Presidential approval ratings and stock market performance in Latin America

Abstract

This paper examines the time-varying causality between presidential approval (PAR) and stock market performance, measured by stock returns and realised volatility, focussing on four prominent Latin American countries: Brazil, Chile, Colombia, and Mexico, from 1990M01 to 2016M05. We use a time-varying causality test, which is robust to structural breaks, to uncover a bidirectional causal relationship between PAR and stock market performance. Our results remain robust when controlling for macroeconomic conditions, PAR in other Latin American countries and US PAR. Our analysis is extended to include the bond market, identifying a significant bivariate causal relationship between PAR and bond market performance.