Study suggests corporate disclosure is made for and by AI robots as EU mulls digitisation of non-financial reporting standards

Meanwhile, Cambridge Judge Business School launches the Regulatory Genome Project

Researchers from three American universities say they have found new evidence of how companies are shaping their disclosures for a predominantly AI audience of algorithmic traders, robot investment advisors and quantitative analysts. 

In a working paper for the National Bureau of Economic Research (NBER), a policy think tank, the researchers argued their findings show that companies are preparing filings in a way that is more friendly to machine parsing and processing, stressing the influence of AI in the financial markets and their potential impact on corporate decisions.

The researchers also argued that it applies not just to filings but earnings conference calls and other interactive events, as companies try to adapt sentiment and tone perception to AI readers, for example, avoiding words perceived negatively by algorithms.

The researchers wrote: “While the literature has shown how investors and researchers apply machine learning and computational tools to extract information from disclosure and news, our study is the first to identify and analyse the feedback effect, i.e., how companies adjust the way they talk knowing that machines are listening.”

The researchers analysed regulatory filings, such as 10-K and 10-Q files, from the Securities and Exchange Commission’s EDGAR database. They are Sean Cao from Georgia State University, Baozhong Yang and Alan L. Zhang from Georgia State University and Wei Jiang from Columbia University.

They cautioned: “While some adaptive behavior, such as making disclosure more machine-reading friendly, is innocuous or even welcome, other algorithm-induced changes, such as the expression of sentiment, highlight the increasing challenge on machine learning to be “manipulation proof” in that the algorithms will learn to anticipate the strategic behavior of informed agents without observing it in the training samples.” 

NBER working papers are circulated for discussion and are not peer-reviewed. 

Meanwhile this week, the University of Cambridge launched a project called the Regulatory Genome Project aimed at sequencing the world’s financial regulation to provide an open-source infrastructure for all countries, particularly in developing regions.

The project is based on research by the Cambridge Centre for Alternative Finance (CCAF) at Cambridge Judge Business School and the Department of Computer Science and Technology of the University of Cambridge.

Robert Wardrop, Faculty (Professor level) in Management Practice at Cambridge Judge and Director of the CCAF, said: “Evolving a de facto standard for the representation of regulatory information in a machine-readable form will enable the interoperability and innovation needed to realise the full potential of digital financial services. We are excited about the vast potential of this project to benefit both public and private sector interests.”

Against the backdrop of these developments, the ongoing preparatory work to potentially set non-financial reporting standards in the EU is studying the development of a data taxonomy for the digitisation of sustainability information. The European Commission will decide next year whether it gives the go-ahead to these standards as part of the review of the Non-financial Reporting Directive. 

The preparatory work is being undertaken by expert groups within the European Financial Reporting Advisory Group (EFRAG). In their latest update report the experts said the lack in the EU of a digital taxonomy of non-financial data results in higher regulatory costs, including time and resources. 

“It also weakens the auditability of information and increases risks due to increased subjectivity, differences in interpretations, misinformation, and inadvertent partial compliance or non-compliance.”

Among EFRAG’s experts with a background and experience in this field are Liv Watson, a co-founder of XBRL and the XBRL International Consortium, and Donato Calace, Innovation Vice-President at ESG risk management software firm Datamaran.

Watson is also a Senior Advisor and Digitisation Lead for the Impact Management Project, which coordinates the alignment project of the Group of Five sustainability standards setters.

See the earlier article Artificial intelligence: The rise of the responsible robots