Summary
This full guide walks through building an AI-powered trading bot for stocks and ETFs without writing code by hand. The creator uses ChatGPT’s Codex app to generate and refine strategies in plain English, paired with the free TradingKit plugin, an MCP server that provides backtesting against a community library of one million tested strategies.
The video is organized into five steps: choosing the tools, connecting them inside ChatGPT, building a strategy with AI, running a full backtest, and automating execution. For live trading, a free middleman service passes signals to the Alpaca broker, which offers a paper trading account for safe testing. The creator notes that this setup removes the need for a TradingView subscription.
A highlight is the scheduled loop, in which a single prompt tells the AI agent to act as a systematic trader and spend two hours researching strategies from sources such as Google Scholar, Quantpedia, Stonehill Forex and TradingView’s open-source indicator library. The video closes with cautionary advice on testing before risking real money. Viewers curious about using AI agents, MCP tools and scheduled research loops for quantitative strategy development will find a clear, reproducible workflow. This is educational content and not financial advice.
📺 Source: Trading with DaviddTech · Published October 05, 2026
🏷️ Format: Tutorial Demo
