Skip to main content

MultiMA_TSL Strategy Analysis

Strategy Number: #20 (20th of 465 strategies)
Strategy Type: Multi Moving Average + Custom Stoploss Trend Following
Timeframe: 5 minutes (5m)

I. Strategy Overview​

MultiMA_TSL is a trend-following strategy combining multiple moving averages with trailing stoploss. The strategy's core feature is using three different moving average types (EMA, ZEMA, TRIMA) with an offset mechanism to capture deep pullbacks, while introducing a custom stoploss function for quick exit when sell signals trigger, achieving the effect of "riding green candles".

Core Features​

FeatureDescription
Entry ConditionsMulti-condition combination (EMA/ZEMA/TRIMA offset + EWO + RSI)
Exit ConditionsOffset sell + custom stoploss coordination
ProtectionCustom stoploss + CooldownPeriod + LowProfitPairs
Timeframe5 minutes
DependenciesTA-Lib, technical
Special FeaturesCustom stoploss function, multiple MA types optional, EWO wave oscillator

II. Strategy Configuration Analysis​

2.1 Base Risk Parameters​

# ROI exit table
minimal_roi = {
    "0": 0.10,    # Immediate exit: 10% profit
    "30": 0.05,   # After 30 minutes: 5% profit
    "60": 0.02,   # After 60 minutes: 2% profit
}

# Stoploss setting
stoploss = -0.15  # -15% hard stoploss

# Trailing stop
trailing_stop = True
trailing_stop_positive = 0.01        # 1% trailing activation
trailing_stop_positive_offset = 0.018  # 1.8% offset trigger

Design Logic:

  • Tiered ROI: 10% → 5% → 2%, longer holding time means lower exit threshold
  • Medium hard stoploss: -15% hard stoploss, gives some room for fluctuation
  • Trailing stop: Activates 1% trailing after 1.8% profit, locks gains

2.2 Protection Mechanism Configuration​

@property
def protections(self):
    return [
        {
            "method": "CooldownPeriod",
            "stop_duration_candles": 2
        },
        {
            "method": "LowProfitPairs",
            "lookback_period_candles": 60,
            "trade_limit": 1,
            "stop_duration": 20,
            "required_profit": -0.05,
        },
    ]

Protection Mechanism Analysis:

  • CooldownPeriod: 2-candle cooldown after entry, prevents frequent trading
  • LowProfitPairs: Pause trading for 20 candles if pair loses over 5% within 60 candles

III. Entry Conditions Details​

3.1 Multi Moving Average Offset Mechanism​

Strategy uses three moving average types with offset mechanism to capture deep pullbacks:

# Buy offset calculation
ema_offset_buy = ta.EMA(close, timeperiod=50) * 0.958    # EMA offset
zema_offset_buy = zema(30) * 0.963                        # ZEMA offset (zero lag)
trima_offset_buy = ta.TRIMA(close, timeperiod=14) * 0.958 # TRIMA offset

Offset Mechanism Analysis:

  • EMA offset: EMA50 × 0.958, price must be below 95.8% of EMA
  • ZEMA offset: ZEMA30 × 0.963, uses zero-lag EMA
  • TRIMA offset: TRIMA14 × 0.958, uses triangular moving average

3.2 EWO Wave Oscillator Filter​

# EWO (Elliott Wave Oscillator) calculation
ewo = EMA(close, 5) - EMA(close, 34)

# EWO filter in entry conditions
(EWO < -20) | ((EWO > 6) & (RSI < 50))

EWO Filter Logic:

  • Deep pullback: EWO < -20 (wave in downward phase)
  • Trend pullback: EWO > 6 and RSI < 50 (pullback opportunity in uptrend)

3.3 Entry Condition Classification​

Condition ModeCore LogicApplicable Scenario
TRIMA modeclose < TRIMA14×0.958 + EWO/RSI filterSmooth trend pullback
ZEMA modeclose < ZEMA30×0.963 + EWO/RSI filterFast trend pullback
EMA modeclose < EMA50×0.958 + EWO/RSI filterStandard trend pullback

