Sparse Index Tracking: Simultaneous Asset Selection and Capital Allocation via $\ell_0$-Constrained Portfolio

Sparse index tracking is one of the prominent passive portfolio management strategies that construct a sparse portfolio to track a financial index. A sparse portfolio is desirable over a full portfolio in terms of transaction cost reduction and avoiding illiquid assets. To enforce the sparsity of the portfolio, conventional studies have proposed formulations based on $\ell_p$-norm regularizations as a continuous surrogate of the $\ell_0$-norm regularization. Although such formulations can be used to construct sparse portfolios, they are not easy to use in actual investments because parameter tuning to specify the exact upper bound on the number of assets in the portfolio is delicate and time-consuming. In this paper, we propose a new problem formulation of sparse index tracking using an $\ell_0$-norm constraint that enables easy control of the upper bound on the number of assets in the portfolio. In addition, our formulation allows the choice between portfolio sparsity and turnover sparsity constraints, which also reduces transaction costs by limiting the number of assets that are updated at each rebalancing. Furthermore, we develop an efficient algorithm for solving this problem based on a primal-dual splitting method. Finally, we illustrate the effectiveness of the proposed method through experiments on the S\&P500 and NASDAQ100 index datasets.