Support Vector Machine-Based GARCH-type Models: Evidence from ASEAN-5 Stock Markets

Abstract

© 2020, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG. Support Vector Machine (SVM) is a semiparametric tool for regression estimation. We will use this tool to estimate the parameters of GARCH models for predicting the conditional volatility of the ASEAN-5 stock market returns. In this study, we aim at comparing the forecasting performance between the Support Vector Machine-based GARCH model and the Maximum likelihood estimation based GARCH model. Four GARCH-type models are considered, namely ARCH, GARCH, EGARCH and GJR-GARCH. The comparison is based on the Mean Absolute Error (MAE), the Mean Squared Error (MSE), and the Root Mean Squared Error (RMSE). The results show that the stock market volatilities of Thailand and Singapore are well forecasted by Support Vector Machine-based-GJR-GARCH model. For the stock market of Malaysia, Indonesia and the Philippines, the Support Vector Machine-based-ARCH model beats all parametric models for all performance comparison criteria.