CMStatistics 2017: Start Registration
View Submission - CFE
A1041
Title: Nowcasting Indian GDP Authors:  Daniela Bragoli - Università Cattolica Milano (Italy) [presenting]
Jack Fosten - University of East Anglia (United Kingdom)
Abstract: A nowcasting model is proposed for the Indian real GDP growth rate which uses the flow of relevant information to update predictions on a daily basis. Based on a dynamic factor model, our method can serve as a timely barometer to track the Indian development process. There are several challenges faced when nowcasting GDP in developing economies such as India. The first challenge is to proxy important missing variables such as international trade in the service sector. Our novel solution augments a baseline model with series on US and Euro-area output which improves predictions, particularly during the 2008-2009 global crisis. The second challenge is the impact of sizeable revisions to the GDP data. We construct a new series for real-time Indian GDP using press releases from the Central Statistics Office (CSO), finding that data revisions have a non-trivial influence on our results. Therefore, caution should be taken when evaluating predictions using the preliminary GDP release.