Strategi Keseimbangan Psikologi Dagangan

Penulis:ChaoZhang, Tarikh: 2024-02-21 14:33:04
Tag:

img

Ringkasan

Tujuan strategi ini adalah untuk menyeimbangkan psikologi dan prestasi peniaga melalui penyesuaian pelbagai parameter, untuk mendapatkan pulangan yang lebih mantap. Ia menggunakan penunjuk seperti purata bergerak, Bollinger Bands dan Saluran Keltner untuk menentukan trend dan turun naik pasaran, bersama dengan penunjuk PSAR untuk mengenal pasti isyarat pembalikan.

Logika Strategi

Logik teras strategi ini adalah seperti berikut:

  1. Penghakiman trend: purata bergerak EMA digunakan untuk menentukan arah trend harga. Harga di atas EMA menandakan trend menaik manakala harga di bawah EMA menunjukkan trend menurun.

  2. Mengenali pembalikan: penunjuk PSAR menjumpai titik pembalikan harga. Titik PSAR yang muncul di atas harga menandakan panjang sementara titik yang muncul di bawah harga memanggil pendek.

  3. Momentum pengukur: penunjuk TTM Squeeze mengukur turun naik dan momentum pasaran. Ia membandingkan Bollinger Bands dan Saluran Keltner untuk mengukur tekanan dan lonjakan turun naik. Squeeze bermaksud turun naik yang sangat rendah sementara pelepasan tekanan menandakan pergerakan harga arah besar yang akan datang.

  4. Menghasilkan isyarat dagangan: isyarat panjang dicetuskan apabila harga menyeberang di atas garis EMA dan titik PSAR, disertai dengan pelepasan Squeeze TTM. Isyarat pendek berlaku apabila harga menyeberang di bawah EMA dan PSAR, bersama-sama dengan pemicu Squeeze TTM.

  5. Kaedah Stop Loss: asas stop loss tinggi-rendah pada harga tinggi/rendah baru-baru ini didarabkan dengan faktor yang ditetapkan.

  6. Kaedah mengambil keuntungan: risiko-balasan mengambil keuntungan secara automatik mengira sasaran keuntungan berdasarkan jarak stop loss dari harga semasa didarabkan dengan nisbah risiko-balasan yang telah ditetapkan.

Pelbagai parameter membolehkan peniaga menyeimbangkan psikologi dengan mengawal kekerapan perdagangan, saiz kedudukan, tahap stop loss dan mengambil mata keuntungan.

Analisis Kelebihan

Keutamaan strategi ini termasuk:

  1. Ketepatan isyarat yang lebih tinggi daripada konsensus pelbagai penunjuk

  2. Terutama berpusat pada pembalikan, mengurangkan kemungkinan pemudarangan palsu

  3. TTM Squeeze mengukur penyatuan untuk mengelakkan perdagangan yang tidak berkesan

  4. Simpel dan boleh diselaraskan kehilangan henti tinggi rendah

  5. Risiko-balasan mengambil keuntungan mengukur nisbah keuntungan untuk penyesuaian mudah

  6. Parameter yang fleksibel untuk menyesuaikan pilihan risiko peribadi

Analisis Risiko

Risiko strategi terdiri daripada:

  1. Peningkatan kemungkinan isyarat masuk yang hilang dari pelbagai penunjuk

  2. Prestasi yang kurang baik di pasaran trend berterusan

  3. Pelanggaran Stop Loss berkala melebihi jangkaan

  4. Potensi pembatalan keluar risiko-balasan oleh whipsaws harga

  5. Penyesuaian parameter yang tidak sesuai boleh menyebabkan kehilangan atau over-stop out

Arahan pengoptimuman

Bidang penambahbaikan yang mungkin meliputi:

