Title:
Review of Time Series Forecasting Methods for Predicting Monetary Policy Impacts in Indonesia

Authors:
Erika Jimena Arilyn, Maris Agung Triandewo, Regi Muzio Ponziani*, Indonesia

Abstract:

This research aims to forecast monetary policy impacts in Indonesia. There are two measures of monetary policy impacts in this research, namely exchange rate and inflation rate. Exchange rate is the USD/IDR exchange rate. Both these measures are resulted from monetary policy taken by the central bank, Bank Indonesia. The data for the monetary policy are extended from January 2015 until April 2024. The data were divided into training and test data. Training data extend from January 2015 until December 2023. Test data extend from January 2024 until April 2024. Training data will be used to generate parameters and forecasts for January until April 2024. The forecasts will be compared with the test data to derive the performance. The forecasting models were ARIMA (1,1,0), ARIMA (1,1,1), ARIMA (1,1,2), ARIMA (2,1,1), ARIMA (2,1,2), and ARIMA (3,1,3). The results showed that for forecasting the exchange rate, ARIMA (1,1,0) and ARIMA (1,1,1) perform best. While for inflation, forecasting inflation, ARIMA (3,1,3) and ARIMA (2,1,2) perform best. This implies that exchange rate forecasting is more random walk in nature, because only autoregressive of order 1 is influential, for inflation, the autoregression order of 2 and 3 are still influential.

DOI: https://doi.org/10.51505/IJEBMR.2024.81005