Large-Dimensional Portfolio Selection with a High-Frequency-Based Dynamic Factor Model
46 Pages Posted: 11 Mar 2020 Last revised: 15 Nov 2022
Date Written: November 11, 2022
Abstract
This paper proposes a new predictive model for large-dimensional realized covariance
matrices. Using high-frequency data, we construct daily realized covariance matrices
for the constituents of the S&P 500 Index and a set of observable factors. Using
a standard decomposition of the joint covariance matrix, we express the covariance
matrix of the individual assets similar to a dynamic factor model. To construct a
conditional covariance model, we suggest to model the components of the covariance
structure using a series of autoregressive time series regressions. A novel feature of
the model, is the use of the data-driven hierarchical clustering algorithm to determine
the structure of the idiosyncratic covariance matrix. In an out-of-sample portfolio
selection exercise, we find that the proposed model outperforms other commonly used
multivariate volatility models in extant literature.
Keywords: Big data, Hierarchical clustering, High-frequency data, Minimum variance portfolio, Multivariate volatility
JEL Classification: C55, C58, G11
Suggested Citation: Suggested Citation