Abstract
Mutual funds are one of the most popular investment avenues in India, yet predicting their performance remains a challenge for both investors and fund managers. Traditional evaluation methods such as regression models, risk-adjusted ratios, and historical NAV analysis often fail to capture the dynamic and non-linear nature of financial markets. This study investigates the application of machine learning (ML) models—specifically Random Forest, XGBoost, and Neural Networks—in forecasting mutual fund returns, using a dataset of 100 equity mutual funds over the period 2015 to 2024. The performance of these models is compared with traditional approaches such as OLS regression and ARIMA. The results reveal that ML models significantly improve predictive accuracy, reducing forecast errors by more than 20 percent relative to conventional benchmarks. Portfolio simulations further demonstrate that ML-driven fund selection strategies deliver higher cumulative returns and superior Sharpe ratios, offering practical value for investors. Directional accuracy analysis shows that ML models correctly predict fund performance trends in more than 80 percent of cases, underscoring their robustness. The findings highlight the potential of machine learning to transform investment decision-making, enhance investor confidence, and unlock smarter investment opportunities in the digital era.
Keywords: Mutual Funds, Machine Learning, Predictive Analytics, Investment Decisions, Portfolio Management
References
Chen, Y., & He, J. (2021). Machine learning in mutual fund performance prediction: Evidence from neural networks. Journal of Financial Data Science, 3(2), 45–61. https://doi.org/10.3905/jfds.2021.1.067
Gupta, R., & Sharma, A. (2022). Artificial intelligence and predictive analytics in Indian mutual funds: An empirical review. International Journal of Finance & Economics, 27(4), 5672–5688. https://doi.org/10.1002/ijfe.2351
Narayan, P. K., & Singh, B. (2020). Predicting stock returns and mutual fund performance using machine learning methods. Economic Modelling, 86, 198–212. https://doi.org/10.1016/j.econmod.2019.11.001
Wermers, R. (2011). Performance evaluation of mutual funds, hedge funds, and institutional accounts. Annual Review of Financial Economics, 3(1), 537–574. https://doi.org/10.1146/annurev-financial-102710-144908
Choudhury, M., & Tripathy, N. (2021). Machine learning in financial markets: A review of applications in mutual funds and portfolio management. Asian Journal of Economics and Finance, 3(1), 23–35.
Jain, A., & Singhal, R. (2020). Predictive analytics in Indian mutual fund industry: Opportunities and challenges. Journal of Emerging Market Finance, 19(2), 189–210. https://doi.org/10.1177/0972652720907062
Yao, S., Wang, J., & Luo, X. (2022). Deep learning models for portfolio optimization and mutual fund forecasting. Finance Research Letters, 47, 102689. https://doi.org/10.1016/j.frl.2021.102689