Hey guys, it’s my birthday! But more importantly, thank you for all the new subscribers and rightful criticism. It has made me realize that the way I’ve been doing things is pretty sloppy and untrustworthy.

For context, I’m 17 (as of today) and started working with Python and finance this summer. I’m still learning, but I take this seriously and want the work to stand on its own, not be excused because of my age. I welcome all feedback and want this project judged by the same standards as anyone else.

I want people to have confidence what I’m showing is genuine and unbiased; I will always post the results whether good or bad. Right now it is going abnormally well, and by no means is this a real investing strategy or advice. Because it’s only one portfolio, obviously luck has attributed to the gains so far.

For transparency, I will be making the following changes from now on:

• Logs of all chat logs going forward
• Post all past and future deep research prompts
• Compare it to the S&P500 rather than Russell 2000 and XBI
• File detailing everywhere I’ve messed up and how I will do better
• Add risk adjusted returns like Sharpe Ratio
• Ensure all prompts and structure are consistent and repeatableI wish I could give a timeline for these changes, but a few fixes should be in at least by the end of the week.Also, if you have any advice or notice an error, PLEASE reach out.I want this to be accurate as it can (minus fees), and can’t do that without your help. Thank you to everyone who subscribed and cared enough to leave justified criticism, it seriously means the world to have so much support and feedback.

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. Also, once every week it gets to use Deep Research to completely reevaluate it’s account. Can ChatGPT carve out consistent alpha in the dangerous world of micro-cap stocks? Lets find out.

Overview

This was yet another staggering week for ChatGPT. Despite facing a downturn on Monday, the growth the following days was unstoppable. Investing profits from Candel Therapeutics Inc. (CADL) into Actuate Therapeutics, Inc. (ACTU) was genius, capturing a 27.37% rise. It’s other stocks however, were unimpressive. Azitra Inc (AZTR) continued to bleed cash and Inspira Technologies (IINN) had a slight gain of 3%.

Performance Graph

Portfolio Review

To see the full report: Click Here

Here was this week’s revaluation summary:

Rationale & Conclusion

This rebalance aggressively reallocates capital toward the highest-upside opportunities while exiting a clear laggard. The strategy effectively trims profits (Inspira), cuts dead-weight (Azitra), and seizes two timely, catalyst-driven trades.

New Additions:

• Axogen (AXGN):
  Near-term binary FDA decision (~6 weeks) for Avance, backed by an already profitable-growth model. Potential for a fast spike if approved, with downside cushioned by ongoing revenue.
• Esperion (ESPR):
  Turnaround cardiovascular story with real revenue growth and multiple quarters of cash runway. Risk/reward skews positive — if profitability or strategic news hits, a double is plausible. Diversifies the portfolio beyond devices/rare diseases.

Key Holdings Retained:

• Abeona (ABEO):
  De-risked biotech with FDA-approved product (ZEVASKYN) launching now. Cash-rich and positioned for alpha as sales come in.
• Inspira Technologies (IINN):
  Riding momentum from major contracts and Nasdaq compliance. Trimmed for profits, but strong catalysts remain — continued upside likely.
• Actuate Therapeutics (ACTU):
  High-risk/high-reward oncology play with excellent data and strong investor backing. Stop-loss protects downside; upside could be a multi-bagger.

---

Risk & Strategy Notes:

• AZTR exited due to weak near-term outlook and high dilution risk.
• AXGN and ESPR added for event-driven setups (FDA approval, profitability inflection).
• Position sizing remains disciplined — no over-concentration, all trades are full-share and within cash means.
• Stop-losses enforced to protect downside; failed trades will self-limit risk.

---

Final Thoughts:

This reallocation sets the portfolio up for a high-alpha second half of 2025. We’re now concentrated in positions with:

• Clear upcoming catalysts
• Positive news flow or earnings momentum
• Institutional or insider alignment

The new mix reflects a focused, evidence-backed, and risk-aware approach — fully aligned with the project’s mission to see if LLM-guided portfolios can consistently generate alpha in volatile micro-cap territory.

My Thoughts

Edit: Because AXGN was way above $7, it decided to just put the capital into ESPR (for a position of 20 shares).

I like the idea of finally cutting off Azitra (AZTR), however I think it’s mostly because the new chat wants to restructure. I am a worried that it will lose conviction with the new picks and won’t wait for the event-setups. Also, I fed it PnL data for all tickers and IINN doesn’t have profits? So I’m unsure where it got that idea, but I’ll tell it before I put in the trades. Can it maintain this level of growth? Subscribe and find out next week!

