Forecasting the Romanian inflation rate: An Autoregressive Integrated Moving-Average (ARIMA) approach

Abstract

The primary objectives of this paper are to empirically create an univariate Autoregressive Integrated Moving-Average (model) using Box-Jenkins methodology to forecast Romanian inflation and inspect the prediction performance of the estimated model between October 2021 and October 2022. This study uses Ordinary Least Squares (OLS) technique for estimation purposes. On the foundation of different selection assessment and diagnostic criteria, the best model is selected to predict inflation in Romania in the short-run. We find that ARIMA (7, 1, 1) model is a suitable one under model identification, parameters estimation, diagnostic checking, and inflation prediction. In-sample forecasting is performed and the estimated ARIMA model reasonably tracks the actual inflation in the sample period.

Keywords

• arima
• box-jenkins
• inflation forecasting
• time-series modelling
• national bank of romania (nbr)