Estimation of the Dynamic Conditional Correlation Model: Two-Step or One-Step Matters?

To understand the potential consequence of the conventional two-step estimation procedure in the dynamic conditional correlation (DCC) model proposed by Engle (2002) on portfolio risk forecasting effectiveness, we develop a novel one-step estimation procedure and compare them. Our one-step approach addresses this issue by simultaneously estimating all the parameters of volatility and correlation in the DCC model. We provide evidence of the economic benefit of this approach which simultaneously estimates both volatilities and correlations in portfolio risk forecasting.