AI Services Market › Use cases › Money, Bills & Benefits
Money, Bills & BenefitsDone with ChatGPT 2026-10-28
Let ChatGPT manage a real micro-cap stock portfolio
An author runs a 6-month experiment where ChatGPT gets daily price data and weekly Deep Research re-evaluations to manage a real micro-cap portfolio; week 5 limit orders mostly didn't fill.
- What They Did
- Each trading day's end, the author gives ChatGPT real market data such as closing price and volume, and it has full control of the account. Once a week it runs Deep Research to re-evaluate the portfolio. In week 5 it placed limit orders, then turned defensive and trimmed ABEO from six shares to four. The author automates results and graphs with code on GitHub.
- How It Went
- Week 5 began with none of the planned orders executing. The AZTR limit sell missed, and the AXGN buy at $7.00 was far below the roughly $13 close. The portfolio ended with $32 cash, a Sharpe ratio of about 0.88 and a Sortino ratio of about 1.87.
- Worth Knowing
- The author accidentally reused the chat for coding debugging and had to start a new chat. He also plans to change the Deep Research prompt, which mostly researched new picks instead of reviewing the real portfolio.
Try it yourself with ChatGPT
I want to run a paper-trading experiment on a [micro-cap / small-cap] portfolio with a starting balance of [amount]. Each week, use Deep Research on my current holdings and a watchlist of [tickers], then recommend buys, sells, or holds with a short reason and a limit price for each. Do not place real trades. End with a summary table of the proposed portfolio and which orders would likely not fill at those limit prices.
Do this with a skill.
These AI Services Market playbooks do this exact job.