Robust Portfolio Selection with Regime Switching and Asymmetric Dependence

34 Pages Posted: 7 May 2020 Last revised: 15 Mar 2021

See all articles by Xiaoshan Su

Xiaoshan Su

Beihang University (BUAA)

Manying Bai

Beihang University (BUAA)

Yingwei Han

China University of Geosciences, Beijing

Date Written: April 30, 2020

Abstract

This paper solves the portfolio selection problem with regime switching and asymmetric dependence in financial markets. Investors sustain substantial loss in times of crisis and expect to reduce their losses. Thus, we consider the uncertainty in hidden states of the economy and define worst-case conditional value-at-risk (WCVaR) to capture extreme portfolio loss during financial crisis. Then, we formulate the portfolio selection problem with WCVaR as the measure of risk. We conduct an empirical study using 13 global equity indices. The results show that for dynamic investments, or during financial crisis, our model outperforms other models that only consider a fixed dependence structure between assets. This is because our model can significantly reduce extreme portfolio loss in times of crisis by selecting the assets with small lower tail dependence. This new portfolio strategy can help risk-averse investors cope with financial crisis.

Keywords: Robust portfolio decisions; Regime switching; Asymmetric dependence; Worst-case CVaR; R-vine copulas; Financial crisis.

JEL Classification: G01, G11, G15

Suggested Citation

Su, Xiaoshan and Bai, Manying and Han, Yingwei, Robust Portfolio Selection with Regime Switching and Asymmetric Dependence (April 30, 2020). Economic Modelling, Forthcoming, Available at SSRN: https://ssrn.com/abstract=3588967 or http://dx.doi.org/10.2139/ssrn.3588967

Xiaoshan Su (Contact Author)

Beihang University (BUAA) ( email )

37 Xue Yuan Road
Beijing 100083
China

Manying Bai

Beihang University (BUAA) ( email )

37 Xue Yuan Road
Beijing 100083
China

Yingwei Han

China University of Geosciences, Beijing ( email )

NO. 29, Xueyuan Road, Haidiao District
Beijing, 100083
China

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
40
Abstract Views
353
PlumX Metrics