Multi-factor Adaptive Momentum Tracking Strategy

Author: ChaoZhang, Date: 2023-12-12 12:02:13
Tags:

Overview

The multi-factor adaptive momentum tracking strategy realizes automated trading of highly volatile assets like cryptocurrencies by identifying market trends and key support/resistance levels through integrating multiple technical indicators. The strategy combines indicators like RSI, MACD, Stochastic to determine entry and exit timing, while also incorporating price percentage change to enable more accurate pattern recognition.

Strategy Principle

The core of the multi-factor adaptive momentum tracking strategy lies in the integration of multiple technical indicators. The main components used in this strategy include:

RSI to judge overbought/oversold conditions. Different parameters can be used to identify regular RSI signals or the tweaked Connors RSI signals to determine if reversal opportunities exist.
MACD to help determine trend direction. Buy and sell signals are generated when the MACD line crosses above or below the signal line.
Stochastic to spot overbought/oversold zones. Golden cross and death cross combinations of the K and D lines indicate whether reversals may occur.
Price percentage change to check if breakouts are real. Calculates the percentage change of highest price, lowest price, close price etc over a certain period to determine if a true breakout has occurred.
EMAs to judge overall trend. Upcrossing of fast EMA above slow EMA gives bullish signals while downcrossing gives bearish signals.

The strategy chooses to go long or short based on market conditions, and sets stop loss and take profit after entering positions to effectively control risks. Exits when reversal signals occur. The entire decision process integrates judgments from multiple factors to realize more accurate results.

Advantage Analysis

The advantages of this strategy include:

Multiple factors drive better judgment. Compared to single indicators, combining multiple indicators enables mutual verification and more reliable results, saving unnecessary trading costs.
Strict conditions avoid bad trades. The strategy sets strict requirements for buy/sell signals, requiring multiple simultaneous signals to filter out noise and avoid bad trades.
Adaptive parameters reduce manual interference. The built-in ability to dynamically calculate indicator parameters avoids the subjectivity of manual parameter selection, making the parameters more scientific and objective.
Stop loss/take profit controls risks. Strategy continuously calculates and plots stop loss/take profit levels after opening positions, effectively capping per trade loss and preventing margin calls.

Risk Analysis

Risks that need to be prevented include:

Probability of incorrect signals from indicators. Although the multiple verification process greatly reduces erroneous signals, some probability remains. This may lead to unnecessary losses.
Risk of stop loss being penetrated. In extreme market conditions, prices may cliff dive and penetrate originally set stop losses easily, leading to above average losses.
Overoptimization from parameter tuning. Although dynamic parameters reduce subjectivity, they may also lead to over-fitting and losing generalizability.

Solutions:

Raise strictness of signal filtering to reduce erroneous signals.
Adopt staged entries to avoid oversized single stop loss.
Enhance sample testing to strictly evaluate parameter stability.

Optimization Directions

This strategy can be further optimized through:

Increasing judgment factors. Combine signals from more indicators of different types, e.g. volatility, volume etc to assist judgment.
Optimizing stop loss algorithms. Introduce more advanced stop loss algorithms like trailing stop loss, volatility stop loss etc to further reduce the probability of stop loss being hit.
Introducing machine learning models. Model historical data using RNN, LSTM etc to aid in buy/sell decisions.
Strategy ensembling. Adopt multiple sub-strategies and use ensemble methods to integrate for more robust overall performance.

Conclusion

The multi-factor adaptive momentum tracking strategy integrates multiple technical indicators to identify trading opportunities. Compared to single indicators, this strategy has more accurate judgments, coupled with built-in parameter adaptation and stop loss mechanisms to control risks. Next steps include introducing more auxiliary judgment factors, advanced stop loss algorithms, machine learning etc to further enhance strategy performance.

