Herding and Google search queries in the Brazilian stock market

PurposeThis study investigates the presence of herding in the Brazilian stock market between 2012 and 2020 and associates it with the volume of searches on the Google platform.Design/methodology/approachFollowing methodologies are used to investigate the presence of herding: the Cross-Sectional Standard Deviation of Returns (CSSD), the Cross-Sectional Absolute Deviation (CSAD) and the Cross-Sectional Deviation of Asset Betas to the Market.FindingsMost of the models detected herding. In addition, there was a causal relationship between peaks in Google search volumes and the incidence of herding across the whole period, especially in 2015 and 2019.Originality/valueThis study suggests that confirmation bias influences investors' decisions to buy or sell assets.