Forecasting Monthly Inflation: An Application To Suriname

An accurate forecast for inflation is mandatory in the conduction of monetary policy. This paper presents models that forecast monthly inflation utilizing various economic techniques for the economy of Suriname. The paper employs Autoregressive Integrated Moving Average models (ARIMA), Vector Autoregressive models (VAR), Factor Augmented Vector Autoregressive models (FAVAR), Bayesian Vector Autoregressive models (BVAR) and Vector Error Correction (VECM) models to model monthly inflation for Suriname over the period from 2004 to 2018. Consequently, the forecast performance of the models is evaluated by comparison of the Root Mean Square Error and the Mean Average Errors. We also conduct a pseudo out-of-sample forecasting exercise. The VECM yields the best results forecasting up to three months ahead, while thereafter, the FAVAR, which includes more economic information, outperforms the VECM, based on the assessment of the pseudo out-of-sample forecast performance of the models.