Doubly Robust Mean-CVaR Portfolio

TL;DR

This paper introduces a Doubly Robust mean-CVaR portfolio optimization method that addresses instability issues in portfolio management, providing a more reliable approach under uncertain market conditions with proven theoretical and empirical advantages.

Contribution

It proposes a novel robust optimization framework for mean-CVaR portfolios that handles parameter uncertainty and estimation errors, formulated as a second-order cone programming problem.

Findings

Outperforms existing strategies in benchmark tests

Provides theoretical error bounds for finite samples

Demonstrates improved stability in volatile markets

Abstract

In this study, we address the challenge of portfolio optimization, a critical aspect of managing investment risks and maximizing returns. The mean-CVaR portfolio is considered a promising method due to today's unstable financial market crises like the COVID-19 pandemic. It incorporates expected returns into the CVaR, which considers the expected value of losses exceeding a specified probability level. However, the instability associated with the input parameter changes and estimation errors can deteriorate portfolio performance. Therefore in this study, we propose a Doubly Robust mean-CVaR Portfolio refined approach to the mean-CVaR portfolio optimization. Our method can solve the instability problem to simultaneously optimize the multiple levels of CVaRs and define uncertainty sets for the mean parameter to perform robust optimization. Theoretically, the proposed method can be formulated as a second-order cone programming problem which is the same formulation as traditional mean-variance portfolio optimization. In addition, we derive an estimation error bound of the proposed method for the finite-sample case. Finally, experiments with benchmark and real market data show that our proposed method exhibits better performance compared to existing portfolio optimization strategies.