日本語 JA
GitHub で見る

Parabolic SAR 距離ブレイクアウト戦略

Parabolic SAR 距離ブレイクアウト戦略は、Parabolic の急速な拡張を観察します。読み値が最近の範囲を超えてジャンプすると、価格はしばしば新しい動きを開始します。

テストでは年平均リターン約118%を示しています。株式市場で最もよく機能します。

インジケーターが最近のデータと偏差乗数から導かれたバンドを突き抜けると、ポジションが開きます。ストップを付けたロングとショートの両取引が可能です。

このシステムは早期ブレイクアウトを求めるモメンタムトレーダーに適しています。Parabolic が平均に戻るとトレードはクローズされます。デフォルト値は Acceleration = 0.02m から始まります。

詳細

  • エントリー条件: インジケーターが偏差乗数分だけ平均を上回る。
  • ロング/ショート: 両方向。
  • エグジット条件: インジケーターが平均に戻る。
  • ストップ: はい。
  • デフォルト値:
    • Acceleration = 0.02m
    • MaxAcceleration = 0.2m
    • LookbackPeriod = 20
    • DeviationMultiplier = 2m
    • CandleType = TimeSpan.FromMinutes(5)
  • フィルター:
    • カテゴリ: ブレイクアウト
    • 方向: 両方
    • インジケーター: Parabolic
    • ストップ: はい
    • 複雑さ: 中級
    • 時間軸: 短期
    • 季節性: いいえ
    • ニューラルネットワーク: いいえ
    • ダイバージェンス: いいえ
    • リスクレベル: 中
using System;
using System.Linq;
using System.Collections.Generic;

using Ecng.Common;
using Ecng.Collections;
using Ecng.Serialization;

using StockSharp.Algo.Indicators;
using StockSharp.Algo.Strategies;
using StockSharp.BusinessEntities;
using StockSharp.Messages;

namespace StockSharp.Samples.Strategies;

/// <summary>
/// Strategy that enters positions when the distance between price and Parabolic SAR
/// exceeds the average distance plus a multiple of standard deviation
/// </summary>
public class ParabolicSarDistanceBreakoutStrategy : Strategy
{
	private readonly StrategyParam<decimal> _acceleration;
	private readonly StrategyParam<decimal> _maxAcceleration;
	private readonly StrategyParam<int> _lookbackPeriod;
	private readonly StrategyParam<decimal> _deviationMultiplier;
	private readonly StrategyParam<DataType> _candleType;

	private ParabolicSar _parabolicSar;
	
	private decimal _avgDistanceLong;
	private decimal _stdDevDistanceLong;
	private decimal _avgDistanceShort;
	private decimal _stdDevDistanceShort;
	
	private decimal _lastLongDistance;
	private decimal _lastShortDistance;
	private int _samplesCount;

	/// <summary>
	/// Initial acceleration factor for Parabolic SAR
	/// </summary>
	public decimal Acceleration
	{
		get => _acceleration.Value;
		set => _acceleration.Value = value;
	}

	/// <summary>
	/// Maximum acceleration factor for Parabolic SAR
	/// </summary>
	public decimal MaxAcceleration
	{
		get => _maxAcceleration.Value;
		set => _maxAcceleration.Value = value;
	}

	/// <summary>
	/// Lookback period for distance statistics calculation
	/// </summary>
	public int LookbackPeriod
	{
		get => _lookbackPeriod.Value;
		set => _lookbackPeriod.Value = value;
	}

	/// <summary>
	/// Standard deviation multiplier for breakout detection
	/// </summary>
	public decimal DeviationMultiplier
	{
		get => _deviationMultiplier.Value;
		set => _deviationMultiplier.Value = value;
	}

	/// <summary>
	/// Candle type
	/// </summary>
	public DataType CandleType
	{
		get => _candleType.Value;
		set => _candleType.Value = value;
	}

	/// <summary>
	/// Constructor
	/// </summary>
	public ParabolicSarDistanceBreakoutStrategy()
	{
		_acceleration = Param(nameof(Acceleration), 0.02m)
			.SetGreaterThanZero()
			.SetDisplay("Acceleration", "Initial acceleration factor for Parabolic SAR", "Indicator Parameters")
			
			.SetOptimize(0.01m, 0.05m, 0.01m);

		_maxAcceleration = Param(nameof(MaxAcceleration), 0.2m)
			.SetGreaterThanZero()
			.SetDisplay("Max Acceleration", "Maximum acceleration factor for Parabolic SAR", "Indicator Parameters")
			
			.SetOptimize(0.1m, 0.5m, 0.1m);

		_lookbackPeriod = Param(nameof(LookbackPeriod), 20)
			.SetGreaterThanZero()
			.SetDisplay("Lookback Period", "Period for statistical calculations", "Strategy Parameters")
			
			.SetOptimize(10, 50, 5);

		_deviationMultiplier = Param(nameof(DeviationMultiplier), 2m)
			.SetGreaterThanZero()
			.SetDisplay("Deviation Multiplier", "Standard deviation multiplier for breakout detection", "Strategy Parameters")
			
			.SetOptimize(1m, 3m, 0.5m);

		_candleType = Param(nameof(CandleType), TimeSpan.FromMinutes(5).TimeFrame())
			.SetDisplay("Candle Type", "Type of candles to use", "General");
	}

	/// <inheritdoc />
	public override IEnumerable<(Security sec, DataType dt)> GetWorkingSecurities()
	{
		return [(Security, CandleType)];
	}

