Genetic algorithms for portfolio selection problems with non-linear objectives

Pandari, Abbas Rezaei

doi:10.5897/ajbm11.2876

Cited by 5 publications

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, they found that each algorithm showed different advantages over different data sets, and they showed that all these algorithms can be applied for efficient portfolio selection. These techniques have also been used by Moral-Escudero, Ruiz-Torrubiano and Suarez (2006), Chang, Yang and Chang (2009), Özdemir (2011), Eshlaghy, Abdolahi, Moghadasi and Maatofi (2011), Gorgulho, Neves and Horta (2011), Chen, Mabu, and Hirasawa (2011), Woodside-Oriakhi, Lucas, and Beasley (2011, Bermúdez, Segura and Vercher (2012), Pandari, Azar and Shavazi (2012), Ackora-Prah, Gyamerah, Andam and Gyamfi (2014), Hsu (2014), Yakut andÇankal (2016), Zeren andBayğın (2015), Chen, Lin, Zeng, Xu, and Zhang (2017), Jalota and Thakur (2018), Sasaki, Laamrani, Yamashiro, Alehegn and Kamoyedji (2018), Garcia, Guijarro and Oliver (2018), Vasiani, Handari and Hertono (2020.…”

Section: Introductionmentioning

confidence: 99%

Untitled

2024

GJEB

View full text Add to dashboard Cite

One of the key concepts in finance is Markowitz's constrained meanvariance model, the number of assets to be included in the portfolio is restricted. The solution of this generalized problem, which belongs to the quadratic and integer programming problem class, as the number of dimensions increases, is difficult to obtain with standard methods. In this study, the simulated annealing (SA) algorithm, which is one of the local search-based meta-heuristic methods, was preferred. The developed SA algorithm was applied to the Hang-Seng benchmark data set, and the results were compared with pioneering studies. According to the experimental results, upon the performance of the algorithm was found to be sufficient, the SA algorithm was applied for the Borsa Istanbul 30 index. The results of the experiments based on the Markowitz mean-variance model demonstrate that, while more assets must be maintained at lower risk levels to converge an unconstrained efficient frontier and the number of assets needed to do so decreases as risk rises

show abstract