gavindiaz/strategies
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
master
strategies/一目均衡云无偏移RSI过滤策略No-Offset-Ichimoku-Cloud-with-RSI-Filter-Strategy.md
T
root 87a415edf7 update
2024-03-03 22:27:55 +08:00

18 KiB
Raw Permalink Blame History

Name

一目均衡云无偏移RSI过滤策略No-Offset-Ichimoku-Cloud-with-RSI-Filter-Strategy

Author

ChaoZhang

Strategy Description

IMG [trans]

概述

这是一个利用一目均衡云指标进行趋势判断,并结合RSI指标进行过滤的趋势跟踪策略。该策略无视观测偏移,能够及时捕捉趋势变化,并通过RSI指标过滤假突破,以控制交易风险。

策略原理

该策略主要基于一目均衡云(Ichimoku Cloud)指标判断趋势方向。一目均衡云由哥转换线、基准线、导引线1、导引线2和延迟线组成。策略使用无偏移的一目均衡云,即转化线、基准线等线段采用未来数值,从而避免因观察偏移造成趋势判断滞后。

策略首先判断价格是否突破云线,如果延迟线上穿云线,则认为上升趋势启动;如果下穿云线,则认为下跌趋势启动。在趋势启动后,策略继续跟踪价格与云线的关系,来判断持续的趋势方向。当延迟线持续维持在云线之上时,认为处于上升趋势;反之,则认为处于下跌趋势。

除了趋势判断,策略还会在转换线和基准线发生黄金交叉时生成买入信号,死亡交叉时生成卖出信号。这些交易信号需要同趋势方向一致时才会被采纳。比如只有在上升趋势中,转换线上穿基准线的黄金交叉才会被采信。

最后,策略还引入RSI指标对信号进行过滤。只有当RSI低于超卖区时,才会采信买入信号;只有RSI高于超买区时,才会采信卖出信号。这在一定程度上可以过滤掉假突破带来的错误信号。

优势分析

  • 使用无偏移的一目均衡云指标,可以及时判断趋势变化,不会错过反转时机
  • 多种条件综合判断,可以有效过滤假突破带来的错误信号
  • RSI指标的引入可以避免超买超卖情况下的不理想入场
  • 策略以考虑未来数据为前提进行了优化,实盘中也可以获得较好的效果

风险分析

  • 考虑未来数据会导致报错,需要对代码进行优化才能在实盘使用
  • 一目均衡云指标对参数比较敏感,不同品种需要调整参数以达到最佳效果
  • 单一交易品种时效果会更好,多品种同时交易时需要考虑品种间相关性
  • 趋势判断规则较多,需要足够的回测周期验证效果

可以通过参数优化找到最佳参数组合。在实盘中可以考虑只交易特定品种或减少开仓手数以控制风险。也可以引入止损策略以限制单笔损失。

优化方向

该策略可以从以下几个方面进行进一步优化:

  1. 对一目均衡云参数进行优化,找到不同交易品种的最佳参数组合

  2. 增加止损策略,可以将单笔损失控制在可承受范围内

  3. 增加仓位管理策略,通过精准调整仓位规模来管理整体风险敞口

  4. 引入更多指标进行综合判断,如波动率指标、交易量等,以提高信号的准确性

  5. 优化入场时机选择,可采用突破确认或回调入场等方式

  6. 进行步移优化,根据不同品种特点选择最佳布林带周期参数

总结

该策略整体来说是一个较为稳健的趋势跟踪策略。它采用一目均衡云指标判断趋势方向,再结合转换线和基准线的交叉来发出交易信号,最后通过RSI指标过滤假突破。策略优化空间还很大,通过 parameter tuning、止损策略、仓位管理等优化可以获得更好的效果。但整体思路清晰易懂,既考虑了趋势判断又控制了风险,是一套值得实盘验证的策略思路。

||

Overview

This is a trend following strategy that utilizes the Ichimoku Cloud indicator for trend identification and the RSI indicator for signal filtering. The strategy uses the non-offset Ichimoku Cloud to capture trend changes in a timely manner and filters fake breakouts with RSI to control trading risks.

Strategy Logic

The strategy primarily relies on the Ichimoku Cloud indicator to determine the trend direction. The Ichimoku Cloud consists of the conversion line, base line, leading span 1, leading span 2, and lagging span. The strategy uses the non-offset Ichimoku Cloud, meaning the lines like conversion line and base line adopt future figures to avoid lagging trend identification due to observation offset.

The strategy first checks if the price breaks through the cloud lines. An upward trend is identified when the lagging span crosses above the cloud, while a downward trend is identified when the lagging span crosses below the cloud. After trend starts, the strategy keeps tracking the price's relationship with the cloud to determine the persistent trend direction. The uptrend remains intact when the lagging span is above the cloud, and vice versa.

