SEC Staff Observations of XBRL Data

The SEC recently posted Staff Observations From Review of Interactive Data Financial Statements. It is an excellent opportunity to understand what the SEC is looking at in these early days and part of Shelly Parratt's promise (subscription required) to help filers succeed.

I think the fact that we're getting this kind of guidance is another indication that the data will play an important role internally at the SEC. The guidance cuts to the heart of a couple of the more common problems seen in filed data that touch both the accounting and technical sides of XBRL. These data quality issues can hurt comparability and automated analysis. More importantly for the issuing community, they will eventually raise red flags inside the SEC and attract attention to your filing.

Over the next few posts I'll take a deep dive into a couple of the juicier sections. For now, here is an overview.

• Rendering -- Relatively short piece. As Walter Hamscher has mentioned before if there's a conflict between the viewable info and the data, get the data right. Mentions using abstracts to create first level sections in a table. Also mentions formatting that neither the SEC's viewer or XBRL can pull off at the moment.
• Element Selection -- The largest part of the document. I'll review each of the important concepts in future posts.
• Context References -- Partially another bit on elements, partially about the funky beginning date thing.
• Negative values and negated labels -- what we lovingly refer to as signage problems. These errors drove at least 20% of calculation inconsistencies.
• Decimals
• Other observations

-- Ed Hodder