GitHub Page and Email:

To see past deep research reports and summaries: Click Here

Have a question? Check out: Q&A

If you’re curious about the code I’m using to automate results and graph, the GitHub page is: ChatGPT-Micro-Cap-Experiment

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. Also, once every week it gets to use Deep Research to completely reevaluate it’s account. Why microcaps? Because I want to see if ChatGPT can make intelligent decisions on picks that have little to no media attention (and also my budget is $100).

Overview

I promise the growth you’re about to see has been double checked, and I can’t believe it either. You may be thinking “Sure, I bet it beat the market by what, 5 percent?“ Nope, just take a look.

Performance Graph

Yes, that is 13.3% PURE ALPHA. One stock in particular has without a doubt been the reason for the massive growth. Since the the start of Week 1, CADL (Candel Therapeutics Inc) has surged 34% while being a major piece of the portfolio. ABEO (Abeona Therapeutics Inc) has also had impressive increase with 11% growth over the last five trading days. Who needs diversification, am I right?

Portfolio Review

This was the current portfolio:

Weekly Portfolio Summary (July 12–18, 2025)

• ABEO: Started at $5.92, ended at $6.59 — up +11.32%
• CADL: Started at $6.43, ended at $6.74 — up +4.82%
• AZTR: Held steady at $0.23 — 0.00%
• IINN: Started at $1.40, ended at $1.41 — up +0.71%

New Updates

Here was this week’s revaluation in ChatGPT’s own words:

“ Summary of Recommended Actions (July 18, 2025)

• Sell Candel Therapeutics (CADL) – SELL to lock in +33% profit. The stock has run up on good news that is now priced in, and no major catalysts are expected until 2026nasdaq.com. We will liquidate the 5 shares of CADL at ~$6.74 each (≈$33.7 total).
• Buy Actuate Therapeutics (ACTU) – NEW BUY. Using the CADL proceeds (and available cash $2.32), purchase ~6 shares of ACTU around $6.00 each. This initiates a position in a promising biotech with positive Phase 2 cancer resultsnasdaq.com and potential short-term upside.
• Hold Abeona Therapeutics (ABEO) – HOLD. Maintain our 6 shares. Abeona’s strong cash position and imminent product launch support continued holdinginvestors.abeonatherapeutics.cominvestors.abeonatherapeutics.com. We expect further value realization as ZEVASKYN sales ramp up.
• Hold Azitra (AZTR) – HOLD (Speculative). Keep our 55 shares of AZTR. Despite current slight loss, the upcoming Phase 1b data in Netherton syndrome (due Q4 2025) is a high-impact catalyst worth waiting forprnewswire.com. (No additional buying due to risk level, but hold existing stake.)
• Hold Inspira Technologies (IINN) – HOLD/Accumulating. Retain our 20 shares. Inspira’s $22.5M order and Nasdaq compliance mark a turning pointglobenewswire.comstocktitan.net. We will hold for anticipated further contracts and consider adding on dips as confidence grows. “

My Thoughts

Safe to say holding through the major drawdown last week (-7% in a day) was move Buffet himself would be proud of. I am really impressed with the decision to lock in profits from CADL, especially considering the volatility in BioTech. Not many advisors would decide to lock in gains with such a high performing piece. Although, I am a little disappointed still holding the hyper speculative Azitra (AZTR). With such a massive lead already on the market, I don’t think investing in risky binary catalyst is necessary. I would rather either hold it in pure cash or find a more solid company. But of course, ChatGPT has done just fine so far.

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

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.“

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

If you missed the first post, I’m running a six month experiment to see whether ChatGPT or Deepseek could beat the market with a micro-cap portfolio. Before the first update, I have to be real with you guys.

DeepSeek did bad. Like -20% first week bad. I realize if nothing changes it’s going to a six month humiliation ritual for DeepSeek. So, I will be cutting DeepSeek out of the experiment. Your 20% downturn in five days will never be forgotten, DeepSeek.

Overall, ChatGPT had a first great week! It even outperformed the Russell 2000. CSAI (Cloudastructure Inc) was the MVP with a 12% increase. However, It’s other two picks stayed stagnant. Of course, we already know how DeepSeek’s went.

