A hybrid two-stage robustness approach to portfolio construction under uncertainty

This paper proposes a hybrid two-stage robustness approach to portfolio construction under data uncertainty. In the first stage, a stock's efficiency performance from candidate stocks is assessed and selected using an integrated dynamic slack-based measure data envelopment analysis model. We discuss the stability of efficiency estimates using the leave-one-out method. In the second stage, a robust stable and scaled mean-variance-Entropic Value-at-Risk model is used to determine the optimal weights allocated to qualified stocks in the presence of proportional transaction costs. The proposed method reduces computational complexity, increases robustness, and provides a comprehensive evaluation of stocks under different financial decisions, thereby increasing conservatism in the investment process. We demonstrate the applicability of the proposed hybrid two-stage approach to stock data from the Shenzhen and Shanghai Stock Exchanges. Results show that with increasing required returns, the proposed method improves the capital amount for investment and lowers transaction costs at the expense of additional risk. The study concludes by comparing the computational performance of the proposed approach to that of existing methods.(c) 2022 The Author(s). Published by Elsevier B.V. on behalf of King Saud University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

This paper proposes a hybrid two-stage robustness approach for portfolio construction that evaluates the efficiency of candidate stocks using a dynamic slack-based measure data envelopment analysis model and determines optimal weights using a robust mean-variance-Entropic Value-at-Risk model. The method reduces computational complexity, increases robustness, and provides a comprehensive evaluation of stocks under different financial decisions.