A Comparative Study of Cumulative Prospect Theory and SVM Approach for Portfolio Optimization

This study investigates portfolio optimization through two distinct methodological frameworks: i) Cumulative Prospect Theory (CPT) integrated with Mean-Variance Theory (MVT) and ii) Support Vector Machine (SVM) with MVT. Empirical asset return data from 2018 to 2023, is used to implement the convex-concave utility structure of CPT with MVT-based portfolio optimization. Additionally, the SVM-based approach employs trading data collected from the Yahoo Finance library to dynamically forecast asset prices and thereby recommend MVT-based portfolio construction. Both methodologies are evaluated to understand their application in real world market situations, focusing on optimization strategies and their underlying concepts. The objective of this research is to demonstrate the adaptability of these techniques in portfolio management, providing practical insight into their implementation for informed investment decision making.