As developers, we love data and automation. So, I ran an experiment: what if we could use the analytical power of LLMs to build strategic investment portfolios?
I fed data from 15 carefully selected dividend-paying ETFs to 5 different AIs (Qwen, Gemini 2.5, ChatGPT, Deepseek, Claude 4.5) and gave them a clear mission: create 5 distinct portfolios for real-life financial goals.
The goals ranged from generating a sustainable income stream for early retirement (FIRE) to creating a high-yield portfolio for a young professional.
The results were fascinating. For instance, for the "Early Retirement" goal, most AIs leaned heavily on low-volatility, high-dividend ETFs like HDLV (Invesco S&P 500 High Dividend Low Volatility) and stable Euro High Yield bonds.
Here’s a snapshot of the portfolio suggested by Gemini 2.5 Pro for the "Anticipated Tranquility" objective:
- 30% JNKE (SPDR Bloomberg Euro High Yield Bond)
- 30% STHY (PIMCO US Short-Term High Yield)
- 25% HDLV (Invesco S&P 500 High Dividend Low Volatility)
- 15% GLDV (SPDR S&P Global Dividend Aristocrats)
The logic is clear: a solid base of bonds for steady income, complemented by high-quality, low-volatility stocks.
Each AI showed a unique "strategic personality." Some were more risk-averse, others more aggressive. This experiment isn't about replacing human judgment but enhancing it with powerful analytical tools.
If you're curious to see all 5 AI-generated portfolios and dive deeper into the strategic thinking behind them, check out the full analysis on my blog.
https://agconsulting.altervista.org/portafogli-di-etf-a-distribuzione-2025/
Top comments (0)