The metal markets have become extremely competitive and highly volatile due to financial globalisation. Therefore, in this study, the Value-at-Risk (VaR) and expected shortfall (ES) are estimated for the metal markets. A two-stage dynamic extreme value theory (EVT) method has been adopted along with the GARCH (1, 1) model to identify the pre-specified threshold for the metal market by using high-frequency returns data of 15-minute intervals ranging from 1st January, 2018 to 24th September, 2021, which provides accurate information about metal market volatility and tail distribution. Moreover, the empirical findings confirm the presence of a high level of volatility persistence in the metal market, especially in the financial returns of gold. Furthermore, silver metal returns exhibit the highest VaR compared to other metals in the market. The empirical results could assist financial investors and portfolio managers to minimise and control the potential risk in the market.