Skip to content

Latest commit

 

History

History
288 lines (206 loc) · 13 KB

三重叠加随机动量策略Triple-Overlapping-Stochastic-Momentum-Strategy.md

File metadata and controls

288 lines (206 loc) · 13 KB

Name

三重叠加随机动量策略Triple-Overlapping-Stochastic-Momentum-Strategy

Author

ChaoZhang

Strategy Description

IMG [trans]

概述

三重叠加随机动量策略是一种典型的短线交易策略。它通过计算三条不同参数设置的随机动量指标,并进行多重叠加,形成交易信号。当三条随机动量指标同时显示超买或超卖信号时,进行买入或卖出。该策略结合了多时间周期分析的优势,可以有效过滤市场噪音,提高信号质量。

策略原理

该策略的核心指标是随机动量指标(SMI)。SMI指标的计算公式如下:

SMI = 100 * EMA(EMA(收盘价-最高价和最低价的中点,N1),N2) / 0.5 * EMA(EMA(最高价-最低价,N1),N2)

其中,N1和N2分别是参数长度。SMI指标的取值范围在-100到100之间,当SMI高于0时表示收盘价位于当日价格范围的上半段,低于0时表示收盘价位于价格范围的下半段。

和传统stoch指标类似,SMI指标超过预设的超买线(如40)和超卖线(如-40)时,表明可能形成反转信号。当SMI指标上穿其移动平均线时,产生买入信号;当SMI指标下穿其移动平均线时,产生卖出信号。

该策略采用三组不同参数设置的SMI指标进行叠加,分别設定为:

  • SMI1: %K长度10周期,%K平滑周期3周期
  • SMI2: %K长度20周期,%K平滑周期3周期
  • SMI3: %K长度5周期,%K平滑周期3周期

当三条SMI指标同时显示超买或超卖时,发出交易信号。这可以有效过滤假Signals,提高信号质量。

策略优势

  • 多时间周期分析,综合判断,有效过滤噪声
  • SMI指标增强了stoch指标的易用性
  • 采用三重叠加,可靠性较单一指标高
  • 参数设置灵活,可调整
  • 适用于高频短线交易

策略风险

  • 多重指标叠加,存在一定滞后
  • 短线操作频繁,交易成本较高
  • 回测数据拟合风险
  • 市场结构变化后参数失效风险

风险缓解措施:

  • 优化参数,降低滞后
  • 适当调整持仓时间,降低交易成本
  • 增加统计检验,检验稳健性
  • 动态调整参数

策略优化

  • 测试不同的SMI参数组合
  • 增加统计指标,评估参数稳定性
  • 结合其他辅助指标,如成交量,布林带等
  • 根据市场环境动态切换参数
  • 优化止损策略

总结

三重叠加随机动量策略通过采用三组不同参数设置的SMI指标,在多个时间周期上进行综合判断,形成高质量的超买超卖交易信号。相比单一指标,该策略可过滤更多噪声,提升稳定性。下一步可通过参数优化、统计检验、辅助指标等方法进行改进,使策略更具鲁棒性。

||

Overview

The Triple Overlapping Stochastic Momentum strategy is a typical short-term trading strategy. It calculates three Stochastic Momentum Index (SMI) indicators with different parameter settings and generates trading signals when all three show overbought or oversold conditions simultaneously. By combining multi-timeframe analysis, this strategy can effectively filter market noise and improve signal quality.

Strategy Logic

The core indicator of this strategy is the Stochastic Momentum Index (SMI). The SMI is calculated as follows:

SMI = 100 * EMA(EMA(Close - Midpoint of High-Low Range, N1), N2) / 0.5 * EMA(EMA(High - Low, N1), N2)

Where N1 and N2 are the parameter lengths. The SMI oscillates between -100 and 100. Values above 0 indicate the close is in the upper half of the daily range, while values below 0 indicate the close is in the lower half.

Similar to the traditional Stochastic Oscillator, overbought (e.g. 40)/oversold (e.g. -40) levels indicate potential reversal signals. Bullish and bearish signals are generated when the SMI crosses above/below its moving average line.

The strategy employs three SMI indicators with different parameter sets, specifically:

  • SMI1: %K Period 10, %K Slowing Period 3
  • SMI2: %K Period 20, %K Slowing Period 3
  • SMI3: %K Period 5, %K Slowing Period 3

Trading signals are generated when all three SMIs concurrently show overbought or oversold conditions. This filters out false signals and improves reliability.

Advantages

  • Multi-timeframe analysis for robust signals
  • SMI enhances usability over traditional Stochastic
  • Triple overlay improves reliability over single indicator
  • Flexible parameters for optimization
  • Well suited for short-term/high frequency trading

Risks

  • Multiple indicators may lag signals
  • High trading frequency increases costs
  • Backtest overfitting
  • Parameters may fail with market regime changes

Risk Mitigations:

  • Optimize parameters to reduce lag
  • Adjust holding period to lower trading costs
  • Perform statistical testing to validate robustness
  • Dynamically adapt parameters

Enhancements

  • Test different SMI parameter combinations
  • Add statistical metrics to evaluate parameter stability
  • Incorporate supporting indicators like volume, Bollinger Bands etc.
  • Dynamic parameter switching based on environment
  • Optimize stop loss strategies

Conclusion

The Triple Overlapping Stochastic Momentum strategy combines robust signal generation across multiple timeframes by overlaying three SMI indicators with unique parameters. Compared to single oscillators, this multi-indicator approach filters more noise and improves consistency. Further refinements can be made going forward through parameter optimization, statistical validation, auxiliary indicators etc. to enhance strategy robustness.

