strategies/基于变化率的量化策略Rate-of-Change-Quantitative-Strategy.md

Name

基于变化率的量化策略Rate-of-Change-Quantitative-Strategy

Author

ChaoZhang

Strategy Description

[trans]

概述

这个策略基于变化率(ROC)指标来判断市场走势和产生交易信号。策略的核心思路是追随长期趋势,通过承担较大的风险来获得超越市场的收益。

策略原理

入场规则

• 如果ROC>0,做多;如果ROC<0,做空。利用ROC指标的正负来判断行情方向。
• 为了过滤震荡,只有当ROC保持在同一边连续两天时,才会发出交易信号。

止损规则

设置了6%的止损。当止损触发后,改变仓位方向。这表示我们可能正处在行情的错误一边,需要及时止损反向操作。

防泡沫机制

如果ROC超过200,则判定为泡沫。当ROC落回泡沫下方时,产生做空信号。同时,要求泡沫至少持续1周。

资金管理

使用固定仓位+递增法。每上涨或下跌400美元,增加或减少200美元的仓位。这样可以利用盈利进行加仓从而获得更大收益,但也增加了回撤。

优势分析

这是一个追踪长期趋势的策略。它的优势有:

1. 遵循趋势交易的哲学,容易获取长期正收益。
2. 利用止损来控制风险,可以减轻短期行情波动的影响。
3. 防泡沫机制可以避免在市场顶部追高。
4. 固定仓位+递增的资金管理方式使其在上涨行情中获得指数型增长。

风险分析

该策略也存在一些风险:

1. ROC指标容易受到震荡的影响,产生错误信号。可以考虑加入其他指标进行组合过滤。
2. 没有考虑交易费用,实际运用时收益会比回测低。
3. 防泡沫参数设置不当也容易错过行情。
4. 固定仓位+递增法增大了损失时的回撤。

优化方向

该策略可以从以下几个方面进行优化:

1. 增加其他指标判断,组成交易系统,以过滤错误信号。例如加入均线、波动率等指标。
2. 优化防泡沫参数,设置更精确的泡沫识别机制。
3. 调整固定仓位和递增参数,获得更好的风险收益平衡。
4. 添加自动止损机制。当发生大幅亏损时自动停损。
5. 考虑交易费用的影响,设置更实际的入场标准。

总结

总的来说,这是一个以ROC指标为核心的长线追踪策略。它通过承担较大的风险获得超越大盘的超额收益,是一个积极进取的策略。我们需要对其进行适当优化,使之能够在实际中运用。关键是要找到适合自己的风险偏好。

||

Overview

This strategy uses the Rate of Change (ROC) indicator to determine market direction and generate trading signals. The core idea of the strategy is to follow long-term trends and outperform the market by taking on greater risk.

Strategy Logic

Entry Rules

• Go long if ROC>0; go short if ROC<0. Use the positive/negative of ROC to judge market direction.
• To filter out volatility, only issue trading signals if ROC stays on the same side for two consecutive days.

Stop Loss

A 6% stop loss is set. When stop loss is triggered, reverse position. This indicates we may be on the wrong side of the market so we exit immediately.

Anti-Bubble Mechanism

If ROC goes above 200, market is considered a bubble. When ROC falls back below bubble territory, go short signal is triggered. Require the bubble persists for at least 1 week.

Money Management

Use fixed position sizing + incremental method. Increase/decrease position by $200 for every $400 gain/loss. This allows us to pyramid profits but also increases drawdown.

Advantage Analysis

Advantages of this strategy:

1. Adheres to trend following philosophy so likely to produce long term positive returns.
2. Use stop loss to control risk and reduce short term volatility.
3. Anti-bubble mechanism avoids chasing tops.
4. Fixed position + incremental method creates exponential growth in uptrends.

Risk Analysis

Some risks also exist:

1. ROC indicator prone to whipsaws which generates false signals. Consider combining with other indicators for filtering.
2. Trading costs not considered which lowers actual returns.
3. Poor anti-bubble parameter tuning also misses trends.
4. Incremental sizing increases drawdown when losing.

