Research Seminar Series in Statistics and Mathematics

Wirtschaftsuniversität Wien, Departments 4 D4.4.00809:00 - 10:30

Art Vortrag/Diskussion
Vortragende/rRodney Strachan (School of Economics, University of Queensland, Australia)
Veranstalter Institut für Statistik und Mathematik
Kontakt katrin.artner@wu.ac.at

Rodney Stra­chan (School of Econo­mics, Univer­sity of Queens­land, Australia) about "Redu­cing Dimen­sions in a Large TVP-VAR"

The Insti­tute for Statis­tics and Mathe­ma­tics (Depart­ment of Finance, Accoun­ting and Statis­tics) cordi­ally invites ever­yone inte­rested to attend the talks in our Rese­arch Seminar Series, where inter­na­tio­nally renowned scho­lars from leading univer­si­ties present and discuss their (working) papers.

The list of talks for the winter term 2018/19 is avail­able via the follo­wing link: https://www.wu.ac.at/en/stat­math/resse­minar


This paper proposes a new approach to esti­ma­ting high dimen­sional time varying para­meter struc­tural vector auto­re­gres­sive models (TVP-S­VARs) by taking advan­tage of an empi­rical feature of TVP-(S)VARs. TVP-(S)VAR models are rarely used with more than 4-5 varia­bles. However recent work has shown the advan­tages of model­ling VARs with large numbers of varia­bles and inte­rest has natu­rally increased in model­ling large dimen­sional TVP-VARs. A feature that has not yet been utilized is that the cova­ri­ance matrix for the state equa­tion, when esti­mated freely, is often near singular. We propose a speci­fi­ca­tion that uses this singu­la­rity to develop a factor-­like struc­ture to esti­mate a TVP-SVAR for 15 varia­bles. Using a gene­ra­liza­tion of the recen­te­ring approach, a rank reduced state cova­ri­ance matrix and judi­cious para­meter expan­sions, we obtain effi­cient and simple compu­ta­tion of a high dimen­sional TVP- SVAR. An advan­tage of our approach is that we retain a formal infe­ren­tial frame­work such that we can propose formal infe­rence on impulse responses, vari­ance decom­po­si­tions and, important for our model, the rank of the state equa­tion cova­ri­ance matrix. We show clear empi­rical evidence in favour of our model and impro­ve­ments in esti­mates of impulse responses.

zurück zur Übersicht