At the U.S. Securities and Exchange Commission’s roundtable on machine-understandable data and superintelligence, Josh White, director of the Division of Economic and Risk Analysis, examined when structured financial disclosures deliver benefits that justify their costs. His remarks distinguished making information machine-readable from enabling automated systems to understand what a number represents, interpret its context and compare it appropriately. Downloads of the SEC’s structured datasets have increased dramatically over the past year, underscoring the need to understand how investors and other users apply the information.

The discussion centered on whether structured data improve investment decisions and reduce analytical costs, and where the same information can be used effectively without a structured format. Key questions included whether superintelligence can reduce issuers’ disclosure preparation costs while improving quality, and how to address automated systems’ reliance on inaccurate, inconsistent or incomplete information. Practical examples of where structured disclosures add value, and where they do not, will help inform the SEC’s approach to disclosure formats and usability as technology evolves.

Related updates