[/trans]

Strategy Arguments

Argument Default Description
v_input_1 10 %K Length
v_input_2 3 %K Smoothing Length
v_input_3 3 %K Double Smoothing Length
v_input_4 10 Signal Length
v_input_5 ema Signal MA Type
v_input_6 40 Overbought Level
v_input_7 -40 Oversold Level
v_input_8 20 %K Length 2
v_input_9 3 %K Smoothing Length 2
v_input_10 3 %K Double Smoothing Length 2
v_input_11 10 Signal Length 2
v_input_12 ema Signal MA Type 2
v_input_13 40 Overbought Level 2
v_input_14 -40 Oversold Level 2
v_input_15 5 %K Length 3
v_input_16 3 %K Smoothing Length 3
v_input_17 3 %K Double Smoothing Length 3
v_input_18 10 Signal Length 3
v_input_19 ema Signal MA Type 3
v_input_20 40 Overbought Level 3
v_input_21 -40 Oversold Level 3
v_input_22 8 From Month
v_input_23 true From Day
v_input_24 2018 From Year
v_input_25 12 To Month
v_input_26 31 To Day
v_input_27 2018 To Year

Source (PineScript)

/*backtest
start: 2023-12-26 00:00:00
end: 2024-01-25 00:00:00
period: 1h
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

//@version=2
strategy("Stochastic Momentum multi strategy", "Stochastic Momentum Index multi strategy", overlay=false)

q = input(10, title="%K Length")
r = input(3, title="%K Smoothing Length")
s = input(3, title="%K Double Smoothing Length")
nsig = input(10, title="Signal Length")
matype = input("ema", title="Signal MA Type")  // possible: ema, sma, wma, trima, hma, dema, tema, zlema
overbought = input(40, title="Overbought Level", type=float)
oversold = input(-40, title="Oversold Level", type=float)

trima(src, length) => sma(sma(src,length),length)
hma(src, length) => wma(2*wma(src, length/2)-wma(src, length), round(sqrt(length)))
dema(src, length) => 2*ema(src,length) - ema(ema(src,length),length)
tema(src, length) => (3*ema(src,length) - 3*ema(ema(src,length),length)) + ema(ema(ema(src,length),length),length)
zlema(src, length) => ema(src,length) + (ema(src,length) - ema(ema(src,length),length))

smi = 100 * ema(ema(close-0.5*(highest(q)+lowest(q)),r),s) / (0.5 * ema(ema(highest(q)-lowest(q),r),s))
sig = matype=="ema" ? ema(smi,nsig) : matype=="sma" ? sma(smi,nsig) : matype=="wma" ? wma(smi,nsig) : matype=="trima" ? trima(smi,nsig) : matype=="hma" ? hma(smi,nsig) : matype=="dema" ? dema(smi,nsig) : matype=="tema" ? tema(smi,nsig) : matype=="zlema" ? zlema(smi,nsig) : ema(smi,nsig)

p_smi = plot(smi, title="SMI", color=aqua)
p_sig = plot(sig, title="Signal", color=red)

// plotchar(crossover(smi, sig), title= "low", location=location.bottom, color=green, char="▲", size= size.tiny)
// plotchar(crossunder(smi, sig), title= "high", location=location.top, color=red, char="▼", size= size.tiny)

/////////////////////////////2
q2 = input(20, title="%K Length 2")
r2 = input(3, title="%K Smoothing Length 2")
s2 = input(3, title="%K Double Smoothing Length 2")
nsig2 = input(10, title="Signal Length 2")
matype2 = input("ema", title="Signal MA Type 2")  // possible: ema, sma, wma, trima, hma, dema, tema, zlema
overbought2 = input(40, title="Overbought Level 2", type=float)
oversold2 = input(-40, title="Oversold Level 2", type=float)

trima2(src2, length2) => sma(sma(src2,length2),length2)
hma2(src2, length2) => wma(2*wma(src2, length2/2)-wma(src2, length2), round(sqrt(length2)))
dema2(src2, length2) => 2*ema(src2,length2) - ema(ema(src2,length2),length2)
tema2(src2, length2) => (3*ema(src2,length2) - 3*ema(ema(src2,length2),length2)) + ema(ema(ema(src2,length2),length2),length2)
zlema2(src2, length2) => ema(src2,length2) + (ema(src2,length2) - ema(ema(src2,length2),length2))

smi2 = 100 * ema(ema(close-0.5*(highest(q2)+lowest(q2)),r2),s2) / (0.5 * ema(ema(highest(q2)-lowest(q2),r2),s2))
sig2 = matype2=="ema" ? ema(smi2,nsig2) : matype2=="sma 2" ? sma(smi2,nsig2) : matype2=="wma 2" ? wma(smi2,nsig2) : matype2=="trima 2" ? trima2(smi2,nsig2) : matype2=="hma 2" ? hma2(smi2,nsig2) : matype=="dema 2" ? dema2(smi2,nsig2) : matype2=="tema 2" ? tema2(smi2,nsig2) : matype2=="zlema 2" ? zlema2(smi2,nsig2) : ema(smi2,nsig2)

p_smi2 = plot(smi2, title="SMI 2", color=aqua)
p_sig2 = plot(sig2, title="Signal2", color=red)

