Robust mean-variance portfolio through the weighted \(L^p\) depth function (Q827128)

The main contribution of this paper is the extension of statistical data depth-based estimators to financial portfolio selection as it was already done with some other well-known robust techniques. To start with, the authors recall the classical mean-variance portfolio model. Then they offer some background on the notion of statistical depth functions along with their main properties and the estimation of the mean and covariance matrix through the weighted \(L^p\)-depth function. Such approach has the advantage to be independent of parametric assumptions, and less sensitive to changes in the asset return distribution than traditional techniques. The exposition is concluded with an out-of-sample evaluation through simulated data and an application to a real data set.