/*backtest
start: 2023-12-04 00:00:00
end: 2023-12-11 00:00:00
period: 3m
basePeriod: 1m
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/
//@version=4

// ██████╗██████╗ ███████╗ █████╗ ████████╗███████╗██████╗     ██████╗ ██╗   ██╗    
//██╔════╝██╔══██╗██╔════╝██╔══██╗╚══██╔══╝██╔════╝██╔══██╗    ██╔══██╗╚██╗ ██╔╝                       
//██║     ██████╔╝█████╗  ███████║   ██║   █████╗  ██║  ██║    ██████╔╝ ╚████╔╝                        
//██║     ██╔══██╗██╔══╝  ██╔══██║   ██║   ██╔══╝  ██║  ██║    ██╔══██╗  ╚██╔╝                         
//╚██████╗██║  ██║███████╗██║  ██║   ██║   ███████╗██████╔╝    ██████╔╝   ██║                          
// ╚═════╝╚═╝  ╚═╝╚══════╝╚═╝  ╚═╝   ╚═╝   ╚══════╝╚═════╝     ╚═════╝    ╚═╝                          
                                                                                                     
//███████╗ ██████╗ ██╗     ██╗   ██╗████████╗██╗ ██████╗ ███╗   ██╗███████╗ ██╗ █████╗ ███████╗ █████╗ 
//██╔════╝██╔═══██╗██║     ██║   ██║╚══██╔══╝██║██╔═══██╗████╗  ██║██╔════╝███║██╔══██╗╚════██║██╔══██╗
//███████╗██║   ██║██║     ██║   ██║   ██║   ██║██║   ██║██╔██╗ ██║███████╗╚██║╚██████║    ██╔╝╚█████╔╝
//╚════██║██║   ██║██║     ██║   ██║   ██║   ██║██║   ██║██║╚██╗██║╚════██║ ██║ ╚═══██║   ██╔╝ ██╔══██╗
//███████║╚██████╔╝███████╗╚██████╔╝   ██║   ██║╚██████╔╝██║ ╚████║███████║ ██║ █████╔╝   ██║  ╚█████╔╝
//╚══════╝ ╚═════╝ ╚══════╝ ╚═════╝    ╚═╝   ╚═╝ ╚═════╝ ╚═╝  ╚═══╝╚══════╝ ╚═╝ ╚════╝    ╚═╝   ╚════╝ 

strategy(shorttitle='Ain1 No Label',title='All in One Strategy no RSI Label', overlay=true, scale=scale.left, initial_capital = 1000, process_orders_on_close=true, default_qty_type = strategy.percent_of_equity, default_qty_value = 100, commission_type=strategy.commission.percent, commission_value=0.18, calc_on_every_tick=true)

kcolor = color.new(#0094FF, 60)
dcolor = color.new(#FF6A00, 60)



// -----------------  Strategy Inputs -------------------------------------------------------------
//Backtest dates with auto finish date of today
start = input(defval = timestamp("01 April 2021 00:00 -0500"), title = "Start Time", type = input.time)
finish = input(defval = timestamp("31 December 2021 00:00 -0600"), title = "End Time", type = input.time)
window()  => true       // create function "within window of time"


// Strategy Selection - Long, Short, or Both
stratinfo = input(true, "Long/Short for Mixed Market, Long for Bull, Short for Bear")
strat = input(title="Trade Types", defval="Long/Short", options=["Long Only", "Long/Short", "Short Only"])
strat_val = strat == "Long Only" ? 1 : strat == "Long/Short" ? 0 : -1