In addition to trend identification, the strategy also generates buy and sell signals when the conversion line and base line have golden cross and death cross. These trading signals are only accepted when they align with the trend direction. For example, the golden cross between conversion line and base line is only considered during an uptrend.

Finally, the RSI indicator is used to filter the trading signals. Only buy signals with RSI below the oversold level and sell signals with RSI above the overbought level are accepted. This helps filter out false breakouts to some extent.

Advantage Analysis

  • The non-offset Ichimoku Cloud can timely identify trend changes without missing reversal opportunities

  • Multiple conditions work together to effectively filter out false breakout signals

  • RSI avoids undesirable market entry under overbought and oversold situations

  • The strategy is optimized with future data in mind and can also achieve good results in live trading

Risk Analysis

  • Using future data may cause errors and needs code optimization before live trading

  • The Ichimoku Cloud is sensitive to parameters and requires parameter tuning for different products

  • Better results when trading a single product. Needs to consider inter-market correlation with multiple products

  • Many trend identification rules require adequate backtesting periods for validation

Parameters can be optimized to find the best combination. Can consider trading specific products or reduce position sizes to control risks in live trading. Stop loss strategies may also be introduced to limit the loss per trade.

Optimization Directions

The strategy can be further optimized in the following aspects:

  1. Optimize Ichimoku Cloud parameters to find the best settings for different trading products

  2. Add stop loss strategies to limit the loss per trade to an acceptable level

  3. Introduce position sizing strategies to manage overall risk exposure precisely

  4. Add more technical indicators like volatility and volume for comprehensive signal verification

  5. Optimize entry techniques like confirmation or pullback entries for better execution

  6. Perform walk-forward optimization to determine the optimal Bollinger Bands lookback period based on product characteristics

Conclusion

In summary, this is quite a robust trend following strategy. It adopts the Ichimoku Cloud for trend identification and the conversion line/base line crosses for trade signals, filtered by RSI. There is still much room for optimization via parameter tuning, stop loss, position sizing etc. The logic is clear and easy to understand. It considers both trend and risks. This is a strategy worth verifying in live trading.

[/trans]

Strategy Arguments

Argument Default Description
v_input_1 false Standard Ichimoku Cloud
v_input_2 true Ichimoku Cloud - no offset - no repaint
v_input_3 9 Conversion Line Period - Tenkan-Sen
v_input_4 26 Base Line Period - Kijun-Sen
v_input_5 52 Base Line Period - Kijun-Sen (auxiliary)
v_input_6 52 Lagging Span 2 Periods - Senkou Span B
v_input_7 26 Displacement: (-) Chikou Span; (+) Senkou Span A
v_input_8 true Displacement: additional bars
v_input_9_close 0 Lagging Span price source - Chikou-Span: close
v_input_10 50 RSI Filter: Overbought level (0 = off)
v_input_11 100 RSI Filter: Oversold level (100 = off)
v_input_12 14 RSI Length
v_input_13_close 0 RSI Price source: close

Source (PineScript)

/*backtest
start: 2022-10-31 00:00:00
end: 2023-02-20 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/
// © KryptoNight

//@version=4
// comment/uncomment Study/Strategy to easily switch modes
// study("Ichimoku Kinko Hyo Cloud - no offset - no repaint - RSI filter - alerts", shorttitle="IchiCloud + RSI - alerts", overlay=true)
// ============================================================================== Strategy mode - uncomment to activate
strategy("Ichimoku Kinko Hyo Cloud - no offset - no repaint - RSI filter - strategy", shorttitle="IchiCloud + RSI - Strategy Tester Mode", overlay=true, pyramiding = 0,
  currency = currency.USD, initial_capital = 10000, default_qty_type = strategy.percent_of_equity, default_qty_value = 100,
  calc_on_every_tick = true, calc_on_order_fills = true, commission_type = strategy.commission.percent, commission_value = 0.075)
// ==============================================================================

// ------------------------------------------------------------------------------

ichiCloud_offset   = input(false, title="Standard Ichimoku Cloud")                  // with the visual offset
ichiCloud_noOffset = input(true,  title="Ichimoku Cloud - no offset - no repaint")  // without the visual offset

conversion_prd = input(9, minval=1, title="Conversion Line Period - Tenkan-Sen")
baseline_prd   = input(26, minval=1, title="Base Line Period - Kijun-Sen")
baselineA_prd  = input(52, minval=1, title="Base Line Period - Kijun-Sen (auxiliary)")
leadingSpan_2prd = input(52, minval=1, title="Lagging Span 2 Periods - Senkou Span B")
displacement = input(26, minval=0, title="Displacement: (-) Chikou Span; (+) Senkou Span A")
extra_bars = input(1, minval=0, title="Displacement: additional bars")
laggingSpan_src = input(close, title="Lagging Span price source - Chikou-Span")

