<![CDATA[This study examines the persistence and conditional uncertainty of Indonesian monthly inflation using an ARIMA(2,0,0)-ARCH(1) model and frames the evidence within Islamic economics, maqashid al-sharia, and sharia-compliant risk management. Monthly inflation data from January 2006 to December 2025 are used, with 2006-2024 as the estimation sample and 2025 as the out-of-sample evaluation period. The Augmented Dickey-Fuller test confirms level stationarity. Among six non-seasonal specifications, ARIMA(2,0,0) is selected because both autoregressive coefficients are significant, the inverse roots are stable, and the AIC is the lowest. ARCH-LM confirms conditional heteroskedasticity, while ARCH(1) removes the remaining first-order ARCH effect. A Student-t distribution improves information criteria and captures fat-tailed price shocks. The 2025 evaluation reports RMSE=0,663449, MAE=0,489019, and Theil U2=0,600060, indicating that the model outperforms a naive benchmark. The Islamic economic contribution lies in interpreting inflation volatility as a risk to hifz al-mal, the real value of contracts, basic consumption, halal business continuity, Islamic banking risk, and zakat-waqf planning. Thus, the model is not merely a forecasting device, but an early-warning instrument for maslahah-oriented price stabilization.]]>
Copyrights © 2026
