Modeling Inflation and Money Supply using Spatial Vector Autoregressive Model with Calendar Variation: Restricted vs Non-restricted Coefficient

Abstract

One of the tasks of the government is to maintain price stability reflected in the stability of inflation through the regulation of the money supply. Therefore the government needs to know the forecast of the inflation rate and the money supply. In modeling inflation and the money supply simultaneously in Indonesia, three things need to be accommodated, namely the relationship between variables, the existence of space-time relationships and the effect of Eid al-Fitr. The spatial vector autoregressive model with calendars variation can accommodate these three things. The purpose of this study is to compare two types of spatial vector autoregressive models with calendar variations, namely restricted and non-restricted coefficient to model inflation and the money supply in Surabaya, Malang, Kediri, and Jember simultaneously. Results of this study indicate that the non-restricted spatial vector autoregressive model with calendar variation is better than the restricted one. This can be seen from the value of MSE of the non-restricted model that is smaller than the restricted model.

Inflation dynamics in the USA and EU: VAR analysis and forecasting

“…Another variant of the VAR model that has been used to model inflation and money supply is the Spatial Vector Autoregression (SVAR) model. Nonrestricted SVAR models have been found to have better predictive power than restricted ones (Sumarminingsih, Suharsono, Ruchjana, & Stiawan, 2019). Additionally, research using VAR has been conducted to investigate the explanatory power of bank credit on inflation rates in Jordan, and it was concluded that there is a positive effect of bank credit on the inflation rate (Al-Oshaibat & Banikhalid, 2019).…”