// plotchar(crossover(smi2, sig2), title= "low2", location=location.bottom, color=green, char="▲", size= size.tiny)
// plotchar(crossunder(smi2, sig2), title= "high2", location=location.top, color=red, char="▼", size= size.tiny)

/////////////////////////////3
q3 = input(5, title="%K Length 3")
r3 = input(3, title="%K Smoothing Length 3")
s3 = input(3, title="%K Double Smoothing Length 3")
nsig3 = input(10, title="Signal Length 3")
matype3 = input("ema", title="Signal MA Type 3")  // possible: ema, sma, wma, trima, hma, dema, tema, zlema
overbought3 = input(40, title="Overbought Level 3", type=float)
oversold3 = input(-40, title="Oversold Level 3", type=float)

trima3(src3, length3) => sma(sma(src3,length3),length3)
hma3(src3, length3) => wma(2*wma(src3, length3/2)-wma(src3, length3), round(sqrt(length3)))
dema3(src3, length3) => 2*ema(src3,length3) - ema(ema(src3,length3),length3)
tema3(src3, length3) => (3*ema(src3,length3) - 3*ema(ema(src3,length3),length3)) + ema(ema(ema(src3,length3),length3),length3)
zlema3(src3, length3) => ema(src3,length3) + (ema(src3,length3) - ema(ema(src3,length3),length3))

smi3 = 100 * ema(ema(close-0.5*(highest(q3)+lowest(q3)),r3),s3) / (0.5 * ema(ema(highest(q3)-lowest(q3),r3),s3))
sig3 = matype3=="ema" ? ema(smi3,nsig3) : matype3=="sma 3" ? sma(smi3,nsig3) : matype3=="wma 3" ? wma(smi3,nsig3) : matype3=="trima 3" ? trima3(smi3,nsig3) : matype3=="hma 3" ? hma3(smi3,nsig3) : matype=="dema 3" ? dema3(smi3,nsig3) : matype3=="tema 3" ? tema3(smi3,nsig3) : matype3=="zlema 3" ? zlema3(smi3,nsig3) : ema(smi3,nsig3)

p_smi3 = plot(smi3, title="SMI 3", color=aqua)
p_sig3 = plot(sig3, title="Signal3", color=red)

// plotchar(crossover(smi3, sig3) and crossover(smi2, sig2) and crossover(smi, sig), title= "low3", location=location.bottom, color=green, char="▲", size= size.tiny)
// plotchar(crossunder(smi3, sig3) and crossunder(smi2, sig2) and crossunder(smi, sig), title= "high3", location=location.top, color=red, char="▼", size= size.tiny)
plotchar (((smi3 < sig3) and (smi2 < sig2) and (smi < sig)), title= "low3", location=location.bottom, color=green, char="▲", size= size.tiny)
plotchar (((smi3 > sig3) and (smi2 > sig2) and (smi > sig)), title= "high3", location=location.top, color=red, char="▼", size= size.tiny)

// === BACKTEST RANGE ===
FromMonth = input(defval = 8, title = "From Month", minval = 1, maxval = 12)
FromDay   = input(defval = 1, title = "From Day", minval = 1, maxval = 31)
FromYear  = input(defval = 2018, title = "From Year", minval = 2014)
ToMonth   = input(defval = 12, title = "To Month", minval = 1, maxval = 12)
ToDay     = input(defval = 31, title = "To Day", minval = 1, maxval = 31)
ToYear    = input(defval = 2018, title = "To Year", minval = 2014)

longCondition = ((smi3 < sig3) and (smi2 < sig2) and (smi < sig))
shortCondition = ((smi3 > sig3) and (smi2 > sig2) and (smi > sig))

// buy = longCondition == 1 and longCondition[1] == 1 ? longCondition : na
buy = longCondition == 1 ? longCondition : na
sell = shortCondition == 1? shortCondition : na

// === ALERTS ===
strategy.entry("L", strategy.long, when=buy)

strategy.entry("S", strategy.short, when=sell)

alertcondition(((smi3 < sig3) and (smi2 < sig2) and (smi < sig)), title='Low Fib.', message='Low Fib. Buy')
alertcondition(((smi3 > sig3) and (smi2 > sig2) and (smi > sig)), title='High Fib.', message='High Fib. Low')

Detail

https://www.fmz.com/strategy/440065

Last Modified

2024-01-26 12:15:20