Equity Options Trading Bot

DISCLAIMER: THIS PROJECT DOES NOT IMPLEMENT A PROFITABLE STRATEGY. FOR REFERENCE ONLY

This is a fully automated equity option trading bot. Provided the user passes valid exchange auth credentials, simply starting the listed programs under Startup will have you kicked back drinking an ice cold beer while the bot trades its strategy. No user interference or guidance necessary.

Strategy

• Stream real-time tick data and build OHLC candles at pre-defined intervals
• Direction is determined by the close of the last candle. Short (long puts) if close is below pre-defined MA else Long (long calls)
• After we have determined direction, we wait for price (does not have to be a candle close) to break out of x candles highs (if long) or lows (if short)
• If these criteria are met, we will enter a position. Once in a position, if price (not candle close) crosses pre-defined MA, flatten position.
• Focused on getting into a position. Currently sending market orders until we stream options MD to decision engine. Limit orders will then be implemented

Startup

Refer to the desired handler's source code to retrieve a list of valid startup parameters and their use cases. Auth credentials will be required if running with a live account.

• Market Data Server: .py311/bin/python md_handler.py --exchange fake --symbols SPY QQQ --timeframe 5m --ma sma --period 9 --lookback 3
• Exchange Router: .py311/bin/python exchange_handler.py --exchange fake Note: Currently, there is no fake exchange handler implemented, but it is rather easy to do so. I will leave that up to you to build out as you please.
• Decision Engine: .py311/bin/python decision_engine.py --symbols SPY QQQ

Architecture

IPC between Decision Engine, Exchange and Market Data Handlers implemented using unix domain sockets

• Decision Engine

  • Subscribes to indicators and price feeds from MDSocketServer
  • Determines whether signal is generated from signal data
  • Requests positions, balances, other relevant info from Exchange Handler
  • Produce decisions based on signals, balances, and positions
  • Sends order parameters to order router for execution on exchange

• Data Handler

  • Supports strategy discussed above
  • Subscribes to MD from exchange
  • Produces signals from MD
  • Sends signals, prices to Decision Engine

• Exchange Handler

  • Tradier Exchange is built out
  • TDA and Webull exchanges paused indefinitely
  • Handles position management, balances, and order entry requests from Decision Engine and sends to exchange

• Book

  • Not built. Bids/Asks will be streamed directly to Decision Engine.
  • If we want to store book data to db, we can implement this
  • Recieves market data from Market Data Feed
  • Builds a book of bids and asks

Instruments

• SPY Options

To Do

• Write MD and exchange task management like i do in decision handler
• Need to (properly) close all processes upon insufficient funds or market close.
• Need to integrate limit orders
• Potentially add some timeout logic to the queues and sockets
• Need to add support running from remote
• Have server respond with list of symbols that fail to subscribe
• If decide to build book, seperate md from signal generator and stream it to both signal generator and book

Notes

• Decision Engine is built to support only one position per underlying. If needed, can make multi
• Incorporate other indicators
• Optimize choosing expiry that minimizes time decay, but ensures a return

Platform

• Tradier Brokerage API
• TD Ameritrade API. Development paused due to TD Ameritrade not supporting asynchronous code
• Webull OpenAPI. Development has not started
• Python 3.11.3