Cointegrated-Based Forecast of Long-Run Relationships

Abstract

The Vector Error Correction Model (VECM) and the Autoregressive Distributed Lag Model (ARDL) are used to estimate the cointegration in the case of the long-run relationship of quarterly GDP and Final Consumption in Romania during the period 1995 – 2019. The actual data of 2020 Q1 and Q2 were used to check the best model’s validity. The static and dynamic approaches of the ARDL model were used to forecast the Final Consumption for Q3 and Q4 of the year 2020. Applying the cointegration model shows the long-term relationship of GDP and Final Consumption, but also the effects of other factors, seen in the differences of Final Consumption from its Long-Run evolution, and comprised in the cointegrating terms.