Author(s):
Joshua C.C. Chan, Economics Discipline Group, UTS Business School, University of Technology, Sydney. Eric Eisenstat, University of Queensland. Chenghan Hou, Hunan University. Gary Koop,University of Strathclyde
Date of publication: May 2018
Working paper number: 44
Abstract: Adding multivariate stochastic volatility of a flexible form to
large Vector Autoregressions (VARs) involving over a hundred variables has
proved challenging due to computational considerations and over-parameterization
concerns. The existing literature either works with homoskedastic models or
smaller models with restrictive forms for the stochastic volatility. In this pa-
per, we develop composite likelihood methods for large VARs with multivariate
stochastic volatility. These involve estimating large numbers of parsimonious
models and then taking a weighted average across these models. We discuss
various schemes for choosing the weights. In our empirical work involving VARs
of up to 196 variables, we show that composite likelihood methods have sim-
ilar properties to existing alternatives used with small data sets in that they
estimate the multivariate stochastic volatility in a exible and realistic manner
and they forecast comparably. In very high dimensional VARs, they are com-
putationally feasible where other approaches involving stochastic volatility are
not and produce superior forecasts than natural conjugate prior homoskedastic
VARs.