Search for collections on Wintec Research Archive

Using data science techniques to create and maintain a globally diversified ETF-based investment portfolio


[thumbnail of Book of abstracts] PDF (Book of abstracts)
V3-Final-Book-of-Abstracts-National-ITP-Research-Symposium.pdf - Supplemental Material

Download (1MB)
[thumbnail of Powerpoint presentation] Microsoft PowerPoint (Powerpoint presentation)
ArthurValle_UsingDataScienceTechniquesCreateMaintainGloballyDiversifiedETFInvestmentPortfolio.pptx - Supplemental Material

Download (5MB)


Investing has been recognised as one way to generate wealth. Investors are particularly interested in chasing short-term top performance while they, in fact, should be concerned with selecting proper financial assets of different asset classes and allocate them to optimize long-term returns. This article explores the use of data science tools and techniques for creating and maintaining (via rebalancing) an investment portfolio that is statistically capable to provide the expect returns while being consistent to the investor profile.
In the article, we quantitatively analysed some of the exchange traded funds (ETFs) from the New Zealand Exchange (NZX) aiming of determining their historical performance and how it can predict future returns. After, we created a portfolio with the selected ETFs as well as an investing approach that would meet the expected performance while maintaining the risks relatively low. Back testing, machine learning, simulation and other data sciences techniques were used to analyse historical and forecasted performance of the portfolio and the potential outcomes of the associate investing strategy.

Item Type: Paper presented at a conference, workshop, or other event which was not published in the proceedings
Uncontrolled Keywords: Data science, investment portfolios
Subjects: H Social Sciences > HG Finance
Divisions: Schools > Centre for Business, Information Technology and Enterprise > School of Information Technology
Depositing User: Arthur Do Valle
Date Deposited: 14 Dec 2021 19:14
Last Modified: 21 Jul 2023 09:27

Actions (login required)

View Item
View Item