This week, the chair of the House Committee on Oversight and Government Reform, Rep. Darrell Issa, sent a letter to SEC Chair Mary Jo White, asking her to appear before the committee later this month to explain why the SEC has moved slowly on using XBRL data.
The letter focuses on three issues:
- Why is the SEC moving slowly to use the XBRL data it collects?
- Why does the SEC buy data from commercial vendors when the XBRL data is higher quality?
- Why doesn’t the SEC move to improve XBRL data quality?
Now, you don’t need a Ph. D in logic to see that there is already a problem with questions 2 and 3. You can’t have it both ways, either XBRL data quality is high (and you shouldn’t need to buy commercial data) or it is low (and you should be sending out comment letters).
The real answer is pretty clear. Commercial databases, such as Capital IQ‘s CompuStat, do have errors that the XBRL doesn’t, as this blog has demonstrated with respect to Amazon (AMZN). But they have something the XBRL data doesn’t have, which is a long backfile of data. With XBRL filing only coming online in 2009, there is less than 5 years of data for almost all filers, and any research needs 10-20 years of data to be valid. That by itself is a clear reason that the SEC should be buying commercial data well into the future.
At the same time, academic research shows that XBRL filed data is of high quality, and it is very unfortunate that Rep. Issa relies on the (itself poor quality) Harris and Morsfield study in questioning filer data quality. Such research was presented this past spring at the Kansas University 6th XBRL Conference.
That leaves question 1. Is the SEC moving slowly to use the data it collects? Only they can answer that. One of the outcomes of the Financial Regulation breakout session at the recent Data Transparency 2013 conference was that participants felt that the SEC should expand the use of XBRL, for example to the proxy statement and the 8-K earnings release.
My impression of the letter itself is that it is classic push marketing. Rather than ask “How are you using XBRL?” the letter assumes a problem. Unnecessary and clumsy, IMHO.
Bottom line: We are all on the same side, here. I thought the letter to be unnecessarily confrontational in its tone, and I look forward to hearing more about the success of XBRL at the SEC, in order to move that success even further forward.