A0178
Title: Identification of possibly nonfundamental Structural VARMA models using higher order moments
Authors: Carlos Velasco - Universidad Carlos III de Madrid (Spain) [presenting]
Abstract: A frequency domain criterion is introduced to identify the parameters of, possibly noncausal and/or noninvertible, structural vector autoregressive moving average (SVARMA) models. We use information from higher order moments to achieve identification on the location of the roots of the VAR and VMA matrix polynomials for possibly non-fundamental non-Gaussian vector time series. This information also provides identification on the rotation of the model errors leading to the structural innovations up to sign and permutation. We develop general representations of the higher order spectral density arrays of vector linear processes and describe sufficient conditions for the parameter identification that rely on both sufficiently rich (linear) dynamics and the independence component structure of the vector of linear innovations. We generalize previous univariate analysis to develop efficient estimates exploiting linear and higher order dynamics.
