I blogged recently about how the year-end report from the SEC’s Investor Advocate urged the Commission to adopt rules that would make companies’ SEC filings machine-readable. An NBER paper points out that corporate disclosure has already been reshaped by machine processors, since those types of downloads have been steadily increasing over the past 15-20 years – and looks at how companies are adjusting their SEC disclosures when they know that machines are doing the reading.
“Machine readability” means that it’s easy for machines to separate and extract tables and numbers from text, it’s easy to identify tabular info because of clear headings, column separators and row separators, the filing contains all the needed info (without relying on external exhibits) and the characters are mostly standard ASCII. See page 31 for examples of high and low machine readability, pulled from actual reports. The researchers say that we humans are starting to make adjustments in our behavior to cater to our robot friends:
Our findings indicate that increasing AI readership motivates firms to prepare filings that are more friendly to machine parsing and processing, highlighting the growing roles of AI in the financial markets and their potential impact on corporate decisions. Firms manage sentiment and tone perception that is catered to AI readers by differentially avoiding words that are perceived as negative by algorithms, as compared to those by human readers.
Such a feedback effect can lead to unexpected outcomes, such as manipulation and collusion (Calvano, Calzolari, Denicolo, and Pastorello, 2019). The technology advancement calls for more studies to understand the impact of and induced behavior by AI in financial economics.
-Liz Dunshee, TheCorporateCounsel.net March 9, 2021