// Risk Management Inputs
sl= input(10.0, "Stop Loss %", minval = 0, maxval = 100, step = 0.01)
stoploss = sl/100
tp = input(20.0, "Target Profit %", minval = 0, maxval = 100, step = 0.01)
TargetProfit = tp/100


useXRSI = input(false, "Use RSI crossing back, select only one strategy")
useCRSI = input(false, "Use Tweaked Connors RSI, select only one")
RSIInfo = input(true, "These are the RSI Strategy Inputs, RSI Length applies to MACD, set OB and OS to 45 for using Stoch and EMA strategies.")
length = input(14, "RSI Length", minval=1)
overbought= input(62, "Overbought")
oversold= input(35, "Oversold")
cl1 = input(3, "Connor's MA Length 1", minval=1, step=1)
cl2 = input(20, "Connor's MA Lenght 2", minval=1, step=1)
cl3 = input(50, "Connor's MA Lenght 3", minval=1, step=1)

// MACD and EMA Inputs
useMACD = input(false, "Use MACD Only, select only one strategy")
useEMA  = input(false, "Use EMA Only, select only one strategy (EMA uses Stochastic inputs too)")
MACDInfo=input(true, "These are the MACD strategy variables")
fastLength = input(5, minval=1, title="EMA Fast Length")
slowLength = input(10, minval=1, title="EMA Slow Length")
ob_min = input(52, "Overbought Lookback Minimum Value", minval=0, maxval=200)
ob_lb = input(25, "Overbought Lookback Bars", minval=0, maxval=100)
os_min = input(50, "Oversold Lookback Minimum Value", minval=0, maxval=200)
os_lb = input(35, "Oversold Lookback Bars", minval=0, maxval=100)
source = input(title="Source", type=input.source, defval=close)
RSI = rsi(source, length)


// Price Movement Inputs
PriceInfo = input(true, "Price Change Percentage Cross Check Inputs for all Strategies, added logic to avoid early sell")
lkbk = input(5,"Max Lookback Period")

// EMA and SMA Background Inputs
useStoch    = input(false, "Use Stochastic Strategy, choose only one")
StochInfo   = input(true, "Stochastic Strategy Inputs")
smoothK     = input(3, "K", minval=1)
smoothD     = input(3, "D", minval=1)
k_mode      = input("SMA", "K Mode", options=["SMA", "EMA", "WMA"])
high_source = input(high,"High Source")
low_source= input(low,"Low Source")
HTF = input("","Curernt or Higher time frame only", type=input.resolution)

// Selections to show or hide the overlays
showZones = input(true, title="Show Bullish/Bearish Zones")
showStoch = input(true, title="Show Stochastic Overlays")
showRSIBS = input(true, title="Show RSI Buy Sell Zones")
showMACD = input(true, title="Show MACD")
color_bars=input(true, "Color Bars")



// ------------------ Dynamic RSI Calculation ----------------------------------------

AvgHigh(src,cnt,val) =>
    total = 0.0
    count = 0
    for i = 0 to cnt
        if src[i] > val
            count := count + 1
            total := total + src[i]
    round(total / count)
    
RSI_high = AvgHigh(RSI, ob_lb, ob_min)

AvgLow(src,cnt,val) =>
    total = 0.0
    count = 0
    for i = 0 to cnt
        if src[i] < val
            count := count + 1
            total := total + src[i]
    round(total / count)

RSI_low = AvgLow(RSI, os_lb, os_min)




// ------------------ Price Percentage Change Calculation -----------------------------------------
perc_change(lkbk) =>
    overall_change = ((close[0] - open[lkbk]) / open[lkbk]) * 100
    highest_high = 0.0
    lowest_low = 0.0
    for i = lkbk to 0
        highest_high := i == lkbk ? high : high[i] > high[(i + 1)] ? high[i] : highest_high[1]
        lowest_low := i == lkbk ? low : low[i] < low[(i + 1)] ? low[i] : lowest_low[1]
    
    start_to_high = ((highest_high - open[lkbk]) / open[lkbk]) * 100
    start_to_low = ((lowest_low - open[lkbk]) / open[lkbk]) * 100
    previous_to_high = ((highest_high - open[1])/open[1])*100
    previous_to_low = ((lowest_low-open[1])/open[1])*100
    previous_bar = ((close[1]-open[1])/open[1])*100
    
