0bps

Strategy vault / Mean reversion

Automated Quantitative Trading Strategy Based on Inside Bar and Moving Average

The core idea of this strategy is to combine inside bar patterns and moving average indicators to implement automated trading. When an inside bar pattern appears, it indicates that the current trend may be reversing…

Pine ScriptMean reversionTrend followingMACDEMASMAATRADX / DMI
Author
ChaoZhang
Complexity
Advanced · 264 lines
Parameters
25
Backtest bar
1h · intraday
Tested on
BTC/USDT · futures
Jump to codeView in vaultOriginal on FMZ

IMG

Overview

The core idea of this strategy is to combine inside bar patterns and moving average indicators to implement automated trading. When an inside bar pattern appears, it indicates that the current trend may be reversing. At this point, we use the position of the moving average line to determine the final trading direction.

Strategy Principle

  1. Identify inside bar patterns. An inside bar refers to a candlestick where both the high and the low are within the real body of the previous bar. Based on the color of the real body, we can judge if it is a bullish or bearish inside bar.

  2. Check the position of the moving average line. When an inside bar is found, if the price is above the moving average line, it is a bullish signal. If the price is below the moving average line, it is a bearish signal.

  3. Combine the inside bar pattern and the moving average signal to determine the final trading direction. That is going short when the bearish inside bar breaks below the moving average line, and going long when the bullish inside bar breaks above the line.

Advantages of the Strategy

  1. Combining technical indicators and price patterns improves the accuracy of trading decisions.

  2. Inside bars themselves contain strong price reversal signals that can identify trend reversal points early.

  3. The moving average filters out some noise and avoids getting caught in range-bound oscillations.

  4. Fully automated trading greatly reduces the time and effort costs of manual trading.

Risks of the Strategy & Solutions

  1. When prices oscillate around the moving average line, more false signals may appear, leading to over-trading. This can be reduced by optimizing the moving average parameters or adding filtering conditions.

  2. This strategy works better in markets with clear trends. The performance may suffer in oscillating markets. Trend-judging indicators like ADX can be added to control the algorithm activation.

  3. There is some time lag. Shortening parameters or optimizing moving average calculation methods may reduce the lag.

  4. The risk of large drawdowns is considerable. Stop loss can control downside risk. Position sizing optimization also helps decrease drawdowns.

Directions for Strategy Optimization

  1. Optimize inside bar determination period parameters to find the best combination.

  2. Try different types of moving averages, like EMA and SMA, to decide the most suitable one.

  3. Add auxiliary indicators like MACD and KDJ to enrich trading signals and improve accuracy.

  4. Incorporate filtering indicators like ADX and ATR to control algorithm activation in unsuitable market environments.

  5. Optimize position management, such as risk-based sizing, pullback sizing etc. to better control risk and pursue higher returns.

Conclusion

This strategy implements a fully-automated quantitative trading solution by dynamically tracking inside bar signals and moving average indicators. The signal generation is simple and clear for easy understanding and tracking. It performs well in markets with obvious trends. Further optimization of parameters and rules can enhance the stability and profitability.

Parameters

NameDefaultWhat it controls
v_input_12Lookback for Inside Bar
v_input_source_1_close0_____ falls above HIGH of inside bar (Long condition): close | high | low | open | hl2 | hlc3 | hlcc4 | ohlc4
v_input_source_2_close0_____ falls below LOW of inside bar (Short condition): close | high | low | open | hl2 | hlc3 | hlcc4 | ohlc4
v_input_source_3_close0EMA Source: close | high | low | open | hl2 | hlc3 | hlcc4 | ohlc4
v_input_250EMA Length
v_input_bool_1false(?ADX)Average Directional Index (ADX)
v_input_314ADX Smoothing
v_input_414DI Length
v_input_525ADX Threshold
v_input_bool_2true(?Date Range)Start
v_input_6timestamp(1 Jan 2019)startPeriodTime
v_input_bool_3trueEnd
v_input_7timestamp(31 Dec 2030)endPeriodTime
v_input_string_10(?Trade Direction)Trade Direction: Long and Short | Long Only | Short Only
v_input_float_110.5(?Take Profit)Take Profit 1 - Target %
v_input_int_125% Of Position
v_input_float_211Take Profit 2 - Target %
v_input_int_225% Of Position
v_input_float_311.5Take Profit 3 - Target %
v_input_int_325% Of Position
v_input_float_412Take Profit 4 - Target %
v_input_float_54(?Stop Loss)Stop Loss (%)
v_input_float_6true(?Leverage)Leverage
v_input_string_2CRYPTANEX_99FTX_Strategy-Name-Here(?ProfitView Alert Syntax)Alert Syntax Prefix
v_input_bool_4false(?Dashboard)Show Dashboard

