| Popis: |
This paper examines the degree to which the benefits of optimal portfolio diversification are still available in the ex-ante sense, when attempts are made to control for estimation risk. We assess the ex-ante performance of applications of mean variance portfolio theory using past data as the inputs. Various portfolio diversification strategies are examined including the use of Bayes-Stein portfolios which control for estimation risk. The results confirm that strategies which control for estimation risk perform better in the ex-ante period than those which do not. In particular, the minimum variance portfolio out-performed all other strategies in the sample period. |
