The Effect of Covariance Matrix Estimation on Portfolio Selection Process: The Application for Different Investment Horizons in ISE

44321

The expected return vector and covariance matrix estimation are important input on the portfolio selection process. In this study, the use of different techniques to estimate the covariance matrix have been focused on the research of impact on portfolio risk. Portfolio selection process is carried out the periods of 1986:01-2009:12. The daily closing price of all stocks listed on the ISE as data set in this period is used. Markowitz portfolio selection model is used as a model. Investment horizon of one day, a week, for fifteen days, one month and one-year periods were chosen. The sample covariance estimator with the shrinkage estimato rdeveloped by Ledoit and Wolf(2004) as covariance estimators is used. According to the research results, the ISE with a lower risk and more easily managed portfolio options have been able to achieve by shrinkage estimator of Ledoit and Wolf(2004).

311

Covariance estimation , portfolio selection , shrinkage estimator , sample covariance estimator , ınvestment horizon.

Ege Academic Review (EAR)

322

Ege Academic Review (EAR)