Robust Covariance Matrix Estimation and Portfolio Allocation: The Case of Non-Homogeneous Assets
Emmanuelle Jay, Thibault Soler, Jean Philippe Ovarlez, Philippe de Peretti, Christophe Chorro
Abstract
This paper presents how the most recent improvements made on covariance matrix estimation and model order selection can be applied to the portfolio optimization problem. Our study is based on the case of the Maximum Variety Portfolio and may be obviously extended to other classical frameworks with analogous results. We focus on the fact that the assets should preferably be classified in homogeneous groups before applying the proposed methodology which is to whiten the data before estimating the covariance matrix using the robust Tyler M-estimator and the Random Matrix Theory (RMT). The proposed procedure is applied and compared to standard techniques on real market data showing promising improvements.
BibTeX
@inproceedings{icassp2020_robustcovariance,
title = {Robust Covariance Matrix Estimation and Portfolio Allocation: The Case of Non-Homogeneous Assets},
author = {Emmanuelle Jay and Thibault Soler and Jean Philippe Ovarlez and Philippe de Peretti and Christophe Chorro},
booktitle = {ICASSP 2020},
year = {2020}
}