📈 Algorithmic Trading Lab (Template)

A professional, reproducible environment for researching, backtesting, and executing algorithmic trading strategies.

This repository bridges the gap between Data Science (Vectorized Backtesting) and Live Execution (Event-Driven Trading). It is designed for developers and researchers who want to move beyond "spaghetti code" scripts into a robust trading infrastructure.

Check out the corresponding blog post:

• Algorithmic Trading with VectorBT and Lumibot

⚡ Key Technologies

• Research (Phase 1): VectorBT — High-performance, vectorized backtesting using NumPy/Pandas broadcasting. Used for hypothesis testing and parameter optimization.
• Execution (Phase 2): Lumibot — Event-driven framework for realistic simulation and live trading. Handles order management, slippage, and broker connection.
• Broker: Alpaca — Commission-free API for stock trading (Paper & Live).
• Environment: VSCode Dev Containers — A fully Dockerized Python environment ensuring code runs exactly the same on every machine.

📂 Project Structure

.
├── .devcontainer/       # Docker configuration for VSCode
├── notebooks/           # Phase 1: Research & VectorBT experiments
│   └── example.ipynb    # Demo: From "Buy & Hold" to "Golden Cross" optimization
├── strategies/          # Phase 2: Production-ready Strategy Classes
│   └── mag_seven.py     # Example: "Magnificent Seven" Rebalancing Strategy
├── run_backtest.py      # Script to simulate strategies with Lumibot
├── run_live.py          # Script to deploy strategies to Alpaca (Paper/Live)
├── .env.example         # Template for API keys
└── requirements.txt     # Python dependencies

🚀 Getting Started

1. Prerequisites

• Docker Desktop (Running)
• VSCode
• VSCode Extension: Dev Containers (ms-vscode-remote.remote-containers)

2. Installation

1. Clone the repository:

   git clone https://github.com/cbrincoveanu/algo-trading-template.git
   cd algo-trading-template

2. Open in VSCode: Open the folder in VSCode. You should see a popup: "Folder contains a Dev Container configuration file. Reopen to folder to develop in a container."
3. Build the Environment: Click "Reopen in Container".
   • Note: This may take a few minutes the first time as it pulls the Docker image and installs Python libraries.

3. Configuration (Optional / For Live Trading)

1. Get your Paper Trading API Keys from the Alpaca Dashboard.
2. Rename .env.example to .env:

   mv .env.example .env

3. Paste your keys into .env:

   ALPACA_API_KEY=PKxxxxxxxxxxxxxxxxxx
   ALPACA_API_SECRET=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
   ALPACA_IS_PAPER=True

---

🧪 Workflow

Phase 1: Research (VectorBT)

Use Jupyter Notebooks to rapidly test hypotheses.

1. Open notebooks/example.ipynb.
2. Select the base (conda) kernel (or /opt/conda/bin/python).
3. Run the cells to see how VectorBT can test thousands of parameter combinations in seconds.

Phase 2: Strategy Development (Lumibot)

Once you have a robust idea, implement it as a class in strategies/.

• See strategies/mag_seven.py for an example of a multi-asset buy-and-hold strategy.

Phase 3: Backtesting (Event-Driven)

Verify your strategy handles realistic constraints (fees, slippage, sequential data).

python run_backtest.py

• Output: A browser window will open with a tear sheet (CAGR, Sharpe, Drawdown).

Phase 4: Live / Paper Trading

Deploy the bot to trade in real-time.

python run_live.py

• Output: The bot will connect to Alpaca and wait for market open.

---

⚠️ Disclaimer

This software is for educational and research purposes only.

• Algorithmic trading involves significant risk.
• Past performance is not indicative of future results.
• Do not trade with money you cannot afford to lose.
• The authors and contributors are not financial advisors.