ChatGPT’s Performance: +6.72%​

DeepSeek’s Performance: -18.06%​

Now for ChatGPT’s portfolio update:

---

ChatGPT’s Changes:

• ABEO (Abeona Therapeutics) – Hold
  FDA-approved gene therapy launching now. $225M cash. Recently added to Russell 2000. Undervalued. Big upside if early sales impress.
• CADL (Candel Therapeutics) – Hold
  Strong Phase 3 cancer data. $100M cash runway. Multiple ongoing trials. Q4 glioma data could be a major catalyst.
• AZTR (Azitra, Inc.) – Buy
  ~$4M market cap. Phase II/III trial for rare skin disease (RDEB) ends Q4. Interim data showed 60% wound healing vs. 15% placebo. High-risk moonshot with potential for 10x+ upside.

---

Sold: CSAI – Lacked short-term catalyst. Capital reallocated into AZTR for higher return potential.
Cash Remaining: ~$20

---

Thought Process:

• Clear Q3–Q4 catalysts across all holdings
• High short interest could fuel squeeze rallies
• Focused on biotech for asymmetric return potential

Will ChatGPT continue to outperform? Will it fall into the depths with DeepSeek? Check back next weekend for the update and thank you for reading!

Edit: Due to DeepSeek lacking present internet knowledge, I decided to pull the plug just on DeepSeek, so its only ChatGPT now. Check out how it’s going on my latest post!

It’s no secret LLMs have exploded in both capacity and popularity over the past few years. Many people have begun wondering what impact this new technology will have on the field of finance. Companies all over the world are racing to develop new techniques for AI to use in markets.

But one question I’ve always had—and never really saw answered online—is:

How well can a language model manage a stock portfolio?

With this experiment, I hope to answer that question.

For our two contenders, we have DeepSeek and ChatGPT’s 4o—two of the most powerful AIs currently available. BOTH AI’S WILL BE GIVEN REAL MONEY. this is so I can stay invested and also ease for calculations. It’s no secret to anyone familiar with both that ChatGPT has a massive edge. While GPT-4o has access to deep, real-time research, DeepSeek is stuck relying on historical data. However, other than that, these models will have equal opportunities.

They’re bound by the same rules, given the same prompt, and judged by the same result: whichever portfolio is worth more on December 27, 2025, wins.

---

Rules Are Pretty Simple

1. $100 Total Portfolio Limit
   No margin, no fractional shares, no overages. Must use full-share positions only.
2. Only U.S.-Listed Micro-Cap Stocks
   Market cap must be under $300 million at the time of selection. OTC stocks are allowed if they’re reasonably tradable.
3. Timeframe: June 27, 2025 → December 27, 2025
   No trades or decisions can be made after June 27. Portfolio is locked in for six months.
4. No Additional Help After Launch
   Models cannot rebalance, re-enter, or adjust anything after the portfolio is finalized. Set it and forget it.
5. All Orders Must Use Full Shares
   No fractional trading. Position sizes must round down to whole shares only.
6. Stop-Losses, Sizing, and Order Types Are Allowed
   Models can implement stop-losses, choose limit vs. market orders, and allocate capital however they want—within the $100 constraint.

Note: Both AIs never only be interacted with when telling data and getting trading decisions. Also, if the AI requests additional data such as the Russell 2000 Index, I will oblige as long as they are not trying to get data on each other.

---

The Prompt Given to Both Models

“You are a professional-grade portfolio strategist. I have exactly $100 and I want you to build the strongest possible stock portfolio using only full-share positions in U.S.-listed micro-cap stocks (market cap under $300M). Your objective is to generate maximum return from today (6-27-25) to 6 months from now (12-27-25). This is your timeframe, you may not make any decisions after the end date. Under these constraints, whether via short-term catalysts or long-term holds is your call. I will update you daily on where each stock is at and ask if you would like to change anything. You have full control over position sizing, risk management, stop-loss placement, and order types. You may concentrate or diversify at will. Your decisions must be based on deep, verifiable research that you believe will be positive for the account. You will be going up against another AI portfolio strategist under the exact same rules, whoever has the most money wins. Now, use deep research and create your portfolio.”

---

Full response: ChatGPT’s report

ChatGPT-4o Portfolio Summary

Total Allocated: $94.85
Strategy: Catalyst-driven, multi-sector, high-upside picks with strict stop-losses and real-time research.

