Forecasting Economic Growth Based on Systemic Risk for a Sample of Emerging Countries: a MIDAS Approach with Almon Polynomial Lag Distribution

This study aims to study the effect of systemic risk on economic growth and provide GDP growth forecasts for five emerging countries. To do so, Mixed Data Sampling (MIDAS) models are used to incorporate the Emerging Market Bond Index (EMBI) and the Composite Leading Indicator (CLI) as explanatory variables. This methodology combines different time series frequencies and various weighting schemes and lag structures. The main empirical results suggest that MIDAS models provide accurate forecasts, in particular during periods of strong recessions, as in the subprime crisis. Moreover, the results show similarities with many previous studies on emerging countries, which demonstrate the accuracy of MIDAS models. This research provides a deeper understanding of the impact of systemic risk on economic growth and the financial factors that influence the latter, thereby contributing substantially to better economic policy design in developing countries.