Πλοήγηση ανά Συγγραφέα "Dedemadi, Maria"

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω

Τώρα δείχνει 1 - 1 από 1

Big data portfolio optimization & an application to the U.S. stock market
(2022-03-16) Dedemadi, Maria; Δεδεμάδη, Μαρία; Athens University of Economics and Business, Department of Economics; Antoniou, Fabio; Varthalitis, Petros; Dendramis, Yiannis
Choosing the best empirical method for optimizing a portfolio is challenging. Especially today multiple methods are developed to incorporate the big volume of financial data available. Under the traditional optimization technique proposed by Markowitz (1952), many computational inefficiencies occur when the sample covariance matrix is large, leading to estimators that carry a lot of error. In this thesis, we provide a detailed analysis of new complementary econometric approaches appropriate for big data optimization, designed to reduce estimation error, and enhance the portfolio’s performance. We also, present some indicative portfolio diversification strategies. After the theoretical analysis, we use a complex dataset containing high-dimensional data of daily stock values of the U.S. stock market. Constructing the mean-variance and the minimum variance portfolios, we perform an empirical application using three optimization approaches - the sample moments, the naïve (1/Ν) approach and the linear shrinkage estimation proposed by Ledoit & Wolf (2003) - comparing the estimated results. Finally, we show that the linear shrinkage technique outperforms the aforementioned ones, in the presence of big data leading to better portfolio performance.