<![CDATA[This study aims to model the daily return volatility of Bank Syariah Indonesia (BSI) in order to understand the characteristics of investment risk in the Islamic banking sector following the merger. Given that financial data often exhibit heteroskedasticity (non-constant variance) and volatility clustering, the use of standard linear models tends to produce biased results. To address this, the study employs Autoregressive Conditional Heteroskedasticity (ARCH) and Generalized ARCH (GARCH) methods, utilizing 948 daily log-return observations. The analytical procedures include the Augmented Dickey-Fuller (ADF) stationarity test, the ARCH effect test, and model selection based on information criteria. The findings indicate that BSI stock returns are stationary at the level form. Diagnostic testing confirms the presence of ARCH effects. Based on the Log Likelihood and Akaike Information Criterion (AIC), the GARCH(1,1) model is identified as the best-fitting specification compared to ARCH(1) and GARCH(1,1)-MA. Parameter estimation reveals a high degree of volatility persistence (α+β=0.92), suggesting that market shocks exert long-term effects on the risk profile of BSI stock.]]>
Copyrights © 2025
