close[1]>close[2] and open[0]<close[1] and open[0]>open[1] and close[0]<close[1] and close[0]>open[1] and high[0]<high[1] and low[0]>low[1] and close[0]<=open[0] ? OnHarami : na

//Piercing Line/Dark Cloud Cover 
piercing_line = close[2]>close[1] and open[0]<low[1] and close[0]>avg(open[1],close[1]) and close[0]<open[1] ? OnPiercingLine : na
dark_cloud_cover = close[2]<close[1] and open[0]>high[1] and close[0]<avg(open[1],close[1]) and close[0]>open[1] ? OnPiercingLine : na

//Morning Star/Evening Star
morning_star = close[3]>close[2] and close[2]<open[2] and open[1]<close[2] and close[1]<close[2] and open[0]>open[1] and open[0]>close[1] and close[0]>close[2] and open[2]-close[2]>close[0]-open[0] ? OnMorningStar : na
evening_star = close[3]<close[2] and close[2]>open[2] and open[1]>close[2] and close[1]>close[2] and open[0]<open[1] and open[0]<close[1] and close[0]<close[2] and close[2]-open[2]>open[0]-close[0] ? OnMorningStar : na

//Belt Hold
bullish_belt_hold = close[1]<open[1] and low[1]>open[0] and close[1]>open[0] and open[0]==low[0] and close[0]>avg(close[0],open[0]) ? OnBeltHold :na
bearish_belt_hold =  close[1]>open[1] and high[1]<open[0] and close[1]<open[0] and open[0]==high[0] and close[0]<avg(close[0],open[0]) ? OnBeltHold :na

//Three White Soldiers/Three Black Crows 
three_white_soldiers = close[3]<open[3] and open[2]<close[3] and close[2]>avg(close[2],open[2]) and open[1]>open[2] and open[1]<close[2] and close[1]>avg(close[1],open[1]) and open[0]>open[1] and open[0]<close[1] and close[0]>avg(close[0],open[0]) and high[1]>high[2] and high[0]>high[1] ? OnThreeWhiteSoldiers : na
three_black_crows =  close[3]>open[3] and open[2]>close[3] and close[2]<avg(close[2],open[2]) and open[1]<open[2] and open[1]>close[2] and close[1]<avg(close[1],open[1]) and open[0]<open[1] and open[0]>close[1] and close[0]<avg(close[0],open[0]) and low[1]<low[2] and low[0]<low[1] ? OnThreeWhiteSoldiers : na

//Three Stars in the South
three_stars_in_the_south = open[3]>close[3] and open[2]>close[2] and open[2]==high[2] and open[1]>close[1] and open[1]<open[2] and open[1]>close[2] and low[1]>low[2] and open[1]==high[1] and open[0]>close[0] and open[0]<open[1] and open[0]>close[1] and open[0]==high[0] and close[0]==low[0] and close[0]>=low[1] ? OnThreeStarsInTheSouth : na

//Stick Sandwich
stick_sandwich = open[2]>close[2] and open[1]>close[2] and open[1]<close[1] and open[0]>close[1] and open[0]>close[0] and close[0]==close[2] ? OnStickSandwich : na

//Meeting Line 
bullish_ml = open[2]>close[2] and open[1]>close[1] and close[1]==close[0] and open[0]<close[0] and open[1]>=high[0] ? OnMeetingLine : na
bearish_ml = open[2]<close[2] and open[1]<close[1] and close[1]==close[0] and open[0]>close[0] and open[1]<=low[0] ? OnMeetingLine : na

//Kicking 
bullish_kicking =  open[1]>close[1] and open[1]==high[1] and close[1]==low[1] and open[0]>open[1] and open[0]==low[0] and close[0]==high[0] and close[0]-open[0]>open[1]-close[1] ? OnKicking : na
bearish_kicking = open[1]<close[1] and open[1]==low[1] and close[1]==high[1] and open[0]<open[1] and open[0]==high[0] and close[0]==low[0] and open[0]-close[0]>close[1]-open[1] ? OnKicking : na

//Ladder Bottom
ladder_bottom = open[4]>close[4] and open[3]>close[3] and open[3]<open[4] and open[2]>close[2] and open[2]<open[3] and open[1]>close[1] and open[1]<open[2] and open[0]<close[0] and open[0]>open[1] and low[4]>low[3] and low[3]>low[2] and low[2]>low[1] ? OnLadderBottom : na