donchian(len) => avg(lowest(len), highest(len))
displ = displacement-extra_bars
// ------------------------------------------------------------------------------
// OFFSET:
conversion = donchian(conversion_prd)   // Conversion Line - Tenkan-Sen (9 Period)
baseline  = donchian(baseline_prd)      // Base Line - Kijun-Sen (26 Period)
baselineA = donchian(baselineA_prd)     // Base Line Period - Kijun-Sen (auxiliary)
leadingSpanA = avg(conversion, baseline)
leadingSpanB = donchian(leadingSpan_2prd)
laggingSpan = laggingSpan_src

// Color - bullish, bearish
col_cloud = leadingSpanA>=leadingSpanB ? color.green : color.red

// Cloud Lines
spanA = plot(ichiCloud_offset? leadingSpanA : na, offset=displ, title="Offset: Lead Line 1 - Senkou Span A cloud", color=color.green)
spanB = plot(ichiCloud_offset? leadingSpanB : na, offset=displ, title="Offset: Lead Line 2 - Senkou Span B cloud", color=color.red)
fill(spanA, spanB, color=col_cloud, transp=80, title="Offset: Ichimoku Cloud - Leading Span 1 & 2 based coloring")

// Other Lines
conversion_p = plot(ichiCloud_offset? conversion : na, title="Offset: Conversion Line - Tenkan-Sen", color=#0496ff)
standard_p = plot(ichiCloud_offset? baseline : na, title="Offset: Base Line - Kijun-Sen", color=#991515)
standardA_p = plot(ichiCloud_offset? baselineA : na, title="Offset: Base Line - Kijun-Sen (auxiliary)", color=color.teal)
lagging_Span_p = plot(ichiCloud_offset? laggingSpan : na, offset=-displ, title="Offset: Chikou Span (Lagging Span)", color=#459915)

// ------------------------------------------------------------------------------
// NO OFFSET:
conversion_noOffset = conversion[displ] // Conversion Line - Tenkan-Sen (9 Period)
baseline_noOffset  = baseline[displ]    // Base Line - Kijun-Sen (26 Period)
baselineA_noOffset = baselineA[displ]   // Base Line Period - Kijun-Sen (auxiliary)
leadingSpanA_noOffset = leadingSpanA[displ*2]
leadingSpanB_noOffset = leadingSpanB[displ*2]
laggingSpan_noOffset = laggingSpan[0]

// Color - bullish, bearish
col_cloud_noOffset = leadingSpanA_noOffset>=leadingSpanB_noOffset ? color.green : color.red

// Cloud Lines
spanA_noOffset = plot(ichiCloud_noOffset? leadingSpanA_noOffset : na, title="No offset: Lead Line 1 - Senkou Span A cloud", color=color.green, transp=0)
spanB_noOffset = plot(ichiCloud_noOffset? leadingSpanB_noOffset : na, title="No offset: Lead Line 2 - Senkou Span B cloud", color=color.red, transp=0)
fill(spanA_noOffset, spanB_noOffset, color=col_cloud_noOffset, transp=80, title="No offset: Ichimoku Cloud - Leading Span 1 & 2 based coloring")

// Other Lines
conversion_p_noOffset = plot(ichiCloud_noOffset? conversion_noOffset : na, title="No offset: Conversion Line - Tenkan-Sen", color=#0496ff, transp=0)
baseline_p_noOffset = plot(ichiCloud_noOffset? baseline_noOffset : na, title="No offset: Base Line - Kijun-Sen", color=#991515, transp=0)
baselineA_p_noOffset = plot(ichiCloud_noOffset? baselineA_noOffset : na, title="No offset: Base Line - Kijun-Sen (auxiliary)", color=color.teal, transp=0)
laggingSpan_p_noOffset = plot(ichiCloud_noOffset? laggingSpan_noOffset : na, title="No offset: Chikou Span (Lagging Span)", color=#459915, transp=0)

// ==============================================================================
// Conditions & Alerts (based on the lines without offset)

maxC = max(leadingSpanA_noOffset,leadingSpanB_noOffset)
minC = min(leadingSpanA_noOffset,leadingSpanB_noOffset)

// Trend start signals: crosses between Chikou Span (Lagging Span) and the Cloud (Senkou Span A, Senkou Span B)
uptrend_start   = crossover(laggingSpan_noOffset,maxC)
downtrend_start = crossunder(laggingSpan_noOffset,minC)

// Trends
uptrend   = laggingSpan_noOffset>maxC // Above Cloud
downtrend = laggingSpan_noOffset<minC // Below Cloud

