| |
TOPIC:
|
A MULTIVARIATE REALIZED GARCH MODEL
|
ABSTRACT
We propose a novel class of multivariate Realized GARCH models that utilize realized measures of volatility and correlations. The key property of the model is a convenient parametrization of the correlation matrix that requires no additional structure to ensure positive definiteness. The correlation matrix is characterized by a vector, that can vary freely in the real vector space. A more parsimonious structure is often desired in practice, in particularly in high dimensional systems, and the framework facilitates simple and intuitive dimension reductions. We apply the model to returns of nine assets and illustrate a dimension reduction that arises from a natural block equicorrelation structure. Interestingly, we find that the empirical distribution of the transformed realized correlations is approximately Gaussian.
Keywords: Financial volatility, realized GARCH, high frequency data, multivariate modeling.
JEL Codes: G11, G17, C32, C58.
Click here to view the paper.
Click here to view the CV.
|
|
|
PRESENTER
Peter Hansen
University of North Carolina
|
RESEARCH FIELDS
Econometrics
|
DATE:
12 March 2020 (Thursday)
|
TIME:
4pm - 5.30pm
|
VENUE:
Meeting Room 5.1, Level 5
School of Economics
Singapore Management University
90 Stamford Road
Singapore 178903
|
|
|
