We propose a methodology that can efficiently measure the Value-at-Risk (VaR) of large portfolios with time-varying volatilities and correlations by bringing together the established historical simulation framework and recent contributions to the Dynamic Factor Models literature. We find that...