FAVAR (Factor-Augmented Vector Autoregression) Modeli Literatür Taraması

Abstract

In the Vector Autoregressive (VAR) models, which are widely used in economic studies and developed by Sims (1980), impulse response functions can only be obtained from variables included only because of the infrequent use of information sets, and the dimensions of structural shocks can not be measured precisely. It is also not possible that for some variables to be represented by a single time series. The VAR estimation is insufficient for parsing operations involving large data sets. FAVAR (Factor Augmented Vector Autoregression) method was developed by Bernanke, Boivin and Eliasz (2005) and this method can use large data sets. In this study, FAVAR method is tried to be explained by comparing with VAR, and a literature search is being conducted in this subject.