Build a master index of SEC filings since 1993 with python-edgar

The SEC filings index is split in quarterly files since 1993 (1993-QTR1, 1993-QTR2...). By using python-edgar and some scripting, you can easily rebuild a master index of all filings since 1993 by stitching quarterly index files together. The master index file can be then feed to a database, a pandas dataframe, stata, etc...

An index file is a csv-like (pipe | separated) file that contains the following information:

• Company name (eg. TWITTER, INC)
• Company CIK (eg. 0001418091)
• Filling date (eg. 2013-10-03)
• Filling type (eg. S1)
• Filling URL on EDGAR (edgar/data/1418091/0001193125-13-390321.txt)

Once python-edgar is finished downloading index files, you can open an index file with csv.csvreader or pandas.read_csv to have the data programmatically usable. Remember that the delimiter character is |!

python-edgar can be used as a library called from another python script, or as a standalone script.

Features

• Compliant: Follows fair access guidelines established by the SEC at https://www.sec.gov/os/accessing-edgar-data
• Efficient: retrieve compressed archives instead of raw index file that are 10 times bigger
• Import as a library in your python project or run as a standalone script
• Python 3 only with 0 external dependencies (Python 3 only as of v3.0.0)

Usage

Using python-edgar as a library

Install from pip in a virtualenv

pip install python-edgar

Call the library

import edgar
edgar.download_index(dest, since_year, user_agent, skip_all_present_except_last=False)

Output

2018-06-23 12:41:46,451 - DEBUG - downloads will be saved to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o
2018-06-23 12:41:46,451 - DEBUG - downloading files since 2017
2018-06-23 12:41:46,451 - INFO - 6 index files to retrieve
2018-06-23 12:41:46,465 - DEBUG - worker count: 4
2018-06-23 12:41:48,359 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR3/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR3.tsv
2018-06-23 12:41:48,611 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2018/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2018-QTR2.tsv
2018-06-23 12:41:48,649 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR4/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR4.tsv
2018-06-23 12:41:48,935 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2018/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2018-QTR1.tsv
2018-06-23 12:41:49,750 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR2.tsv
2018-06-23 12:41:50,237 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR1.tsv
2018-06-23 12:41:50,376 - INFO - complete
2018-06-23 12:41:50,377 - INFO - Files downloaded in /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o

Using python-edgar as a standalone script

• Download this repository as a zip ("Clone or Download" green button, > Download as zip.)
• Open your terminal inside that directory and run python run.py -h. You can specify a destination directory for downloaded index files like -d edgar-idx (defaults to a temporary directory) and/or specify the year from which you want to build the index with -y 2017 (defaults to current year).

 $ python run.py -y 2017 -ua "MyCompany edward@mycompany.com"
2018-06-23 12:41:46,451 - DEBUG - downloads will be saved to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o
2018-06-23 12:41:46,451 - DEBUG - downloading files since 2017
2018-06-23 12:41:46,451 - INFO - 6 index files to retrieve
2018-06-23 12:41:46,465 - DEBUG - worker count: 4
2018-06-23 12:41:48,359 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR3/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR3.tsv
2018-06-23 12:41:48,611 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2018/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2018-QTR2.tsv
2018-06-23 12:41:48,649 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR4/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR4.tsv
2018-06-23 12:41:48,935 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2018/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2018-QTR1.tsv
2018-06-23 12:41:49,750 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR2.tsv
2018-06-23 12:41:50,237 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o/2017-QTR1.tsv
2018-06-23 12:41:50,376 - INFO - complete
2018-06-23 12:41:50,377 - INFO - Files downloaded in /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpr2Nk3o

Common issues

RuntimeError: An attempt has been made to start a new process before the current process has finished its bootstrapping phase.

See https://pytorch.org/docs/stable/notes/windows.html#multiprocessing-error-without-if-clause-protection

You will need to wrap your code in main() function to be able to run it on Windows:

def main(): 
    import edgar
    edgar.download_index(".", 2020, skip_all_present_except_last=False)    

if __name__ == '__main__':
    main()    

I am using python 2 ...

Python 2 support has been dropped as of October 2019. See https://pythonclock.org.

Stitch quarterly files to a master file

python-edgar does only one thing and does it well: getting and cleaning uncompressed quarterly index files to your computer. Use command line tools, in the spirit of unix philosophy, to stitch these index files together and create our master index file.

