ABSTRACT
The main aim of this paper is to investigate the stylized facts associated with the volatility of precious metals before and during the COVID-19 pandemic using GARCH-type models. In particular, we employ an ARMA-GARCH, ARMA-EGARCH and ARMA-FIGARCH framework to account for volatility clustering, asymmetry and long memory in the volatility of gold, silver, platinum and palladium. Based on structural breaks, we divide the whole sample into sub-samples and find that the breakpoints occurred after the declaration of COVID-19 pandemic. Our results show a very distinct behaviour in the memory of the four metals before and during the crisis. While there is a moderate persistence in the full sample and in the pre-COVID-19 sub-period for the four metals, this effect vanishes after the crisis outburst. Positive asymmetric effects are also found in gold and silver volatilities, which intensify during COVID-19 phase. We ascribe this phenomenon to the hedge/safe-haven properties of these metals. By contrast, a diverse pattern is observed in the palladium and platinum volatilities, which display negative asymmetries before the pandemic, in tandem of financial markets. After the crisis, these metals show mixed evidence. Moreover, we argue that COVID-19 significantly affects the volatility of precious metals. © 2022 Elsevier B.V.