Holdings:

• Abeona Therapeutics (ABEO) – 6 shares @ $6.00 = $36.00
  FDA-approved gene therapy launching in Q3 2025.
  $3.1 million per treatment, massive addressable market.
  Over $250 million in cash from a recent voucher sale.
  Low downside, high upside anchor stock.
• Candel Therapeutics (CADL) – 5 shares @ $5.60 = $28.00
  Phase 1b brain cancer data due in Q4 2025.
  Prior data release caused 100%+ spike.
  High-risk, high-reward clinical-stage biotech.
• Cloudastructure (CSAI) – 15 shares @ $1.94 = $29.10
  AI-powered video surveillance firm.
  212% YoY revenue growth, multiple major contracts signed.
  Potential under-the-radar multi-bagger.
• Cash Reserve – $6.00
  Held for flexibility, fees, or future buy-ins on confirmation.

Risk Management:

• Stop-losses ranging from 15–25% depending on volatility.
• All trades made via limit orders.
• Trailing stops applied if a stock gains significantly.
• Portfolio monitored daily.

Expected Return:

• Base Case: $150–$180
• Bull Case: $220–$280
• Bear Case: $65–$85 (if stop-losses are triggered)

---

Full response: DeepSeek’s Report

DeepSeek Portfolio Summary

Total Allocated: $100
Strategy: High-volatility, high-potential trades based on historical catalysts and technical indicators.

Holdings:

• Pasithea Therapeutics (KTTA) – $40 allocation (~40%)
  Phase 2 PTSD trial results expected Q3 2025.
  Possible 3–5x upside if data is positive.
  Very high volatility, binary outcome.
  Stop-loss: $0.33
• ToughBuilt Industries (TBLT) – $35 allocation (~35%)
  120% YoY e-commerce growth.
  50% short interest and RSI at 28 signal potential squeeze.
  Q2 earnings in August could be the trigger.
  Stop-loss: $0.40
• SaverOne 2014 Ltd (SVRE) – $25 allocation (~25%)
  Developing anti-distracted driving technology.
  Potential Ford partnership expected by September 2025.
  Trading at 1.2x sales (vs. sector avg. of 8x).
  Stop-loss: $0.30

Risk Management:

• 20% stop-loss on each position.
• Momentum-based strategy tied to binary catalysts.
• Model uses only static historical data—no access to real-time developments.

Expected Return:

• Base Case: $140–$160
• Bull Case: $250+ (if KTTA or SVRE hits)
• Bear Case: $50–$70 (if catalysts flop or no triggers hit)

---

My Thoughts

I don’t believe LLMs today can truly discover hidden micro-cap gems on their own. But I do think they’re capable of making rational, informed decisions based on what’s already publicly available online. However, I know sometimes LLMs like to use fabricated data to make decisions.

In that context, I think a 10% to 20% return over six months is very reasonable if the models play it smart.

That said, I’m concerned about DeepSeek. Since it’s working purely off historical data, it may end up selecting micro-cap names that looked promising in the past but are now decaying in 2025.

For me, the most interesting part of this challenge isn’t just who wins, it’s how far ahead ChatGPT ends up compared to DeepSeek. If GPT-4o can convincingly outperform a strong historical-data-only model, it says a lot about how real-time research capabilities will define the future of AI-driven investing.

Final Notes

Because of the wide spreads of microcaps, it’s likely the prices will be different than what the model expects. However, of course I will consult with each model for compromises.

I will be updating every week on Friday so be sure to check in to see how each model is performing!

Note: this experiment is no longer active. I recently had a better idea to put ChatGPT and DeepSeek against eachother with the same concept. Check it out on my profile!

AI Trading Experiment – Daily Update (6/25/2025)

Status: Orders Executed — The Experiment Is Now Live

Due to wide bid-ask spreads common in micro-cap stocks, the original limit orders weren’t realistic. After re-evaluating with the AI, we adjusted entries to reflect actual liquidity, and all orders were successfully executed on 6/25 using real capital.

Executed Positions:

• MYO – Myomo Inc.
  14 shares @ $2.37 → Total: $33.18
  Close: $2.30 (−2.54%) → Value: $32.20
• CADL – Candel Therapeutics
  5 shares @ $4.93 → Total: $24.65
  Close: $4.85 (−1.62%) → Value: $24.25
• CPSH – CPS Technologies
  10 shares @ $3.02 → Total: $30.20
  Close: $2.89 (−4.30%) → Value: $28.90

Portfolio Summary:

• Total Invested: $88.03
• Current Portfolio Value: $85.35
• Cash Remaining: $11.97
• Unrealized PnL: −$2.68

These are real filled positions using real money. The experiment is now officially underway.