    [overall_change, start_to_high, start_to_low, previous_to_high, previous_to_low, previous_bar]
    
// Call the function    
[overall, to_high, to_low, last_high, last_low, last_bar] = perc_change(lkbk)

// Plot the function
//plot(overall*50, color=color.white, title='Overall Percentage Change', linewidth=3)
//plot(to_high*50, color=color.green,title='Percentage Change from Start to High', linewidth=2)
//plot(to_low*50, color=color.red, title='Percentage Change from Start to Low', linewidth=2)
//plot(last_high*100, color=color.teal, title="Previous to High", linewidth=2)
//plot(last_low*100, color=color.maroon, title="Previous to Close", linewidth=2)
//plot(last_bar*100, color=color.orange, title="Previous Bar", linewidth=2)
//hline(0, title='Center Line', color=color.orange, linewidth=2)

true_dip = overall < 0 and to_high > 0 and to_low < 0 and last_high > 0 and last_low < 0 and last_bar < 0
true_peak = overall > 0 and to_high > 0 and to_low > 0 and last_high > 0 and last_low < 0 and last_bar > 0

alertcondition(true_dip, title='True Dip', message='Dip')
alertcondition(true_peak, title='True Peak', message='Peak')

// ------------------ Background Colors based on EMA Indicators -----------------------------------
// Uses standard lengths of 9 and 21, if you want control delete the constant definition and uncomment the inputs
haClose(gap) => (open[gap] + high[gap] + low[gap] + close[gap]) / 4
rsi_ema = rsi(haClose(0), length)
v2 = ema(rsi_ema, length)                                                
v3 = 2 * v2 - ema(v2, length)  
emaA = ema(rsi_ema, fastLength)                                     
emaFast = 2 * emaA - ema(emaA, fastLength)
emaB = ema(rsi_ema, slowLength)                                     
emaSlow = 2 * emaB - ema(emaB, slowLength) 

//plot(rsi_ema, color=color.white, title='RSI EMA', linewidth=3)
//plot(v2, color=color.green,title='v2', linewidth=2)
//plot(v3, color=color.red, title='v3', linewidth=2)
//plot(emaFast, color=color.teal, title="EMA Fast", linewidth=2)
//plot(emaSlow, color=color.maroon, title="EMA Slow", linewidth=2)

EMABuy = crossunder(emaFast, v2) and window()
EMASell = crossover(emaFast, emaSlow) and window()


alertcondition(EMABuy, title='EMA Buy', message='EMA Buy Condition')
alertcondition(EMASell, title='EMA Sell', message='EMA Sell Condition')



// bullish signal rule: 
bullishRule =emaFast > emaSlow
// bearish signal rule: 
bearishRule =emaFast < emaSlow

// current trading State
ruleState = 0
ruleState := bullishRule ? 1 : bearishRule ? -1 : nz(ruleState[1])
ruleColor = ruleState==1 ? color.new(color.blue, 90) : ruleState == -1 ? color.new(color.red, 90) : ruleState == 0 ? color.new(color.gray, 90) : na
bgcolor(showZones ? ruleColor : na, title="Bullish/Bearish Zones")


// ------------------  Stochastic Indicator Overlay -----------------------------------------------

// Calculation
// Use highest highs and lowest lows
h_high = highest(high_source ,lkbk)
l_low = lowest(low_source ,lkbk)

stoch = stoch(RSI, RSI_high, RSI_low, length)
k =
 k_mode=="EMA" ? ema(stoch, smoothK) :
 k_mode=="WMA" ? wma(stoch, smoothK) :
 sma(stoch, smoothK)
d = sma(k, smoothD)
k_c = change(k)
d_c = change(d)
kd = k - d

