Efficient mean-variance portfolio selection by double regularization

This paper addresses the estimation issue that exists when estimating the traditional mean-variance portfolio. More precisely, the efficient mean-variance is estimated by a double regularization. These regularization techniques namely the ridge, the spectral cut-off, and Landweber-Fridman involve a regularization parameter or penalty term whose optimal value needs to be selected efficiently. A data-driven method has been proposed to select the tuning parameter. We show that the double regularized portfolio guarantees to investors the maximum expected return with the lowest risk. In empirical and Monte Carlo experiments, our double regularized rules are compared to several strategies, such as the traditional regularized portfolios, the new Lasso strategy of Ao, Yingying, and Zheng (2019), and the naive 1/N strategy in terms of in-sample and out-of-sample Sharpe ratio performance, and it is shown that our method yields significant Sharpe ratio improvements and a reduction in the expected utility loss.