A higher order portfolio optimization model incorporating information entropy

Gonçalves, Guilherme de Sousa Bezerra; Tan, Yong

doi:10.1016/j.iswa.2022.200101

Cited by 8 publications

(1 citation statement)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Engaging higher moments for portfolio optimization introduces a heightened level of complexity compared to standard methods like the mean-variance approach. This complexity stems from the delicate balance between conflicting objectives; namely, investors strive to maximize expected returns and skewness (Arditti, F. D. (1975)), all the while minimizing variance and kurtosis (See Gonçalves, G. W. (2022). To tackle this complex portfolio challenge involving multiple goals, we need to employ certain numerical methods, as often there's no straightforward answer available.…”

mentioning

confidence: 99%

Python for financial engineers: Mastering four moments in portfolio management

Chafik,

Benslimane,

Boussedra

2024

RMd

View full text Add to dashboard Cite

This research conducts a comprehensive analysis aimed at optimizing portfolios comprising 14 stocks listed on the Moroccan stock exchange. Our journey culminates in the construction of portfolios that are meticulously designed to maximize returns while prudently managing risk. These portfolios are the result of an exhaustive Monte Carlo simulation that explored over three million unique portfolio combinations. The simulations take into account the skewness and kurtosis of the return distributions, offering investors a robust framework for decision-making. We collected historical data for theses 14 stocks on the Moroccan market exchange by accessing 5 years' worth of historical data from investing.com.We explore the concepts of Modern Portfolio Theory (MPT), which forms the backbone of our approach, and we employ the power of mathematics and Python programming to bring forth insights that can inform sound investment decisions. The primary focus of this study centers on the incorporation of higher statistical moments from the returns of key financial indices, with a particular emphasis on their skewness and kurtosis characteristics. To achieve this goal, various evaluative criteria derived from these statistical parameters are introduced and thoroughly investigated. Within this research framework, we confront a spectrum of optimization challenges, including the maximization of skewness, and minimization of kurtosis.

show abstract

mentioning

confidence: 99%