Having read Toward XBRL Fluency: Common Mistakes Observed By The SEC, And How To Avoid Them what do you see as the main message in the SEC's observations?
The SEC's main message to filers is that although a significant effort has been put forth to submit XBRL data, certain matters remain and need to be addressed. The SEC's guidance is to consider these issues when submitting subsequent XBRL filings, so as to improve quality. Overall, the SEC wants to collect high-quality data with tagging applied consistently, which will enhance the usability of the data.
What mistake seems to bother the SEC the most?
Any of the issues that the SEC has identified bothers the SEC, but the hot topics now seem to be the improper entry of negative values and the unnecessary use of extensions. Improperly entering a number in the XBRL with the wrong positive/negative sign is bothersome because it conveys an erroneous value for the given item. Equally problematic is the use of unnecessary extensions, because it reduces the ability to compare the XBRL data across companies. If certain items are truly comparable across companies but one filer uses an unnecessary extension, then comparability is compromised.
Do SEC observations since 2009 reflect improvement in XBRL filings?
It is interesting to see that some of the same observations the SEC made in 2009 are repeated in 2010 and again in 2011. I expect this is because each year there is a new group of registrants submitting XBRL for the first time, and the same issues are repeated by the new group of filers. As companies have submitted multiple XBRL filings, however, the quality of those filings has improved. The reason for this is that over time, as mistakes are identified, they are corrected and are not repeated in subsequent filings.
How easy or hard is it to avoid these mistakes?
Most of the mistakes are relatively easy to avoid. I was surprised initially at the frequency and the type of errors that were appearing. Many of these errors could have been prevented by a review of the XBRL by someone with a basic knowledge of accounting and XBRL. For example, the problem of negative values being entered incorrectly is very easy to solve, given an understanding of the elements in the US GAAP taxonomy and how they are applied. Having the right skill set of resources for preparing and reviewing XBRL data is key to avoiding these errors. The right skill set combines knowledge of accounting, XBRL, and SEC requirements. Fortunately, the SEC has pointed out the errors and, since most are easily avoided, I expect their occurrence will diminish significantly.
One of the SEC observations is not as simple, however, and will take extra effort from the registrants to get it right. I'm referring to the observation about the proper use of axes and members to represent financial information. Unfortunately, there are many situations where XBRL filings include axis and member combinations that are misaligned or poorly structured. When the pieces of the data are misaligned or some pieces are missing, the user is provided with incomplete, inconsistent, or inaccurate information. This is part of the growing pains, but these problems can be and must be corrected. Although not as simple to correct as the other SEC observations, proper construction of the XBRL by knowledgeable individuals can get this back on track.
How realistic is the advice given by SEC staff?
The SEC staff's advice is relatively easy to implement. Companies should review their filings for each of the items pointed out by the SEC staff to determine if changes need to be made. For example, for the observation that the XBRL rendering does not have to match the HTML exactly, companies simply need to adopt the mindset that XBRL is about the data and not the presentation. Getting the data tagged properly in XBRL is important. The way the XBRL displays for human-readable purposes is of secondary importance.