EMA Oscillation Reversal System Strategy

This strategy uses a moving average system to determine trend direction and combines volatility index to avoid low volatility oscillating markets, with chandelier exit to manage trades.

Pine ScriptMean reversion Machine learning Trend following EMA SMA Bollinger Bands ATR

Author: ChaoZhang
Complexity: Intermediate · 91 lines
Parameters: 12
Backtest bar: 1d · swing
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Overview

This strategy uses a moving average system to determine trend direction and combines volatility index to avoid low volatility oscillating markets, with chandelier exit to manage trades.

Principle

The strategy judges trend direction by comparing fast and slow moving averages. It goes long when fast MA crosses above slow MA, and goes short when fast MA crosses below slow MA. To avoid oscillating markets, the strategy also incorporates Bollinger Bands. It generates trading signals when the rate of change of BB width exceeds a threshold. Finally, the strategy uses Chandelier exits as stop loss to avoid being trapped in range-bound markets.

Specifically, the trading logic is as follows:

Calculate fast MA (default 20-day) and slow MA (default 50-day).
Calculate the rate of change of Bollinger Band (default 40-day, 2 standard deviations) width.
Go long when fast MA crosses above slow MA, and BB width change rate exceeds default threshold 9%.
Go short when fast MA crosses below slow MA, and BB width change rate exceeds default threshold 9%.
Calculate Chandelier long and short stops.
Long stop is highest high - ATR * multiplier. Short stop is lowest low + ATR * multiplier.

Advantages

MA system effectively tracks trends.
BB width change filters out oscillation, reducing unnecessary trades.
Chandelier exits timely stop losses avoiding being trapped.
Multiple adjustable parameters for optimization.
Clear logic, easy to understand and implement.

Risks

MA lag may miss fast reversals.
Improper BB parameters may filter valid signals.
Excessive chandelier exits cause over-trading.
Inadequate parameter optimization leads to holding risks.
Unable to adapt to extreme market changes from major events.

Optimization

Test different MA combinations to find optimal parameters.
Test different BB periods for best volatility filter.
Add other indicators for entry confirmation.
Introduce dynamic stops to better track markets.
Utilize machine learning to auto-optimize for changing markets.

Summary

This strategy integrates MA system, BB indicator and chandelier exits to form a relatively stable trend following system. Proper parameter optimization can achieve good results. But risks of trend reversal and oscillation remain. Machine learning can further enhance robustness. Overall a good strategy for learning.

Parameters

Name	Default	What it controls
v_input_1	22	highest high o lowest low period
v_input_2	4	multiplicador
v_input_3	50	moving averages period
v_input_4	20	moving averages period2
v_input_5	0	moving averages type: ema \| sma
v_input_6	40	Bollinger bands 1 period
v_input_7_close	0	Bollinger band 1 source: close \| high \| low \| open \| hl2 \| hlc3 \| hlcc4 \| ohlc4
v_input_8	2	Bollinger Bands 1 factor
v_input_9	false	Show Bollinger bands 1
v_input_10	9	widen %
v_input_11	true	enable long
v_input_12	true	enable short

Source Pine Script

/*backtest
start: 2022-10-11 00:00:00
end: 2023-10-17 00:00:00
period: 1d
basePeriod: 1h
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

// This source code is subject to the terms of the Mozilla Public License 2.0 at https://mozilla.org/MPL/2.0/
// © juanchez

//@version=4
strategy("CHI", overlay = true, close_entries_rule = "ANY")

n = input(title= "highest high o lowest low period", defval= 22)
f= input(title= "multiplicador", defval= 4)
long = highest(high, n) - atr(n)*f
short= lowest(low, n) + atr(n)*f
plot(long, color= color.red)
plot(short, color= color.green)

//moving averages
period= input(title= "moving averages period", defval= 50)
period2= input(title= "moving averages period2", defval= 20)
type= input(title= "moving averages type", options= ["sma", "ema"], defval= "ema")

//moving average function
mo(p, t) =>
    if t == "sma"
        sma(close[barstate.islast ? 1: 0], p)
    else  if t== "ema"
        ema(close[barstate.islast ? 1: 0], p)

m= mo(period, type)
m2= mo(period2, type)

trend= m2 > m 

plot(m, color = color.maroon, linewidth = 3)
plot(m2, linewidth= 3)


//BOLLINGER BANDS ENTRIES
bb1_period= input(title= "Bollinger bands 1 period", defval=40, minval=1)
bb1_source=input(title="Bollinger band 1 source", defval=close)
bb1_multi=input(title="Bollinger Bands 1 factor", defval=2, minval=1, step=0.1)
show_bb1= input(title="Show Bollinger bands 1", defval=false)
//BOLLINGER BANDS
_bb(src, lenght, multi)=>
    float moving_avg= sma(src[barstate.islast? 1: 0], lenght)
    float deviation= stdev(src[barstate.islast? 1: 0], lenght)
    float lowerband = moving_avg - deviation*multi
    float upperband = moving_avg + deviation*multi
    
    [moving_avg, lowerband, upperband]
    
[bb1, lowerband1, upperband1]= _bb(bb1_source,  bb1_period, bb1_multi)

//FIRST BAND    
plot(show_bb1? bb1 : na, title="BB1 Moving average", linewidth= 3, color= color.fuchsia)
plot(show_bb1? upperband1 : na, title="BB1 Upper Band", linewidth= 3, color= color.green)
plot(show_bb1? lowerband1 : na, title="BB1 Lower Band", linewidth= 3, color= color.red)

//BB's Width threshold 
thresh= input(title= "widen %", defval= 9, minval = 0, step = 1, maxval= 100)

widht= (upperband1 - lowerband1)/bb1
roc= change(widht)/widht[1]*100
cross=crossover(roc, thresh)

// entry
//long
elong= input(true, title= "enable long")
longcondition= m2 > m and cross and elong

//short
eshort= input(true, title= "enable short")
shortcondition= m2 < m and cross and eshort


plotshape(longcondition? true: false , location= location.belowbar, style= shape.labelup, size= size.small, color= color.green, text= "Buy", textcolor= color.white)
plotshape(shortcondition? true: false , location= location.abovebar, style= shape.labeldown, size= size.small, color= color.red, text= "Sell", textcolor= color.white)

out= crossunder(close, long)
outt= crossover(close, short)

strategy.entry("long", strategy.long, when = longcondition)
strategy.close("long", when = out)

strategy.entry("short", strategy.short, when = shortcondition)
strategy.close("short", when = outt)

Published by ChaoZhang on the FMZ strategy square · last modified 2023-10-18. 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/EMA-Oscillation-Reversal-System-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.