Cross-validated covariance estimators for high-dimensional minimum-variance portfolios

The global minimum-variance portfolio is a typical choice for investors because of its simplicity and broad applicability. Although it requires only one input, namely the covariance matrix asset returns, estimating optimal solution remains challenge. In presence high-dimensionality in data, sample estimator becomes ill-conditioned leads to suboptimal portfolios out-of-sample. To address this issue, we review recently proposed efficient estimation methods extend literature by suggesting multi-fold cross-validation technique selecting necessary tuning parameters within each method. Conducting an extensive empirical analysis with four datasets based on S&P 500, show that data-driven specific improves out-of-sample performance portfolio. addition, identify estimators are strongly influenced parameter detect clear relationship between selection criterion evaluated measure.