This is the home for Trading Tommy! Tommy is a very simple trading bot that find opporunities to trade equity securities (stocks) on the NYSE. Currently, Tommy only makes decisions based on the price and the Relative Strength Index (RSI) of a symbol.

To install all required Python modules as specified in requirements.txt, navigate to python/ and run:

python -m pip install -r requirements.txt

Follow the instruction on the Alpaca website to create an account.

For the bot to retrieve data from the Alpaca API, you must set the APCA_API_KEY_ID and APCA_API_SECRET_KEY environment variables to the API Key and API Secret Key listed in the account you create from step 2.

Note: We highly recommend you use a paper account (that simulates trades instead of placing real orders) while you get started!

Navigate inside python/ and run:

python3 main.py

In addition, you can pass in the following arguments:

• Run the custom function and exit - for testing purpose

-c, --custom

• Log debug messages

-d, --debug

• Suppresses all output except critical, overrides debug

-q, --quiet

• Store all output in the specified file

-o, --output

• Specify the timeframe over which to calculate signals and make trades, in MINUTES - [1, 1440]

-t, --timeframe

The bot will identify optimal buying/selling points, and generate graphs to display the detected buying/selling points. The following is an example graph:

[Image to insert]

Tommy is split up into three main pieces:

• Core.py is responisble for making web requests to Alpaca, for actions like fetching data and placing orders. These are almost entirely done through Alpaca's Python API.
• Brain.py calculates quantitative finance signals, like the RSI, as well as determines whether a symbol should be bought or sold.
• Platform.py decides the order in which different elements are run and the flow of the program. Currently, there are two main threads. The first thread loops through high-volume symbols (obtained from Util.py) and determines whether each should be bought or sold based on answers from Brain.py. The second thread fetches data from Alpaca in the background, which Brain.py uses to produce as up-to-date signals as possible.