I agree. Plus it's way too short a timeframe to evaluate any trading activity seriously.

But I still think the experiment is interesting because it gives us insight into how LLMs approach risk management, and what effects on that we can have with prompting.