Nowcasting the Real GDP Growth of Rwanda

Abstract

The main contribution of this paper is to develop a new set of real Gross Domestic Product (GDP) nowcasting
tools, namely, the Bridge equations, Mixed Frequency Data Sampling (MIDAS) models and the combined forecasting
technique, and to compare their performance against the benchmark models currently used at the National Bank of
Rwanda (NBR), namely, the Autoregressive Moving Average (ARMA) models and the Dynamic Factors Model (DFM).
Our empirical findings indicate that all the three new nowcasting models outperform the benchmark models, with the
bridge equations taking the lead. We therefore recommend the inclusion of MIDAS, Bridge and combined forecasting
models as part of the GDP nowcasting system for the NBR, to complement the existing models as this can help to
improve the forecast accuracy.