The democratization of investment research and the informativeness of retail investor trading

Abstract

We study the effects of social media on the informativeness of retail trading. Our identification strategy exploits the editorial delay between report submission and publication on Seeking Alpha, a popular crowdsourced investment research platform. We find the ability of retail order imbalances to predict the cross-section of stock returns and cash-flow news increases sharply in the intraday post-publication window relative to the pre-publication window. The findings are robust to controlling for report tone and stronger for reports authored by more capable contributors. The evidence suggests that recent technology-enabled innovations in how individuals share information help retail investors become better informed.

Introduction

Investing has always been social, and a large literature highlights the influence of peers on investment decisions (e.g., Shiller and Pound 1989, Duflo and Saez 2002, 2003, Ivković and Weisbenner 2007, Ouimet and Tate 2020). In recent years, improvements in technology have greatly expanded the scope for sharing information. Retail investors have embraced finance social media sites as users as well as creators of content, discussing news events, sharing investment research, and debating investment strategies (Grennan and Michaely, 2020). While innovations in social media offer the potential for improved access to investment information, theory suggests that peer interactions may also exacerbate behavioral biases (Han et al., 2021).

Empirical evidence on the effects of social media on retail investors is limited. Heimer (2016), Cookson et al. (2020), and Chawla et al. (2017) suggest that social media intensifies behavioral biases and spreads stale news. On the other hand, several recent studies find evidence that certain types of social media can provide investment value (Chen et al., 2014; Jame et al., 2016; Bartov et al., 2018; Crawford et al., 2018), yet it is unclear the extent to which social media informs retail traders. In this article, we study a popular investor social media site, Seeking Alpha, to identify when individual investors produce and share investment research, and we examine whether these activities increase the informativeness of their trading.

The Seeking Alpha platform, which curates crowdsourced investment research from non-professional analysts, offers several features that make it a natural setting to examine this question. First of all, Seeking Alpha (SA) provides broader access to in-depth investment analysis than most other social media platforms.¹ Consistent with this view, SA research reports and the comments they engender have been shown to predict future stock returns and earnings surprises (Chen et al., 2014).²

Seeking Alpha research reports provide investment analysis rather than break news, and their publication process includes an editorial review to ensure quality.³ This review-induced publication delay permits us to separate the impact of SA research from earlier news events that may also influence trading. Specifically, we use the intraday window immediately after SA report publication to measure the level of social-network-induced trading, and we use the intraday window prior to publication (but after potential information events that may have influenced the report) to capture the counterfactual level of trading that would have occurred in the absence of the SA report. The review-induced delay injects an element of randomness into the intraday timing of publication, and consistent with our identifying assumption, we find no evidence that media articles, brokerage research, or earnings announcements systematically precede or follow SA research publications over intraday windows.

We begin our analysis by documenting that Seeking Alpha's investor-authored research caters to retail investor information demand. We analyze roughly 180,000 research reports discussing 4900 stocks and find that after controlling for other firm characteristics, SA coverage is higher among firms with low institutional ownership and greater breadth of ownership, whereas the opposite is true for brokerage research coverage. The research coverage evidence confirms Seeking Alpha's emphasis on providing an investment analysis platform for retail investors.

Our analysis points toward a causal relation between Seeking Alpha research and retail investor trading. We analyze retail trading using ten half-hour intraday event windows around Seeking Alpha report publication using trade and quote data from NYSE TAQ and the method of (Boehmer et al., 2020) (BJZZ) to identify retail investor trades. Our regression approach includes individual report fixed effects, which benchmark the post-publication intraday period to the pre-publication intraday period. The results indicate that retail trading is markedly higher after the publication of Seeking Alpha research. For example, aggregate retail trading in the first half-hour after Seeking Alpha report publication is 7.68% higher than in the half-hour before publication. Moreover, measures of report sentiment that predict future returns, such as report tone and contributors’ investment positions (Campbell et al., 2019; Chen et al., 2014), explain retail investor trade order imbalances in the post-publication period. In contrast, we find no evidence of an increase in retail trading or report-sentiment-driven order flows prior to report publication, which is inconsistent with retail investors reacting to unobserved information events. The evidence suggests that Seeking Alpha has a distinct influence on the intensity and direction of retail trading.

