👉🏻 Learn AI trading bots and algorithmic trading with us:
https://www.skool.com/daviddtech/about
I built an AI trading bot to trade stocks and ETFs, and this full guide shows you how to build and test the workflow yourself.
Using ChatGPT through OpenAI Codex, I connect AI strategy research, backtesting and automated alerts with TradingKit, Trigger.trade and an Alpaca paper trading account.
You describe what you want in plain English. AI writes the code, while TradingKit provides the backtesting engine and strategy alerts.
The big change in this setup: I create the alerts directly inside TradingKit, removing the need for a paid TradingView subscription for the automation shown in this video.
I’ll walk through the research prompt, a two-hour strategy research schedule, the results dashboard and the broker connection. You’ll also see why I give strategies an incubation period before risking capital, and why I currently trade only seven bots after testing thousands of ideas and settings.
The broker demonstration uses Alpaca paper trading with simulated funds. The strategy results discussed are historical backtests.
IN THIS VIDEO, YOU’LL DISCOVER:
✅ How I use ChatGPT and OpenAI Codex to build AI trading bots for stocks and ETFs
✅ How to install and connect the TradingKit plugin
✅ Where I research strategies using Google Scholar, Quantpedia and open-source indicators
✅ How to schedule an AI strategy research and backtesting loop
✅ How I review a dashboard showing 1,368 strategy tests and 277 marked as passed
✅ How I assess equity curves, profit factor and historical test periods
✅ Why backtests still need incubation and forward testing
✅ How to create automated strategy alerts directly in TradingKit
✅ How to connect Trigger.trade to an Alpaca paper trading account
✅ How API keys, webhook URLs and alert templates fit together
✅ Where stop losses and take profits fit into the automation
✅ What I’ve learned about diversification, trade samples and switching bots off
⏰ VIDEO CHAPTERS
0: 00 – Building an AI Trading Bot for Stocks and ETFs
0: 34 – The Build, Backtest and Automation Roadmap
1: 51 – Choosing ChatGPT and OpenAI Codex
2: 27 – The TradingKit Update: Removing a Subscription
4: 11 – Installing and Connecting the TradingKit Plugin
6: 51 – First Backtest: Checking the Connection
7: 47 – Setting Up the AI Strategy Research Loop
8: 33 – My Stock Strategy Research Prompt
11: 09 – Reviewing 1,368 Strategy Tests
12: 28 – Incubation and Testing on New Data
13: 59 – Creating Strategy Alerts Inside TradingKit
14: 48 – Trigger.trade and the Alpaca Paper Connection
15: 30 – Alpaca API Keys and Webhook Setup
16: 15 – Activating the Automated Strategy Alert
17: 17 – Backtest Limitations, Fees and Slippage
17: 59 – Diversifying Your Trading Bots
18: 23 – Giving Strategies Enough Trades
19: 06 – When to Turn a Trading Bot Off
19: 44 – Prompts, Strategy Resources and the Community
TOOLS AND RESOURCES
TradingKit:
https://tradingkit.com/
Trigger.trade:
https://trigger.trade/
Alpaca:
https://alpaca.markets/
TradingView, discussed as an indicator source and optional charting platform:
https://www.tradingview.com/
The walkthrough uses ChatGPT through OpenAI Codex. Claude is discussed as an alternative. Other resources mentioned include Google Scholar, Quantpedia, Stonehill Forex, Pine Script and MCP tools.
The prompts shown and the stock, ETF and bond strategy resources discussed are available through my Skool community linked above.
💬 Comment “STOCKS” and tell me which stock or ETF you’d like me to test next.
Subscribe for more AI trading bot builds, ChatGPT trading bot tutorials, strategy backtests and lessons from running automated trading systems.
IMPORTANT NOTES
This setup removes the TradingView subscription from the demonstrated alert workflow. AI subscriptions, tool access, community membership and any applicable data or trading costs should be checked separately.
⚠️ This video is for education and entertainment only, not financial advice. Backtests and paper trading are not real-money results or guarantees of future returns. A strategy passing a historical test does not establish a lasting trading edge. Account for fees, slippage, drawdowns and the possibility of losing capital.
Trade smarter.
Test everything.
Let the data do the talking.
#AITrading #TradingBot #ChatGPT