Speaker: Federico Carlini, Lugano USI
Title: Vector Autoregressive Model with Dynamic Factor
Abstract: We study a Vector Autoregressive model of order one augmented by unobservable factors with a dynamic described by a vector autoregression of order one, proposed by e.g. Darolles, Dubecq and Gourieroux (2014). A detailed discussion of diff erent identi cation strategies is provided. We estimate the model parameters by means of a 3-step procedure such that each step is in closed-form (up to eigenvalues-eigenvectors decomposition of matrices of small dimension). We study the asymptotic and fi nite-sample properties of the estimators and of a rank test for the number of unobservable factors.