On Portfolio Optimization: Imposing the Right Constraints

We develop a shrinkage theory based framework for determining optimal port-folio weight constraints for minimum-variance portfolios in presence of parameteruncertainty. We propose to impose the set of constraints that yields the opti-mal trade-o between sampling error reduction and bias for the variance-covariancematrix. Our empirical results for ve data sets show that our approach yields sig-nicantly lower variances and a consistently higher Sharpe ratios than the equallyor value weighted portfolio strategies.[...]