When the Securities and Exchange Commission mandated the tagging of financial filings with eXtensible Business Reporting Language (XBRL), its stated intent was to help investors and analysts access corporate data more easily, and standardize filings so that they could be compared across companies. But as we approach the fourth anniversary of the SEC mandate on January 30, a new study from Columbia Business School reports that investors and analysts in fact have not been taking advantage of data generated from XBRL financial reports.
The study found that less than 10% of investors and analysts surveyed use XBRL-tagged SEC data, and none have attempted to access XBRL data from corporate websites. Further, the few that have tried to use the data expressed disappointment both at the usability and quality of the underlying data, blaming regulators, filers, and the XBRL development community.
The study’s authors said they could not identify any users who said they were comfortable with the reliability of the XBRL-tagged data currently available. “Our hope is that this will provide a reality check as to what’s going on out there. There’s so much investment taking place, and it could be done so much more efficiently,” says Columbia Prof. Trevor Harris, who coauthored the study with Suzanne Morsfield, research scholar in the Faculty of Business. “There are companies that built this into their systems, and there are others doing minimal compliance and spitting out all sorts of silly errors. And if you’re a user, there are no tools that make this very easy to use.”
Users who attempt to analyze the XBRL data tend to find so many errors that they revert to their original data aggregators, Harris says. In particular, investors reported observing a large number of company-specific tags that make it difficult to search for information.
For example, when companies repatriate foreign earnings, there is a very simple one-or-two-line disclosure: the dollar value of what it has left abroad, and what it has brought home. “This would result in two standard, normal [XBRL] tags,” Morsfield explains. But instead of using the standard tags, some companies create their own, defeating the standardization intended by the SEC. “Each tag, by its nature, has a series of characteristics that help define it as an asset, or a current asset, and so on,” Harris says. But if companies create their own tag for the amount of money that’s been repatriated, the investor searching for the standard tag will retrieve no results. Instead, says Harris, “[The investor] now has all these other tags and has to sort out where they fit.”
In the report, the authors suggested several solutions to reduce the number of errors and unnecessary tags, including providing greater regulatory oversight for tagging or requiring filers to resolve errors.
Another problem, the study found, is the perception among some filers that investors are not interested in interactive data. Most filers, according to the report, doubt whether any investors are using their XBRL data, and believe they are bearing an unnecessary incremental cost, the benefits of which go to data aggregators who resell the data while reducing their own data-collection costs.
Harris suggests in the report that filers should “spend the effort they are investing in attempting to destroy the SEC’s XBRL regulation instead to improving the quality of their own data, as well as to making their own data more useful and accessible to users.” Not surprisingly, he has found that filers have reacted strongly.
Filers “have been arguing that investors don’t want it. What we’re saying is, XBRL might have issues, but there are plenty of investors who want interactive data that is available to them in a timely manner. And companies want to make that data available to shareholders, so I think everybody wins. But we should try to find a way to make it easier to use.”
To conduct the qualitative study, researchers gathered constituents (filers, analysts, technologists, and representatives from the SEC and the Financial Accounting Standards Board) for a three-hour roundtable discussion about the state of XBRL, and then followed up that discussion with an in-depth survey of 25 analysts and investors.
The study was organized around the original vision for interactive data: that it would provide incrementally more relevant, timely, and reliable information to more end-users, who could then manipulate and organize the data according to their own purposes at a lower cost.
Clearly, that vision has yet to be realized.