  1. Tambah atau sesuaikan berat penunjuk untuk ketepatan isyarat yang lebih tinggi

  2. Mengoptimumkan parameter pembalikan dan trend untuk menangkap keuntungan yang lebih baik

  3. Memperbaiki tahap kehilangan hentian tinggi-rendah untuk kecekapan maksimum

  4. Uji nisbah risiko-balasan yang berbeza untuk hasil yang optimum

  5. Penyesuaian saiz kedudukan untuk meminimumkan kesan kerugian perdagangan tunggal

Ringkasan

Ringkasnya, melalui kombinasi penunjuk dan tetapan yang boleh disesuaikan, strategi ini mampu menyeimbangkan psikologi perdagangan dan memastikan hasil positif yang stabil. Walaupun beberapa kenaikan yang masih ada, ia telah menunjukkan penerapan praktikal. maklum balas dan kalibrasi pasaran langsung yang lebih lanjut mungkin akan meningkatkannya menjadi alat yang berkesan untuk menguruskan emosi dan mencapai keuntungan stabil jangka panjang.


/*backtest
start: 2024-01-01 00:00:00
end: 2024-01-31 23:59:59
period: 1h
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/


//@version=5
// This source code is subject to the terms of the Mozilla Public License 2.0 at https://mozilla.org/MPL/2.0/
// © simwai
strategy('Octopus Nest Strategy 🐙', shorttitle='🐙', overlay=true )

// -- Colors --
color maximumYellowRed = color.rgb(255, 203, 98) // yellow
color rajah = color.rgb(242, 166, 84) // orange
color magicMint = color.rgb(171, 237, 198)
color languidLavender = color.rgb(232, 215, 255)
color maximumBluePurple = color.rgb(181, 161, 226)
color skyBlue = color.rgb(144, 226, 244)
color lightGray = color.rgb(214, 214, 214)
color quickSilver = color.rgb(163, 163, 163)
color mediumAquamarine = color.rgb(104, 223, 153)
color carrotOrange = color.rgb(239, 146, 46)

// -- Inputs --
float src = input.source(close, 'Choose Source', group='General', inline='1')
bool isSignalLabelEnabled = input.bool(title='Show Signal Labels?', defval=true, group='General', inline='2')
bool isPsarAdaptive = input.bool(title='Is PSAR Adaptive?', defval=false, group='General', inline='2')

float highLowStopLossMultiplier = input.float(defval=0.98,  step=0.01, minval=0, maxval=1, title='Multiplier', group='High Low Stop Loss', inline='1')
float highLowStopLossBackupMultiplier = input.float(defval=0.98, step=0.01, minval=0, maxval=1, title='Backup Multiplier', group='High Low Stop Loss', inline='1')
int highLowStopLossLookback = input.int(defval=20, step=5, minval=1, title='Lookback', group='High Low Stop Loss', inline='2')
float automaticHighLowTakeProfitRatio = input.float(defval=1.125, step=0.1, minval=0, title='Risk Reward Ratio', group='Automatic High Low Take Profit', inline='2')

int emaLength = input.int(100, minval=2, title='Length', group='EMA', inline='1')
int ttmLength = input.int(title='Length', defval=20, minval=0, group='TTM Squeeze', inline='1')

float psarStart = input.float(0.02, 'Start', step=0.01, minval=0.0, group='PSAR', inline='1')
float psarInc = input.float(0.02, 'Increment', step=0.01, minval=0.01, group='PSAR', inline='1')
float psarMax = input.float(0.2, 'Max', step=0.05, minval=0.0, group='PSAR', inline='2')

startAFactor = input.float(0.02, 'Starting Acceleration Factor', step = 0.001, group='Adaptive PSAR', inline='1')
minStep = input.float(0.0, 'Min Step', step = 0.001, group='Adaptive PSAR', inline='1')
maxStep = input.float(0.02, 'Max Step', step = 0.001, group='Adaptive PSAR', inline='2')
maxAFactor = input.float(0.2, 'Max Acceleration Factor', step = 0.001, group='Adaptive PSAR', inline='2')  

hiloMode = input.string('On', 'HiLo Mode', options = ['Off', 'On'], group='Adaptive PSAR')
adaptMode = input.string('Kaufman', 'Adaptive Mode', options = ['Off', 'Kaufman', 'Ehlers'], group='Adaptive PSAR')
adaptSmth = input.int(5, 'Adaptive Smoothing Period', minval = 1, group='Adaptive PSAR')
filt = input.float(0.0, 'Filter in Pips', group='Adaptive PSAR', minval = 0)
minChng = input.float(0.0, 'Min Change in Pips', group='Adaptive PSAR', minval = 0)
SignalMode = input.string('Only Stops', 'Signal Mode', options = ['Only Stops', 'Signals & Stops'], group='Adaptive PSAR')