Source Pine Script

/*backtest
start: 2023-11-20 00:00:00
end: 2023-12-20 00:00:00
period: 1h
basePeriod: 15m
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/
// © myn

//@version=5
strategy('Strategy Myth-Busting #10 - InsideBar+EMA - [MYN]', max_bars_back=5000, overlay=true, pyramiding=0, initial_capital=20000, currency='USD', default_qty_type=strategy.percent_of_equity, default_qty_value=100.0, commission_value=0.075, use_bar_magnifier = false)

/////////////////////////////////////
//* Put your strategy logic below *//
/////////////////////////////////////
//short if: inside bar and bearish & below 50 ema & price falls below low of inside bar. Opposite for long. on 4H TF

// Inside Bar
//░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

f_priorBarsSatisfied(_objectToEval, _numOfBarsToLookBack) => 
    returnVal = false
    for i = 0 to _numOfBarsToLookBack
        if (_objectToEval[i] == true)
            returnVal = true

i_numLookbackBars = input(2,title="Lookback for Inside Bar")



// This source code is subject to the terms of the GNU License 2.0 at https://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html
// © cma

//@version=5
//indicator('Inside Bar Ind/Alert', overlay=true)

bullishBar = 1
bearishBar = -1

isInside() =>
    previousBar = 1
    bodyStatus = close >= open ? 1 : -1
    isInsidePattern = high < high[previousBar] and low > low[previousBar]

    isInsidePattern ? bodyStatus : 0

barcolor(isInside() == bullishBar ? color.green : na)
barcolor(isInside() == bearishBar ? color.red : na)

// When is bullish bar paint green
plotshape(isInside() == bullishBar, style=shape.triangleup, location=location.abovebar, color=color.new(color.green, 0))

// When is bearish bar paint red
plotshape(isInside() == bearishBar, style=shape.triangledown, location=location.belowbar, color=color.new(color.red, 0))

isInsideBarMade = isInside() == bullishBar or isInside() == bearishBar

alertcondition(isInsideBarMade, title='Inside Bar', message='Inside Bar came up!')

i_srcInsideBarLong = input.source(close, title = "_____ falls above HIGH of inside bar (Long condition)")
i_srcInsideBarShort = input.source(close, title = "_____ falls below LOW of inside bar (Short condition)")


//if: inside bar and falls below low of inside bar. I think.
insideBarLongEntry = f_priorBarsSatisfied(isInside() == bullishBar,i_numLookbackBars) and i_srcInsideBarLong > high[i_numLookbackBars]     //isInside() == bullishBar
insideBarShortEntry = f_priorBarsSatisfied(isInside() == bearishBar,i_numLookbackBars)  and i_srcInsideBarShort < low[i_numLookbackBars]    //isInside() == bearishBar


// EMA
//░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

i_src = input.source(close, title = "EMA Source")
i_emaLength = input(50,title="EMA Length")
ema = ta.ema(i_src, i_emaLength)
emaPlot = plot(series=ema,color=color.blue, linewidth=2)

emaLongEntry = i_src > ema
emaShortEntry = i_src < ema

//////////////////////////////////////
//* Put your strategy rules below *//
/////////////////////////////////////

longCondition = insideBarLongEntry and emaLongEntry
shortCondition = insideBarShortEntry and emaShortEntry
//define as 0 if do not want to use
closeLongCondition = 0
closeShortCondition = 0


// ADX
//░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

adxEnabled = input.bool(defval = false , title = "Average Directional Index (ADX)", tooltip = "", group ="ADX" ) 
adxlen = input(14, title="ADX Smoothing", group="ADX")
adxdilen = input(14, title="DI Length", group="ADX")
adxabove = input(25, title="ADX Threshold", group="ADX")

adxdirmov(len) =>
	adxup = ta.change(high)
	adxdown = -ta.change(low)
	adxplusDM = na(adxup) ? na : (adxup > adxdown and adxup > 0 ? adxup : 0)
	adxminusDM = na(adxdown) ? na : (adxdown > adxup and adxdown > 0 ? adxdown : 0)
	adxtruerange = ta.rma(ta.tr, len)
	adxplus = fixnan(100 * ta.rma(adxplusDM, len) / adxtruerange)
	adxminus = fixnan(100 * ta.rma(adxminusDM, len) / adxtruerange)
	[adxplus, adxminus]
adx(adxdilen, adxlen) =>
	[adxplus, adxminus] = adxdirmov(adxdilen)
	adxsum = adxplus + adxminus
	adx = 100 * ta.rma(math.abs(adxplus - adxminus) / (adxsum == 0 ? 1 : adxsum), adxlen)

adxsig = adxEnabled ? adx(adxdilen, adxlen) : na
isADXEnabledAndAboveThreshold = adxEnabled ? (adxsig > adxabove) : true

