Preparers and service providers making some of the most common mistakes in their XBRL filings are typically unaware that they are the target of the frequent calls from the Securities and Exchange Commission for corrections.
That’s the view of Mike Willis, a partner at PricewaterhouseCoopers and an active leader with XBRL International. He likens the scene to a parent investigating a broken lamp. “When something is broken, you ask the kids: who did it?” Willis said. “What do they say? Not me. That’s what a lot of the companies are saying right now. They think the SEC is talking about someone else. They perceive that their review process is adequate.”
Many preparers and those working with XBRL still don’t yet understand the difference between the physical rendering and the underlying metadata, according to Willis. They believe as long as their filing looks OK, it must be OK. “That’s like faxing an Excel worksheet and asking someone to validate the formula,” he said. “You can’t see the formula on a sheet of paper.”
Consumers want to take the data and put it into their own presentation programs, Willis said, but they need properly structured data to rely on it. As companies continue making routine errors, they inhibit the ability of consumers to use and rely on the data, he said. “In many cases, the errors are incredibly simple and wildly obvious,” he said.
Companies are still having a lot of problems with negative values, inadvertently expressing them in opposite terms, Willis said. It leads to expense items looking like revenue or dividends paid looking like dividends received, for example. “That’s the most common, prolific error in company reports, and many company reports have dozens of those errors,” he said. “It’s not like they have one or two.”
Most companies filing in XBRL are now liable for the accuracy of their filings, he said. The SEC’s limited liability protection for the first two years of filing was meant to give preparers time to get accustomed to working in XBRL. The SEC has reminded companies repeatedly that they are still making a great many small but important mistakes, and they’re now exposed to liability for those errors, Willis said.
If the metadata is not correct, consumers of XBRL data are putting erroneous data into their models, and arriving at the wrong answers as a result, he said. “So they relied on something that’s an error,” he said. That’s one of the big reasons investors and other consumers of XBRL data have been slow to make greater use of structured data. he said.
“The SEC has said they’re not going to hold people accountable,” Willis said. “The market is. The market is a more efficient model for holding people accountable for their quality issues.” The market has so far not brought a significant case forward, he said, but it’s likely just a matter of time. “An investor could say I relied on your information and the value went down, so write me a check,” he said.
Willis said companies need to make greater use of any number of validation tools that are available in the market to spot and correct their errors. “The error issue is preventing greater use of the data,” he said. “The analyst community has complained about the quality of the data. They like XBRL and want to use it, but what hinders they from using it all out is the level of errors. It requires them to do a lot of manual rework.”
Willis suggests a number of resources companies should explore to get a handle on the common errors that persist in their XBRL filings:
Webinar: Avoiding Common Errors in XBRL
Webinar: XBRL Data Quality: Beyond Validation and the Importance of XBRL
Avoiding Common Errors of XBRL Implementation
Avoiding Common Errors in XBRL Creation
XBRL US Best Practices
Recent SEC Staff Observations on Common Errors:
Three common reasons for Not using XBRL: What CFOs should really know
The New Math for Transparency
Webinar archive: XBRL: Finding a better way:
An explanation of the liability provisions of the SEC XBRL mandate:http://merrillcompliancesolutions.wordpress.com/2012/08/21/liability-for-xbrl-filings/