// Plot
signalColor = k>oversold and d<overbought and k>d and k_c>0 and d_c>0 ? kcolor : 
 k<overbought and d>oversold and k<d and k_c<0 and d_c<0 ? dcolor : na
kp = plot(showStoch ? k : na, "K", color=kcolor)
dp = plot(showStoch ? d : na, "D", color=dcolor)
fill(kp, dp, color = signalColor, title="K-D")
signalUp = showStoch ? not na(signalColor) and kd>0 : na
signalDown = showStoch ? not na(signalColor) and kd<0 : na
//plot(signalUp ? kd : na, "Signal Up", color=kcolor, transp=90, style=plot.style_columns)
//plot(signalDown ? (kd+100) : na , "Signal Down", color=dcolor, transp=90, style=plot.style_columns, histbase=100)

//StochBuy = crossover(k, d) and kd>0 and to_low<0 and window()
//StochSell = crossunder(k,d) and kd<0 and to_high>0 and window()

StochBuy = crossover(k, d) and window()
StochSell = crossunder(k, d) and window()

alertcondition(StochBuy, title='Stoch Buy', message='K Crossing D')
alertcondition(StochSell, title='Stoch Sell', message='D Crossing K')


// -------------- Add Price Movement -------------------------
// Calculations
h1 = vwma(high, length)
l1 = vwma(low, length)
hp = h_high[1]
lp = l_low[1]

// Plot
var plot_color=#353535
var sig = 0
if (h1 >hp)
    sig:=1
    plot_color:=color.lime
else if (l1 <lp)
    sig:=-1
    plot_color:=color.maroon
//plot(1,title = "Price Movement Bars", style=plot.style_columns,color=plot_color)
//plot(sig,title="Signal 1 or -1",display=display.none)



// --------------------------------------- RSI Plot ----------------------------------------------
// Plot Oversold and Overbought Lines
over = hline(oversold, title="Oversold", color=color.green)
under = hline(overbought, title="Overbought", color=color.red)
fillcolor = color.new(#9915FF, 90)
fill(over, under, fillcolor, title="Band Background")


// Show RSI and EMA crosses with arrows and RSI Color (tweaked Connors RSI)
// Improves strategy setting ease by showing where EMA 5 crosses EMA 10 from above to confirm overbought conditions or trend reversals
// This shows where you should enter shorts or exit longs

// Tweaked Connors RSI Calculation
connor_ob = overbought
connor_os = oversold
ma1 = sma(close,cl1)
ma2 = sma(close, cl2)
ma3 = sma(close, cl3)

// Buy Sell Zones using tweaked Connors RSI (RSI values of 80 and 20 for Crypto as well as ma3, ma20, and ma50 are the tweaks)
RSI_SELL = ma1 > ma2 and open > ma3 and RSI >= connor_ob and true_peak and window()
RSI_BUY = ma2 < ma3 and ma3 > close and RSI <= connor_os and true_dip and window()

alertcondition(RSI_BUY, title='Connors Buy', message='Connors RSI Buy')
alertcondition(RSI_SELL, title='Connors Sell', message='Connors RSI Sell')

// Color Definition
col = useCRSI ? (close > ma2 and close < ma3 and RSI <= connor_os ? color.lime : close < ma2 and close > ma3 and RSI <= connor_ob ? color.red : color.yellow ) : color.yellow

// Plot colored RSI Line
plot(RSI, title="RSI", linewidth=3, color=col)


//------------------- MACD Strategy -------------------------------------------------
[macdLine, signalLine, _] = macd(close, fastLength, slowLength, length)

bartrendcolor = macdLine > signalLine and k > 50 and RSI > 50 ? color.teal : macdLine < signalLine and k < 50 and RSI < 50 ? color.maroon : macdLine < signalLine ? color.yellow : color.gray
barcolor(color = color_bars ? bartrendcolor : na)


MACDBuy = macdLine>signalLine and RSI<RSI_low and overall<0 and window()
MACDSell = macdLine<signalLine and RSI>RSI_high and overall>0 and window()

