Sparse portfolio optimization via ℓ1 over ℓ2 regularization

Zhongming Wu, Kexin Sun, Zhili Ge, Zhihua Allen-Zhao, Tieyong Zeng

Abstract: Highlights•An ℓ1 over ℓ2 regularization was introduced to promote sparse portfolio selection.•Financial insights regarding short positions and estimation errors were explored.•An ADMM-based solving method with convergence analysis was proposed.•An extension of the model with a more general regularization was discussed.•The in-sample and out-of-sample numerical experiments were conducted.