Estratégia progressiva de lucro

Autora:ChaoZhangData: 24 de outubro de 2023 14:00
Tags:

Estratégia progressiva de lucro

Resumo

Esta estratégia combina o indicador RSI e a média móvel de preços para identificar oportunidades de sobrevenda quando o preço cai abaixo da linha média móvel. À medida que o preço continua a diminuir, a estratégia irá progressivamente piramidizar posições mais longas com base em porcentagens pré-definidas para alcançar a média de custos. Quando o lucro das posições atinge a porcentagem de lucro configurada, a estratégia fechará as posições. Também introduz um mecanismo progressivo de lucro que ajusta dinamicamente o preço de stop-profit global com base nos lucros realizados por posição. Isso pode efetivamente reduzir o risco de perdas e alcançar uma saída gradual.

Estratégia lógica

Quando o RSI cair abaixo da linha de sobrevenda de 29 e o preço de fechamento estiver abaixo da média móvel, abra a primeira posição longa.
Quando o preço cai 2% abaixo do primeiro preço de entrada, adicione uma segunda posição longa, e assim por diante até um máximo de 8 entradas.
Após cada entrada, registre o preço de entrada. Estes preços servem como preços de referência para as entradas.
Após as entradas, calcule o preço médio de detenção, utilizando 3% do preço médio como lucro para cada posição e 4% para a posição global.
Quando o preço sobe acima do preço de lucro de uma posição, feche essa posição.
Calculo progressivo de lucro: após o fechamento de cada posição, deduzir o lucro realizado do preço total de lucro. Isso lentamente arrasta a linha de lucro. Somente quando o lucro total cobre a perda máxima, a estratégia obterá lucro completamente.
Quando o preço atingir a linha de lucro, feche todas as posições.

Vantagens

O RSI é bom em identificar zonas de sobrevenda/supercompra, permitindo boas entradas para reversões.
As entradas múltiplas permitem a média de custos a preços baixos.
O progressive take profit reduz o risco e obtém saídas graduais.
O índice de lucro e as etapas de entrada personalizáveis permitem o ajustamento do risco.
As linhas de entrada e de lucro planeadas oferecem orientação visual sobre as posições.

Riscos

Os mercados de Whipsaw podem desencadear entradas e saídas excessivas, causando deslizamento.
A má configuração das etapas de entrada e proporções pode causar excesso de negociação.
Continuar a pirâmide durante os declínios traz riscos ilimitados de perdas.
Otimizar com base em dados de backtest.

Melhorias

Adicione filtros como o MACD para evitar maus sinais do RSI.
Incorporar stop loss baseado no ATR para limitar eventos de perdas extremas.
Otimizar a entrada, tirar lucro e outros parâmetros para diferentes ativos.
Ajuste dinâmico de lucro com base na volatilidade.

Conclusão

A estratégia utiliza totalmente o RSI para identificar a sobrevenda, combinando com o MA para negociação de reversão. Os mecanismos de pirâmide e progressiva de lucro controlam o risco, permitindo entradas longas eficazes.

/*backtest
start: 2023-09-23 00:00:00
end: 2023-10-23 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/
//@version=5
// © A3Sh

// RSI Strategy that buys the dips, uses Price Averaging and Pyramiding.
// When the price drops below specified percentages of the price (8 PA layers), new entries are openend to average the price of the assets.
// Open entries are closed by a specified take profit.
// Entries can be reopened, after closing and consequently crossing a PA layer again.
// This strategy is based on the RSI+PA+DCA strategy I created earlier. The difference is the way the Take Profit is calculated.
// Instead of directly connecting the take profit limit to the decreasing average price level with an X percent above the average price, 
// the take profit is calculated for a part on the decreasing average price and for another part on the deduction 
// of the profits of the individual closed positions.
// The Take Profit Limit drop less significant then the average price level and the full position only completely exits 
// when enough individual closed positions made up for the losses.
// This makes it less risky and more conservative and great for a long term trading strategy
// RSI code is adapted from the build in Relative Strength Index indicator
// MA Filter and RSI concept adapted from the Optimized RSI Buy the Dips strategy, by Coinrule
// https://www.tradingview.com/script/Pm1WAtyI-Optimized-RSI-Strategy-Buy-The-Dips-by-Coinrule/
// Pyramiding entries code adapted from Pyramiding Entries on Early Trends startegy, by Coinrule
// Pyramiding entries code adapted from Pyramiding Entries on Early Trends startegy, by Coinrule
// https://www.tradingview.com/script/7NNJ0sXB-Pyramiding-Entries-On-Early-Trends-by-Coinrule/
// Plot entry layers code adapted from HOWTO Plot Entry Price by vitvlkv
// https://www.tradingview.com/script/bHTnipgY-HOWTO-Plot-Entry-Price/


strategy(title='RSI+PA+PTP', pyramiding=16, overlay=true, initial_capital=400, default_qty_type=strategy.percent_of_equity, default_qty_value=15, commission_type=strategy.commission.percent, commission_value=0.075, close_entries_rule='FIFO')

port = input.float(12, group = "Risk", title='Portfolio % Used To Open The 8 Positions', step=0.1, minval=0.1, maxval=100)
q    = strategy.equity / 100 * port / open