IV. Exit Logic Explained​

4.1 Offset Sell Mechanism​

# Sell offset calculation
ema_offset_sell = ta.EMA(close, timeperiod=32) * 1.002   # EMA offset sell
trima_offset_sell = ta.TRIMA(close, timeperiod=48) * 1.085  # TRIMA offset sell

# Sell conditions
(close > ema_offset_sell) | (close > trima_offset_sell)

Sell Logic Analysis:

  • EMA sell: Price > EMA32 × 1.002 (price slightly above EMA)
  • TRIMA sell: Price > TRIMA48 × 1.085 (price 8.5% above TRIMA)
  • Either triggers: Sell signal issued when either condition is met

4.2 Custom Stoploss Function​

def custom_stoploss(self, pair, trade, current_time, current_rate, current_profit, **kwargs) -> float:
    sl_new = 1  # Default: don't trigger stoploss
    if not self.config["runmode"].value in ("backtest", "hyperopt"):
        dataframe, _ = self.dp.get_analyzed_dataframe(pair, self.timeframe)
        last_candle = dataframe.iloc[-1]
        if (last_candle["sell_copy"] == 1) & (last_candle["buy_copy"] == 0):
            sl_new = 0.001  # Almost immediate stoploss
    return sl_new

Working Mechanism:

  • Live mode: Only effective in live/dry-run mode
  • Sell signal trigger: When sell signal triggers and no buy signal
  • Quick exit: Sets stoploss to 0.001, achieves "riding green candles"

4.3 Sell Signal Summary​

Signal TypeTrigger ConditionFunction
Offset sellPrice > MA×offsetBase sell signal
ROI exitProfit reaches ROI table thresholdTiered take-profit
Trailing stopProfit pullback exceeds 1%Lock gains
Custom stoplossSell signal triggersQuick exit

V. Technical Indicator System​

5.1 Core Indicators​

Indicator CategorySpecific IndicatorParametersPurpose
TrendEMA50, 32 periodsBuy/sell offset baseline
TrendZEMA30 periodsZero-lag EMA buy offset
TrendTRIMA14, 48 periodsTriangular MA buy/sell offset
MomentumEWO5/34 period differenceWave oscillator filter
MomentumRSI14 periodsOverbought/oversold confirmation
MomentumRSI Fast4 periodsFast RSI signal

5.2 Moving Average Type Comparison​

MA TypeCharacteristicsAdvantagesDisadvantages
EMAExponential moving averageFaster responseHas lag
ZEMAZero-lag EMAAlmost no lagMay have noise
TRIMATriangular moving averageHigh smoothnessSlower response

VI. Risk Management Features​

6.1 Custom Stoploss Coordination​

# When sell signal triggers, stoploss set to 0.001
sl_new = 0.001

Design Philosophy:

  • Custom stoploss function coordinates with sell signals
  • Quick exit when sell signal triggers
  • Avoids "roller coaster" profit giveback

6.2 CooldownPeriod Protection​

{
    "method": "CooldownPeriod",
    "stop_duration_candles": 2
}

Purpose: Prohibits re-entry within 2 candles after buying, prevents excessive fee consumption from high-frequency trading.

6.3 LowProfitPairs Protection​

{
    "method": "LowProfitPairs",
    "lookback_period_candles": 60,
    "trade_limit": 1,
    "stop_duration": 20,
    "required_profit": -0.05,
}

Purpose: Pauses trading for 20 candles if pair loses over 5% within 60 candles, avoids repeated losses on "problem pairs".

