RSI Divergence Trading Strategy

This strategy identifies regular and hidden bullish/bearish RSI divergence signals to determine long and short positions.

Pine ScriptMachine learning RSI Pivots / S&R

Author: ChaoZhang
Complexity: Intermediate · 166 lines
Parameters: 10
Backtest bar: 1h · intraday
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Strategy Name

RSI Bullish/Bearish Divergence Trading Strategy

Overview

This strategy identifies regular and hidden bullish/bearish RSI divergence signals to determine long and short positions.

Principles

When price makes a new high and RSI fails to make a new high, it forms bullish divergence, which is treated as selling signal. When price makes a new low and RSI fails to make a new low, it forms bearish divergence, which is treated as buying signal. Regular divergence refers to the obvious divergence between price and RSI while hidden divergence means relatively concealed divergence between them. Long or short positions are determined based on identified regular or hidden bullish/bearish divergence signals.

Advantage Analysis

Divergence signals have relatively high reliability with higher winning rate.
Identification of both regular and hidden bullish/bearish divergence provides extensive coverage.
Adjustable RSI parameters make it adaptable to different market environments.

Risk Analysis

Hidden divergence signals have higher probability of misjudgment.
Manual review is needed to filter out misjudged signals.
Effectiveness depends on RSI parameter settings.

Optimization Directions

Optimize RSI parameters to find best parameter combinations.
Add machine learning algorithms for automatic identification of true signals.
Incorporate more indicators to verify signal reliability.

Summary

This strategy identifies RSI divergence trading signals based on regular and hidden bullish/bearish divergence to determine long or short positions, which provides relatively higher winning rate. Further improvements on strategy effectiveness could be achieved by optimizing RSI parameters, adding other validating indicators.

Parameters

Name	Default	What it controls
v_input_int_1	14	RSI Period
v_input_1_close	0	RSI Source: close \| high \| low \| open \| hl2 \| hlc3 \| hlcc4 \| ohlc4
v_input_2	5	Pivot Lookback Right
v_input_3	5	Pivot Lookback Left
v_input_4	60	Max of Lookback Range
v_input_5	5	Min of Lookback Range
v_input_6	true	Plot Bullish
v_input_7	true	Plot Hidden Bullish
v_input_8	true	Plot Bearish
v_input_9	true	Plot Hidden Bearish

Source Pine Script

/*backtest
start: 2023-12-22 00:00:00
end: 2024-01-21 00:00:00
period: 1h
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=5
strategy(title="Divergence Indicator")
len = input.int(title="RSI Period", minval=1, defval=14)
src = input(title="RSI Source", defval=close)
lbR = input(title="Pivot Lookback Right", defval=5)
lbL = input(title="Pivot Lookback Left", defval=5)
rangeUpper = input(title="Max of Lookback Range", defval=60)
rangeLower = input(title="Min of Lookback Range", defval=5)
plotBull = input(title="Plot Bullish", defval=true)
plotHiddenBull = input(title="Plot Hidden Bullish", defval=true)
plotBear = input(title="Plot Bearish", defval=true)
plotHiddenBear = input(title="Plot Hidden Bearish", defval=true)
bearColor = color.red
bullColor = color.green
hiddenBullColor = color.new(color.green, 80)
hiddenBearColor = color.new(color.red, 80)
textColor = color.white
noneColor = color.new(color.white, 100)
osc = ta.rsi(src, len)

plot(osc, title="RSI", linewidth=2, color=#2962FF)
hline(50, title="Middle Line", color=#787B86, linestyle=hline.style_dotted)
obLevel = hline(70, title="Overbought", color=#787B86, linestyle=hline.style_dotted)
osLevel = hline(30, title="Oversold", color=#787B86, linestyle=hline.style_dotted)
fill(obLevel, osLevel, title="Background", color=color.rgb(33, 150, 243, 90))

plFound = na(ta.pivotlow(osc, lbL, lbR)) ? false : true
phFound = na(ta.pivothigh(osc, lbL, lbR)) ? false : true
_inRange(cond) =>
	bars = ta.barssince(cond == true)
	rangeLower <= bars and bars <= rangeUpper

//------------------------------------------------------------------------------
// Regular Bullish
// Osc: Higher Low

oscHL = osc[lbR] > ta.valuewhen(plFound, osc[lbR], 1) and _inRange(plFound[1])

