Please use this identifier to cite or link to this item:
|Title:||Financial portfolio optimization||Authors:||Chen, Nannan||Keywords:||Engineering::Electrical and electronic engineering::Computer hardware, software and systems||Issue Date:||2021||Publisher:||Nanyang Technological University||Source:||Chen, N. (2021). Financial portfolio optimization. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/149903||Abstract:||Financial markets provide platforms where businesses can gather funds from individual investors and investors can in turn gain from the growth of businesses. The objectives of an investment are always maximizing its return and minimizing its risk by allocating limited funds among a range of assets. This can be characterized as a multi-objective optimization problem (MOP). Multi-objective evolutionary algorithm (MOEA) is an effective tool to identify multiple Pareto-optimal solutions which represent best possible trade-offs for different risk-return preferences among different objectives. Each solution defines a portfolio. Investors with different risk tolerance can thus choose different portfolios to fit their needs. In addition, clustering technique is applied before MOEA to gather the assets with high correlation. The portfolio risk is thus reduced by holding combinations of assets from different clusters. In this report, I adapt the clustering technique and integrate into the framework of MOEAs to enhance the diversity of the evolutionary process. I can thus obtain Pareto-optimal solutions which are close to global optimums. As a result, empowered by the above combination techniques, investors can find portfolios that fulfil their investment requirements. Since each MOEA has its own advantages and disadvantages in a particular situation, I choose four MOEAs that are either representative or promising as a technique. Experimentally, I am particularly interested in stock markets since it is one major financial market. With extensive experiments on real-world datasets, I show the performance of each MOEA by analyzing each Pareto-optimal solution. The experimental framework developed in this report can be easily applied to other financial markets.||URI:||https://hdl.handle.net/10356/149903||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Student Reports (FYP/IA/PA/PI)|
Updated on May 18, 2022
Updated on May 18, 2022
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.