// Long position PA entry layers. Percentage from the entry price of the the first long
ps2 = input.float(2,  group = "Long Position Entry Layers", title='2nd Long Entry %', step=0.1)
ps3 = input.float(3,  group = "Long Position Entry Layers", title='3rd Long Entry %', step=0.1)
ps4 = input.float(5,  group = "Long Position Entry Layers", title='4th Long Entry %', step=0.1)
ps5 = input.float(10, group = "Long Position Entry Layers", title='5th Long Entry %', step=0.1)
ps6 = input.float(16, group = "Long Position Entry Layers", title='6th Long Entry %', step=0.1)
ps7 = input.float(25, group = "Long Position Entry Layers" ,title='7th Long Entry %', step=0.1)
ps8 = input.float(40, group = "Long Position Entry Layers", title='8th Long Entry %', step=0.1)


// Calculate Moving Averages
plotMA               = input.bool(group = "Moving Average Filter", title='Plot Moving Average', defval=false)
movingaverage_signal = ta.sma(close, input(100, group = "Moving Average Filter", title='MA Length'))

plot (plotMA ? movingaverage_signal : na, color = color.new (color.green, 0))


// RSI inputs and calculations
rsiLengthInput = input.int(14, minval=1, title="RSI Length", group="RSI Settings")
rsiSourceInput = input.source(close, "Source", group="RSI Settings")

up   = ta.rma(math.max(ta.change(rsiSourceInput), 0), rsiLengthInput)
down = ta.rma(-math.min(ta.change(rsiSourceInput), 0), rsiLengthInput)
rsi  = down == 0 ? 100 : up == 0 ? 0 : 100 - (100 / (1 + up / down))

overSold = input.int(29, title="Oversold, Trigger to Enter First Position", group = "RSI Settings")

// Long trigger (co)
co = ta.crossover(rsi, overSold) and close < movingaverage_signal


// Store values to create and plot the different PA layers
long1 = ta.valuewhen(co, close, 0)
long2 = ta.valuewhen(co, close - close / 100 * ps2, 0)
long3 = ta.valuewhen(co, close - close / 100 * ps3, 0)
long4 = ta.valuewhen(co, close - close / 100 * ps4, 0)
long5 = ta.valuewhen(co, close - close / 100 * ps5, 0)
long6 = ta.valuewhen(co, close - close / 100 * ps6, 0)
long7 = ta.valuewhen(co, close - close / 100 * ps7, 0)
long8 = ta.valuewhen(co, close - close / 100 * ps8, 0)

eps1 = 0.00
eps1 := na(eps1[1]) ? na : eps1[1]

eps2 = 0.00
eps2 := na(eps2[1]) ? na : eps2[1]

eps3 = 0.00
eps3 := na(eps3[1]) ? na : eps3[1]

eps4 = 0.00
eps4 := na(eps4[1]) ? na : eps4[1]

eps5 = 0.00
eps5 := na(eps5[1]) ? na : eps5[1]

eps6 = 0.00
eps6 := na(eps6[1]) ? na : eps6[1]

eps7 = 0.00
eps7 := na(eps7[1]) ? na : eps7[1]

eps8 = 0.00
eps8 := na(eps8[1]) ? na : eps8[1]

plot(strategy.position_size > 0 ? eps1 : na, title='Long entry 1', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps2 : na, title='Long entry 2', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps3 : na, title='Long entry 3', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps4 : na, title='Long entry 4', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps5 : na, title='Long entry 5', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps6 : na, title='Long entry 6', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps7 : na, title='Long entry 7', style=plot.style_linebr)
plot(strategy.position_size > 0 ? eps8 : na, title='Long entry 8', style=plot.style_linebr)


// Take Profit Settings
ProfitTarget_Percent     = input.float(3.0,   group = "Take Profit Settings", title='Take Profit % (Per Position)')
ProfitTarget_Percent_All = input.float(4.0,   group = "Take Profit Settings", title='Take Profit % (Exit All, Progressive Take Profit Limit')
TakeProfitProgression    = input.float(12,    group = "Take Profit Settings", title='Take Profit Progression', tooltip = 'Progression is defined by the position size. By default 12% of the start equity (portfolio) is used to open a position, see Risk. This same % percentage is used to calculate the profit amount that will be deducted from the Take Profit Limit.')
entryOn                  = input.bool (true,  group = "Take Profit Settings", title='New entries affect Take Profit limit', tooltip = 'This option changes the behaviour of the Progressive Take Profit. When switchted on, the difference between the former and current original Take Profit is deducted from the Progressive Take Profit. When switchted off, the Progressive Take Profit is only affected by the profit deduction or each closed position.')
avPricePlot              = input.bool (false, group = "Take Profit Settings", title='Plot Average Price (FIFO)')
// Original Take Profit Limit
tpLimit                  = strategy.position_avg_price + (strategy.position_avg_price / 100 * ProfitTarget_Percent_All) 


