PROYEKSI DATA PRODUK DOMESTIK BRUTO (PDB) DAN FOREIGN DIRECT INVESTMENT (FDI) MENGGUNAKAN VECTOR AUTOREGRESSIVE (VAR)

Abstract

Gross Domestic Product (GDP) and Foreign Direct Investment (FDI) is an economic instrument that has an attachment and is often used for economic development of a country. To predict these two variables there are several methods that can be used, one of which is a method of Vector Autoregressive (VAR). VAR method has some assumptions that the data to be foreseen must have an attachment, stationary in the mean and variance and the resulting error must meet the test of independence and normal distribution. In the early identification stage of trials conducted with respect Augmented Dickey Fuller (ADF) as a stationary test, granger causality test as a test of attachment between the variable and the value of AIC as a determinant of the optimal lag value. In the parameter estimation phase used Ordinary Least Square method (OLS). After getting the model it is necessary to do verification on condition that the residuals must meet the test of independence and multivariate normal test. With a second fulfillment verification tests it was determined the optimal model of VAR (4) as the model used for forecasting. Based on data of GDP and FDI were used as a case study, the results obtained projection of the next 5 years with values declining in 2015 and 2017 and then increased in 2018 for the variable FDI to GDP is increasing every year with an average rise of up to 200% Keywords: FDI, GDP, VAR, causality, independency, multivariate normal, R-Square