The way in which the Securities and Exchange Commission (SEC)
has launched its Electronic Data Gathering, Analysis and Retrieval
System (EDGAR) is an intriguing tale that demonstrates the ways
individuals with a vision can affect the net. One player in the tale is
Jamie Love, of Ralph Nader's Taxpayer Assets Project (TAP). Love has
been lobbying for years to make publicly funded governmental data
available to the public. The SEC EDGAR database was listed as one of
the TAP "Crown Jewels" that Love argued should be made available to
the general public.
Meanwhile, another net personality, Carl Malamud, had wrangled a
book contract to travel the world and see how the net was being used
in different countries. While some of us talk about visiting Singapore
or Switzerland via the net, Malamud has done the opposite: he has
traveled in person to those countries. In the process, he also promoted
an agenda for making standards freely accessible on the net. He talked
the International Telecommunications Union (ITU) into granting him
permission to open an anonymous FTP site for the free distribution of
the ITU "Blue Book," a major source of many telecommunications
standards. It was so successful that the ITU pulled the plug on the
project to reevaluate if they really wanted to make their standards so
easily accessible. The whole detailed story of the trip and the
standards fight can be found in _Exploring the Internet_ [1].
Malamud found new network endeavors after his journeys. He
formed the Internet Multicasting Service (IMS) and started the Internet
Talk Radio. IMS broadcasts the weekly "Geek of the Week" interview
over the Internet. It is sent as an audio file. Since the interview can
generate a rather large file, and only certain hardware/software
combinations can play the file, Internet Talk Radio has had a rather
small audience. But, with the opening of the Internet site
town.hall.org, IMS is reaching a much larger audience with EDGAR and
other databases.
The availability of EDGAR on the Town Hall demonstrates the
collaborative opportunities the Internet offers. Teaming up with New
York University, Malamud's Internet Multicasting Service received a
National Science Foundation grant to make EDGAR available on the
Internet. (They also received additional support from Sun
Microsystems, HarperCollins, Persoft, and UUNET Technologies.) The
Internet Town Hall provides a variety of mechanisms for access to SEC
filings, from the low-end e-mail interface to high-end Mosaic
connections.
WARNINGS
There is a vast amount of useful data available through the EDGAR
project. However, the help files make clear that this free public
dissemination of EDGAR files is an experiment. To quote:
_"Because this is a research project you should expect the data
formats to change, the access methods to change, and you should
expect fluctuations in the numbers and types of users and services
supported. Do not become addicted to your favorite protocol or you may
have to go cold turkey. This data is provided as-is with no warranties
or guarantees of any kind" _[2].
Many other networked information sources could use a similar
disclaimer in their introductory materials. However, despite the lack
of guarantees, IMS is generally dependable with their services, and
they have been constantly improving means of access and ease of use.
And it has certainly proved popular. Just about a month after starting
up the service, Malamud reported sending out over 100,000 files and
that between 10,000 and 20,000 documents were delivered each week
[3].
E-MAIL and FTP
To their credit, the Internet Town Hall is committed to providing a
user-friendly front end for the low-end user with only e-mail access
to the Internet. Those with access to a gopher or WWW (World-Wide
Web) client will be more likely to use those interactive connections,
but those limited to e-mail and/or FTP with their Internet connection
can still retrieve all of the same data.
Send e-mail to either mail@town. hall.org or edgar@town.hall.org to
use the e-mail interface. The first two files that should be requested
are general.txt, the general introduction, and the searching help file.
To retrieve both, e-mail to either of the preceding addresses and
include help on the first line and send edgar/general.txt on the second
line within the body of the e-mail message. The subject line can be
left blank, although for mailers that demand something in the subject
line, putting something there will not cause a problem.
Finding and correctly transcribing the correct filename(s) for
retrieval is the major disadvantage to both the e-mail and FTP
approach. Instead of names such as microsoft.10k, the various filings
will be more like 0000884219-94-000033.txt. Thus the first task to
using either e-mail or FTP with EDGAR is to find what files are
available for specific companies and what the filenames are. This can
be done with the e-mail search engine or by retrieving one of the index
files.
The help file, which is retrieved with the help request, describes
the e-mail search capabilities. In brief, it requires sending an e-mail
message to either address and including a search argument in the text.
The search argument should begin with the search command, followed
by edgar, and then the keyword(s). Keywords can be joined by Boolean
operators and or or; otherwise, the default connector for multiple
keywords is an "and," which also requires that the keywords be listed
in the order given. For example, the search statement search edgar
jefferson pilot 10 k will retrieve the 10-K reports available for
Jefferson Pilot Corp.; whereas search edgar pilot jefferson will
retrieve nothing.
Old instructions for using the mail interface suggest retrieving the
company index file, which lists the company name and form of all the
filings available. To retrieve this file, e-mail to either of the
addresses mentioned earlier and include send edgar/full-
index/company.idx in the body of the e-mail message. This is a rather
large file, arranged alphabetically by company name. In addition to
company name, it lists form type, date filed, and the filename.
Whatever method is used to find appropriate files, be sure to avoid
typos in the retrieval command. To retrieve a report, include the send
command followed by the long filename in the next e-mail message.
For example, to retrieve the Jefferson Pilot Corp. 10-K report, use the
command send /edgar/data1/53347/0000053347-94-000001.txt. One
typo in the command will result in a null set or the wrong file.
