eduzhai > Applied Sciences > Computer Science >

Identification and Estimation of SVARMA models with Independent and Non-Gaussian Inputs

  • KanKan
  • (0) Download
  • 20210425
  • Save

... pages left unread,continue reading

Document pages: 65 pages

Abstract: This paper analyzes identifiability properties of structural vectorautoregressive moving average (SVARMA) models driven by independent andnon-Gaussian shocks. It is well known, that SVARMA models driven by Gaussianerrors are not identified without imposing further identifying restrictions onthe parameters. Even in reduced form and assuming stability and invertibility,vector autoregressive moving average models are in general not identifiedwithout requiring certain parameter matrices to be non-singular. Independenceand non-Gaussianity of the shocks is used to show that they are identified upto permutations and scalings. In this way, typically imposed identifyingrestrictions are made testable. Furthermore, we introduce a maximum-likelihoodestimator of the non-Gaussian SVARMA model which is consistent andasymptotically normally distributed.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×