Several contributions have shown the statistical advantages of the Bayesian approach for estimating RSGARCH models (Ardia, 2008; Bauwens et al., 2010; Bauwens et al., in press). However, none has attempted to determine the economic and social advantages of the Bayesian approach. Our research project aims at filling this gap by providing risk managers and regulators with new methodologies for improving risk forecasts of their portfolios. As financial institutions invest in thousands of assets, models and estimation methods have to be tested on a large universe to reach relevant conclusions. Our study will thus be conducted on the basis of hundreds of stocks and asset classes worldwide.