Vector Autoregressive X (VARX) Modeling for Indonesian Macroeconomic Indicators and Handling Different Time Variations with Cubic Spline Interpolation

Abstrak

This study discusses data handling that has different time variations (for example, data available in quarterly form but the desired data is monthly) in this case the GDP variable in the quarter series, while the other five variables use monthly series, whereas in multivariate analysis the data condition must be the same, then an approach is taken to reduce monthly data from quarterly data using the interpolation method. Therefore, before conducting the VARX analysis the author interpolated GDP data from the quarter to monthly by interpolation. After the data is ready, VARX modeling of the exchange rate, economic growth (GDP), interest rates on Bank Indonesia Certificates (SBI), and inflation as endogenous variables and US interest rates (FFR) and US inflation as exogenous variables. The purpose of this study is to implement and evaluate the performance of Cubic Spline interpolation methods for time series data that have different time variations. Build VARX models and predict exchange rates, economic growth (GDP), SBI interest rates, and inflation based on US interest rates (FFR) and US inflation with the best models. Meanwhile, the interpolation method used by researchers to estimate the monthly value of the GDP variable based cubic spline interpolation. Based on the AIC value of the smallest VARX model obtained at 240.6668 so the best model obtained is the VARX (4.0) model.