Application of Machine Learning Models for Predicting Mutual Fund Performance in India: A SmartPLS Structural Model Analysis

Abstract

In the last decade, the Indian mutual fund industry has seen a tremendous boom, driven by financial awareness, digital investment platforms, and systematic investment planning. There are, however, some behavioural and serviced related factors that continue to affect investor decisions in recommending and adoption of mutual funds. This study aims to explore the factors influencing the willingness to recommend the mutual funds in India using the structural model method with the help of SmartPLS.  This study has been conducted on a sample of 400 investors and potential investors from the Indian market. This research explores the impact of the satisfaction levels with customer service, perceived difference between mutual funds and traditional investments, and perceived level of risk on willingness to recommend the mutual fund. The analysis of the structural relationship of the variables was done using partial Least Squares Structural Equation Modelling (PLS-SEM) with the SmartPLS methodology. The results indicate that the attitude towards mutual funds relative to traditional investment significantly affects the recommendation intention, and customer service satisfaction has a moderately significant positive relationship. There is a non-significant correlation between perceived risk level and recommendations behaviour. This work adds to the literature of financial behavior and decision making on investments, by incorporating behavioral finance concepts and concepts of modern predictive modelling. The findings are valuable to financial institutions, mutual funds companies, policy makers and researchers to understand the investors' advocacy and enhance the marketing approaches of mutual funds in India.