Inhalt und Ziele des Forschungsprojekts
In zwei Teilprojekten werden die Tweets der 1500 U.S.-Unternehmen des S&P Index von 2013 bis 2020 ausgewertet. Im ersten Teilprojekt wird die Verbreitung von Pro-forma-Kennzahlen untersucht, d.h. Kennzahlen, die „dem Anschein nach“ reglementierten Kennzahlen entsprechen, jedoch unternehmensindividuell bereinigte Effekte aufweisen. Aufgrund fehlender Regulierung können Firmen über Twitter stark verkürzte und somit weniger transparente Informationen verbreiten. Im zweiten Teilprojekt liegt der Fokus auf Management Guidance, d.h. die Gewinnprognosen des Managements. Twitter ermöglicht die zeitnahe Verbreitung von Gewinnprognosen, die sich durch einen höheren Informationsgehalt positiv auf Investoren auswirken können. Es wird untersucht inwiefern Pro-forma-Kennzahlen und Management Guidance verstärkt über Twitter verbreitet werden, wie stark strategische Beweggründe dahinter liegen und welche Kapitalmarkteffekte beobachtet werden können.
Wissenschaftlicher und gesellschaftlicher Kontext des Forschungsprojekts
Die zunehmende Relevanz sozialer Medien ist unbestritten. Zunehmend nutzen auch private Aktieninvestoren mit geringem Finanzmarktvorwissen diese Quelle zur Informationsbeschaffung. Das Projekt leistet einen Beitrag zur allgemeinen kritischen Diskussion hinsichtlich Nutzen und Herausforderungen der sozialen Medien aus Sicht der Kapitalmarktkommunikation.
Firms increasingly use social media, particularly Twitter, to disseminate information (Jung et al., 2018; Zhou et al., 2015). While social media has become a common disclosure channel for firms (Best and Caylor, 2019), research on Twitter to date is largely generic with little insights into firms’ motives to use social media and the actual content of the disseminated information.
In this project, we aim to analyze S&P 1500 firms’ disclosures via Twitter from 2013-2020 for informational and strategic purposes. We focus on two specific types of firm disclosures in two separate manuscripts, namely reporting of management-adjusted earnings (non-GAAP) and firms’ forecasts (management guidance), respectively.
Twitter, as a corporate communication channel, has very specific characteristics that provides fruitful ground for investigation. Namely, the brief publishing format of Twitter allows managers to emphasize disclosures and timely updates about the firm can be immediately released to a large, diverse set of investors. Results from our project are expected to contribute to the current regulatory debate on non-GAAP measures and the use of social media for information dissemination of management guidance.
The proliferation of non-GAAP reporting has been steadily increasing (Black et al., 2018), which led to specific non-GAAP regulations for traditional communication channels. In the U.S., Reg G of the 2002 Sarbanes-Oxley Act requires a reconciliation from non-GAAP to GAAP earnings in firms’ disclosures. To date, social media is less regulated than other information channels (e.g., Bartov et al. 2018) and firms can provide shortened and/or less transparent information by using Twitter. Hence, non-GAAP disclosures via social media may be inherently different from earnings releases.
In the first paper, we aim to provide detailed insights into the determinants of non-GAAP tweets. Further, by comparing tweeted non-GAAP information with non-GAAP information disseminated through traditional channels, we analyze the nature of the non-GAAP disclosures (e.g., earnings vs. other information, exclusion types etc.). We further investigate the extent to which non-GAAP tweets appear to be associated with strategic disclosure behavior.
In the second paper on management guidance via Twitter, we first analyze firms’ Twitter information using natural language processing (NLP) and optical character recognition (OCR) to categorize the content of the tweets and to identify guidance-related tweets. We then analyze the timing of the guidance-related tweets to identify whether and which firms follow a guidance policy via Twitter (i.e., with regular, persistent guidance), and how this relates to their “former” established guidance policy documented using more traditional channels (e.g., IBES Guidance from press releases and conference calls). We further compare the accuracy of the guidance information in the tweets with the other/prior forecast information in the guidance sources. By that we analyze the accuracy of the highlighted guidance in the tweets and whether firms provide such information via Twitter strategically or whether it acts as another practical mechanism of information dissemination.
Finally, we analyze whether there is a differential market response, as well as liquidity effect, to guidance dissemination via Twitter compared to (or in addition to) traditional guidance disclosures.
Our project addresses recent calls to analyze the increasing volume of unstructured data (Lewis and Young, 2019). Not only have the narrative parts of annual reports, analyst reports, and conference calls increased, but also social media posts further add to the (unstructured) information available to firm stakeholders. Finally, understanding how guidance through social media channels can help managers’ efforts to provide an early outlook to firms’ financial performance will likely be informative and consequential to investors, analysts, policymakers, and researchers in the accounting and finance areas.
Last updated:29.05.2022