// Price: Lower Low

priceLL = low[lbR] < ta.valuewhen(plFound, low[lbR], 1) 
// bull : 상승 Condition : 조건
bullCond = plotBull and priceLL and oscHL and plFound // 상승다이버전스?
strategy.entry("상승 다이버전스 진입", strategy.long, when = bullCond)
strategy.close("상승 다이버전스 진입", when = ta.crossover(osc, 50)) 
plot(
     plFound ? osc[lbR] : na,
     offset=-lbR,
     title="Regular Bullish",
     linewidth=2,
     color=(bullCond ? bullColor : noneColor)
     )

plotshape(
	 bullCond ? osc[lbR] : na,
	 offset=-lbR,
	 title="Regular Bullish Label",
	 text=" Bull ",
	 style=shape.labelup,
	 location=location.absolute,
	 color=bullColor,
	 textcolor=textColor
	 )

//------------------------------------------------------------------------------
// Hidden Bullish
// Osc: Lower Low

oscLL = osc[lbR] < ta.valuewhen(plFound, osc[lbR], 1) and _inRange(plFound[1])

// Price: Higher Low

priceHL = low[lbR] > ta.valuewhen(plFound, low[lbR], 1)
hiddenBullCond = plotHiddenBull and priceHL and oscLL and plFound
// strategy.entry("히든 상승 다이버전스 진입", strategy.long, when = hiddenBullCond)
// strategy.close("히든 상승 다이버전스 진입", when = ta.crossover(osc, 50))
plot(
	 plFound ? osc[lbR] : na,
	 offset=-lbR,
	 title="Hidden Bullish",
	 linewidth=2,
	 color=(hiddenBullCond ? hiddenBullColor : noneColor)
	 )

plotshape(
	 hiddenBullCond ? osc[lbR] : na,
	 offset=-lbR,
	 title="Hidden Bullish Label",
	 text=" H Bull ",
	 style=shape.labelup,
	 location=location.absolute,
	 color=bullColor,
	 textcolor=textColor
	 )

//------------------------------------------------------------------------------
// Regular Bearish
// Osc: Lower High

oscLH = osc[lbR] < ta.valuewhen(phFound, osc[lbR], 1) and _inRange(phFound[1])

// Price: Higher High

priceHH = high[lbR] > ta.valuewhen(phFound, high[lbR], 1)
// bear : 하락 
bearCond = plotBear and priceHH and oscLH and phFound
// strategy.entry("하락 다이버전스 진입", strategy.short, when = bearCond)
// strategy.close("하락 다이버전스 진입", when = ta.crossunder(osc, 50)) 
plot(
	 phFound ? osc[lbR] : na,
	 offset=-lbR,
	 title="Regular Bearish",
	 linewidth=2,
	 color=(bearCond ? bearColor : noneColor)
	 )

plotshape(
	 bearCond ? osc[lbR] : na,
	 offset=-lbR,
	 title="Regular Bearish Label",
	 text=" Bear ",
	 style=shape.labeldown,
	 location=location.absolute,
	 color=bearColor,
	 textcolor=textColor
	 )

//------------------------------------------------------------------------------
// Hidden Bearish
// Osc: Higher High

oscHH = osc[lbR] > ta.valuewhen(phFound, osc[lbR], 1) and _inRange(phFound[1])

// Price: Lower High

priceLH = high[lbR] < ta.valuewhen(phFound, high[lbR], 1)

hiddenBearCond = plotHiddenBear and priceLH and oscHH and phFound
// strategy.entry("히든 하락 다이버전스 진입", strategy.short, when = hiddenBearCond)
// strategy.close("히든 하락 다이버전스 진입", when = ta.crossunder(osc, 50)) 
plot(
	 phFound ? osc[lbR] : na,
	 offset=-lbR,
	 title="Hidden Bearish",
	 linewidth=2,
	 color=(hiddenBearCond ? hiddenBearColor : noneColor)
	 )

plotshape(
	 hiddenBearCond ? osc[lbR] : na,
	 offset=-lbR,
	 title="Hidden Bearish Label",
	 text=" H Bear ",
	 style=shape.labeldown,
	 location=location.absolute,
	 color=bearColor,
	 textcolor=textColor
	 )

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/RSI-Divergence-Trading-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.