ICT Trading Strategy Framework

A sophisticated algorithmic trading framework built on NautilusTrader that implements Inner Circle Trader (ICT) concepts and Turtle Soup strategies for systematic trading.

Features

• Rule-Based Trading System: Flexible, modular rule engine for building complex trading strategies
• ICT Concepts Implementation:
  • Fair Value Gaps (FVG)
  • Smart Pivot Points
  • Liquidity Pools
  • Market Structure Analysis
  • Daily/Weekly Context Rules
• Turtle Soup Strategy: Multi-timeframe implementation with position management
• Advanced Position Management:
  • Trailing stop loss
  • Partial close rules
  • Move to breakeven
  • SMA-based exit signals
• Risk Management: Fixed risk sizing and reward-risk ratio validation
• Backtesting: High-performance backtesting engine with detailed analytics

Performance

Latest backtest results on E-mini S&P 500 Futures (ESZ5):

============================================================
BACKTEST RESULTS
============================================================
Trader ID:        BACKTESTER-001
Run ID:           f48cfad3-2e2c-4c67-bd15-503910f1c56f
Backtest Period:  2025-01-01 23:01:00 → 2025-10-20 00:00:00
Elapsed Time:     0.03 seconds
Iterations:       282,241
Total Orders:     246
Total Positions:  79

============================================================
PNL STATISTICS (USD)
============================================================
PnL (total):      $4,055.50
PnL% (total):     13.52%
Expectancy:       $51.34
Win Rate:         35.44%
Max Winner:       $1,320.00
Avg Winner:       $713.21
Min Winner:       $588.00
Max Loser:        $-389.25
Avg Loser:        $-312.04
Min Loser:        $-199.50

============================================================
RETURN STATISTICS
============================================================
Sharpe Ratio (252d):   1.8668
Sortino Ratio (252d):  3.5886
Profit Factor:         1.3416
Returns Volatility:    0.1345
Risk Return Ratio:     0.1176
Avg Return:            0.0731%
Avg Win Return:        0.8105%
Avg Loss Return:       -0.3317%

Installation

Prerequisites

• Python 3.11 or higher
• pip or uv package manager

Setup

1. Clone the repository:

git clone <repository-url>
cd "ICT NT"

2. Create and activate a virtual environment:

python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies:

Option A: Using pip and requirements files

# Core dependencies
pip install -r requirements.txt

# For development (includes linting, testing tools, etc.)
pip install -r requirements-dev.txt

Option B: Using pip with pyproject.toml

# Core dependencies only
pip install -e .

# With Jupyter notebooks support
pip install -e ".[jupyter]"

# With development tools
pip install -e ".[dev]"

# Everything
pip install -e ".[all]"

Project Structure

.
├── core/                    # Core framework components
│   ├── rules/              # Base rule classes
│   ├── strategies/         # Base strategy classes
│   └── enums/              # Enums and constants
├── rules/                   # Trading rule implementations
│   ├── entry_turtle_soup_rule.py
│   ├── search_liquidity_pools_rule.py
│   ├── market_structure_rule.py
│   ├── weekly_context_rule.py
│   ├── daily_bias_rule.py
│   └── sma_exit_rule.py
├── strategies/              # Strategy implementations
│   ├── turtle_soup/        # Turtle Soup strategy
│   └── ict/                # ICT-based strategies
├── indicators/              # Custom indicators
│   ├── pivot_points_high_low.py
│   ├── smart_pivot_points.py
│   ├── fair_value_gap.py
│   └── fibonacci_levels.py
├── risk/                    # Risk management modules
├── tests/                   # Unit tests
├── notebooks/              # Jupyter notebooks for analysis
└── data/                   # Historical data storage

Usage

Running a Backtest

from nautilus_trader.backtest.node import BacktestNode
from nautilus_trader.backtest.config import BacktestRunConfig

# Configure and run backtest
config = BacktestRunConfig(...)
node = BacktestNode(configs=[config])
results = node.run()

See backtest_benchmark.py for a complete example.

