The annual report provides an opportunity for public companies to communicate their financial health, promote their culture and brand, and engage with a full spectrum of stakeholders. More and more companies are realizing that the target audience for disclosure is no longer just human analysts and investors, but also robots and algorithms that process information and recommend what shares to buy and sell. Indeed, machine downloads of corporate annual and quarterly filings have increased exponentially, from around 361,000 in 2003 to around 165 mn in 2016.
Given this development, companies that wish to communicate and engage with investors need to adjust how they disclose information in the age of artificial intelligence (AI). To help with this, we conducted a study to explore the feedback effect of technology on Given this development, companies that wish to communicate and engage with investors need to adjust how they disclose information in the age of artificial intelligence (AI). To help with this, we conducted a study to explore the feedback effect of technology on corporate disclosure. We start with a diagnostic test that connects the expected extent of AI readership for a company’s SEC filings on Edgar, measured by machine downloads (MDs) with how machine-friendly its disclosure is, measured by machine readability (MR).
The MR measure builds on five elements identified by recent literature as affecting the ease with which a machine can parse and synthesize. Our test shows that the expected MD is positively and significantly associated with the MR of the filing, suggesting that managers do adjust their reporting styles and cater to machine readers.
We find that institutional machine downloaders are more likely to be hedge funds or banking conglomerates that use big data and AI technologies in their trading. We further show that machine processing can have real impact on information dissemination. Trades in a company’s shares happen more quickly after a filing becomes public when MDs are higher, especially for firms with higher MR.
We next explore how firms manage sentiment and tone, as perceived by machines. The publication of Loughran and McDonald (LM) in the Journal of Finance in 2011 presents an instrumental event to investigate sentiment management pertaining to machine readers, because LM presented a new, specialized finance dictionary of positive/negative words that has served as a leading lexicon for algorithms to sort out sentiment in both the industry and academia.
We establish that firms that expect many MDs avoid LM-negative words but only after 2011 (the year of publication of the LM dictionary). Such a structural change is absent with respect to words deemed negative by the Harvard Psychological Dictionary, which was known to human readers for many years, suggesting that the change in disclosure style is indeed driven by the publication of the LM dictionary in association with machine readership.
While our analyses thus far focus on the textual information, the application of the underlying theme to speech serves as a further test. Applying a machine-learning software to extract emotional features on managerial speech in conference calls, we find that managers of companies with higher expected machine readership speak in more positive and excited tones, supporting the anecdotal evidence that managers increasingly seek professional voice coaches.
Our paper is the first to show how corporate disclosure in writing and orally has been reshaped by machine readership employed by algorithmic traders and quantitative analysts, or the ‘feedback effect’, which can lead not only to better dissemination of information, but also unexpected outcomes, such as manipulation and collusion. The technology advancement calls for more studies to understand the impact of and induced behavior by AI for corporations and the broader society.
Wei Jiang is professor of free and competitive enterprise at Columbia Business School. Sean Cao is an assistant professor of accounting, Baozhong Yang is associate professor of finance and Alan Zhang is a PhD student at the J Mack Robinson College of Business at Georgia State University. This opinion is based on the article ‘How to talk when a machine is listening: Corporate disclosure in the age of AI‘