Do Non-Professional Investors Influence Managers and Financial Analysts?

Abstract

With the advent of online social networks, individuals' sentiment and opinions about firms are more accessible and processable. These individuals are often non - professional investors (NPIs). This thesis investigates whether NPIs' sentiment and opinions are noise or information on two types of social networks-namely social media and crowdsourcing platforms, and whether their sentiment and opinions can help management and financial analysts predict the disclosures. The first objective of this thesis is to investigate whether managers consider NPI sentiment towards their firms in issuing optimistic earnings guidance to create desirable market reactions. I infer firm-level NPI sentiment from social media discussions on StockTwits, comprising over 17 million tweets of 118,685 users concerning 3,212 distinct firms between May 2008 and January 2017. Following Aboody et al. (2018), I adopt the overnight stock return to measure NPI reaction and find that NPI reaction to positive guidance is stronger when their sentiment is high and that managers are more likely to issue positive guidance at these times. This association between the likelihood of issuing positive guidance and NPI sentiment is stronger in firms in which NPIs have greater proportionate shareholdings and where managers' equity incentives are highly contingent on short-term stock price increases. The findings are consistent with managers opportunistically manipulating guidance to exploit NPI sentiment and contribute to research by opening a research avenue on the role of NPI sentiment in management decision making. The second objective of this thesis is to investigate whether financial analysts exploit NPIs' earnings expectations towards a firm to 'walk down' their forecasts with an aim to create beatable forecasts. I infer NPIs' earnings expectation from crowdsourced earnings estimates on Estimize, comprising 879,015 'street earnings' estimates submitted by 70,926 investors in the period from January 2012 through to September 2018. I document a positive association between analyst forecast revision and the change in investors' earnings expectations. I also find that the likelihood of analysts issuing forecasts that generate pessimistic errors is high when they revise their forecasts down. The findings are consistent with analysts opportunistically manipulating earnings forecasts to exploit investors' expectations of future earnings. This study extends the literature on how market forces constrain analysts' conflicts of interest by demonstrating that new crowdsourcing technologies disrupt the market for traditional earnings forecast providers. It also informs the regulators on how the sharing of opinions by NPIs on these public platforms may potentially expose themselves to exploitation by more sophisticated market participants who possess the processing power to harness these scattered pieces of information of enormous volume.