Finding XBRL Files in SEC Company Filings
Summary
This exchange explains why an SEC filing download may look partly unreadable and how to locate the structured XBRL documents. The file in the question is described as a complete submission text file, which can contain several filing components, including embedded material that appears as binary or otherwise garbled text. The response directs readers to the filing index and archive directory, where the XBRL files are listed separately from the complete submission.
The practical lesson is to use the filing index to identify the document type before trying to parse a downloaded file. The exchange does not explain how to extract or analyze XBRL facts, and its example uses an older filing. It also includes a software suggestion and a commercial service mention; neither is evidence about data quality or a substitute for understanding the SEC filing structure.
Key ideas
- A complete SEC submission text file is not itself an XBRL data file.
- The filing index identifies the individual XBRL documents associated with a submission.
- Garbled sections can occur when embedded non-text material is viewed as plain text.
- Locating the correct files is a separate step from parsing XBRL facts.
Tags
Full text
# Reading XBRL Data from the SEC FTP SITE # Reading XBRL Data from the SEC FTP SITE After I ftp into the SEC Edgar site (ftp.sec.gov) I am able to pull the appropriate financial statements (i.e., 10-k, 10-q, 8-k, etc.) onto my local computer. However, when I go to open these files, I can't seem to make much sense of them. There are large parts of the files which look like garbage. These parts usually occur after the following tag: begin 644 Financial_Report.xls. This looks like it corresponds to the an excel file that for some reason is then embedded into the xbrl. Am I grabbing the right files and am I looking at them correctly? As an example, one of the 8-k's that I grabbed was the following (you can paste this into a browser) and if you go halfway down, there's quite a bit of garbage there. ftp://ftp.sec.gov/edgar/data/1112412/0001193125-13-028328.txt Thanks! BK ## Answer by Dirk Eddelbuettel (score 1) https://quant.stackexchange.com/a/10594 There is a promising-looking R package to read XBRL data which I have yet to try myself. But it looks like it does the legwork for you. ## Answer by Bondifrench (score 0) https://quant.stackexchange.com/a/10699 You are looking at the wrong place. For your particular example, check this link: http://www.sec.gov/Archives/edgar/data/1112412/000119312513028328/0001193125-13-028328-index.htm You can see that the link that you gave us is the complete submission text file, which is not XBRL. The XBRL files are the 6 files at the bottom of the link I gave you. Alternatively you should check the bottom of this link: http://www.sec.gov/Archives/edgar/data/1112412/000119312513028328/ Hope this helps. P.S. You should check www.finviews.com, we have already done all the work. We can provide all the XBRL reports directly into Excel, and as soon as a company files its reports (10Q, 10K etc..) with the SEC, it takes us less than 2 minutes to have all the data (main statements to start with) in Excel as the company published, so you can do your financial modelling afterwards.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.