A1902
Title: Rolling window selection in FAR models with structural instabilities
Authors: Antoine Djogbenou - York University (Canada) [presenting]
Abstract: A theory for rolling window selection for generating out-of-sample forecasts is developed using factor-augmented regression (FAR) models in the presence of structural instabilities. It shows how a rolling window can be selected by minimizing the conditional mean square forecast error (MSFE) while accounting for factor estimation uncertainty. Because the conditional MSFE is unobserved and the factors are latent, a feasible version of the criterion is proposed and conditions are derived under which the new method is asymptotic loss efficient. A simulation experiment and an empirical application are used to document the performance of the procedure.
