Prediction of Equity Mutual Fund Performance in South Africa using Regression Modelling

Abstract

Studies have found that mutual funds tend to underperform the relevant benchmark index in the long term. The failure to outperform the index may be partly attributed to the fund managers’ lack of tools that may assist in providing reliable estimates of how the funds are likely perform. The study uses a sample of nine South African equity mutual funds to test the predictive ability of linear and support vector regression models in predicting future mutual fund performance. Fund-specific and macroeconomic variables were used as predictor variables for predicting the year-to- date performance of the fund . The results of the study indicate that based on RMSE, support vector regression outperforms the linear regression models in predicting mutual fund performance and fund-specific and macroeconomic factors may be considered as predictor variables for fund performance