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 daily closing 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 consistent alpha in the dangerous world of micro-cap stocks? Lets find out.
Overview
The week began with a drawdown as ATYR pulled back, followed by a flat trading day with little movement. Throughout the week, I placed a series of limit orders to gradually trim the position by selling 2 shares at a time. Only one of those orders executed, leaving the position at 10 shares and raising available cash to $23.77.
With the company’s key catalyst approaching, sentiment shifted sharply on Friday. ATYR surged 14% in a single session, which drove the overall portfolio up 8% in one day.
Adding to the anticipation, ATYR’s Implied Volatility Percentile Rank now sits at 100% over the past year (Market Chameleon), meaning options are pricing in the largest expected move seen all year. The company has guided topline Phase 3 results for “mid-September 2025”, and I personally expect the readout to come this week.
Performance Graph
Metrics:
[ Risk & Return ]
Max Drawdown: -7.11% on 2025-07-11
Sharpe Ratio (period): 1.3872
Sharpe Ratio (annualized): 2.8911
Sortino Ratio (period): 2.5010
Sortino Ratio (annualized): 5.2123
[ CAPM vs Benchmarks ]
Beta (daily) vs ^GSPC: 1.4233
Alpha (annualized) vs ^GSPC: 161.87%
R² (fit quality): 0.082 Obs: 52
Note: Short sample and/or low R² — alpha/beta may be unstable.
Portfolio Review
To see the full report: Click Here
Here was this week’s revaluation summary:
Exact Orders
Sell ATYR
- Shares: 4
- Order Type: Limit @ $6.00
- Time in Force: DAY
- Intended Exec. Date: 2025-09-15
- Special Instructions: Place at market open or early in session
- Stop Loss: N/A
- Rationale: Trim ~40% of position before Phase 3 data readout. Locks in ~15% profit while reducing binary event risk.
Buy FBIO
- Shares: 5
- Order Type: Limit @ $3.85
- Time in Force: DAY
- Intended Exec. Date: 2025-09-15
- Special Instructions: “Open” order only; seek fill near opening price
- Stop Loss: $2.00
- Rationale: Increase exposure ahead of FDA decision. $3.85 limit is just below prior close to capture a dip. Stop at $2.00 protects downside if approval fails. Target upside from CUTX-101 approval and PRV value, while keeping size modest.
Buy SNGX
- Shares: 5
- Order Type: Limit @ $3.10
- Time in Force: DAY
- Intended Exec. Date: 2025-09-15
- Special Instructions: Small lots only; avoid chasing above $3.10
- Stop Loss: $2.50
- Rationale: Initiating a speculative position near consolidation (~$3). Stop set below support to limit downside. Positioning for possible breakout on positive Phase 2a data in rare disease and low-float dynamics.
Thesis Review Summary
To wrap up, our portfolio strategy for Week 12 remains focused on catalyst-driven alpha, balanced by tightened risk controls. We have repositioned the portfolio for upcoming inflection points.
Core Holdings
aTyr Pharma (ATYR)
- Continues to be a cornerstone due to the transformative potential of efzofitimod’s Phase 3 results.
- Trimmed the stake to protect gains and reduce binary risk exposure.
- High conviction: if positive, ATYR could be a game-changer (large unmet market).
- Risk management: hedged downside to ensure portfolio outperformance vs. the S&P 500 even in a negative scenario.
Abeona Therapeutics (ABEO)
- Now a commercial-stage company with an FDA-approved therapy and >2 years of cash runway.
- Patiently holding unless technicals force an exit.
- Thesis: capitalize on RDEB gene therapy approval and potentially attract acquisition interest.
- Watching upcoming conference updates for launch traction.
- Along with ATYR, forms the “backbone” of the portfolio:
- ATYR → near-term catalyst.
- ABEO → medium-term commercial ramp.
Tactical Positions
Fortress Biotech (FBIO)
- Approaching binary FDA catalyst at month’s end.
- Increased exposure: approval + PRV sale could significantly boost value.
- A calculated, event-driven risk consistent with alpha mandate.
4D Molecular Therapeutics (FDMT)
- Already delivered on initial thesis (positive data, stock rebound).
- Now a longer-term bet on ophthalmology gene therapy.
- Strong balance sheet (~$417M cash) + fast-tracked Phase 3 trials.
- Acts as a “life jacket” in a sea of binary risks: intrinsic value + multiple shots on goal.
- Plan to hold through Q4 milestones unless fundamentals shift.
Opportunistic Bet
Soligenix (SNGX)
- New addition with exciting early data and tiny float.
- Small allocation = minimal risk, high upside optionality.
- Treated as an opportunistic trade (momentum play), not a long-term hold.
- Adds tactical “spice” to boost returns.
Performance Recap
- After 11 weeks, portfolio is up ~35% vs. S&P’s ~6%.
- The coming week is crucial:
- Successful aTyr readout → potential major upside.
- Failure scenario → cuts into outperformance, but portfolio still retains substantial lead due to:
- Gains already banked.
- Diversification across catalysts and risk tiers.
My Thoughts
I respect the choice to reduce ATYR’s position after its strong spike, and I also like that the proceeds are being redeployed into a more diversified mix of fundamental and momentum-driven plays. That said, I think the current stop-loss level on aTyr Pharma (set at $4.60, about ~24% below the current price) leaves too much downside risk, especially with the imminent binary catalyst. A tighter stop in the 10–15% range would better balance protection. Overall, though, I believe this is one of the strongest portfolio rebalances ChatGPT has made so far.
This project is purely educational and research-focused. Nothing here should be taken as financial advice. Full disclaimer: Here
GitHub Page and Email:
To see past deep research reports and summaries: Click Here
Past and present chats: 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.
If you have any suggestions or advice, Gmail is : [email protected]