Do Opinion Polls Move Stock Prices? Evidence from the US Presidential Election in 2016
Posted: 29 Jun 2020 Last revised: 12 Apr 2021
Date Written: April 08, 2021
This paper analyzes stock returns before and during the 2016 US presidential election. This election stands out because both nominees Clinton and Trump were frequently characterized to have very different profiles and the outcome was considered as a major surprise. We explore whether the stock market reaction on the election could have been predicted with polling data that has been released in the pre-election period. To measure the impact of polling data on stock prices, we develop the “Trump Head-to-Head Momentum” (THHM) reflecting increasing poll scores for Trump during election uncertainty. Implementing a GARCH-in-Mean model, we find that the measure is significantly related to industries representing about 30% of the market capitalization. Based on THHM, we identify active trading strategies that bet on the event “Trump wins the election” that outperform a passive benchmark strategy after the election.
Keywords: Presidential election uncertainty, Pre-election information, Opinion polls, Industry returns, GARCH-M, Market efficiency
JEL Classification: G12, G14
Suggested Citation: Suggested Citation