// -- Functions --
tr(_high, _low, _close) => math.max(_high - _low, math.abs(_high - _close[1]), math.abs(_low - _close[1]))

// -- Calculation --
var string lastTrade = 'initial'

float _low = low
float _high = high
float _close = close

// -- TTM Squeeze – Credits to @Greeny --
bband(ttmLength, mult) =>
    ta.sma(src, ttmLength) + mult * ta.stdev(src, ttmLength)
keltner(ttmLength, mult) =>
    ta.ema(src, ttmLength) + mult * ta.ema(tr(_high, _low, _close), ttmLength)

e1 = (ta.highest(_high, ttmLength) + ta.lowest(_low, ttmLength)) / 2 + ta.sma(src, ttmLength)
osc = ta.linreg(src - e1 / 2, ttmLength, 0)
diff = bband(ttmLength, 2) - keltner(ttmLength, 1)
osc_color = osc[1] < osc[0] ? osc[0] >= 0 ? #00ffff : #cc00cc : osc[0] >= 0 ? #009b9b : #ff9bff
mid_color = diff >= 0 ? color.green : color.red

// -- PSAR --
// Credits to @Bjorgum
calcBaseUnit() =>
    bool  isForexSymbol = syminfo.type     == 'forex'
    bool  isYenPair     = syminfo.currency == 'JPY'
    float result = isForexSymbol ? isYenPair ? 0.01 : 0.0001 : syminfo.mintick

// Credits to @loxx
_afact(mode,input, per, smooth) =>
    eff = 0., seff = 0.
    len = 0, sum = 0., max = 0., min = 1000000000.
    len := mode == 'Kaufman' ? math.ceil(per) : math.ceil(math.max(20, 5 * per))
    for i = 0 to len 
        if (mode == 'Kaufman') 
            sum += math.abs(input[i] - input[i + 1])
        else
            max := input[i] > max ? input[i] : max
            min := input[i] < min ? input[i] : min
    if (mode == 'Kaufman' and sum != 0) 
        eff := math.abs(input - input[len]) / sum
    else
        if (mode == 'Ehlers' and (max - min) > 0) 
            eff := (input - min) / (max - min)
    seff := ta.ema(eff, smooth)
    seff

hVal2 = nz(high[2]), hVal1 = nz(high[1]), hVal0 = high
lowVal2 = nz(low[2]), lowVal1 = nz(low[1]), lowVal0 = low
hiprice2 = nz(high[2]), hiprice1 = nz(high[1]), hiprice0 = high
loprice2 = nz(low[2]), loprice1 = nz(low[1]), loprice0 = low

upSig = 0., dnSig = 0.
aFactor = 0., step = 0., trend = 0.
upTrndSAR = 0., dnTrndSAR = 0.
length = (2 / maxAFactor - 1)

if (hiloMode == 'On') 
    hiprice0 := high
    loprice0 := low
else
    hiprice0 := src
    loprice0 := hiprice0

if bar_index == 1
    trend := 1
    hVal1 := hiprice1
    hVal0 := math.max(hiprice0, hVal1)
    lowVal1 := loprice1
    lowVal0 := math.min(loprice0, lowVal1)
    aFactor := startAFactor
    upTrndSAR := lowVal0
    dnTrndSAR := 0.
else
    hVal0 := hVal1
    lowVal0 := lowVal1
    trend := nz(trend[1])
    aFactor := nz(aFactor[1])
    inputs = 0.
    inprice = src
    if (adaptMode != 'Off')
        if (hiloMode == 'On') 
            inprice := src
        else 
            inprice := hiprice0
        if (adaptMode == 'Kaufman') 
            inputs := inprice
        else
            if (adaptMode == 'Ehlers') 
                if (nz(upTrndSAR[1]) != 0.)
                    inputs := math.abs(inprice - nz(upTrndSAR[1]))
                else
                    if (nz(dnTrndSAR[1]) != 0.) 
                        inputs := math.abs(inprice - nz(dnTrndSAR[1]))
        step := minStep + _afact(adaptMode, inputs, length, adaptSmth) * (maxStep - minStep)
    else 
        step := maxStep
        
    upTrndSAR := 0., dnTrndSAR := 0., upSig := 0., dnSig := 0.
    
