%0 Journal Article
%@ ISSN: 0972-0871
%A Supandi, Epha Diana
%A Rosadi, Dedi
%A Abdurakhman, Abdurakhman
%D 2017
%F digilib:37242
%I Pushpa Publishing House, Allahabad, India
%J Far East Journal of Mathematical Sciences
%K Portfolio, error estimation, loss function
%N 3
%P 643-657
%T THE OPTIMAL PORTFOLIO WEIGHTS USING THE PROPORTIONAL TYPE ESTIMATORS
%U https://digilib.uin-suka.ac.id/id/eprint/37242/
%V 101
%X Error estimation in both the expected returns and the covariance matrix hamper the construction of optimal mean-variance portfolio model. In order to overcome this problem, we consider the class of proportional type estimators. The sensitivity of the proposed estimators to errors is measured by the expected loss function (the risk function). The simulation study is conducted when multivariate returns are normally distributed and serially independent. Furthermore, simulation study is complemented by an investigation of the ex post excess returns for empirical datasets, i.e., average, standard deviation, Sharpe ratio, and utility. It turns out that the unbiased proportional estimator and the maximum likelihood estimator are underperformed compared to “the dominant” estimator.
%Z Prof Dedi Rosadi, Ph.D. UGM. Yogyakarta. Indonesia  Dr. Abdurakhman, M.Si. UGM. Yogyakarta. Indonesia