Optimization Directions

Some ways to optimize the strategy:

1. Add other indicators to filter signals, such as MA, Volatility etc.
2. Optimize anti-bubble parameters for better bubble detection.
3. Adjust fixed position and incremental ratios for better risk/reward balance.
4. Add automatic stop loss when large loss occurs.
5. Consider trading costs and set entry rules accordingly.

Conclusion

In summary, this is a long term trend following strategy centered around the ROC indicator. It aims to generate alpha by taking on higher risk. Further optimizations can improve its viability. The key is finding suitable risk tolerance.
[/trans]

Strategy Arguments

Argument	Default	Description
v_input_int_1	365	(?Technical parameters)ROC Length
v_input_int_2	200	ROC Bubble signal
v_input_float_1	10	(?Risk Management)Stop Loss (in %)
v_input_int_3	400	(?Money Management)Fixed Ratio Value ($)
v_input_int_4	200	Increasing Order Amount ($)
v_input_1	timestamp(1 Jan 2017 00:00:00)	(?Backtesting Period)Start Date
v_input_2	timestamp(1 July 2024 00:00:00)	End Date

Source (PineScript)

/*backtest
start: 2022-12-05 00:00:00
end: 2023-12-11 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/
// © gsanson66


//This strategy use the Rate of Change (ROC) of the closing price to send enter signal. 
//@version=5
strategy("RATE OF CHANGE BACKTESTING", shorttitle="ROC BACKTESTING", overlay=false, precision=3, initial_capital=1000, default_qty_type=strategy.cash, default_qty_value=950, commission_type=strategy.commission.percent, commission_value=0.18)


//--------------------------------FUNCTIONS-----------------------------------//

//@function Displays text passed to `txt` when called.
debugLabel(txt, color, loc) =>
    label.new(bar_index, loc, text = txt, color=color, style = label.style_label_lower_right, textcolor = color.black, size = size.small)

//@function which looks if the close date of the current bar falls inside the date range
inBacktestPeriod(start, end) => (time >= start) and (time <= end)


//----------------------------------USER INPUTS----------------------------------//

//Technical parameters
rocLength = input.int(defval=365, minval=0, title='ROC Length', group="Technical parameters")
bubbleValue = input.int(defval=200, minval=0, title="ROC Bubble signal", group="Technical parameters")
//Risk management
stopLossInput = input.float(defval=10, minval=0, title="Stop Loss (in %)", group="Risk Management")
//Money management
fixedRatio = input.int(defval=400, minval=1, title="Fixed Ratio Value ($)", group="Money Management")
increasingOrderAmount = input.int(defval=200, minval=1, title="Increasing Order Amount ($)", group="Money Management")
//Backtesting period
startDate = input(title="Start Date", defval=timestamp("1 Jan 2017 00:00:00"), group="Backtesting Period")
endDate = input(title="End Date", defval=timestamp("1 July 2024 00:00:00"), group="Backtesting Period")


//-------------------------------------VARIABLES INITIALISATION-----------------------------//

roc = (close/close[rocLength] - 1)*100
midlineConst = 0
var bool inBubble = na
bool shortBubbleCondition = na
equity = strategy.equity - strategy.openprofit
strategy.initial_capital = 50000
var float capital_ref = strategy.initial_capital
var float cashOrder = strategy.initial_capital * 0.95
bool inRange = na


//------------------------------CHECKING SOME CONDITIONS ON EACH SCRIPT EXECUTION-------------------------------//

//Checking if the date belong to the range
inRange := true

//Checking if we are in a bubble
if roc > bubbleValue and not inBubble
    inBubble := true

//Checking if the bubble is over
if roc < 0 and inBubble
    inBubble := false

//Checking the condition to short the bubble : The ROC must be above the bubblevalue for at least 1 week
if roc[1]>bubbleValue and roc[2]>bubbleValue and roc[3]>bubbleValue and roc[4]>bubbleValue and roc[5]>bubbleValue and roc[6]>bubbleValue and roc[7]>bubbleValue
    shortBubbleCondition := true

