In this paper, we propose a formal and unified statistical framework in order to help performing reliable nowcasts, short- and long-term forecasts of the real GDP growth in the Tunisian context. To do so, we use a set of available monthly macrofinancial data and takes into account the sovereign ratings assigned to Tunisia by the four biggest Credit Rating Agencies since 1994. These data are used to estimate an appropriate multivariate unobserved componants times series model. The empirical results clearly show that combining of macro-financial indicators and sovereign ratings data seems to produce valid and consistent latent factors which track well GDP growth realizations throughout the estimation period. The short- and long-term factor-based forecasts of the real GDP growth based on our econometric framework are also more accurate than those produced by other classical benchmark forecasting models. This framework, which we consider as a first step in setting up a real-time monitoring system for the macro-financial situation in Tunisia, already allows us to highlight the main flash indicators capturing the dynamics of the real GDP growth. The mixture of different types of data in a formal statistical framework thus allows to provide the real-time monitoring of current economic conditions in Tunisia. Overall, such findings can be insightful for policymakers by providing reliable and early analysis of the ongoing economic situation. As well, further information about the future real activity can help business entrepreneurs to plan better their business projects.

References are not ready for this file yet, please refer to reference from the PDF file

Keywords

Nowcasting; Forecasting; GDP growth; flash indicators; factors’ valtidity