The largest 500 companies regulated by the U.S. Securities and Exchange Commission (SEC) are poised to submit quarterly financial reports that, for the first time, will be tagged using XBRL code — which will allow computers to “read” their content and make it easier for people to find and analyze financial data contained in the reports. However, a new study by researchers at North Carolina State University finds that XBRL filings submitted voluntarily as part of an SEC pilot for the program contained significant flaws. If the accuracy of the upcoming filings is not significantly improved, the researchers say, these errors will undermine confidence in the XBRL program from the very beginning.
The goal of XBRL is to make quarterly and annual reports computer-readable, allowing investors, companies, finance professionals and academics to sort and access data more efficiently. “Rather than going through a report page by page to find the information they’re looking for, users can plug their requests into the computer and have it pull up all relevant data,” explains Dr. Eileen Taylor, an assistant professor of accounting at NC State and co-author of the recent study on the accuracy of the voluntary XBRL filings.
Another benefit of XBRL is that it requires companies to use standardized tags for financial statement items, making it easier for users to compare data from different companies. This is significant because companies often use different terms to refer to the same thing. “For example,” Taylor says, “one company may refer to its ‘operating revenue,’ while another company may use the term ‘sales of goods net,’ and both mean the same thing. By using standardized tags, users can compare apples to apples, which really levels the playing field for individual investors,” who may not have the time or expertise to find and accurately compare data from these reports on their own.
But, while the XBRL concept is promising, the study from NC State found that reports from companies that participated in the voluntary pilot program contained multiple errors. “They were poorly tagged,” Taylor says, “and there were fundamental errors of accounting. One report, for example, contained too many zeros — turning millions into billions.” In their abstract, the researchers note that “These errors are serious because since XBRL data is computer-readable, users will not visually recognize the errors, especially when using XBRL analysis software.” In other words, users won’t be able to spot that something is wrong.
Now the SEC is requiring that companies file their reports in XBRL, as well as through traditional methods. The mandate is being phased in, with the 500 largest companies required to submit XBRL filings for the quarter ending June 15. The researchers are concerned that, if the upcoming XBRL filings do not represent a significant improvement from the voluntary reports, stakeholders in the financial community will not have any faith in the XBRL program — and it will be rendered relatively ineffective.
The study, “A Comparison of XBRL Filings to Corporate 10-Ks ? Evidence from the Voluntary Filing Program,” examined XBRL filings by 22 companies that participated in the SEC’s voluntary pilot program in 2006. The study was co-authored by Taylor, Drs. Al Y. S. Chen and Jon Bartley, who are both professors of accounting at NC State. The study will be presented at the American Accounting Association Annual Meeting being held in New York City, Aug. 2-5.