// --- Plotting Patterns ---

plotshape(bullish_engulfing, text='Engulfing', style=shape.triangleup, color=#1FADA2, editable=true, title="Bullish Engulfing Text")
plotshape(bearish_engulfing,text='Engulfing', style=shape.triangledown, color=#F35A54, editable=true, title="Bearish Engulfing Text")
plotshape(bullish_harami,text='Harami', style=shape.triangleup, color=#1FADA2, editable=true, title="Bullish Harami Text")
plotshape(bearish_harami,text='Harami', style=shape.triangledown, color=#F35A54, editable=true, title="BEarish Harami Text")
plotshape(piercing_line,text='Piercing Line', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(dark_cloud_cover,text='Dark Cloud Cover', style=shape.triangledown, color=#F35A54, editable=false)
plotshape(morning_star,text='Morning Star', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(evening_star,text='Evening Star', style=shape.triangledown, color=#F35A54, editable=false)
plotshape(bullish_belt_hold,text='Belt Hold', style=shape.triangleup, color=#1FADA2, editable=false)    
plotshape(bearish_belt_hold,text='Belt Hold', style=shape.triangledown, color=#F35A54, editable=false)
plotshape(three_white_soldiers,text='Three White Soldiers', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(three_black_crows,text='Three Black Crows', style=shape.triangledown, color=#F35A54, editable=false)
plotshape(three_stars_in_the_south,text='3 Stars South', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(stick_sandwich,text='Stick Sandwich', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(bullish_ml,text='Meeting Line', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(bearish_ml,text='Meeting Line', style=shape.triangledown, color=#F35A54, editable=false)
plotshape(bullish_kicking,text='Kicking', style=shape.triangleup, color=#1FADA2, editable=false)
plotshape(bearish_kicking,text='Kicking', style=shape.triangledown, color=#F35A54, editable=false)
plotshape(ladder_bottom,text='Ladder Bottom', style=shape.triangleup, color=#1FADA2, editable=false)

// --- STRATEGY ---

SignalUp = bullish_engulfing or bullish_harami or piercing_line or morning_star or bullish_belt_hold or three_white_soldiers or three_stars_in_the_south or stick_sandwich or bullish_ml or bullish_kicking or ladder_bottom
SignalDown = bearish_engulfing or bearish_harami or dark_cloud_cover or evening_star or bearish_belt_hold or three_black_crows or bearish_ml or bearish_kicking

PointOfEntry = SignalUp ? high[0] + indent : SignalDown ? low[0] - indent : na

bu = strategy.position_avg_price
shlo = strategy.position_size
stL = shlo > 0 and close [0] > bu + bm ? bu : shlo < 0 and close [0] < bu - bm ? bu : na
du = sma(close, maPer)
smaF = SignalUp and high[0]>du[0] ? true : SignalUp and high[0]<du[0] ? false : SignalDown and high[0]>du[0] ? false : SignalDown and high[0]<du[0] ? true : na 
smaFilter = InpSmaFilter ? smaF : true
duplot = InpSmaFilter ? du : na
plot(duplot, color=red)

// -- Orders --

strategy.order("buy", true, stop = PointOfEntry, oca_name = "trade",  when = SignalUp and session_open and smaFilter)
strategy.cancel("buy", when = not session_open or SignalDown or strategy.opentrades > 0)
strategy.order("stop sell", false, stop = bu-sl, oca_name = "trade",  when = strategy.opentrades > 0 and shlo > 0 and session_open)
strategy.cancel("stop sell", close [0] > bu + bm or not session_open or strategy.opentrades == 0 )
strategy.order("breakeven sell", false, stop = bu + 1*tick, oca_name = "trade",  when = strategy.opentrades > 0 and shlo > 0 and close [0] > bu + bm and session_open)
strategy.cancel("breakeven sell", when = strategy.opentrades == 0 or not session_open)

strategy.order("sell", false, stop = PointOfEntry, oca_name = "trade",  when = SignalDown and session_open and smaFilter)
strategy.cancel("sell", when = not session_open or SignalUp or strategy.opentrades > 0)
strategy.order("stop buy", true, stop = bu+sl, oca_name = "trade",  when = strategy.opentrades > 0 and shlo < 0 and session_open)
strategy.cancel("stop buy", when = close [0] < bu - bm or not session_open or  strategy.opentrades == 0 )
strategy.order("breakeven buy", true, stop = bu - 1 * tick, oca_name = "trade",  when = strategy.opentrades > 0 and shlo < 0 and close [0] < bu - bm and session_open)
strategy.cancel("breakeven buy", when = strategy.opentrades == 0 or not session_open)

strategy.close("buy", when = not session_open or SignalDown )
strategy.close("sell", when = not session_open or SignalUp)
strategy.exit("Take Profit", from_entry = "buy", profit = tp)
strategy.exit("Take Profit", from_entry = "sell", profit = tp)

Published by ChaoZhang on the FMZ strategy square · last modified 2023-09-22. 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/Candlestick-Patterns-Based-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.