The SEC tots up the value of tagged data
The US Securities and Exchange Commission (SEC) has published its latest semi-annual report to Congress on the use of machine-readable data in corporate disclosures, a requirement of the Financial Data Transparency Act. It is a useful stock-take that demonstrates the reach and value of structured data.
On the benefits, the report draws together a large body of research to show that machine-readable disclosures reduce the cost of processing information, narrow the information gap between firms and investors, make share prices more informative and level the field between large and small players. Issuers gain too, through higher liquidity, a lower cost of capital and reduced audit fees. The costs, by contrast, fall mainly on the issuers who apply the tags and on the Commission that maintains the infrastructure, though the report notes compliance costs tend to fall as filers and software mature.
The freshest analysis here is on artificial intelligence. The report highlights findings that feeding large language models structured data rather than PDFs sharply reduces error rates, and points to emerging tools, including XBRL-based Model Context Protocol servers, that let analysts interrogate tagged filings through plain-language questions. With XBRL eliminating many of the errors that trip AI as it reads financial statements, the value of structured data continues to compound as the tools to use it improve.
Read the report here.
