%20Results.png){kind=link}
Welcome to the repo behind my 6-month live trading experiment where ChatGPT manages a real-money micro-cap portfolio.
Everyday, I kept seeing the same ad about having an some A.I. pick undervalued stocks. It was obvious it was trying to get me to subscribe to some garbage, so I just rolled my eyes. Then I started wondering, "How well would that actually work?".
So, starting with just $100, I wanted to answer a simple but powerful question:
Can powerful large language models like ChatGPT actually generate alpha (or at least make smart trading decisions) using real-time data?
-
I provide it trading data on the stocks in it's portfolio.
-
Strict stop-loss rules apply.
-
Everyweek I allow it to use deep research to reevaluate it's account.
-
I track and publish performance data weekly on my blog. SubStack Link
- Currently stomping on the Russell 2K.
Live trading scripts — Used to evaluate prices and update holdings daily
LLM-powered decision engine — ChatGPT picks the trades
Performance tracking — CSVs with daily PnL, total equity, and trade history
Visualization tools — Matplotlib graphs comparing ChatGPT vs Index
Logs & trade data — Auto-saved logs for transparency
AI is being hyped across every industry, but can it really manage money without guidance?
This project is an attempt to find out, with transparency, data, and a real budget.
Basic Python
Pandas + yFinance for data & logic
Matplotlib for visualizations
ChatGPT 4o for decision-making
The experiment runs June 2025 to December 2025. Every trading day I will update the portfolio CSV file. If you feel inspired to do something simiar, feel free to use this as a blueprint.
Updates are posted weekly on my blog — more coming soon!
One final shameless plug: (https://substack.com/@nathanbsmith?utm_source=edit-profile-page)
Find a mistake in the logs or have advice? Please Reach out here: nathanbsmith.business@gmail.com