Network models to improve robot advisory portfolios

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 asset allocation services, making user’s involvement more transparent. Against this background, there exists the possibility that robot advisors underestimate market risks, especially during crisis times, when high order interconnections arise. This may lead to a mismatch between investors’ expected and actual risk. The aim of this paper is to overcome this issue, taking into account not only investors’ risk preference but also their attitude towards interconnectdness. To achieve this aim, we combine random matrix theory with correlation networks and extend the Markowitz’ optimisation problem to a third dimension. To demonstrate the practical advantage of our proposed approach we employ daily returns of a large set of Exchange Traded Funds, which are representative of the financial products employed by robot-advisors.