Using Social Media to Identify the Effects of Congressional Viewpoints on Asset Prices

Abstract

This paper examines the extent to which individual politicians affect asset prices using a high-frequency identification approach. We exploit the regular flow of viewpoints contained in a large volume of tweets from members of US Congress. Congressional tweets targeting individual firms are collected and classified based on their tone. Supportive (critical) tweets increase (decrease) stock prices of the targeted firm in minutes around the tweet. The price response persists for several days, during which analysts revise their forecasts about the firm cash flows. Selected politician tweets linked to legislation affect the stock prices of firms in the same industry as the targeted firm. The timeline of politician viewpoints within a particular bill exhibits surges in relevant news that predict roll call votes months before the signing of the bill. We highlight how the social media accounts of politicians are a valuable source of political news.

Keywords: Asset pricing, political news, social media, high-frequency identification.

JEL Classification: D72, G14