Sharia Stock Return Volatility Model through Bayesian MSGARCH with Asymmetric Effects and Data Structure Changes

Abstract

The Indonesian Sharia stock market has experienced significant growth over the past year, accompanied by an increase in market capitalization. However, high volatility remains a critical challenge for investors when deciding to invest in Sharia stocks. Modeling Sharia stock return volatility is essential to help investors minimize investment risks. This study aims to identify the most effective model for measuring volatility in the Sharia stock market by comparing classical models such as Generalized Autoregressive Conditional Heteroscedasticity (GARCH), asymmetric models including Exponential GARCH (EGARCH), Threshold GARCH (TGARCH), and Asymmetric Power GARCH (APGARCH), along with the Markov Switching GARCH (MSGARCH) and a Bayesian MSGARCH model that incorporates structural regime changes. The results indicate that the Bayesian MSGARCH model outperforms the other models in capturing the volatility of the Jakarta Islamic Index Sharia stock returns, achieving the lowest prediction error and improved accuracy in parameter estimation. Moreover, the study reveals that investment activities influence the volatility structure during periods of market appreciation and depreciation, with identifiable durations, thereby providing valuable insights for the formulation of effective investment strategies.