New Query Language and Features

We’re excited to announce the release of a significant new set of search features allowing you to find more documents of interest and to filter out documents not of interest. Previously we treated any input text query as just a series of terms (punctuation ignored, etc.) and matched any document that containing all of those terms. Now we support a complete Query Language with several specially interpreted operators and expressions. The linked page will remain the definitive reference, while I’ll elaborate and give some additional usage examples below.

This feature alone has been commonly requested and in fact, 6% of user saved searches already include a quoted phrase. Most of these searches are now more precise. A few may need to be broadened based on the results now returned.

There are many cases where company names or source-specific terminology (financial, legal, etc.) is composed of phrases of multiple words that can also appear individually with a much more common and less interesting meaning. For these cases, phrase search gives better results. As an example, compare results for the query Lands' End (without quotes) to "Lands' End".

Phrases can also improve person name searches. Note however that in the structured formatting of SEC filings, for example insiders forms 3-5, names are listed in surname-first order. The order of terms in phrases is significant, so for example: "Cook Timothy".

Boolean Queries

It’s now possible to broaden a search using the OR operator to match any of multiple alternative terms or phrases, or to narrow a search using NOT to eliminate documents matching other undesired terms or phrases. See the detailed boolean operator descriptions and examples in the Query Language.

Searching in Specific Federal Courts

We’ve had many users take advantage of searching for dockets at specific federal courts since we advertised that court codes were searchable and provided a listing last year. This approach wasn’t completely reliable however, so as part of the new query language, we’ve added a COURT: scope allowing you to select one or even multiple courts.

See the updated list of available federal courts.

Your existing saved searches that used raw (as text) court codes have now been converted to the new syntax. The raw court code method and syntax continues to work, its just not as reliable.

Standard Industrial Classification (SIC)

The SEC classifies companies by a primary industrial topic hierarchy known as the Standard Industrial Classification (SIC). We’ve added the ability to select SEC documents based on the SIC group or code of the filing company by using the SIC: scope. Some users will find there is a SIC group or code that closely relates to their area of interest. This system was established in 1937 and can be challenging for certain topics, for example involving recent technological progress. Its also limited to public companies and only allows for one class per company. Some examples:

SIC Companies Classified
3571 Electronic Computers Apple, Cray, Dell
7372 Services–Prepackaged Software Microsoft, Salesforce, Ellie Mae
7370 Services–Computer Programming, Data Processing, Etc. Google, Facebook, Red Hat
5961 Retail–Catalog and Mail-Order Houses Amazon, CDW, Wayfair
5331 Retail–Variety Stores Wal-Mart, Target, Fred’s

See our complete Standard Industrial Classification list.

8-K Events

The SEC defines a variety of events which require a company to file form 8-K with an associated item. The new query language makes it possible to reliably select 8-Ks with the item’s listed and linked searches below.

Note that we display these items as part of the 8-K summaries in search results and detail pages, but for several items we use our own simplified title for brevity. For example item 5.02 is displayed as:

§ Departure/Appointment/Compensation of Officers/Directors

Searching for this as a phrase won’t work, because it is indexed only in the original form. Use the above linked query instead. These queries can also be combined with other SEC filters like a geographic location and/or a SIC or other text query terms.

Conclusion

As with all prior features and enhancements, this work was the result of your feedback. Please let us know how well (or not) these features are working for your needs, or if you would like to see other features or data sources!