Network Models to Improve Robot Advisory Portfolio Asset Allocation
25 Pages Posted: 18 Mar 2020
Date Written: March 22, 2019
Robot advisory services are rapidly expanding, responding to a growing interest people have in directly managing their savings. Robot advisors may reduce costs and improve the quality of the service, making user involvement more transparent. However, they may underestimate market risks, especially when highly correlated assets are being considered, leading to a mismatch between investors' expected and actual risk. The aim of the paper is to enhance robot advisory portfolio allocation, taking users' preference into account. In particular, we demonstrate how Random Matrix Theory and Network models can be combined to construct investment portfolios that provide lower risks with respect to standard Markovitz portfolios. To demonstrate the advantages of this approach we employ the observed returns of a large set of ETFs, which is representative of the financial products at the ground of the activity of robot advisors.
Keywords: Correlation networks, Network centrality, Portfolio optimization, Random Matrix theory
JEL Classification: C01, C32, C58, G21, G32
Suggested Citation: Suggested Citation