	/// <inheritdoc />
	protected override void OnReseted()
	{
		base.OnReseted();
		_avgDistanceLong = 0;
		_stdDevDistanceLong = 0;
		_avgDistanceShort = 0;
		_stdDevDistanceShort = 0;
		_lastLongDistance = 0;
		_lastShortDistance = 0;
		_samplesCount = 0;
	}

	/// <inheritdoc />
	protected override void OnStarted2(DateTime time)
	{
		_parabolicSar = new ParabolicSar
		{
			Acceleration = Acceleration,
			AccelerationMax = MaxAcceleration
		};


		var subscription = SubscribeCandles(CandleType);
		subscription
			.Bind(_parabolicSar, ProcessCandle)
			.Start();

		// Setup chart visualization if available
		var area = CreateChartArea();
		if (area != null)
		{
			DrawCandles(area, subscription);
			DrawIndicator(area, _parabolicSar);
			DrawOwnTrades(area);
		}

		// Set up position protection using the dynamic Parabolic SAR
		StartProtection(
			takeProfit: null, // We'll handle exits via strategy logic
			stopLoss: null,   // The dynamic SAR will act as our stop
			isStopTrailing: true
		);

		base.OnStarted2(time);
	}

	private void ProcessCandle(ICandleMessage candle, decimal sarValue)
	{
		// Skip unfinished candles
		if (candle.State != CandleStates.Finished)
			return;

		// Check if strategy is ready for trading
		if (!IsFormedAndOnlineAndAllowTrading())
			return;

		// Calculate distances
		decimal longDistance = 0;
		decimal shortDistance = 0;
		
		// If SAR is below price, it's in uptrend
		if (sarValue < candle.ClosePrice)
			longDistance = candle.ClosePrice - sarValue;
		// If SAR is above price, it's in downtrend
		else if (sarValue > candle.ClosePrice)
			shortDistance = sarValue - candle.ClosePrice;
		
		// Update statistics
		UpdateDistanceStatistics(longDistance, shortDistance);
		
		// Trading logic
		if (_samplesCount >= LookbackPeriod)
		{
			// Long signal: distance exceeds average + k*stddev and we don't have a long position
			if (longDistance > 0 && 
				longDistance > _avgDistanceLong + DeviationMultiplier * _stdDevDistanceLong && 
				Position <= 0)
			{
				// Cancel existing orders
				CancelActiveOrders();
				
				// Enter long position
				var volume = Volume + Math.Abs(Position);
				BuyMarket(volume);
				
				LogInfo($"Long signal: Distance {longDistance} > Avg {_avgDistanceLong} + {DeviationMultiplier}*StdDev {_stdDevDistanceLong}");
			}
			// Short signal: distance exceeds average + k*stddev and we don't have a short position
			else if (shortDistance > 0 && 
					 shortDistance > _avgDistanceShort + DeviationMultiplier * _stdDevDistanceShort && 
					 Position >= 0)
			{
				// Cancel existing orders
				CancelActiveOrders();
				
				// Enter short position
				var volume = Volume + Math.Abs(Position);
				SellMarket(volume);
				
				LogInfo($"Short signal: Distance {shortDistance} > Avg {_avgDistanceShort} + {DeviationMultiplier}*StdDev {_stdDevDistanceShort}");
			}
			
			// Exit conditions - when price crosses SAR
			if (Position > 0 && candle.ClosePrice < sarValue)
			{
				// Exit long position
				SellMarket(Math.Abs(Position));
				LogInfo($"Exit long: Price {candle.ClosePrice} crossed below SAR {sarValue}");
			}
			else if (Position < 0 && candle.ClosePrice > sarValue)
			{
				// Exit short position
				BuyMarket(Math.Abs(Position));
				LogInfo($"Exit short: Price {candle.ClosePrice} crossed above SAR {sarValue}");
			}
		}
		
		// Store current distances for next update
		_lastLongDistance = longDistance;
		_lastShortDistance = shortDistance;
	}
	
	private void UpdateDistanceStatistics(decimal longDistance, decimal shortDistance)
	{
		_samplesCount++;
		
		// Simple calculation of running average and standard deviation
		if (_samplesCount == 1)
		{
			// Initialize with first values
			_avgDistanceLong = longDistance;
			_avgDistanceShort = shortDistance;
			_stdDevDistanceLong = 0;
			_stdDevDistanceShort = 0;
		}
		else
		{
			// Update running average
			decimal oldAvgLong = _avgDistanceLong;
			decimal oldAvgShort = _avgDistanceShort;
			
			_avgDistanceLong = oldAvgLong + (longDistance - oldAvgLong) / _samplesCount;
			_avgDistanceShort = oldAvgShort + (shortDistance - oldAvgShort) / _samplesCount;
			
			// Update running standard deviation using Welford's algorithm
			if (_samplesCount > 1)
			{
				_stdDevDistanceLong = (1 - 1.0m / (_samplesCount - 1)) * _stdDevDistanceLong + 
									   _samplesCount * ((_avgDistanceLong - oldAvgLong) * (_avgDistanceLong - oldAvgLong));
				
				_stdDevDistanceShort = (1 - 1.0m / (_samplesCount - 1)) * _stdDevDistanceShort + 
										_samplesCount * ((_avgDistanceShort - oldAvgShort) * (_avgDistanceShort - oldAvgShort));
			}
			
			// We only need last LookbackPeriod samples
			if (_samplesCount > LookbackPeriod)
			{
				_samplesCount = LookbackPeriod;
			}
		}
		
		// Calculate square root for final standard deviation
		_stdDevDistanceLong = (decimal)Math.Sqrt((double)_stdDevDistanceLong / _samplesCount);
		_stdDevDistanceShort = (decimal)Math.Sqrt((double)_stdDevDistanceShort / _samplesCount);
	}
}