//Checking performances of the strategy
if equity > capital_ref + fixedRatio
    spread = (equity - capital_ref)/fixedRatio
    nb_level = int(spread)
    increasingOrder = nb_level * increasingOrderAmount
    cashOrder := cashOrder + increasingOrder
    capital_ref := capital_ref + nb_level*fixedRatio
if equity < capital_ref - fixedRatio
    spread = (capital_ref - equity)/fixedRatio
    nb_level = int(spread)
    decreasingOrder = nb_level * increasingOrderAmount
    cashOrder := cashOrder - decreasingOrder
    capital_ref := capital_ref - nb_level*fixedRatio

//Checking if we close all trades in case where we exit the backtesting period
if strategy.position_size!=0 and not inRange
    debugLabel("END OF BACKTESTING PERIOD : we close the trade", color=color.rgb(116, 116, 116), loc=roc)
    strategy.close_all()


//-------------------------------LONG/SHORT CONDITION-------------------------------//

//Long condition
//We reduce noise by taking signal only if the last roc value is in the same side as the current one
if (strategy.position_size<=0 and ta.crossover(roc, midlineConst)[1] and roc>0 and inRange)
    //If we were in a short position, we pass to a long position
    qty = cashOrder/close
    strategy.entry("Long", strategy.long, qty)
    stopLoss = close * (1-stopLossInput/100)
    strategy.exit("Long Risk Managment", "Long", stop=stopLoss)

//Short condition
//We take a short position if we are in a bubble and roc is decreasing
if (strategy.position_size>=0 and ta.crossunder(roc, midlineConst)[1] and roc<0 and inRange) or 
     (strategy.position_size>=0 and inBubble and ta.crossunder(roc, bubbleValue) and shortBubbleCondition and inRange)
    //If we were in a long position, we pass to a short position
    qty = cashOrder/close
    strategy.entry("Short", strategy.short, qty)
    stopLoss = close * (1+stopLossInput/100)
    strategy.exit("Short Risk Managment", "Short", stop=stopLoss)


//--------------------------------RISK MANAGEMENT--------------------------------------//

//We manage our risk and change the sense of position after SL is hitten
if strategy.position_size == 0 and inRange
    //We find the direction of the last trade
    id = strategy.closedtrades.entry_id(strategy.closedtrades-1)
    if id == "Short"
        qty = cashOrder/close
        strategy.entry("Long", strategy.long, qty)
        stopLoss = close * (1-stopLossInput/100)
        strategy.exit("Long Risk Managment", "Long", stop=stopLoss)
    else if id =="Long"
        qty = cashOrder/close
        strategy.entry("Short", strategy.short, qty)
        stopLoss = close * (1+stopLossInput/100)
        strategy.exit("Short Risk Managment", "Short", stop=stopLoss)


//---------------------------------PLOTTING ELEMENTS---------------------------------------//

//Plotting of ROC
rocPlot = plot(roc, "ROC", color=#7E57C2)
midline = hline(0, "ROC Middle Band", color=color.new(#787B86, 25))
midLinePlot = plot(0, color = na, editable = false, display = display.none)
fill(rocPlot, midLinePlot, 40, 0, top_color = strategy.position_size>0 ? color.new(color.green, 0) : strategy.position_size<0 ? color.new(color.red, 0) : na, bottom_color = strategy.position_size>0 ? color.new(color.green, 100) : strategy.position_size<0 ? color.new(color.red, 100) : na,  title = "Positive area")
fill(rocPlot, midLinePlot, 0,  -40,  top_color = strategy.position_size<0 ? color.new(color.red, 100) : strategy.position_size>0 ? color.new(color.green, 100) : na, bottom_color = strategy.position_size<0 ? color.new(color.red, 0) : strategy.position_size>0 ? color.new(color.green, 0) : na, title = "Negative area")

Detail

https://www.fmz.com/strategy/435134

Last Modified

2023-12-12 15:56:56