Multitrend Conditional Value at Risk for Portfolio Optimization.

Area of Science:

• Quantitative Finance
• Machine Learning in Finance
• Risk Management

Background:

• Trend representation is increasingly used in portfolio optimization (PO) with machine learning.
• Current methods primarily focus on estimating asset returns, neglecting risk measurement.
• Existing approaches often rely on strict statistical assumptions or prior knowledge.

Purpose of the Study:

• To address the gap in using trend representation for risk measurement in PO.
• To propose a novel risk metric, multitrend conditional value at risk (MT-CVaR).
• To develop a new PO model incorporating MT-CVaR and an efficient solving algorithm.

Main Methods:

• Development of the multitrend conditional value at risk (MT-CVaR) metric.
• Formulation of a novel portfolio optimization model using MT-CVaR.
• Design of a solving algorithm based on the interior point method.

Main Results:

• MT-CVaR effectively embeds multiple trends and their influences into risk assessment.
• The proposed PO model with MT-CVaR demonstrates superior performance.
• Experiments show state-of-the-art results in investing performance and risk management across diverse datasets.

Conclusions:

• MT-CVaR offers a significant advancement in measuring risk for portfolio optimization.
• The proposed method provides enhanced investment performance and robust risk management.
• This approach is effective across various financial markets and data frequencies.