VII. Strategy Pros & Cons​

✅ Advantages​

  1. Multiple MA types: EMA, ZEMA, TRIMA three options, adapts to different market rhythms
  2. Offset trading: Captures deep pullbacks, better entry prices
  3. Custom stoploss: Quick exit when sell signal triggers, locks profits
  4. EWO filter: Wave oscillator identifies trend phases and pullbacks
  5. Protection mechanisms: Cooldown + LowProfitPairs dual protection
  6. Hyperparameter optimization: Supports Hyperopt for key parameters

⚠️ Limitations​

  1. High complexity: Multiple MA + custom stoploss, difficult to debug and optimize
  2. No BTC correlation: Doesn't detect Bitcoin market trend, easy to lose in market crashes
  3. Parameter sensitivity: Offset amounts and MA periods need optimization for different pairs
  4. Medium computation: Multiple MA types increase computational burden
  5. Backtest differences: Custom stoploss not effective in backtest mode, may differ from live

VIII. Applicable Scenarios​

Market EnvironmentRecommended ConfigurationNote
Ranging marketDefault configurationOffset trading suitable for ranging pullback entries
Slow bull trendDefault configurationCustom stoploss can ride green candles
Sharp up/downAdjust stoplossMay need to adjust stoploss and trailing parameters
Single-sided declinePause or light positionNo long-term trend filter, easy to lose consecutively
Sideways consolidationReduce positionSignal frequency may be lower

IX. Applicable Market Environments Explained​

MultiMA_TSL is a trend-following strategy based on the core philosophy of "multi moving average offset + custom stoploss". Based on its code architecture and community long-term live trading verification experience, it's most suitable for ranging trend markets, and performs poorly in single-sided crashes.

9.1 Strategy Core Logic​

  • Multi MA offset: Buy when price below MA×offset, captures deep pullbacks
  • EWO filter: Wave oscillator identifies trend phases, distinguishes deep pullbacks from trend pullbacks
  • Custom stoploss: Quick exit when sell signal triggers, avoids profit giveback

9.2 Performance in Different Market Environments​

Market TypePerformance RatingReason Analysis
📈 Slow bull/ranging up★★★★☆Offset entry + custom stoploss works well
🔄 Wide ranging★★★★☆Pullback entry mechanism suitable for ranging
📉 Single-sided crash★★☆☆☆No trend filter, may consecutively buy and get stuck
⚡️ Extreme sideways★★★☆☆Signals reduce, but offset mechanism can filter false signals

9.3 Key Configuration Recommendations​

ConfigurationRecommended ValueNote
Number of pairs15-30Moderate signal frequency
Max positions3-5Control risk
Position modeFixed positionRecommended fixed position
Timeframe5mMandatory requirement

X. Important Note: The Cost of Complexity​

10.1 Moderate Learning Curve​

Strategy code is about 120 lines, requires understanding:

  • Differences and selection of three moving averages
  • Principles of offset mechanism
  • Custom stoploss function working mechanism

10.2 Moderate Hardware Requirements​

Multiple MA calculations increase computation:

Number of PairsMinimum RAMRecommended RAM
15-30 pairs1GB2GB
30-60 pairs2GB4GB

10.3 Backtest vs Live Differences​

Custom stoploss only effective in live mode:

  • Backtest results may be overly optimistic
  • Live trading may have different exit behavior
  • Need dry-run verification before live trading

10.4 Manual Trader Recommendations​

Manual traders can reference this strategy's offset trading approach:

  • Observe multiple MA types simultaneously
  • Buy when price below MA×offset
  • Use custom stoploss to protect profits

XI. Summary​

MultiMA_TSL is a well-designed trend-following strategy. Its core value lies in:

  1. Multiple MA types: EMA, ZEMA, TRIMA three options, adapts to different market rhythms
  2. Offset trading: Captures deep pullbacks, better entry prices
  3. Custom stoploss: Quick exit when sell signal triggers, locks profits
  4. EWO filter: Wave oscillator identifies trend phases
  5. Protection mechanisms: Cooldown + LowProfitPairs dual protection

For quantitative traders, this is an excellent trend-following strategy template. Recommendations:

  • Use as an advanced case for learning multi-MA strategies
  • Understand offset mechanism and custom stoploss
  • Can add BTC trend filter and other protection mechanisms
  • Note backtest/live differences, verify with dry-run before live trading