Presidential Politics and Investor Behavior in the Stock Market: Evidence from a Century of Stock Market Data

This paper examines the role of political cycles on investor behavior in the U.S. stock market with a particular focus on herding behavior. Using long history of stock market data spanning from 1926 to 2022 and employing the return-based herding tests of Christie and Huang (1995) and Chang et al. (2000), we first show that herding is more pronounced during periods of high market volatility or crisis periods across all industry groups examined. While herding is primarily driven by fundamental information regarding the stock market, we also find evidence of intentional herding, particularly during crisis periods, due to non-fundamental market factors. More importantly, we establish a link between herd behavior and political cycles. We show that herding is more prevalent during Democratic administrations, in line with the risk-based model of Pastor and Veronesi (2020) that associates Democratic presidencies with high risk aversion and thus high risk premia in equities. Furthermore, we present evidence of spurious herding during periods of high approval of presidential economic policies, while intentional anti-herding is also present mostly for production-based industries during such periods, implying heterogeneous effects of partisan politics on industry herding. Our findings provide novel insight for the link between political dynamics and investor behavior with important implications for policy makers and investors.