Wpływ pandemii COVID-19 na zmienność i ryzyko ekstremalnych stóp zwrotu metali szlachetnych

Investors and researchers continue to question whether occurrences such as the COVID-19 pandemic significantly affect the level of investment risk, and whether pre-pandemic risk models can also be applied to the analysis of financial market returns during this crisis. The aim of the study presented in this article is to check whether a GARCH volatility model with a heavy-tailed conditional error distribution is adequate for the measurement of the risk levels of extreme returns rates of three precious metals, i.e. gold, silver and platinum, during the COVID-19 pandemic. The analysis is based on daily data from the London Metal Exchange for two periods: the pre-pandemic phase (2nd January 2018–11th March 2020) and the COVID-19 pandemic (12th March 2020–17th May 2022). The study employs GARCH(1,1) models with three error distributions, namely the skewed Student’s t, the skewed GED and the ?-stable distributions. Extreme risk measures have been estimated using Value-at-Risk (VaR) and Expected Shortfall (ES), and their accuracy evaluated through Kupiec and Acerbi-Szekely’s tests. The results indicate that during the COVID-19 pandemic, the risk of extreme losses increased for all the analysed metals, but the volatility was the highest in the case of silver and platinum. The volatility of the analysed assets was best captured by GARCH models with the ?-stable distribution. However, the risk of extreme return rates of precious metals was influenced not only by the effects of the pandemic, but other factors as well, such as the geopolitical situation following Russia’s aggression against Ukraine.