LEAN Money Printer - Quantitative Trading Analysis

A Python-based quantitative trading analysis project using the LEAN engine to backtest and analyze trading algorithms. Experiment with market algorithms, analyze performance metrics (alpha, drawdown, risk/reward ratios), and research trading strategies using Jupyter notebooks.

Prerequisites

• Miniforge (minimal conda distribution with conda-forge)
• VS Code (recommended IDE)
• Git (version control)

Install Miniforge

If you don't have Miniforge installed, download it from miniforge.pydata.org:

macOS (via Homebrew):

brew install miniforge
conda init zsh  # or bash, depending on your shell

Or download the installer directly from the repository.

Quick Start

1. Clone & Navigate to Project

git clone <repository-url>
cd lean_mny_printer

2. Set Up Python Environment with Miniforge

Create and activate the conda environment:

conda env create -f environment.yml
conda activate lean-mny-printer

This creates a conda environment with:

• LEAN - Backtesting engine (from conda-forge)
• Jupyter - Interactive notebooks
• Data tools - pandas, numpy, scipy, scikit-learn
• Visualization - matplotlib, plotly, seaborn
• Development tools - black, flake8, pytest

3. Open in VS Code

code .

VS Code will prompt you to install recommended extensions (Jupyter, Python, Pylance, Data Wrangler, Copilot, etc.). Click "Install All" to set up your IDE.

After extensions are installed, select the conda kernel:

• Open the command palette (Cmd+Shift+P on macOS)
• Search for "Python: Select Interpreter"
• Choose the lean-mny-printer environment

4. Start Experimenting

Open research.ipynb in VS Code:

• Click on the notebook file
• The Python kernel should auto-detect your conda environment
• Run cells to analyze trading strategies

Project Structure

lean_mny_printer/
├── main.py                    # QCAlgorithm implementation
├── config.json                # Algorithm configuration
├── research.ipynb             # Jupyter notebook for analysis
├── copilot-instructions.md    # AI assistant instructions
├── environment.yml            # Conda environment specification
├── .vscode/
│   └── extensions.json        # Recommended VS Code extensions
└── .venv/                     # (Not used - conda handles environments)

Usage

Running the Jupyter Notebook

1. Activate environment (if not already active):

   conda activate lean-mny-printer

2. Start Jupyter Lab:

   jupyter lab

3. Or use VS Code's native Jupyter support:

   • Open research.ipynb in VS Code
   • Select the lean-mny-printer conda kernel
   • Run cells

Running the Algorithm

The main.py contains a simple Leanmnyprinter algorithm that backtests on SPY data.

Using LEAN CLI (see optional section below):

lean backtester

Notebook Analysis

research.ipynb includes:

• QuantBook - Interactive analysis tool
• Indicator Analysis - Bollinger Bands example
• Historical Data - Access to market data
• Visualization - Charts and performance metrics

LEAN CLI (Optional but Recommended)

Many advanced scenarios require the LEAN CLI:

• Downloading additional market data
• Running live paper trading
• Optimizing algorithms
• Managing multiple strategies
• Local backtesting with full customization

Install LEAN CLI

pip install lean-cli

Or download from: https://www.quantconnect.com/docs/v2/lean-cli

Verify Installation

lean --version

Download Sample Data (if needed)

lean data download --dataset-file <config-file>

Run Backtest Locally

lean backtester --algorithm main.py --data-dir ./data

Key LEAN Concepts

Algorithm Structure

class MyAlgorithm(QCAlgorithm):
    def initialize(self):
        """Setup: Set dates, cash, subscriptions"""
        self.set_start_date(2020, 1, 1)
        self.set_end_date(2023, 12, 31)
        self.set_cash(100000)
        self.add_equity("SPY", Resolution.DAILY)
    
    def on_data(self, data: Slice):
        """Main event: Called on each data point"""
        if not self.portfolio.invested:
            self.set_holdings("SPY", 1)

Performance Metrics

• Alpha - Excess returns vs. benchmark
• Sharpe Ratio - Risk-adjusted returns (target > 1.0)
• Maximum Drawdown - Peak-to-trough decline
• Volatility - Standard deviation of returns
• Profit Factor - Gross profit / Gross loss (target > 1.5)
• CAGR - Compound annual growth rate

Resources

• LEAN Documentation: https://www.quantconnect.com/docs/v2/writing-algorithms/
• QuantConnect Getting Started: https://www.quantconnect.com/docs/v2/writing-algorithms/key-concepts/getting-started
• LEAN CLI Docs: https://www.quantconnect.com/docs/v2/lean-cli
• LEAN GitHub: https://github.com/QuantConnect/Lean
• uv Documentation: https://astral.sh/uv

Common Commands

# Activate environment
conda activate lean-mny-printer

# Add new dependency
conda install -c conda-forge <package>

# Update environment
conda env update -f environment.yml

# Run Jupyter Lab
jupyter lab

# Run Jupyter Notebook in VS Code
# (Open .ipynb file directly in VS Code)

# List installed packages
conda list

# Deactivate environment
conda deactivate

Troubleshooting

"Python not found"

• Ensure Miniforge is installed and initialized: conda --version
• Verify conda shell integration: conda init zsh (or your shell)

"Module not found" in Jupyter

• Verify the kernel uses your conda environment: Click kernel selector in VS Code, choose lean-mny-printer
• Reinstall dependencies: conda env update -f environment.yml

LEAN import errors

• Ensure LEAN is installed: conda list | grep lean
• Update environment: conda env update -f environment.yml --prune

Jupyter not opening

• Ensure environment is activated: conda activate lean-mny-printer
• Verify Jupyter is installed: conda list | grep jupyter
• Check kernel: jupyter kernelspec list

Next Steps

1. ✅ Install Miniforge
2. ✅ Create conda environment from environment.yml
3. ✅ Open project in VS Code
4. ✅ Select conda kernel in VS Code
5. ✅ Open research.ipynb and run analysis
6. 🔄 Modify algorithms in main.py or notebook
7. 📊 Analyze performance metrics
8. (Optional) Install LEAN CLI for advanced features

---

Happy backtesting! 📈

Other Quant Tools

https://github.com/wilsonfreitas/awesome-quant