The article presents a method to estimate the risk of extreme financial losses in a volatile market. By combining GARCH models and extreme value theory, the researchers accurately predict Value at Risk (VaR) and expected shortfalls. Their approach outperforms traditional methods by considering heavy-tailed innovations and stochastic volatility. Through backtesting, they show that their procedure provides better 1-day estimates. Additionally, using Monte Carlo simulations, they find that their models are more accurate for estimating future returns over multiple days compared to simpler methods.