Recap Hey guys! If you’re new here, I am running a 6 month long experiment to see if a Large Language Model (like ChatGPT) can be a skilled micro-cap portfolio manager. I give it real time data at the end of every trading day and it has full control over its assets. Why microcaps? Because I want to see if ChatGPT can  make intelligent decisions on picks that have little to none media attention (and also my budget is $100). The Rise Normally I would just throw up a graph of the week and talk a bit; I don’t think I could have ever predicted how this week went. ChatGPT came out very confident. Out of the blue, it decided to sell Cloudastructure Inc. (CSAI) and buy Inspira Technologies (IINN) on Monday. It skyrocketed the next two days. Wednesday, it went to a staggering +6.6% alpha returns. You have no idea how prepared I was to brag on ChatGPT. I even made a GitHub repo on Wednesday and included this (very ironic) graph: Everything was going a little too perfectly. The Fall Thursday was a slight downturn, but it was just normal market movement. Friday is self explanatory. One day wiped out -6.7% of the account. ONE SINGLE DAY. I know BioTech is notoriously rough, but I was honestly devested. ChatGPT took it on the chin though. Portfolio Review: Forgot to mention, once every week ChatGPT is able able to use deep research to review it’s account. Here was this week’s revaluation: Current Holdings (ABEO, CADL, AZTR, IINN):All still viable. No position is fundamentally broken. Some are in pullback but thesis remains.New Deep Research:I screened over 25+ micro-caps with real catalysts. Most were weak. No new stock beat the current risk/reward balance of what we hold.Decision:No trades.We hold all 4 current positions (plus cash). No change to stop-losses. No new buys.Reason:The portfolio is diversified, catalyst-loaded, and structurally sound. There is no better opportunity available with high enough conviction right now. Holding is optimal. Current Holdings (ABEO, CADL, AZTR, IINN): Current Holdings (ABEO, CADL, AZTR, IINN): All still viable. No position is fundamentally broken. Some are in pullback but thesis remains. New Deep Research: New Deep Research: I screened over 25+ micro-caps with real catalysts. Most were weak. No new stock beat the current risk/reward balance of what we hold. Decision: Decision: No trades. No trades. We hold all 4 current positions (plus cash). No change to stop-losses. No new buys. Reason: Reason: The portfolio is diversified, catalyst-loaded, and structurally sound. There is no better opportunity available with high enough conviction right now. Holding is optimal. diversified, catalyst-loaded, and structurally sound My Thoughts Honestly, I thought it would completely breakdown after Friday, but it still feels confident in it’s thesis; not many traders can lose ~7% of their account in a day and make zero changes. I was a little worried it would fall into a cycle of picking companies and giving up after like a week, but that hasn’t happened (at least yet). I’m excited to see if there will be any lasting effects on ChatGPT’s mentality after such a big drawdown. Thanks for checking out this week’s update and see you next time! GitHub Page and Email: If you’re curious about the code I’m using to automate results and graph, the GitHub page is:  ChatGPT-Micro-Cap-Experiment. ChatGPT-Micro-Cap-Experiment If you have any suggestions or advice, my Gmail is : nathanbsmith729@gmail.com nathanbsmith729@gmail.com Thanks for reading! Subscribe for to get an update every week and support my work.

Can ChatGPT Outperform the Market? Week 2

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

Can ChatGpt Outperform the Market? Week 7

The Hidden Architecture Powering AI Search: Vector Databases Under the Hood

The Quiet Bottleneck in Your Symfony App: Fragmented Indexes and Random UUIDs

I Spent 30 Days “Vibe Coding” an MVP — Burned $127, Broke Everything, and Still Found Product-Market

Proactive QA Monitoring in Production: Catching Production Issues Before the Customer Does

The Pragmatic Guide to Federated AI: Building Compliant LLM/XGBoost Pipelines for Sensitive Data

Can ChatGpt Outperform the Market? Week 7

The Hidden Architecture Powering AI Search: Vector Databases Under the Hood

The Quiet Bottleneck in Your Symfony App: Fragmented Indexes and Random UUIDs

I Spent 30 Days “Vibe Coding” an MVP — Burned $127, Broke Everything, and Still Found Product-Market

Proactive QA Monitoring in Production: Catching Production Issues Before the Customer Does

The Pragmatic Guide to Federated AI: Building Compliant LLM/XGBoost Pipelines for Sensitive Data

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps