The Effectiveness of Multivariate Garch Models in Portfolio Selection in the Context of the Covid-19 Pandemic

Abstract

Uncertainty modeling is a critical aspect of quantitative finance, utilized in the three primary domains: portfolio allocation, risk management, and valuation of financial instruments. This modeling offers statistical characteristics of price fluctuations, therefore identifying methods to enhance predictions in the context of the COVID-19 Pandemic.This paper examines the efficacy of optimal portfolios through experiments utilizing two representatives multivariate GARCH models, CCC-GARCH and DCC-GARCH, within the framework of the COVID-19 epidemic in Vietnam. The methodology ofthe study was conducted based on data from the financial market in Vietnam from 2019 to 2023. The application of multivariate GARCH methods in portfolio selection continues to be a subject of considerable debate.The research findings indicate that multivariate GARCH models substantially exceed the broader market (VN-Index) across six portfolio performance criteria during the evaluation period. Only the CCC-GARCH model demonstrates more extraordinary performance than conventional estimate models, whereas the DCC-GARCH model exhibits inferior results. This may stem from constraints associated with portfolio turnover rates, which significantly influence transaction costs, and the comparatively prolonged computing time, hindering the actual execution of strategies. The unique contribution ofthe research contributions is anticipated to enhance investors' comprehension of the efficacy of multivariate GARCH models in portfolio selection within the Vietnamese financial market.