    if (nz(trend[1]) > 0) 
        if (nz(trend[1]) == nz(trend[2]))
            aFactor := hVal1 > hVal2 ? nz(aFactor[1]) + step : aFactor
            aFactor := aFactor > maxAFactor ? maxAFactor : aFactor
            aFactor := hVal1 < hVal2 ? startAFactor : aFactor
        else 
            aFactor := nz(aFactor[1])
            
        upTrndSAR := nz(upTrndSAR[1]) + aFactor * (hVal1 - nz(upTrndSAR[1]))
        upTrndSAR := upTrndSAR > loprice1 ? loprice1 : upTrndSAR
        upTrndSAR := upTrndSAR > loprice2 ? loprice2 : upTrndSAR
    else
        if (nz(trend[1]) == nz(trend[2])) 
            aFactor := lowVal1 < lowVal2 ? nz(aFactor[1]) + step : aFactor
            aFactor := aFactor > maxAFactor ? maxAFactor : aFactor
            aFactor := lowVal1 > lowVal2 ? startAFactor : aFactor
        else
            aFactor := nz(aFactor[1])
            
        dnTrndSAR := nz(dnTrndSAR[1]) + aFactor * (lowVal1 - nz(dnTrndSAR[1]))
        dnTrndSAR := dnTrndSAR < hiprice1 ? hiprice1 : dnTrndSAR
        dnTrndSAR := dnTrndSAR < hiprice2 ? hiprice2 : dnTrndSAR
    
    hVal0 := hiprice0 > hVal0 ? hiprice0 : hVal0
    lowVal0 := loprice0 < lowVal0 ? loprice0 : lowVal0
        
    if (minChng > 0) 
        if (upTrndSAR - nz(upTrndSAR[1]) < minChng * calcBaseUnit() and upTrndSAR != 0. and nz(upTrndSAR[1]) != 0.)
            upTrndSAR := nz(upTrndSAR[1])
        if (nz(dnTrndSAR[1]) - dnTrndSAR < minChng * calcBaseUnit() and dnTrndSAR != 0. and nz(dnTrndSAR[1]) != 0.)
            dnTrndSAR := nz(dnTrndSAR[1])

    dnTrndSAR := trend < 0 and dnTrndSAR > nz(dnTrndSAR[1]) ? nz(dnTrndSAR[1]) : dnTrndSAR
    upTrndSAR := trend > 0 and upTrndSAR < nz(upTrndSAR[1]) ? nz(upTrndSAR[1]) : upTrndSAR
    
    if (trend < 0 and hiprice0 >= dnTrndSAR + filt * calcBaseUnit())
        trend := 1
        upTrndSAR := lowVal0
        upSig := SignalMode == 'Signals & Stops' ? lowVal0 : upSig
        dnTrndSAR := 0.
        aFactor := startAFactor
        lowVal0 := loprice0
        hVal0 := hiprice0
    else if (trend > 0 and loprice0 <= upTrndSAR - filt * calcBaseUnit())
        trend := -1
        dnTrndSAR := hVal0
        dnSig := SignalMode == 'Signals & Stops' ? hVal0 : dnSig
        upTrndSAR := 0.
        aFactor := startAFactor
        lowVal0 := loprice0
        hVal0 := hiprice0
    
psar = upTrndSAR > 0 ? upTrndSAR : dnTrndSAR
psar := isPsarAdaptive ? psar : ta.sar(psarStart, psarInc, psarMax) 
plot(psar, title='PSAR', color=src < psar ? rajah : magicMint, style=plot.style_circles)