To assess the effect of Seeking Alpha research on the informativeness of retail investor trading, we compare the ability of aggregate retail order imbalances to predict the cross-section of five-day ahead stock returns in the period immediately before and after SA report publication.⁴ The estimates indicate SA research publication leads to more informed retail trading. For example, the increase in future returns predicted by a one standard deviation increase in post-publication retail order imbalances is 0.26 percentage points larger than that predicted by pre-publication retail order imbalances. We do not find any increase in the informativeness of retail order imbalances over the five pre-publication half-hour windows, which suggests that the documented post-publication increase is not the continuation of pre-event trend.

Several additional tests suggest that the relation between retail order imbalances and future returns following SA research is at least partially attributable to informed trading, rather than price pressure or liquidity provision. First, we do not find evidence that the documented post-publication return predictability reverses over the subsequent quarter, alleviating concerns about price pressure. Second, similar to Boehmer et al., 2020, we decompose retail order imbalances into a persistent component which captures price pressure, a contrarian component which captures liquidity provision, and a residual component which captures informed trading. We find that the residual (informed) component remains a highly significant predictor of returns. Finally, retail order imbalances’ ability to predict analyst earnings forecast revisions and traditional media sentiment over the subsequent five days strengthens in the five half-hours after SA research is published, further supporting the view that retail trades reveal fundamental information.

We observe thathe incremental information revealed by retail trades after Seeking Alpha research is largely orthogonal to the information revealed by SA research report tone and contributor investment position, consistent with retail investors actively gleaning valuable information rather than passively following opinions expressed by social media contributors. We hypothesize that higher quality research reports will lead to more informed trading, and we explore whether reports that are authored by more accomplished or capable contributors offer more opportunities for extracting valuable information. Consistent with our conjecture, we find that retail order imbalances predict future stock returns and cash flows news more convincingly after reports that receive more comments and those authored by contributors with strong academic backgrounds or a track record of impactful reports.⁵

Kogan et al. (2020), Mitts (2020), and Dyer and Kim (2021) find that a small percentage of Seeking Alpha research reports, identified ex ante as misleading or “fake”, distort market prices. Drawing on these studies, we identify fake reports as those that are posted anonymously or have a low textual authenticity score and investigate whether they affect retail trading differently. We find that fake reports influence retail trading intensity and direction similarly or more than non-fake reports. In addition, retail order imbalances after the publication of fake reports predict the cross-section of one-week returns but not five-week returns, whereas retail order imbalances after non-fake reports predict the cross-section of five-week returns more strongly than one-week returns. These results are consistent with a small subset of SA research inducing uninformed retail trading that pushes prices from fundamentals over short horizons.

Our study contributes to the debate about the role of social media in capital markets. Since its arrival in the late 90s, regulators have repeatedly expressed concerns about social media impeding market efficiency and harming retail investors.⁶ While a host of recent studies provide evidence that different types of social media contain investment value (Chen et al., 2014; Jame et al., 2016; Bartov et al., 2018), there is little evidence to suggest that it leads to more informative retail trading. To the contrary, existing evidence emphasizes that social media can exacerbate behavioral biases harmful to performance (Heimer, 2016; Cookson et al., 2020; Ammann and Schaub, 2020)⁷. Our results establish the role of crowdsourced investment research in informing retail investor decision-making, while at the same time validating concerns about misleading research content (Kogan et al., 2020; Mitts, 2020).

Our analysis also advances the literature that studies the informativeness of retail trading. Early studies conclude that individual investors are unsophisticated “noise” traders who tend to suffer from behavioral biases and may push prices away from fundamentals (e.g., Barber and Odean 2000, Kumar and Lee 2006, Frazzini and Lamont 2008, Hvidkjaer 2008, Barber et al., 2009). In contrast, more recent work finds evidence of informed trading by individuals and speculates that retail investors gain insights from geographic proximity to firms, relations with employees, or insights into consumer preferences (e.g., Kaniel et al. 2012, Kelley and Tetlock 2013, 2017, Boehmer et al., 2020). Our findings highlight a specific mechanism, technology-enabled improvements in how retail investors produce and share investment research, as a likely channel by which individual investors become better informed.

Another stream of literature examines the use of technology by regulators to level the informational playing field between institutional investors and retail investors.⁸ Seeking Alpha is a technology-enabled market innovation whose ostensible purpose is to democratize the flow of investment analysis. Our findings illustrate how technological change enables new business models that can improve retail investors’ access to investment research and level the informational playing field among investors.