//Backtesting Time Period (Input.time not working as expected as of 03/30/2021.  Giving odd start/end dates
//░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
useStartPeriodTime = input.bool(true, 'Start', group='Date Range', inline='Start Period')
startPeriodTime = input(timestamp('1 Jan 2019'), '', group='Date Range', inline='Start Period')
useEndPeriodTime = input.bool(true, 'End', group='Date Range', inline='End Period')
endPeriodTime = input(timestamp('31 Dec 2030'), '', group='Date Range', inline='End Period')

start = useStartPeriodTime ? startPeriodTime >= time : false
end = useEndPeriodTime ? endPeriodTime <= time : false
calcPeriod = true

// Trade Direction 
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
tradeDirection = input.string('Long and Short', title='Trade Direction', options=['Long and Short', 'Long Only', 'Short Only'], group='Trade Direction')

// Percent as Points
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
per(pcnt) =>
    strategy.position_size != 0 ? math.round(pcnt / 100 * strategy.position_avg_price / syminfo.mintick) : float(na)

// Take profit 1
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
tp1 = input.float(title='Take Profit 1 - Target %', defval=10.5, minval=0.0, step=0.5, group='Take Profit', inline='Take Profit 1')
q1 = input.int(title='% Of Position', defval=25, minval=0, group='Take Profit', inline='Take Profit 1')

// Take profit 2
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
tp2 = input.float(title='Take Profit 2 - Target %', defval=11, minval=0.0, step=0.5, group='Take Profit', inline='Take Profit 2')
q2 = input.int(title='% Of Position', defval=25, minval=0, group='Take Profit', inline='Take Profit 2')

// Take profit 3
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
tp3 = input.float(title='Take Profit 3 - Target %', defval=11.5, minval=0.0, step=0.5, group='Take Profit', inline='Take Profit 3')
q3 = input.int(title='% Of Position', defval=25, minval=0, group='Take Profit', inline='Take Profit 3')

// Take profit 4
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
tp4 = input.float(title='Take Profit 4 - Target %', defval=12, minval=0.0, step=0.5, group='Take Profit')

/// Stop Loss
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
stoplossPercent = input.float(title='Stop Loss (%)', defval=4, minval=0.01, group='Stop Loss') * 0.01
slLongClose = close < strategy.position_avg_price * (1 - stoplossPercent)
slShortClose = close > strategy.position_avg_price * (1 + stoplossPercent)

/// Leverage
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
leverage = input.float(1, 'Leverage', step=.5, group='Leverage')
contracts = math.min(math.max(.000001, strategy.equity / close * leverage), 1000000000)


/// Trade State Management
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

isInLongPosition = strategy.position_size > 0
isInShortPosition = strategy.position_size < 0

/// ProfitView Alert Syntax String Generation
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

alertSyntaxPrefix = input.string(defval='CRYPTANEX_99FTX_Strategy-Name-Here', title='Alert Syntax Prefix', group='ProfitView Alert Syntax')
alertSyntaxBase = alertSyntaxPrefix + '\n#' + str.tostring(open) + ',' + str.tostring(high) + ',' + str.tostring(low) + ',' + str.tostring(close) + ',' + str.tostring(volume) + ','


/// Trade Execution
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░

longConditionCalc = (longCondition and isADXEnabledAndAboveThreshold)
shortConditionCalc = (shortCondition and isADXEnabledAndAboveThreshold)

if calcPeriod
    if longConditionCalc and tradeDirection != 'Short Only' and isInLongPosition == false
        strategy.entry('Long', strategy.long, qty=contracts)

        alert(message=alertSyntaxBase + 'side:long', freq=alert.freq_once_per_bar_close)

    if shortConditionCalc and tradeDirection != 'Long Only' and isInShortPosition == false
        strategy.entry('Short', strategy.short, qty=contracts)

        alert(message=alertSyntaxBase + 'side:short', freq=alert.freq_once_per_bar_close)
    
