An Analysis of Machine Learning Risk Factors and Risk Parity Portfolio Optimization

Many strategies are used to optimize portfolios to achieve high-profit returns while minimizing risk. Portfolio building is critical to achieving this aim, and risk budgeting significantly influences portfolio optimization techniques. This research looks at how machine learning and factor-based portfolio optimization may create risk parity portfolios with risk factor contributions. In terms of risk reduction, auto-encoder neural networks and Minimum-variance portfolios with latent components outperform more straightforward benchmarks. Finally, the Mean Conditional Value at Risk (Mean-CVaR) is utilized to maximize returns and minimize risks. The findings show that our method works in a real-world scenario of constructing an essential resource distribution based on financial components for an annuity reserve with risk constraints. In times of high volatility or tail risk, these effects are amplified for financial supporters