A playground for algorithmic trading.

Alpaca is a commission free trading platform similar to Robinhood but has an API to allow for the development of algorithmic trading and a variety of app integrations. It provides both live trading accounts and paper trading accounts to test out trading strategies. More information can be found here. In order to use this files in this repo, you must create an account and generate a set of API keys.

This repository currently has folders for different types of algorithms. Currently there are two. Long-short takes positions in the S&P 100 based on market movements. SP-Dips looks for market dips in all of the companies in the S&P 500 in order to take long positions based on the theory that S&P 500 stocks will always trend towards long term growth. Both of these algorithms are examples provided by Alpaca for testing purposes, with the modification of the stock space on Long-short being the S&P 100. NOTE: if you plan on running these on a live trading account, be sure to keep the FINRA pattern day trader rules in mind. A pattern day trader is someone who executes four or more day trades in a five business day period. In order to day trade, an account minimum of $25,000 must be maintained.

Both programs can be built and run from the Dockerfiles using the following commands, executed in the root of both folders:

docker build -td <image_name> .

docker run -e KEY_ID=<alpaca_key_id> -e SECRET_ID=<alpaca_secret_key> -td <algo_image_name>

Additionally, they can be pulled from my Docker Hub here.

Due to the importance of up-time, k8s is a good choice for running trading algorithms. Both have been provided as Docker images for this purpose. Attached is a ReplicaSet YAML for each algorithm to spin up a single pod of either of the algorithms.

In the future, I would like to improve these algorithms because currently they are operating at a loss. I also would like to add an HFT (High Frequency Trading) algorithm to take advantage of small market shifts.