ChatGPTTradingBot ⭐

This is my AI high frequency trading algorithm, a Node.js project that uses the Alpaca Trade API for trading activities. When a live news event is released via Benzinga News, it will send the important data from the article to a tailored GPT 4o model and receive a sentiment analysis score on if the event will have a positive or negative effect on the company stock on a scale from 0-100, 0=worst news, 100=best news, all within 20ms. Based on the score, an algorithm determines if and how much it will BUY/SELL or SHORT/COVER the stock, the algorithm scales based on funds in the account and will adjust if margin trading is wanted. Any quantity of money used up to $10,000,000 will have similar profitable results. Location used has largest impact on success.

Additional Features:

• Adjustable Fail safe mechanism at different thresholds that will sell all positions in the event of
• Adjustable requirements for the volume/liquidity of a stock to be able to buy or short.
• Adjustable profit taking thresholds, 2%, 4%, 6%, 10% - Along with % of stock to be sold/covered after profiting from those thresholds.
• Instant and efficient News Reception, parsing, data feed into LLM and consistent accurate responses.
• From news -> Purchased stock = 5-70ms
• Live Logs all scores and prices of stock to compare against portfolio to make any adjustments/optimizations as needed at all times.
• Supports pre-market and after-market trading
• Supports pre-market and after-market live pricing updates to accurate adjust portfolio at all times.
• In the case of internet outage, bot will continue to retry and run to save and update positions as soon as problems are resolved.
• Console log updates on thought process and execution of bot
• Bot will not invest in Crypto currencies, Adjustable

BOT INFO

3 - "Time in beats timing" 5% emergency sell off - sell off lowest on new opportunity - Margin/Short If anything drops 5% from highest logged price, sell (LONG) If anything drops 5% from highest logged price, buy (SHORT) Buy/Short tradeValueThreshold = 5000000 : /1 Buy/Short limit_price = current_price * 1.007 : .993 Sell/Cover limit_price = current_price * .99 : 1.01 emergency sell-off/cover @3.5% loss of position from avg purchase price

RegT Threshold @ 98%

Getting Started

Prerequisites

• Make sure Node.js is installed on your machine
• npm install
• Configuration - Create your own ChatGPT API on https://platform.openai.com/
• Configuration - Create your own Alpaca account and obtain your API Keys on https://app.alpaca.markets/
• Configuration - Create your own Beninga News account and obtain your API Keys on https://www.benzinga.com/apis/ (14 day free trial)
• Configuration - Create your own Financial Modeling Prep account and obtain your API Keys on https://site.financialmodelingprep.com/developer/docs/pricing (for pre-after market live pricing and trading)
• Configuration - Inside the .env file enter your API Keys

Running the Server

This script uses ECMAScript (ES) modules, which is a modern JavaScript module system that differs from CommonJS

• node s.mjs

Dependancies

• @alpacahq/alpaca-trade-api
• dotenv
• fs
• moment-timezone
• node-fetch
• openai

Tech Stack

• Node.js (version 12 +)
• JavaScript (ES Modules)
• npm (usually comes with Node.js)(npm install)
• OpenAI
• dotenv
• node-fetch
• @alpacahq/alpaca-trade-api
• fs & fs/promises
• moment-timezone npm install dotenv node-fetch @alpacahq/alpaca-trade-api moment-timezone openai