Recent XBRL data improvement seen as first of more to come

XBRL leaders are making note of significant recent reductions in data errors, but they’re also planning new measures to bring about even more improvement in data quality.

The XBRL US Data Quality Committee says public companies reduced the number of errors in their filings by 64 percent in the first quarter of 2016 compared with the first quarter of 2015. The committee attributes the improvement to a set of validation rules the committee developed and released in November 2015 that are intended to help companies detect inconsistencies or errors in their XBRL-formatted financial statement data. That way, mistakes can be corrected before the data is filed with the Securities and Exchange Commission, which has called much attention to errors but has done little proactively to address the issue.

Mike Starr, vice president at service provider Workiva and chair of the Data Quality Committee, says it’s not clear how many companies voluntarily adopted and followed the committee’s validation rules, which are not required by the SEC or any other regulatory body. “I know there were 6,000 hits or viewings of the XBRL US link that connects you to the API (application program interface) that lets you run the rules,” he says.

The recent data error analysis revealed many companies reduced their number of errors, but not all companies, says Starr. “We don’t know that all service providers used those rules,” he says. “We know that those that are members of the Center for Data Quality adopted them.”

Workiva observed the validation rules and noticed its do-it-yourself customer base, which performs its own XBRL tagging by accessing software through the firm, appears to have done so as well, says Starr. “We saw a pretty significant improvement in our DIY group,” he says.

As exciting as the improvement is to Starr, he’s already anticipating even bigger improvements ahead due to continued initiatives of the Data Quality Committee. The committee plans soon to release guidance and validation rules for public comment focused on negative values and axis-member combinations.

Another release soon to follow, says Starr, is a proposed framework for element selection and extension use. Extensions in particular have led to problems with comparability of XBRL-formatted data, although extension use has tapered somewhat from the earliest days of XBRL submissions. “When employed, we think the framework is going to dramatically reduce use of extensions,” says Starr.

Staff at the Financial Accounting Standards Board, which maintains the GAAP Taxonomy companies used to complete their data tagging in XBRL, also is taking measures to improve data quality. FASB recently release the latest in its series of implementation guides, this one focused on dimensional modeling for financing receivables, and has published a new proposed guide to help steer decisions on when to use hierarchical domains versus distinct domains for dimensional modeling.