DTA

Digital Theses Archive

 

Tesi etd-11032021-174742

Type of thesis
Corsi integrativi di II livello
Author
CALEMMA, CORRADO
URN
etd-11032021-174742
Title
On Higher Moments Identification for Financial VAR Models
Structure
Cl. Sc. Sociali - Scienze Economiche
Course
SCIENZE ECONOMICHE E MANAGERIALI - SCIENZE ECONOMICHE E MANAGERIALI
Committee
Relatore Prof. MONETA, ALESSIO
Tutor Prof. BOTTAZZI, GIULIO
Presidente Prof. BARONTINI, ROBERTO
Membro Prof. FREY, MARCO
Membro Prof. TURCHETTI, GIUSEPPE
Membro Prof. MOSCHELLA, DANIELE
Keywords
  • connectedness
  • Financial VAR
  • shock identification
Exam session start date
02/12/2021;
Availability
parziale
Abstract
Following recent literature about the caveats of using higher moments to identify financial VAR models – arguing that these techniques do not simply exploit more information in the data, but require stronger assumptions on the data generating process than conventional methods – we compare the results provided by a set of data-driven identification strategies employed within the structural version of the framework from Diebold and Yılmaz (2014) to measure financial spillovers among the US, the Eurozone, and selected emerging market economies. We provide economic identification by maximum shock contribution. While results obtained with daily financial data are almost all consistent with a recursive structure, spillover indices obtained with 5-day returns vary significantly across different methods when considering either total spillover indicators or country-specific ones, showing the need not only for strong identification of the models, but also for clear economic interpretations of the differences between similar identification methods1.
Files