// -- EMA --
float ema = ta.ema(src, emaLength)
plot(ema, title='EMA', color=languidLavender)

// -- Signals --
var string isTradeOpen = ''
var string signalCache = ''

bool enterLong = src > ema and ta.crossover(src, psar) and ta.crossover(osc, 0)
bool enterShort = src < ema and ta.crossunder(src, psar) and ta.crossunder(osc, 0)
// bool exitLong = ta.crossunder(src, ema)
// bool exitShort = ta.crossover(src, ema)

if (signalCache == 'long entry')
    signalCache := ''
    enterLong := true
else if (signalCache == 'short entry')
    signalCache := ''
    enterShort := true

if (isTradeOpen == '')
    if (enterLong)
        isTradeOpen := 'long'
    else if (enterShort)
        isTradeOpen := 'short'
else if (isTradeOpen == 'long')
    if (enterLong)
        enterLong := false
else if (isTradeOpen == 'short')
    if (enterShort)
        enterShort := false

plotshape((isSignalLabelEnabled and enterLong and (isTradeOpen == 'long')) ? psar : na, title='LONG', text='L', style=shape.labelup, color=mediumAquamarine, textcolor=color.white, size=size.tiny, location=location.absolute)
plotshape((isSignalLabelEnabled and enterShort and (isTradeOpen == 'short')) ? psar : na, title='SHORT', text='S', style=shape.labeldown, color=carrotOrange, textcolor=color.white, size=size.tiny, location=location.absolute)

// -- High Low Stop Loss and Take Profit --
bool isHighLowStopLossEnabled = true
bool isAutomaticHighLowTakeProfitEnabled = true
bool recalculateStopLossTakeProfit = false
bool isStrategyEntryEnabled = false
bool isLongEnabled = true
bool isShortEnabled = true
bool isStopLossTakeProfitRecalculationEnabled = true

bool longStopLossTakeProfitRecalculation = isStopLossTakeProfitRecalculationEnabled ? true : (lastTrade == 'short' or lastTrade == 'initial')
bool shortStopLossTakeProfitRecalculation = isStopLossTakeProfitRecalculationEnabled ? true : (lastTrade == 'long' or lastTrade == 'initial')

var float longHighLowStopLoss = 0
var float shortHighLowStopLoss = 0

float highLowStopLossLowest = ta.lowest(_low, highLowStopLossLookback)
float highLowStopLossHighest = ta.highest(_high, highLowStopLossLookback)

if (isHighLowStopLossEnabled)
    if (((enterLong and longStopLossTakeProfitRecalculation) or recalculateStopLossTakeProfit) and (isStrategyEntryEnabled ? not(strategy.position_size > 0) : true))
        if (highLowStopLossLowest == _low)
            longHighLowStopLoss := _high * highLowStopLossBackupMultiplier
        else if (highLowStopLossLowest > 0)
            longHighLowStopLoss := highLowStopLossLowest * highLowStopLossMultiplier
            
    if (((enterShort and shortStopLossTakeProfitRecalculation) or recalculateStopLossTakeProfit) and (isStrategyEntryEnabled ? not(strategy.position_size < 0) : true))
        if (highLowStopLossHighest == _high)
            shortHighLowStopLoss := _high * (1 + (1 - highLowStopLossBackupMultiplier))
        else if (highLowStopLossHighest > 0)
            shortHighLowStopLoss := highLowStopLossHighest * (1 + (1 - highLowStopLossMultiplier))
        
plot((isLongEnabled and isHighLowStopLossEnabled and (isTradeOpen == 'long')) ? longHighLowStopLoss : na, 'Long High Low Stop Loss', color=magicMint, style=plot.style_circles, trackprice=false)
plot((isShortEnabled and isHighLowStopLossEnabled and (isTradeOpen == 'short')) ? shortHighLowStopLoss : na, 'Short High Low Stop Loss ', color=rajah, style=plot.style_circles, trackprice=false)

