Frank Smith1, Ajay K. Aggarwal2
(1. Colorado State University, Fort Collins, CO 80523, USA; 2. Henderson State University, Arkadelphia, AR 71999, USA)

 

Abstract: Using a dataset of more than 70 industries over a 40 year presidential election cycle, we find that by observing who is ahead in the Gallop poll 90 days prior to a presidential election, positive economically significant abnormal returns can be earned. While other time frames are also studied, the consistency of who wins based on Gallop leads in excess of the margin of error of the poll can return annualized returns that exceed 80%.

 

Key words: presidential elections; election year returns

JEL codes: G1, M2