Creating Custom Rules

from core.rules.rule_base import RuleBase
from core.enums.rule_signal import RuleSignal

class MyCustomRule(RuleBase):
    def check(self) -> RuleSignal:
        # Implement your trading logic
        return RuleSignal.LONG if condition else RuleSignal.NO_SIGNAL

Building a Strategy

from core.strategies.rule_based_strategy import RuleBasedStrategy

strategy = RuleBasedStrategy(
    entry_rules=[entry_rule1, entry_rule2],
    exit_rules=[exit_rule],
    position_management_rules=[trailing_stop, partial_close]
)

Jupyter Notebooks

The notebooks/ directory contains interactive Jupyter notebooks for analysis, backtesting, and data preparation:

Strategy Backtesting

• turtle_soup_strategy_backtest.ipynb: Complete backtest of the Turtle Soup strategy with detailed performance metrics, position analysis, and visualizations. Generates comprehensive reports including:

  • PnL statistics (total return, win rate, expectancy)
  • Risk metrics (Sharpe ratio, Sortino ratio, drawdowns)
  • Position reports and cumulative returns charts
  • Commission analysis by liquidity side

• ict_strategy_backtest.ipynb: Full backtest implementation of the ICT-based strategy with:

  • Multi-timeframe analysis (weekly, daily, intraday)
  • Equity curve and drawdown visualization
  • Detailed performance statistics
  • Position-by-position analysis

Data Management

• databento_dbn_zst_to_parquet_catalog.ipynb: Converts Databento DBN compressed files (.dbn.zst) to NautilusTrader-compatible Parquet data catalog format. Handles:

  • Schema mapping and normalization
  • Timezone conversions
  • Partitioned Parquet file writing

• databento_csv_to_parquet_catalog.ipynb: Converts CSV historical data to Parquet catalog format

• databento_viewer.ipynb: Utility notebook for viewing and processing Databento data files:

  • Remove calendar spreads
  • Filter front-month contracts by daily volume
  • Export processed data to CSV

• write_instrument_to_parquet_catalog.ipynb: Utility for writing instrument definitions to the Parquet catalog

Running Notebooks

# Start Jupyter
jupyter notebook

# Or JupyterLab
jupyter lab

Navigate to the notebooks/ directory and open any notebook to run backtests or process data.

Configuration

Create a .env file in the project root for environment-specific settings:

# Add your configuration here

Testing

Run the test suite:

pytest

Run specific tests:

pytest tests/test_smart_pivot_points.py -v

Continuous Integration

This project uses GitHub Actions for automated testing:

• Tests: Automatically run on all pushes and pull requests across multiple OS (Ubuntu, macOS, Windows) and Python versions (3.11, 3.12)
• Code Quality: Automated linting, formatting checks, type checking, and security scans
• Dependency Updates: Dependabot automatically creates PRs for dependency updates

Check the status badges at the top of this README for current build status.

Key Concepts

ICT Methodology

• Fair Value Gaps: Price imbalances that act as magnets
• Liquidity Pools: Areas of accumulated stop losses
• Market Structure: Higher highs, higher lows analysis
• Order Blocks: Key institutional price levels

Turtle Soup

A counter-trend strategy that fades false breakouts from consolidation ranges, originally designed to exploit failures of the Turtle Trading System breakout signals.

Development

Adding New Rules

1. Create a new file in rules/ directory
2. Inherit from RuleBase or QuoteTickRuleBase
3. Implement the check() method
4. Add tests in tests/ directory

Adding New Indicators

1. Create indicator in indicators/ directory
2. Follow NumPy/Pandas vectorized operations for performance
3. Add unit tests with edge cases

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

License

See LICENSE file for details.

Acknowledgments

• Built on NautilusTrader - High-performance algorithmic trading platform
• ICT concepts by Michael J. Huddleston
• Turtle Soup strategy by Linda Raschke

Disclaimer

This software is for educational and research purposes only. Trading financial instruments carries risk. Past performance does not guarantee future results. Use at your own risk.

Support

For questions, issues, or feature requests, please open an issue on GitHub.