Kalshi Market Making Algorithm

This project implements a market making algorithm for Kalshi markets, capable of running multiple strategies in parallel. Its most involved algorithm implementaiton is the Avellaneda-Stoikov model.

Local Setup

1. Clone the repository
2. Install dependencies:

   pip install -r requirements.txt
   

3. Create a .env file with your Kalshi credentials:

   KALSHI_EMAIL=your_email
   KALSHI_PASSWORD=your_password
   KALSHI_BASE_URL=https://trading-api.kalshi.com/trade-api/v2
   

4. Create or modify the config.yaml file with your market making configurations. Each configuration in this file will run as a separate strategy.
5. Run the script:

   python runner.py --config config.yaml
   

Configuration

The config.yaml file should contain one or more strategy configurations. Each strategy should have a unique name and include the following sections:

• api: Specifies the market ticker and trade side.
• market_maker: Defines parameters for the Avellaneda market maker algorithm.
• dt: The time step for the market maker's main loop.

Example:

strategy_name:
  api:
    market_ticker: MARKET-TICKER
    trade_side: "yes"
  market_maker:
    max_position: 5
    order_expiration: 28800
    gamma: 0.1
    k: 1.5
    sigma: 0.001
    T: 28800
    min_spread: 0.0
    position_limit_buffer: 0.1
    inventory_skew_factor: 0.001
  dt: 2.0

You can define multiple strategies in the same file. The runner will execute all strategies in parallel.

Deploying on fly.io

1. Install the flyctl CLI: https://fly.io/docs/hands-on/install-flyctl/
2. Login to fly.io:

   flyctl auth login
   

3. Navigate to your project directory and initialize your fly.io app:

   flyctl launch
   

   Follow the prompts, but don't deploy yet.
4. Set your Kalshi credentials and base URL as secrets:

   flyctl secrets set KALSHI_EMAIL=your_email
   flyctl secrets set KALSHI_PASSWORD=your_password
   flyctl secrets set KALSHI_BASE_URL=https://trading-api.kalshi.com/trade-api/v2
   

5. Ensure your config.yaml file is in the project directory and contains all the strategies you want to run.
6. Deploy the app:

   flyctl deploy
   

The deployment will use the runner.py script, which will run all strategies defined in your config.yaml file in parallel.

Monitoring

Each strategy will log its activities to a separate log file named after the strategy (e.g., strategy_name.log). You can monitor these logs using the fly.io logging system:

flyctl logs

For more detailed instructions on monitoring and managing your deployment, refer to the fly.io documentation.