Skip to content

A robust backtesting and live trading engine designed for seamless strategy development and deployment. It supports user-defined strategies, multi-threaded execution, and integrations with brokers and data sources.

License

Notifications You must be signed in to change notification settings

midassystems/midastrader

Repository files navigation

MidasTrader

License pypi-version

MidasTrader is a robust trading system designed for seamless transitions between backtesting and live trading without requiring changes to user code. It integrates a flexible architecture combining a data engine, execution engine, and core components that streamline the strategy development process. The system is built with a multi-threaded design where each component communicates via a shared message bus.

Key Components

  1. Core Engine:

    • Central to the system, the Core Engine includes:
      • Order Book: Tracks market depth and price movements.
      • Portfolio Server: Manages and tracks portfolio allocations and positions.
      • Performance Tracking: Calculates and monitors key trading metrics.
      • Order Management System: Handles order placement, modifications, and cancellations.
      • Base Strategy: A foundation for user-defined strategies.
  2. Data Engine:

    • Connects to user-defined data sources:
      • Midas Server: Access historical data via the Midas ecosystem.
      • Binary Data Files: Handles local files encoded with the Midas Binary Encoding Library.
      • External Sources: Currently supports Databento, with more integrations planned.
  3. Execution Engine:

    • Facilitates live trading by connecting to brokers:
      • Currently supports Interactive Brokers.
      • Users can configure broker details in the config.toml file.

Installation

You can install midastrader directly from PyPI:

pip install midastrader

Configuration

Configuration File (config.toml)

Define system parameters, including data sources, execution settings, and strategy configuration.

Strategy File (logic.py)

Strategies are implemented by extending the BaseStrategy class. Define your custom logic in Python.

Usage

CLI Mode

Run the system using the following commands:

# Backtest Mode
midas path/to/config.toml backtest

# Live Mode
midas path/to/config.toml live

Application Mode

Alternatively, you can use the system programmatically in your application:

from midas.cli import run

# Backtest Mode
run("path/to/config.toml", "backtest")

# Live Mode
run("path/to/config.toml", "live")

Supported Features

Data Vendors
Databento
Brokers
Interactive Brokers

Future Plans

  • Add more data sources.
  • Integrate additional brokers.

Contributing

Contributions are welcome! Feel free to open an issue or submit a pull request with suggestions or improvements.

License

This project is licensed under the Apache 2.0 License. See the LICENSE file for details.

About

A robust backtesting and live trading engine designed for seamless strategy development and deployment. It supports user-defined strategies, multi-threaded execution, and integrations with brokers and data sources.

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published