Forcaseting Stock Market Volatility with Support Vector Machine

Abstract

Support Vector Machine (SVM) is one type of machine learning that has been applied in various fields. In the financial field, although some of SVM’s applications have been studied, they remain largely unexplored, especially in Thailand. There are two main objectives in this study; the first is to evaluate the accuracy of the predicted SET Index’s volatility by using an SVM model; the second is to compare the accuracy of the SVM model’s prediction with GARCH-type models’ prediction. In this paper, the author applies the SVM method to the Stock Exchange of Thailand Index (SET index) by using adjusted close prices in the timeframe from 1st January 2010 to 31st December 2021. The comparison of RMSE between SVM models and GARCH-type models show that SVM models have more accuracy in volatility forecasting than GARCH-type models.