Flexible shrinkage in portfolio selection

How to quantify estimation risk is important in portfolio selection. For this purpose we derive the flexible shrinkage estimator for the optimal portfolio weights, which allows dynamic adjustments of model structure. Our estimator is based on grouping the assets in order to capture non-homogeneity of estimation risk. The assets are assigned to groups using a clustering procedure with the number of groups determined from the data. The proposed flexible shrinkage approach exhibits sound and robust performance compared to the popular portfolio selection alternatives.