Research on regularized mean-variance portfolio selection strategy with modified Roy safety-first principle

Area of Science:

• Quantitative Finance
• Financial Risk Management
• Portfolio Optimization

Background:

• Traditional portfolio optimization often relies on variance alone, potentially overlooking downside risk.
• The safety-first principle offers a valuable alternative for risk-averse investors.
• Integrating these approaches can lead to more robust investment strategies.

Purpose of the Study:

• To develop a consolidated risk measure for portfolio optimization.
• To enhance the safety-first principle for improved financial portfolio selection.
• To create stable and sparse portfolios using norm regularization.

Main Methods:

• A novel mean-risk portfolio optimization framework integrating variance and a modified safety-first principle.
• Application of norm regularization techniques to achieve portfolio stability and sparsity.
• Comparative analysis of the proposed model's cumulative wealth against the S&P 500 index.

Main Results:

• The proposed portfolio strategies exhibit superior out-of-sample performance compared to existing literature benchmarks.
• The model effectively controls the downside risk of portfolio returns.
• Norm regularization successfully yields near-optimal, stable, and sparse portfolios.

Conclusions:

• The consolidated risk measure and enhanced safety-first principle provide a robust framework for portfolio optimization.
• The proposed models offer a competitive alternative to traditional methods, particularly in managing downside risk.
• The study highlights the benefits of norm regularization for creating efficient and stable investment portfolios.