Constructing an optimal portfolio on the Bulgarian stock market using hybrid genetic algorithm for pre- and post-COVID-19 periods

Abstract

In the aftermath of the COVID-19 pandemic, global financial markets have seen growing uncertainty and volatility and as a consequence, precise prediction of stock price trend has emerged to be extremely challenging. In this background, we propose two time frameworks wherein the Hybrid genetic algorithm (HGA) is used to set up an optimal portfolio included ten stocks traded in Bulgarian stock market during pre and post COVID-19 periods. The fitness function values of constructed HGA during pre- and post-COVID-19 periods were −7.194e[Formula: see text] and −7.014e[Formula: see text], respectively. The estimated nonzero portfolio weights during pre-COVID-19 period were ALCM (0.025), HNVK (0.253), HVAR (0.378), MSH (0.204), NEOH (0.038), and SFT (0.102) while during post-COVID-19 period were AGH (0.003), ALCM (0.015), HNVK (0.272), HVAR (0.460), MSH (0.142), NEOH (0.057), SFT (0.031), and SPH (0.021). The corresponding expected portfolio return and portfolio risk during pre-COVID-19 period were 9.825e[Formula: see text] and 7.163e[Formula: see text] while during post-COVID-19 period were 9.656e[Formula: see text] and 6.895e[Formula: see text], respectively.