PENDEKATAN MODEL VECTOR AUTOREGRESSIVE (VAR) UNTUK MERAMALKAN FAKTOR-FAKTOR YANG MEMPENGARUHI INFLASI DI PROVINSI GORONTALO

Abstract

The vector autoregressive (VAR) model is a simultaneous equation modeling used to construct forecasting systems from interrelated time-series data. This study intends to predict factors that significantly influence inflation in the province of Gorontalo. Moreover, the data used in this study involved inflation data and factors that influence inflation every month in the province in the period of January 2009 - December 2018. The results of inflation forecasting in Gorontalo in 2019 show that at the beginning of 2019, the inflation was considered to be very low at around -0.48% to -0.40%. However, the inflation surged in March with -0.25% (the highest inflation rate). The percentage decreased to -0.30% and -0.33% in April and May. After the decline in April and May, in the middle of the year (June) inflation returned to -0.31% and did not experience a significant change until the end of the year, which was still in the range of -0.32%. The accuracy of the prediction results seen in the MAPE value from out sample data of variables Y1 to Y8 is on the average below 10%, indicating that VAR is a significant forecasting model.