VECTOR AUTOREGRESSIVE STABILITY CONDITION CHECK UNTUK PEMODELAN DAN PREDIKSI SUMBER PENERIMAAN PABEAN BELAWAN

Abstract

Customs Intermediate are an institution that is responsible for regulating the flow of export and import trade activities in the Customs Area with the revenue coming from import duties and export duties. The time series data from the customs acceptance component import dan export which have a relationship between variables. Vector Autoregressive is a statistical method used in predicting and evaluating interrelationships between variables. The purpose of this study is to obtain a model for predicting import and export by using the VAR model and detecting the stability of the model. Model requirements are said to be stable if all modulus values from roots characteristic of coefficient matrices ≤ 1 that the predicted results can be verified. The data is divided into in sample data starting from January 2010 to June 2018 and out sample data starts from July 2018 until December 2018. The results of data analysis in this study, the model obtained for prediction is the VAR model (4) and there is a direct relationship between both variables. The VAR (4) residual model fulfills the assumption of white noise, while the assumption of multivariate normality is not fulfilled. Based on out sample the value of MAPE for import variables 18.42%, export 12.94% shows the VAR model (4) has good predictive capabilities that can be used for predicting future periods. Predicted results on import show fluctuations during the period of January to December 2019 while in the export shows increase during the period of January to December 2019.