In this example, we called python run.py without arguments. It'll download every quarterly index file since 1993.

 python run.py -y 1993
 
2018-06-23 13:00:16,855 - DEBUG - downloads will be saved to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7
2018-06-23 13:00:16,855 - DEBUG - downloading files since 1993
2018-06-23 13:00:16,856 - INFO - 102 index files to retrieve
2018-06-23 13:00:16,879 - DEBUG - worker count: 4
2018-06-23 13:00:18,814 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR4/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2017-QTR4.tsv
2018-06-23 13:00:19,026 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR3/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2017-QTR3.tsv
2018-06-23 13:00:19,157 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2018/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2018-QTR2.tsv
2018-06-23 13:00:19,543 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2018/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2018-QTR1.tsv
2018-06-23 13:00:20,521 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2017-QTR2.tsv
2018-06-23 13:00:20,719 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2016/QTR4/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2016-QTR4.tsv
2018-06-23 13:00:21,016 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2016/QTR3/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2016-QTR3.tsv
2018-06-23 13:00:21,134 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2017/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2017-QTR1.tsv
2018-06-23 13:00:22,099 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/2016/QTR2/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/2016-QTR2.tsv
(...)
dcw07x6zrrr0000gn/T/tmpcF1rx7/1993-QTR2.tsv
2018-06-23 13:00:54,378 - INFO - > downloaded https://www.sec.gov/Archives/edgar/full-index/1993/QTR1/master.zip to /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7/1993-QTR1.tsv
2018-06-23 13:00:54,423 - INFO - complete
2018-06-23 13:00:54,424 - INFO - Files downloaded in /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7

Inspect the directory where our files where downloaded:

$ ls -lh /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7
total 4964656
drwx------  104 eswiac  staff   3.3K Jun 23 13:00 .
drwxr-xr-x  342 eswiac  staff    11K Jun 23 13:01 ..
-rw-r--r--    1 eswiac  staff   585B Jun 23 13:00 1993-QTR1.tsv
-rw-r--r--    1 eswiac  staff   580B Jun 23 13:00 1993-QTR2.tsv
-rw-r--r--    1 eswiac  staff   1.0K Jun 23 13:00 1993-QTR3.tsv
-rw-r--r--    1 eswiac  staff   2.8K Jun 23 13:00 1993-QTR4.tsv
-rw-r--r--    1 eswiac  staff   2.9M Jun 23 13:00 1994-QTR1.tsv
-rw-r--r--    1 eswiac  staff   2.3M Jun 23 13:00 1994-QTR2.tsv
(...)
-rw-r--r--    1 eswiac  staff    27M Jun 23 13:00 2017-QTR3.tsv
-rw-r--r--    1 eswiac  staff    27M Jun 23 13:00 2017-QTR4.tsv
-rw-r--r--    1 eswiac  staff    41M Jun 23 13:00 2018-QTR1.tsv
-rw-r--r--    1 eswiac  staff    31M Jun 23 13:00 2018-QTR2.tsv

Head to that directory so we can merge these files into a master file using cat

$ cd  /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpcF1rx7
$ cat *.tsv > master.tsv
$ du -h master.tsv
2.3G	master.tsv

Now you have this master index file. It's not sorted but that's easy to do (hint: Look into the sort command)

Grab filings from a specific company

Now that we have downloaded the index files it becomes easy, with a bit of command line scripting, to quickly filter by company and extract URLs to the filings we want with grep . In the following example we grep by CIK (1000045), store the output in an intermediate text file, which we re-open with cat and grep again by form 10-K. Prefix the paths with https://www.sec.gov/Archives/ and you'll get the full URL.

eswiac@mbp python-edgar (master) $ grep -h 1000045 /var/folders/bv/2zbdkyyj14766dcw07x6zrrr0000gn/T/tmpvwOzOU/* > 1000045.txt
eswiac@mbp python-edgar (master) $ cat 1000045.txt | grep -h 10-K
1000045|NICHOLAS FINANCIAL INC|10-K|2015-06-15|edgar/data/1000045/0001193125-15-223218.txt|edgar/data/1000045/0001193125-15-223218-index.html
1000045|NICHOLAS FINANCIAL INC|10-K|2016-06-14|edgar/data/1000045/0001193125-16-620952.txt|edgar/data/1000045/0001193125-16-620952-index.html
1000045|NICHOLAS FINANCIAL INC|10-K|2017-06-14|edgar/data/1000045/0001193125-17-203193.txt|edgar/data/1000045/0001193125-17-203193-index.html
1000045|NICHOLAS FINANCIAL INC|10-K|2018-06-27|edgar/data/1000045/0001193125-18-205637.txt|edgar/data/1000045/0001193125-18-205637-index.html

Query the master index with q

https://github.com/harelba/q allows you to run SQL directly on tabular data.

Use with caution: q does not use indexes so running queries against the master index will be very slow since it's rather large. Sorting the master index or narrowing the data to a smaller subset will make search faster. Ultimately you want to load the master index file into a proper database that's able to handle the size.

Some queries you may want to try

• q "SELECT COUNT(1) FROM 1999-QTR4.tsv"
• q -d"|" "SELECT * FROM master.tsv where c1 = 1418091 and c3 = '10-Q' order by c4"

License

MIT