Analysis of Wholesale Price Indexes with the Vector Error Correction Model (VECM) Approach

Abstract

To obtain the wholesale price index (WPI) model and see the causal relationship between variables with the VECM approach based on the period January 2000 to August 2019. In this study, 3 variables are used; (i) Agriculture, (ii) Industry and (iii) Oil and Gas. This research begins with the stationarity test by looking at the plot and unit root test. When the variables are not stationary, a differencing is necessarily needed. The lag assessment is performed by observing at the smallest Aikake Information Criteria (AIC) and Schwarzt Information Criteria (SIC). A cointegration test using Johansen test is conducted to find whether there is a long-term relationship between WPI variables as a condition to proceed to VECM modeling. To see the dynamic behavior of the VECM model the Impulse Response Function (IRF) is used.    The best model obtained in this study is VECM(3) and based on the Granger causality analysis we obtained that Industry is affected by Oil and Gas, and Oil and Gas is affected by Industry and Agriculture.