Optimal portfolio selection using a simple double-shrinkage selection rule

In the field of risk management, it is of great importance to obtain an efficient portfolio when market participants invest in a variety of assets. In this study, we propose a simple double-shrinkage portfolio selection rule to improve the out-of-sample performance of the portfolio. The double-shrinkage portfolio is obtained by a convex combination between highly structured covariance matrices and sample covariance matrix. Using various real datasets we show that the proposed portfolio strategy is found to be comparatively stable and yields higher values of Sharpe-ratio and lower values of conditional value at risk. Thus, the double-shrinkage selection rule improves the performances of the portfolios significantly.