The creator experiments with a $1,000 AI-driven trading agent that uses sentiment analysis from social media and stock volatility data to trade CFDs, focusing on high-volatility stocks like Beyond Meat and SoFi. Leveraging Cloud Code for efficient data processing, the project aims to evaluate the agent’s real-time performance over a week, with plans to expand into cryptocurrency trading and provide ongoing updates.
In this video, the creator embarks on an experiment to build and test a trading AI agent using $1,000 to trade stocks over the course of a week. Inspired by recent research and developments in AI-driven trading, particularly in cryptocurrency markets, the creator aims to develop a similar system for CFD (Contract for Difference) trading in stocks. The setup involves leveraging sentiment analysis from social media platforms like Reddit (especially Wall Street Bets) and X (formerly Twitter), combined with volatility data to identify promising trading opportunities. The goal is to see how well the AI agent can perform in real-time trading using this data-driven approach.
The AI agent is powered by Cloud Code and uses multiple MCP servers to gather and analyze data. These servers pull stock prices from Yahoo Finance, sentiment data from social media, and trending stock information from websites like apewisdom.io and Swaggy Stocks. The creator focuses on stocks with high volatility and strong sentiment signals, deliberately avoiding large-cap stocks like Google, Apple, and Nvidia due to their lower volatility, which limits profit potential in CFD trading. The agent’s workflow includes extracting tickers from sentiment images, checking account balance, and filtering stocks to identify viable trading candidates.
The trading strategy is straightforward: the AI agent acts as a professional trader specializing in CFD trading, using sentiment signals and volatility to suggest trades. The creator sets up specific entry points, leverage levels, stop-loss, and take-profit targets for each trade. For this initial run, two trades were selected: Beyond Meat with 5x leverage and SoFi with 2.5x leverage. The creator plans to monitor these trades throughout the week to evaluate the agent’s performance, emphasizing that this is a simple and manual setup intended as a starting point for further iterations and improvements.
The creator highlights the advantages of using Cloud Code for this project, particularly its ability to handle parallel tool calls efficiently, which is crucial for processing multiple data sources simultaneously. The video also mentions plans to expand the project by creating a similar AI agent for cryptocurrency trading using Hyperliquid, which offers an API for more autonomous trading. The creator envisions this as an ongoing series, providing updates at the start and end of each trading week to track progress and refine strategies based on results.
In conclusion, the video presents an experimental approach to AI-driven stock trading using sentiment analysis and volatility data, with a modest initial investment of $1,000. While acknowledging that some success may be due to luck, the creator is optimistic about the potential of this method and intends to continue developing and improving the trading agents. Viewers are invited to follow along for weekly updates and to see whether the AI agent can generate consistent profits in the volatile world of CFD trading.