Title: Sparse graphical modelling for minimum variance portfolios Authors:  Riccardo Riccobello - University of Trento (Italy)
Giovanni Bonaccolto - University of Padova (Italy)
Philipp Johannes Kremer - EBS Universitaet fuer Wirtschaft und Recht (Germany)
Malgorzata Bogdan - Lund University (Sweden)
Sandra Paterlini - University of Trento (Italy) [presenting]
Abstract: Graphical models have shown remarkable performance in uncovering the conditional dependence structure across a set of given variables. Two new graphical modelling approaches are introduced, called Gslope and Tslopeto, the portfolio selection literature for directly estimating the inverse of the covariance matrix, using the so-called Sorted L1-Penalized Estimator (SLOPE). Spanning Gaussian and non-Gaussian distributed data, the new methods directly acknowledge asset returns' empirically observed distributional characteristics. Extensive simulation analysis and real-world applications highlight the superiority of the new methods, especially with regard to clustering and stability characteristics, compared to state-of-the-art covariance matrix estimation techniques.