AI Portfolio Reaction (Based on 6/25 close):

1. MYO – Myomo:
   Slight dip, no red flags. Volume and news were quiet.
   Decision: Hold
2. CADL – Candel:
   Held price in a weak biotech market. Shows relative strength.
   Decision: Hold
3. CPSH – CPS Technologies:
   Down 4.6%, but light volume — normal volatility for illiquid micro-caps.
   Decision: Hold

Stop-Loss Review:
No stops triggered. Each position still has 10–15% room before triggering stops. No adjustments needed.

Plan for 6/26:
No trades, no rebalances, no stop adjustments. The AI recommends letting the initial entries play out and watching volume for signs of abnormal moves.

Important Note:
After this first red day, I asked the AI whether it needed more than just closing prices to make smart decisions. It responded with the following checklist:

• Portfolio summary (shares, execution price, current price, cash)
• Volume vs average volume
• Relevant news or earnings updates
• Russell 2000 % change (as benchmark)
• Optional technicals: price/volume notes, bid/ask spread, RSI, moving averages

After giving it all that, the AI simply responded:

“Hold all positions.
No stop-loss changes.
Watch CPSH tomorrow — could be make-or-break day.
Cash stays parked for now. No new entry meets our standard.”

That’s it for today. No changes. Check back tomorrow for the next update!

Note: this experiment is no longer active. I recently had a better idea to put ChatGPT and DeepSeek against eachother with the same concept. Check it out on my profile!

Welcome to my real-money experiment: Can an LLM beat the market with just $100?

For the next 6 months, I’m testing whether a large language model like ChatGPT (4o to be specific) can generate real profit using only U.S.-listed micro-cap stocks. I’ve set strict rules:

• Only full-share positions
• Market cap under $300M
• $100 total budget
• 6-month time frame (6/23/25 – 12/23/25)

I’m giving the AI complete control over what to buy, how much, when to sell, and how to manage risk. This isn’t a paper-trading experiment — I’m using real money so I stay invested and accountable. I will only reach out to provide data and get decisions, nothing else.

I’ll post regular updates with price movements and ask the AI if any changes are needed.

Here’s the original prompt I gave:

“You are a professional-grade portfolio strategist. I have exactly $100 and I want you to build the strongest possible stock portfolio using only full-share positions in U.S.-listed micro-cap stocks (market cap under $300M). Your objective is to generate maximum return from today (6-23-25) to 6 months from now (12-23-25). This is your timeframe, you may not make any decisions after the end date. Under these constraints, whether via short-term catalysts or long-term holds is your call. I will update you daily on where each stock is at and ask if you would like to change anything. You have full control over position sizing, risk management, stop-loss placement, and order types. You may concentrate or diversify at will. Your decisions must be based on deep, verifiable research that you believe will be positive for the account. Now, use deep research and create your portfolio.”

And here’s what the AI came up with after about 10 minutes of deep research:

The full report is pretty long, (Full Reply: Response), but here’s a summary.

Portfolio:

MYO – Myomo Inc.
14 shares @ $3.00 → Limit Buy | Total $42
Medtech company with surging sales from Medicare expansion. Q2/Q3 earnings could confirm breakout growth.

CADL – Candel Therapeutics
5 shares @ $5.60 → Limit Buy | Total $28
Biotech with positive Phase 3 cancer data. Major survival results due Q4 — high upside if results hold.

CPSH – CPS Technologies
10 shares @ $3.00 → Limit Buy | Total $30
Industrial turnaround play. Record revenue, new defense contracts, and real profitability.

So far I haven’t bought anything, but I have put the orders in.

My initial thoughts are pretty positive. While I think the AI won’t be able to find truly hidden undervalued stocks, I think it will make pretty thought-out decisions about what’s already online. My estimate is that the total value will grow to $105 – $110 (5% – 10%). Considering the chosen assets are BioTech, much bigger swings either way are likely.

Stay tuned! Every day, I’ll update the prices and let the AI react: rebalancing, taking profit, or holding.

Let’s see if AI can be a hedge fund manager.