// No trend: choppy trading - the price is in transition
notrend = maxC>=laggingSpan_noOffset and laggingSpan_noOffset>=minC

// Confirmations
uptrend_confirm   = crossover(leadingSpanA_noOffset,leadingSpanB_noOffset)
downtrend_confirm = crossunder(leadingSpanA_noOffset,leadingSpanB_noOffset)

// Signals - crosses between Conversion Line (Tenkan-Sen) and Base Line (Kijun-Sen)
bullish_signal = crossover(conversion_noOffset,baseline_noOffset)
bearish_signal = crossunder(conversion_noOffset,baseline_noOffset)

// Various alerts
alertcondition(uptrend_start,   title="Uptrend Started",   message="Uptrend Started")
alertcondition(downtrend_start, title="Downtrend Started", message="Downtrend Started")

alertcondition(uptrend_confirm,   title="Uptrend Confirmed",   message="Uptrend Confirmed")
alertcondition(downtrend_confirm, title="Downtrend Confirmed", message="Downtrend Confirmed")

alertcondition(bullish_signal, title="Buy Signal",  message="Buy Signal")
alertcondition(bearish_signal, title="Sell Signal", message="Sell Signal")

rsi_OBlevel = input(50, title="RSI Filter: Overbought level (0 = off)")
rsi_OSlevel = input(100,title="RSI Filter: Oversold level (100 = off)")
rsi_len = input(14,title="RSI Length")
rsi_src = input(close,title="RSI Price source")
rsi = rsi(rsi_src,rsi_len)

// Strategy -------------------------------
long_signal  = bullish_signal and uptrend   and rsi<=rsi_OSlevel // breakout filtered by the rsi
exit_long    = bearish_signal and uptrend
short_signal = bearish_signal and downtrend and rsi>=rsi_OBlevel // breakout filtered by the rsi
exit_short   = bullish_signal and downtrend

// Strategy alerts
alertcondition(long_signal, title="Long Signal - Uptrend",      message="Long Signal - Uptrend")
alertcondition(exit_long,   title="Long Exit Signal - Uptrend", message="Long Exit Signal - Uptrend")

alertcondition(short_signal, title="Long Signal - Downtrend",       message="Long Signal - Downtrend")
alertcondition(exit_short,   title="Short Exit Signal - Downtrend", message="Short Exit Signal - Downtrend")

// Plot areas for trend and transition
color_trend = uptrend? #00FF00 : downtrend? #FF0000 : notrend? color.new(#FFFFFF, 50) : na
fill(spanA_noOffset, spanB_noOffset, color=color_trend, transp=90, title="No offset: Ichimoku Cloud - Lagging Span & Cloud based coloring")

plotshape(ichiCloud_noOffset?uptrend_start:na, title="No offset: Uptrend Started", color=color.green, style=shape.circle, location=location.belowbar, size=size.tiny, text="Up")
plotshape(ichiCloud_noOffset?downtrend_start:na, title="No offset: Downtrend Started", color=color.red, style=shape.circle,location=location.abovebar, size=size.tiny, text="Down")

plotshape(ichiCloud_noOffset?uptrend_confirm:na, title="No offset: Uptrend Confirmed", color=color.green, style=shape.circle, location=location.belowbar, size=size.small, text="Confirm Up")
plotshape(ichiCloud_noOffset?downtrend_confirm:na, title="No offset: Downtrend Confirmed", color=color.red, style=shape.circle, location=location.abovebar, size=size.small, text="Confirm Down")

plotshape(ichiCloud_noOffset?long_signal:na, title="No offset: Long Signal", color=#00FF00, style=shape.triangleup, location=location.belowbar, size=size.small, text="Long")
plotshape(ichiCloud_noOffset?exit_long:na, title="No offset: Exit Long Signal", color=color.fuchsia, style=shape.triangledown, location=location.abovebar, size=size.small, text="Exit long")

plotshape(ichiCloud_noOffset?short_signal:na, title="No offset: Short Signal", color=#FF0000, style=shape.triangledown, location=location.abovebar, size=size.small, text="Short")
plotshape(ichiCloud_noOffset?exit_short:na, title="No offset: Exit Short Signal", color=color.fuchsia, style=shape.triangleup, location=location.belowbar, size=size.small, text="Exit short")

// ============================================================================== Strategy Component - uncomment to activate
if (long_signal)
    strategy.entry("Long",strategy.long)
if (exit_long)
    strategy.close("Long")
if (short_signal)
    strategy.entry("Short",strategy.short)
if (exit_short)
    strategy.close("Short")
// ==============================================================================

Detail

https://www.fmz.com/strategy/431397

Last Modified

2023-11-07 15:31:06

Reference in New Issue View Git Blame Copy Permalink