Skip to main content

Obelisk_3EMA_StochRSI_ATR Strategy Analysis

Table of Contents​

  1. Strategy Overview
  2. Theoretical Foundation
  3. Technical Indicator Details
  4. Entry Signal Analysis
  5. Exit Mechanism Design
  6. Risk Management Framework
  7. Timeframe Strategy
  8. Parameter Configuration
  9. Backtesting and Optimization
  10. Live Trading Notes
  11. Summary

1. Strategy Overview​

1.1 Background​

Obelisk_3EMA_StochRSI_ATR, released April 2021 by Obelisk, draws inspiration from Trade Pro's YouTube video "76% Win Rate Highly Profitable Trading Strategy." The strategy combines EMA for trend direction, StochRSI for timing, and ATR for dynamic stop-loss/take-profit sizing.

⚠️ Important: Code explicitly states "DO NOT RUN LIVE" — for research and backtesting only.

1.2 Core Philosophy​

Multi-indicator synergy: EMA identifies trend, StochRSI confirms momentum turning points, ATR dynamically sizes stops. Trend-following with momentum confirmation.

1.3 Positioning​

Long-only trend-following strategy. Suitable for:

  • Quantitative strategy research and learning
  • Backtesting analysis
  • Hyperparameter optimization experiments
  • Risk management mechanism study

2. Theoretical Foundation​

2.1 Trend-Following Theory​

Assumption: "Trends persist" — once a clear up/down trend forms, it likely continues for a period. Triple EMA arrangement identifies trend via multi-period MA stacking.

2.2 Momentum Oscillator Theory​

StochRSI applies stochastic formula to RSI, providing higher sensitivity than plain RSI. Identifies momentum turning points for precise entries.

2.3 Volatility Measurement Theory​

ATR measures true range, capturing actual market volatility. Dynamic stops adapt to volatility — wider in volatile markets, tighter in calm markets.

3. Technical Indicator Details​

3.1 Triple EMA​

  • EMA8: Fast line, short-term momentum (N=8)
  • EMA14: Middle line, medium-term smoothing (N=14)
  • EMA50: Slow line, main trend filter (N=50)

EMA formula: EMA(t) = α × Price(t) + (1-α) × EMA(t-1), where α = 2/(N+1)

3.2 Stochastic RSI​

Custom implementation:

  • RSI period: 14
  • StochRSI period: 14
  • K smoothing: 3
  • D smoothing: 3
df['rsi'] = ta.RSI(dataframe, timeperiod=14)
stochrsi = (rsi - rsi_min) / (rsi_max - rsi_min)
df['srsi_k'] = stochrsi.rolling(3).mean() * 100
df['srsi_d'] = df['srsi_k'].rolling(3).mean()

3.3 ATR​

dataframe['atr'] = ta.ATR(dataframe, timeperiod=14)

Stop-loss: entry - 3 × ATR Take-profit: entry + 2 × ATR

4. Entry Signal Analysis​

4.1 Entry Conditions​

Both conditions must be true simultaneously:

Condition 1: EMA Bullish Arrangement

(dataframe['ema8'] > dataframe['ema14']) & (dataframe['ema14'] > dataframe['ema50'])

EMA8 > EMA14 > EMA50 = clear uptrend.

Condition 2: StochRSI Golden Cross

qtpylib.crossed_above(dataframe['srsi_k'], dataframe['srsi_d'])

K-line crosses above D-line = short-term momentum starting to strengthen.

4.2 Signal Generation Logic​

dataframe.loc[
    (ema8 > ema14) & (ema14 > ema50) &
    qtpylib.crossed_above(srsi_k, srsi_d),
    'go_long'
] = 1

4.3 Signal Filtering​

Multi-confirmation: EMA arrangement filters trend, StochRSI confirms momentum — two conditions must coincide.

5. Exit Mechanism Design​

5.1 Dynamic Stop-Loss/Take-Profit​

dataframe['take_profit'] = close + atr * 2
dataframe['stop_loss'] = close - atr * 3
dataframe['stop_pct'] = (atr * 3) / close

5.2 Custom Stop-Loss Function​

def custom_stoploss(...):
    # If current rate > take_profit → close for profit
    if current_rate > take_profit:
        return 0.001
    # Otherwise maintain stop distance from entry
    return stoploss_from_open(-stop_pct, current_profit)

5.3 Exit Signal​

def populate_exit_trend(...):
    dataframe['sell'] = 0
    return dataframe

No active exit signals — all exits through stop-loss/take-profit mechanism.

6. Risk Management Framework​

6.1 Stop-Loss Analysis​

  • Dynamic: Based on ATR, adapts to volatility
  • Fixed ratio: Always 3× ATR from entry
  • Upper bound: stoploss = -0.10 caps max loss at 10%

6.2 Take-Profit Analysis​

  • Fixed ratio: 2× ATR from entry
  • Fixed target: Does not move with price

6.3 Risk/Reward Ratio​

Assuming ATR ≈ 2% of price:

  • Stop distance = 3 × 2% = 6%
  • Target distance = 2 × 2% = 4%
  • Risk/reward = 4%/6% ≈ 0.67

Strategy needs ≈ 60% win rate to break even.

7. Timeframe Strategy​

7.1 Dual Timeframe Design​

Backtest timeframe: 5 minutes Signal timeframe: 1 hour (via informative pairs)

7.2 Rationale​

5-minute backtesting: more precise price simulation, avoids "whipsaw" issues where a single candle crosses both stop and target.

1-hour signals: reduces noise, prevents overtrading.

7.3 Timeframe Conversion​

time_factor = int(60 / 5)  # = 12

8. Parameter Configuration​

8.1 Core Parameters​

  • startup_candle_count = 500
  • minimal_roi = {"0": 1} (100% — only ATR take-profit is used)
  • stoploss = -0.10

8.2 Indicator Parameters​

  • EMA periods: 8, 14, 50
  • RSI: 14
  • StochRSI: 14, smoothing 3/3
  • ATR: 14

9. Backtesting and Optimization​

  • Use 5m or 1m timeframe
  • At least 6 months data
  • Include bull, bear, and ranging periods
  • Set realistic slippage (0.5%-1%)

9.2 Optimization Suggestions​

Optimizable: EMA periods, StochRSI parameters, ATR multiples, stop-loss cap.

Avoid overfitting — use out-of-sample validation.

10. Live Trading Notes​

⚠️ Developer Warning: "DO NOT RUN LIVE"

Limitations for live use:

  • Trend-dependent: poor in ranging/volatile markets
  • Lag: EMA-based indicators lag in fast reversals
  • Long-only: cannot profit in downtrends

Suggested Improvements for Live​

  • Add RSI filter
  • Add volume confirmation
  • Implement true trailing stop
  • Add shorting logic

11. Summary​

A structurally clean trend-following strategy demonstrating:

  • EMA trend + StochRSI momentum combination
  • ATR-based dynamic risk management
  • Dual-timeframe implementation

⚠️ Not recommended for live trading per developer. Excellent for learning.

Key learning points: multi-indicator synergy, dynamic stops, multi-timeframe design, risk/reward balance.

Strategy Author: Obelisk (brookmiles) Reference: Trade Pro "76% Win Rate Strategy" Disclaimer: For learning only, not investment advice.