    //Inspired from Multiple %% profit exits example by adolgo https://www.tradingview.com/script/kHhCik9f-Multiple-profit-exits-example/
    strategy.exit('TP1', qty_percent=q1, profit=per(tp1))
    strategy.exit('TP2', qty_percent=q2, profit=per(tp2))
    strategy.exit('TP3', qty_percent=q3, profit=per(tp3))
    strategy.exit('TP4', profit=per(tp4))

    strategy.close('Long', qty_percent=100, comment='SL Long', when=slLongClose)
    strategy.close('Short', qty_percent=100, comment='SL Short', when=slShortClose)

    strategy.close_all(when=closeLongCondition or closeShortCondition, comment='Close Postion')

/// Dashboard
// ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
// Inspired by https://www.tradingview.com/script/uWqKX6A2/ - Thanks VertMT

showDashboard = input.bool(group="Dashboard", title="Show Dashboard", defval=false)

f_fillCell(_table, _column, _row, _title, _value, _bgcolor, _txtcolor) =>
    _cellText = _title + "\n" + _value
    table.cell(_table, _column, _row, _cellText, bgcolor=_bgcolor, text_color=_txtcolor, text_size=size.auto)

// Draw dashboard table
if showDashboard
    var bgcolor = color.new(color.black,0)
    
    // Keep track of Wins/Losses streaks
    newWin  = (strategy.wintrades  > strategy.wintrades[1]) and (strategy.losstrades == strategy.losstrades[1]) and (strategy.eventrades == strategy.eventrades[1])
    newLoss = (strategy.wintrades == strategy.wintrades[1]) and (strategy.losstrades  > strategy.losstrades[1]) and (strategy.eventrades == strategy.eventrades[1])

    varip int winRow     = 0
    varip int lossRow    = 0
    varip int maxWinRow  = 0
    varip int maxLossRow = 0

    if newWin
        lossRow := 0
        winRow := winRow + 1
    if winRow > maxWinRow
        maxWinRow := winRow
        
    if newLoss
        winRow := 0
        lossRow := lossRow + 1
    if lossRow > maxLossRow
        maxLossRow := lossRow


    // Prepare stats table
    var table dashTable = table.new(position.bottom_right, 1, 15, border_width=1)
    
   
    if barstate.islastconfirmedhistory
        // Update table
        dollarReturn = strategy.netprofit
        f_fillCell(dashTable, 0, 0, "Start:", str.format("{0,date,long}", strategy.closedtrades.entry_time(0)) , bgcolor, color.white) // + str.format(" {0,time,HH:mm}", strategy.closedtrades.entry_time(0)) 
        f_fillCell(dashTable, 0, 1, "End:", str.format("{0,date,long}", strategy.opentrades.entry_time(0)) , bgcolor, color.white) // + str.format(" {0,time,HH:mm}", strategy.opentrades.entry_time(0))
        _profit = (strategy.netprofit / strategy.initial_capital) * 100
        f_fillCell(dashTable, 0, 2, "Net Profit:", str.tostring(_profit, '##.##') + "%", _profit > 0 ? color.green : color.red, color.white)
        _numOfDaysInStrategy = (strategy.opentrades.entry_time(0) - strategy.closedtrades.entry_time(0)) / (1000 * 3600 * 24)
        f_fillCell(dashTable, 0, 3, "Percent Per Day", str.tostring(_profit / _numOfDaysInStrategy, '#########################.#####')+"%", _profit > 0 ? color.green : color.red, color.white)
        _winRate = ( strategy.wintrades / strategy.closedtrades ) * 100
        f_fillCell(dashTable, 0, 4, "Percent Profitable:", str.tostring(_winRate, '##.##') + "%", _winRate < 50 ? color.red : _winRate < 75 ? #999900 : color.green, color.white)
        f_fillCell(dashTable, 0, 5, "Profit Factor:", str.tostring(strategy.grossprofit / strategy.grossloss,  '##.###'), strategy.grossprofit > strategy.grossloss ? color.green : color.red, color.white)
        f_fillCell(dashTable, 0, 6, "Total Trades:", str.tostring(strategy.closedtrades), bgcolor, color.white)
        f_fillCell(dashTable, 0, 8, "Max Wins In A Row:", str.tostring(maxWinRow, '######') , bgcolor, color.white)
        f_fillCell(dashTable, 0, 9, "Max Losses In A Row:", str.tostring(maxLossRow, '######') , bgcolor, color.white)

Published by ChaoZhang on the FMZ strategy square · last modified 2023-12-21. 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/Automated-Quantitative-Trading-Strategy-Based-on-Inside-Bar-and-Moving-Average.md
  1. npx agenkit install engineering-kit in your project
  2. Paste the command into your agent and approve each phase
  3. Get tests, backtest, risk controls and a paper-trading switch
Get AgenKitHow /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.

Related specs

esc

Try , , or .

↑↓ move↵ open