Let ChatGPT rebalance a live micro-cap portfolio

A writer let ChatGPT make portfolio calls on a tiny real-money stock account using daily updates and weekly Deep Research reviews.

Done with ChatGPT· Advanced

What they did
The writer sent ChatGPT end-of-day figures like closing prices and trading volume after each market day, then let it make the portfolio decisions. Every week, it also ran Deep Research to reassess the holdings and adjust the plan. In week 5, the agent moved into limit-based trading and cut ABEO from six shares to four as part of a more cautious stance.
How it went
That week started with no fills on the planned trades. The AZTR limit sale did not go through, and the AXGN purchase at $7.00 sat well below the stock’s near-$13 close. By the end, the account held $32 in cash, with a Sharpe ratio around 0.88 and a Sortino ratio around 1.87.
Tips
If you want to try this, keep the agent’s scope tight so it stays on the portfolio instead of drifting into other tasks. Also check limit levels carefully, because prices that are too far from the market may not execute.

Try it yourself with ChatGPT

I’m running a [paper / real-money] portfolio of [micro-cap / small-cap] stocks with [amount] to start. After each trading day, I’ll give you the updated prices, volume, and any other market data you need for the holdings and watchlist. Each week, review the current positions with Deep Research and tell me whether to buy, sell, trim, or keep [tickers], with a short reason and a limit price for each idea. Before any irreversible step such as submitting orders, selling shares, or moving money, pause and ask for my approval. If a limit price looks unlikely to execute, call that out and explain why. Finish with a plain summary table of the portfolio and any orders that may not fill.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.