| Abstract: | Given the growing need for managing financial risk, risk predictionplays an increasing role in banking and finance. In this studywe compare the out-of-sample performance of existing methodsand some new models for predicting value-at-risk (VaR) in aunivariate context. Using more than 30 years of the daily returndata on the NASDAQ Composite Index, we find that most approachesperform inadequately, although several models are acceptableunder current regulatory assessment rules for model adequacy.A hybrid method, combining a heavy-tailed generalized autoregressiveconditionally heteroskedastic (GARCH) filter with an extremevalue theory-based approach, performs best overall, closelyfollowed by a variant on a filtered historical simulation, anda new model based on heteroskedastic mixture distributions.Conditional autoregressive VaR (CAViaR) models perform inadequately,though an extension to a particular CAViaR model is shown tooutperform the others. |
