<![CDATA[Inflation is a key macroeconomic indicator that reflects price stability and forms the basis for monetary policy formulation. This study aims to model and evaluate the ability to forecast Indonesia's inflation rate using monthly data from 2015 to 2025 with the Augmented Integrated Moving Average (ARIMA) approach. The ARIMA method is used to capture the dynamics of the inflation time series and identify the existence of inflation inertia. The analysis stages include stationarity testing, model identification based on the Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) patterns, parameter estimation, and residual diagnostic and forecasting accuracy evaluation. The analysis results show that inflation data is stationary at a level that does not require differencing. The best model obtained is ARIMA (1,0,0) with a statistically significant first-order autoregressive component. Model performance evaluation using Mean Absolute Percentage Error (MAPE) yields a value of 15.37%, indicating that the model has a fairly good forecasting accuracy. Empirically, these findings support the theory of inflation inertia, in which previous periods of inflation play an important role in explaining current inflation. The results of this study show that the simple ARIMA model is still relevant and effective for short-term inflation forecasting in Indonesia, especially in conditions of relatively stable monetary policy.]]>
Copyrights © 2025
