Designing Fund Recommendation Systems for Novice Investors: Insights from Portfolio Construction, Sentiment Analysis and Investor Suitability
Year: 2026
Type: article
Abstract: This study presents a financial application design and development process for novice investors which provides insights into portfolio management, financial sentiment and investor suitability. Aimed to compare different assets performance, eight portfolios are constructed, including equity, cryptocurrency, and combined asset groups using equal-weighting, minimum-variance, maximum-Sharpe ratio and risk-parity methods. Their performance is showed through annual return, volatility, Sharpe ratio, maximum drawdown and the growth of a once-dollar investment. The application also uses finVADER and tests whether sentiment tilt can improve portfolio performance. However, the sentiment tilt has a week improvement in portfolio performance, suggesting that sentiment is better used as a supporting indicator than primary driver of portfolio allocation. Overall, the findings suggest that fund recommendations for novice investors should consider both portfolio performance and investor risk tolerance when supporting informed investment decision.
DOI: 10.54097/kfmnss92
Author: Yu Shen
Institution: UNSW Sydney
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7237333404
