HomeArticlesServicePriceAbout

A recent study explores ChatGPT’s ability to construct stock portfolios tailored to investors' risk preferences. The researchers asked ChatGPT to create both low-risk and high-risk portfolios using stocks from the S&P 500 and STOXX Europe 600 indices and compared their performance to benchmark indices. The results show that ChatGPT can effectively adjust portfolio risk and returns based on an investor’s risk appetite, but performance varies by market and GPT model version. Additionally, errors in stock selection and weight calculations highlight the need for human oversight.

Stock Selection Using ChatGPT:

• ChatGPT was asked to select stocks from the S&P 500 and STOXX Europe 600 based on risk profiles.

Weight Allocation:

• ChatGPT was instructed to assign stock weightings for a GPT-weighted portfolio and a equal-weighted portfolio.

Portfolio Optimization:

• Applied cardinality constraints (limiting the number of assets) to keep portfolios practical for real-world execution.

Performance Comparison:

• Benchmarked against S&P 500, STOXX Europe 600, Dow Jones, NASDAQ, and popular investment funds.

1. ChatGPT Can Construct Risk-Adjusted Portfolios

Low-Risk Portfolio:

• Focused on defensive sectors such as consumer staples, healthcare, and utilities, reflecting a conservative investment approach that provides stability during market downturns.

High-Risk Portfolio:

• Concentrated in growth-oriented sectors, including technology, consumer discretionary, and biotech, aligning with aggressive investment strategies that aim for higher returns but come with greater volatility.

Risk Customization Capability:

• ChatGPT successfully adjusted portfolio risk levels based on investor preferences, demonstrating its potential as a personalized investment assistant.

2. Market-Specific Performance Variability

• GPT-4o performed best for U.S. markets (S&P 500).

• GPT-4 showed higher returns in European markets (STOXX Europe 600).

• These results suggest that GPT models may be more effective in certain markets due to differences in training data and market dynamics.

3. ChatGPT’s Errors & Limitations

Incorrect Stock Inclusion:

• ChatGPT mistakenly included stocks (e.g., BioNTech, Sunrun Inc., Square Inc.) that were not actually in the S&P 500.

Weight Calculation Issues:

• Some portfolios had weighting errors, where the sum of weights did not equal 100% (e.g., totaling 98% instead of 100%).

Takeaway:

• Investors must validate ChatGPT’s stock selections and calculations before relying on them.

4. Portfolio Constraints for Practical Execution

• Cardinality Constraints Applied:

  • Limited the number of stocks to prevent over-diversification and excessive complexity.

  • Helped create realistic, actionable investment strategies.

The study confirms that ChatGPT can construct portfolios tailored to investors' risk profiles, adjusting for low-risk and high-risk preferences. However, GPT model performance varies by market, and errors in stock selection and weight calculations require human oversight. While ChatGPT can serve as a useful portfolio assistant, investors should not fully automate investment decisions without verifying AI-generated outputs.

[Compliance Note]

• All posts by Sellsmart are for informational purposes only. Final investment decisions should be made with careful judgment and at the investor’s own risk.

• The content of this post may be inaccurate, and any profits or losses resulting from trades are solely the responsibility of the investor.

• Core16 may hold positions in the stocks mentioned in this post and may buy or sell them at any time.