Are conflict and uncertainty measures useful for macroeconomic nowcasting? An application for Latin America

• Activity Number: 345 - Advances in Macroeconomic Nowcasting and Forecasting: Role of Traditional and Nontraditional Indicators and Big Data
• Type: Topic-Contributed
• Date/Time: Thursday, August 12, 2021 : 10:00 AM to 11:50 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Javier J. Perez* and Hannes Mueller and Marina Diakonova and Luis Molina and Christopher Rauh
• Companies: Bank of Spain and Barcelona GSE and Bank of Spain and Bank of Spain and University of Cambridge
• Keywords: uncertainty; forecasting; nowcasting; latin america; political risk; conflict

Abstract

Political-policy uncertainty and social unrest affect macroeconomic developments. Having measurable indicators of such type of events may help anticipate economic shocks in real-time. We test such hypothesis for Colombia, Brazil, and Mexico. First, we assemble a database of “traditional” macro-financial monthly indicators and standard political risk variables, and nowcast quarterly GDP for the three countries. The modeling framework is the time series mixed-frequencies MIDAS approach. Next, we incorporate in the models text-based variables: (i) conflict variables; (ii) economic policy uncertainty (EPU) indicators; (iii) Google-trends-based, searches of social-political conflict. We show that policy-political and social unrest indicators are useful to nowcast country-specific macroeconomic variables in the selected Latin American economies.