// Create variables to calculate the Take Profit Limit Progresssion
var endVal   = 0.0   
var startVal = 0.0

// The value at the the start of the loop is the value of the end of the previous loop
startVal := endVal 

// Set variable to the original Take Profit Limit when the first position opens.
if strategy.position_size > 0 and strategy.position_size[1] ==0
    endVal := tpLimit  

// Everytime a specific position opens, the difference of the previous (original) Take Profit price and the current (original) Take Profit price will be deducted from the Progressive Take Profit Limit
// This feature can be toggled on and off in the settings panel. By default it is toggled on.
entryAmount = 0.0
for i = 1 to strategy.opentrades
    entryAmount := i
    if entryOn  and strategy.position_size > 0 and strategy.opentrades[1] == (entryAmount) and strategy.opentrades == (entryAmount + 1)
        endVal := startVal - (tpLimit[1] - tpLimit)

// Everytime a specific position closes, the amount of profit from that specific position will be deducted from the Progressive Take Profit Limit.
exitAmount = 0.0
for id = 1 to strategy.opentrades
    exitAmount := id
    if strategy.opentrades[1] ==(exitAmount + 1) and strategy.opentrades == (exitAmount)
        endVal := startVal - (TakeProfitProgression / 100 * strategy.opentrades.entry_price (id - 1) / 100 * ProfitTarget_Percent )

// The Final Take Profit Price
tpn = (strategy.position_avg_price + (strategy.position_avg_price / 100 * ProfitTarget_Percent_All))  - (strategy.position_avg_price + (strategy.position_avg_price / 100 * ProfitTarget_Percent_All) - endVal)
plot  (strategy.position_size > 0 ? tpn : na, title = "Take Profit Limit", color=color.new(color.red, 0), style = plot.style_linebr, linewidth = 1) 

// Plot position average price as reference
plot  (avPricePlot ? strategy.position_avg_price : na, title= "Average price", color = color.new(color.white, 0), style = plot.style_linebr, linewidth = 1) 


// When to trigger the Take Profit per position or the Progressive Take Profit
tpl1 = close < tpn ? eps1 + close * (ProfitTarget_Percent / 100) : tpn
tpl2 = close < tpn ? eps2 + close * (ProfitTarget_Percent / 100) : tpn
tpl3 = close < tpn ? eps3 + close * (ProfitTarget_Percent / 100) : tpn
tpl4 = close < tpn ? eps4 + close * (ProfitTarget_Percent / 100) : tpn
tpl5 = close < tpn ? eps5 + close * (ProfitTarget_Percent / 100) : tpn
tpl6 = close < tpn ? eps6 + close * (ProfitTarget_Percent / 100) : tpn
tpl7 = close < tpn ? eps7 + close * (ProfitTarget_Percent / 100) : tpn
tpl8 = close < tpn ? eps8 + close * (ProfitTarget_Percent / 100) : tpn



// Submit Entry Orders
if co and strategy.opentrades == 0
    eps1 := long1
    eps2 := long2
    eps3 := long3
    eps4 := long4
    eps5 := long5
    eps6 := long6
    eps7 := long7
    eps8 := long8

    strategy.entry('Long1', strategy.long, q)

if strategy.opentrades == 1
    strategy.entry('Long2', strategy.long, q, limit=eps2)

if strategy.opentrades == 2
    strategy.entry('Long3', strategy.long, q, limit=eps3)

if strategy.opentrades == 3
    strategy.entry('Long4', strategy.long, q, limit=eps4)

if strategy.opentrades == 4
    strategy.entry('Long5', strategy.long, q, limit=eps5)

if strategy.opentrades == 5
    strategy.entry('Long6', strategy.long, q, limit=eps6)

if strategy.opentrades == 6
    strategy.entry('Long7', strategy.long, q, limit=eps7)

if strategy.opentrades == 7
    strategy.entry('Long8', strategy.long, q, limit=eps8)



// Submit Exit orders
if strategy.position_size > 0
    strategy.exit(id='Exit 1', from_entry='Long1', limit=tpl1)
    strategy.exit(id='Exit 2', from_entry='Long2', limit=tpl2)
    strategy.exit(id='Exit 3', from_entry='Long3', limit=tpl3)
    strategy.exit(id='Exit 4', from_entry='Long4', limit=tpl4)
    strategy.exit(id='Exit 5', from_entry='Long5', limit=tpl5)
    strategy.exit(id='Exit 6', from_entry='Long6', limit=tpl6)
    strategy.exit(id='Exit 7', from_entry='Long7', limit=tpl7)
    strategy.exit(id='Exit 8', from_entry='Long8', limit=tpl8)


// Make sure that all open limit orders are canceled after exiting all the positions 
longClose = strategy.position_size[1] > 0 and strategy.position_size == 0 ? 1 : 0
if longClose
    strategy.cancel_all()

Mais.