Performance comparison of multifractal techniques and artificial neural networks in the construction of investment portfolios

被引：0

作者：

de Oliveira, Alexandre Silva ^{[1
]}

Ceretta, Paulo Sergio

Albrecht, Peter ^{[2
,3
]}

机构：

[1] Fed Univ Pampa, Technol Ctr Alegrete CTA, BR-96400590 Alegrete, Brazil

[2] Fed Univ St Maria, Human Social Sci Ctr CCSH, BR-97105900 Santa Maria, Brazil

[3] Mendel Univ Brno, Fac Business & Econ, Dept Informat, Zemedelska 1, Brno 61300, Czech Republic

来源：

FINANCE RESEARCH LETTERS | 2023年 / 55卷

关键词：

Artificial neural networks; Asymmetric probability; Portfolio management; STOCK-MARKET EFFICIENCY; SELECTION; ASSETS;

D O I：

10.1016/j.frl.2023.103814

中图分类号：

F8 [财政、金融];

学科分类号：

0202 ;

摘要：

This work aims to compare the performance of the traditional portfolios of the S & P500, Markowitz, and Sharpe with the multifractal trend fluctuation portfolios (MF-DFA) and portfolios of artificial neural networks with Student's asymmetric probability classification (ANN-t). In this study, we use daily data for S & P500 stocks between January 18, 2018, and July 12, 2022, where we backtest return and risk metrics such as annual volatility, Value at Risk, Sharpe Ratio, Sortino Ratio, Beta, and Jensen & PRIME;s Alpha. For both return and risk, we obtain the results confirming that the ANN-t technique might indicate better investment entries, which contradicts the Efficient Market Hypothesis (EMH).

引用

页数：7