For FTP access to the records, use anonymous FTP to connect to
ftp.town.hall.org. EDGAR files have been appropriately placed in the
edgar directory. The general.txt file can be retrieved from that
directory. The rest of the information is contained in the
subdirectories. Look in the /edgar/full-index directory for the
company.idx file; the /edgar/forms directory for descriptions of the
different SEC forms; and the /edgar/data1 directory for the actual
filings.
The most time-efficient method for retrieval with e-mail and FTP
is to combine the approaches. First, send the search request via e-
mail. When the results come back, then initiate the anonymous FTP
session. With the Jefferson Pilot example, the long filename suggests
that after connecting, you should change directories to the
/edgar/data1/53347 directory and retrieve the file named
0000053347-94-000001.txt.
GOPHER AND WWW
For those with access to a gopher or WWW client, the interface is
even easier than the e-mail connection, with the added bonus of being
interactive. With a gopher client, connect to gopher.town.hall.org, port
70, then choose the SEC EDGAR option in the menu. For direct access
with command line gopher clients, try gopher -p 1/edgar
gopher.town.hall.org 70.
Figure 1 shows the IMS Town Hall gopher server menu for EDGAR.
The search option works with the same principles as the e-mail search
option. The "About the EDGARÉ" directory includes a number of files
describing the project, including the general.txt file. The "SEC EDGAR
Documentation" includes some of the technical documentation that
comes with the distribution tapes. Of more interest is the listing of
various SEC forms that may be required. These can be found in the
"Guide to SECÉ" section. The other options contain the index files.
The interactive nature of the gopher connection offers a significant
increase in ease-of-use over the e-mail or FTP connections. Using the
search option will result in a list of hits, which can then be directly
retrieved, saved, printed, downloaded, or e-mailed to a client. The
explanations for the forms and the general information on the project
are also all readily available with gopher.
The Web provides a very similar list of options to the gopher
connection. Point Lynx, Mosaic, or another WWW client to the URL of
http://town.hall.org/edgar/edgar. html. The directory structure is a bit
different (Figure 2) from the gopher listing, but for most intents it
works about the same. Since the filings are tagged with SGML, IMS may
convert that to HTML in the future, which would create better looking
documents for Mosaic, but that has not yet been implemented.
ADVANTAGES AND DISADVANTAGES
Financial data from SEC filings have long been available in many
other online and CD-ROM sources. Mead Data Central makes the filings
available on Nexis. Disclosure produces Laser D, a CD-ROM set with
images of the complete filings. Data from the 10-K and other reports
are routinely available in a wide variety of electronic data sources.
For libraries without access to the full SEC filings, Disclosure's
Compact D/SEC has been a staple of many libraries' CD-ROM
collections. What then are the advantages and disadvantages of using
the Internet route to such data?
Certainly, the most cited rationale for checking out the Internet
Town Hall's EDGAR archives is cost. EDGAR access is free. Even with
the cost of Internet access figured into the total cost, the IMS' Town
Hall is significantly cheaper than most other sources of SEC filings.
However, EDGAR may not be as comprehensive or include additional
value-added information available through other sources.
Compact D/SEC from Disclosure, Inc. is a standard source for
company information in many libraries. As with most online sources,
EDGAR beats out Compact D/SEC for currency. On the other hand, EDGAR
is not yet as comprehensive and does not include all of the companies
available in D/SEC. Not all companies file their reports with the SEC
via EDGAR at this time. While all corporations are scheduled to be
online with EDGAR by May 1996, they are not all there yet.
The data contained in the different sources vary greatly. EDGAR
includes the reports just as they are filed. Compact D/SEC and other
services often add significant value to their products by combining
SEC filing information with data from other sources. A quick
comparison of an EDGAR 10-K report and a complete D/SEC record
show that in sheer mass, the EDGAR record is about twice as long.
For those with electronic access to the commercial versions of the
SEC filings data or to Disclosure's Laser D, EDGAR may not be as
comprehensive or useful. But for those with no ready access to this
information, other than a product such as D/SEC, EDGAR provides a
valuable supplementary source of corporate studies material. A more
detailed study of the data elements available in the different reports
and from the different services needs to be done. _(Editor's Note:
Watch for an upcoming article on this in DATABASE. --PH)_
Malamud, and all the others involved in bringing the SEC to the net,
should be congratulated. The Town Hall's offering of the EDGAR
database continues the trend of adding major database and information
repositories to resources available on the net.
REFERENCES
[1] Malamud, Carl. _Exploring the Internet: A Technical Travelogue_.
Englewood Cliffs, NJ: PTR Prentice Hall, 1992.
[2] _General Information about EDGAR Data_. Feb. 23, 1993 [sic,
actually 1994] revision. URL: ftp://town.hall.org/edgar/general.txt
[3] "EDGAR Internet Project is Racking up Impressive Numbers."
_Electronic Public Information Newsletter_ 4, No. 4, (Feb. 25, 1994).
[Contact epin@access.digex.net about availability information for the
newsletter.]
Communications to the author should be addressed to Greg R. Notess,
Montana State University Libraries, Bozeman, MT 59717-0332;
406/994-6563; greg@notess.com ; http://www.notess.com.
Copyright © 1995, Online Inc. All rights reserved.