// -- Automatic High Low Take Profit --
var float longAutomaticHighLowTakeProfit = na
var float shortAutomaticHighLowTakeProfit = na

if (isAutomaticHighLowTakeProfitEnabled)
    if (((enterLong and longStopLossTakeProfitRecalculation) or recalculateStopLossTakeProfit) and (isStrategyEntryEnabled ? not(strategy.position_size > 0) : true))
        longHighLowStopLossPercentage = 1 - (longHighLowStopLoss / _close)
        longAutomaticHighLowTakeProfit := _close * (1 + (longHighLowStopLossPercentage  * automaticHighLowTakeProfitRatio))
    if (((enterShort and shortStopLossTakeProfitRecalculation) or recalculateStopLossTakeProfit) and (isStrategyEntryEnabled ? not(strategy.position_size > 0) : true)) 
        shortHighLowStopLossPercentage = 1 - (_close / shortHighLowStopLoss)
        shortAutomaticHighLowTakeProfit := _close * (1 - (shortHighLowStopLossPercentage * automaticHighLowTakeProfitRatio))

plot((isAutomaticHighLowTakeProfitEnabled and isHighLowStopLossEnabled and (isTradeOpen == 'long')) ? longAutomaticHighLowTakeProfit : na, 'Long Automatic High Low Take Profit', color=magicMint, style=plot.style_circles, trackprice=false)
plot((isAutomaticHighLowTakeProfitEnabled and isHighLowStopLossEnabled and (isTradeOpen == 'short')) ? shortAutomaticHighLowTakeProfit : na, 'Short Automatic High Low Take Profit', color=rajah, style=plot.style_circles, trackprice=false)

// log.info('Automatic Long High Low Take Profit: ' + str.tostring(longAutomaticHighLowTakeProfit))
// log.info('Automatic Short High Low Take Profit: ' + str.tostring(shortAutomaticHighLowTakeProfit))

// log.info('Long High Low Stop Loss: ' + str.tostring(longHighLowStopLoss))
// log.info('Short High Low Stop Loss: ' + str.tostring(shortHighLowStopLoss))

bool longHighLowStopLossCondition = ta.crossunder(_close, longHighLowStopLoss)
bool shortHighLowStopLossCondition = ta.crossover(_close, shortHighLowStopLoss)

bool longAutomaticHighLowTakeProfitCondition = ta.crossover(_close, longAutomaticHighLowTakeProfit)
bool shortAutomaticHighLowTakeProfitCondition = ta.crossunder(_close, shortAutomaticHighLowTakeProfit)

bool exitLong = (longHighLowStopLossCondition or longAutomaticHighLowTakeProfitCondition) and strategy.position_size > 0
bool exitShort = (shortHighLowStopLossCondition or shortAutomaticHighLowTakeProfitCondition) and strategy.position_size < 0

plotshape((isSignalLabelEnabled and exitLong and (isTradeOpen == 'long')) ? psar : na, title='LONG EXIT', style=shape.circle, color=magicMint, size=size.tiny, location=location.absolute)
plotshape((isSignalLabelEnabled and exitShort and (isTradeOpen == 'short')) ? psar : na, title='SHORT EXIT', style=shape.circle, color=rajah, size=size.tiny, location=location.absolute)

// Long Exits
if (exitLong)
    strategy.close('long', comment=longAutomaticHighLowTakeProfitCondition ? 'EXIT_LONG_TP' : 'EXIT_LONG_SL')
    isTradeOpen := ''

// Short Exits
if (exitShort)
    strategy.close('short', comment=shortAutomaticHighLowTakeProfitCondition ? 'EXIT_SHORT_TP' : 'EXIT_SHORT_SL')
    isTradeOpen := ''

// Long Entries
if (enterLong and (strategy.position_size == 0))
    strategy.entry('long', strategy.long, comment='ENTER_LONG')

// Short Entries
if (enterShort and (strategy.position_size == 0))
    strategy.entry('short', strategy.short, comment='ENTER_SHORT')

// Save last trade state
if (enterLong or exitLong)
    lastTrade := 'long'
if (enterShort or exitShort)
    lastTrade := 'short'

barcolor(color=isTradeOpen == 'long' ? mediumAquamarine : isTradeOpen == 'short' ? carrotOrange : na)

Lebih lanjut