//plotshape(showMACD ? MACDBuy: na, title = "MACD Buy", style = shape.arrowup, text = "MACD Buy", color=color.green, textcolor=color.green, size=size.small)
//plotshape(showMACD ? MACDSell: na, title = "MACD Sell", style = shape.arrowdown, text = "MACD Sell", color=color.red, textcolor=color.red, size=size.small)
MACColor = MACDBuy ? color.new(color.teal, 50) : MACDSell ? color.new(color.maroon, 50) : na
bgcolor(showMACD ? MACColor : na, title ="MACD Signals")


// -------------------------------- Entry and Exit Logic ------------------------------------


// Entry Logic
XRSI_OB = crossunder(RSI, overbought) and overall<0 and window()
RSI_OB = RSI>overbought and true_peak and window()
XRSI_OS = crossover(RSI, oversold) and overall>0 and window()
RSI_OS = RSI<oversold and true_dip and window()

alertcondition(XRSI_OB, title='Reverse RSI Sell', message='RSI Crossing back under OB')
alertcondition(XRSI_OS, title='Reverse RSI Buy', message='RSI Crossing back over OS')

alertcondition(RSI_OS, title='RSI Buy', message='RSI Crossover OS')
alertcondition(RSI_SELL, title='RSI Sell', message='RSI Crossunder OB')


// Strategy Entry and Exit with built in Risk Management
GoLong = strategy.position_size==0 and strat_val > -1 and rsi_ema > RSI and k < d ? (useXRSI ? XRSI_OS : useMACD ? MACDBuy : useCRSI ? RSI_BUY : useStoch ? StochBuy : RSI_OS) : false

GoShort = strategy.position_size==0 and strat_val < 1 and rsi_ema < RSI and d < k ? (useXRSI ? XRSI_OB : useMACD ? MACDSell : useCRSI ? RSI_SELL : useStoch ? StochSell : RSI_OB) : false

if (GoLong)
    strategy.entry("LONG", strategy.long)

if (GoShort) 
    strategy.entry("SHORT", strategy.short)


longStopPrice  = strategy.position_avg_price * (1 - stoploss)
longTakePrice  = strategy.position_avg_price * (1 + TargetProfit)
shortStopPrice = strategy.position_avg_price * (1 + stoploss)
shortTakePrice = strategy.position_avg_price * (1 - TargetProfit)

//plot(series=(strategy.position_size > 0) ? longTakePrice : na, color=color.green, style=plot.style_circles, linewidth=3, title="Long Take Profit")
//plot(series=(strategy.position_size < 0) ? shortTakePrice : na, color=color.green, style=plot.style_circles, linewidth=3, title="Short Take Profit")
//plot(series=(strategy.position_size > 0) ? longStopPrice : na, color=color.red, style=plot.style_cross, linewidth=2, title="Long Stop Loss")
//plot(series=(strategy.position_size < 0) ? shortStopPrice : na, color=color.red, style=plot.style_cross, linewidth=2, title="Short Stop Loss")

if (strategy.position_size > 0)
    strategy.exit(id="Exit Long", from_entry = "LONG", stop = longStopPrice, limit = longTakePrice)
    
if (strategy.position_size < 0)
    strategy.exit(id="Exit Short", from_entry = "SHORT", stop = shortStopPrice, limit = shortTakePrice)


CloseLong = strat_val > -1 and strategy.position_size > 0 and rsi_ema > RSI and d > k ? (useXRSI ? XRSI_OB : useMACD ? MACDSell : useCRSI ? RSI_SELL : RSI_OB) : false

if(CloseLong)
    strategy.close("LONG")
        
CloseShort = strat_val < 1 and strategy.position_size < 0 and rsi_ema < RSI and k > d ? (useXRSI ? XRSI_OS : useMACD ? MACDBuy : useCRSI ? RSI_BUY : RSI_OS) : false

if(CloseShort)
    strategy.close("SHORT")

More

Forums
|
API
|
Github
|
About
|
SiteMap