Digital Financial Reporting

What technical syntax do you use to implement process automation? Roll-your-own proprietary syntax?  OK, so you define your syntax; what about specifying the semantics?  You could use the semantic web stack; but that is only a syntax and you still have to define your semantics.

What if there was a better way?  I used XBRL to implement accounting process automation and financial report creation automation.  Basically, I stole all the best ideas I could find and added a few pieces that no one else provided.

In the document Accounting Process Automation Using XBRL I point out that XBRL is a well-thought-out flexible global standard syntax for representing facts and describing relations between facts. XBRL can handle financial and non-financial facts. XBRL excels at handling complex sets of facts such as an entire financial report. But XBRL can also be used for smaller sets of facts, binding information together in a tidy package, bound together with the strong “steel wires” of business logic.

Check it out and let me know what you think.

This is a prototype of transactions processing using machine intelligence to enable accounting processing automation.  More information to come later.  Just getting my thoughts together.

This is a prototype of an information repository.  I want to turn this into a complete knowledgebase.  More to come; stay tuned!

I mentioned some Excel-based extraction and validation tools that I created.

Well, I created an improved version.  There are two significant improvements to this fundamental accounting concepts validation tool:

1. Validate one file at a time.  Rather than validating many different files and then comparing them like the other tools, this one validates ONE filing at a time to let you focus on that filing.  Again, you can use a URL from the SEC EDGAR repository or a local file.
2. I added what I call a "line of reasoning" output to help you understand what the tool is doing.

Let me know what you think.  I believe the line of reasoning turned out nice.  I know that there is a lot of room for improvement such as formatting the numbers, providing additional information, etc.  You can add that functionality yourself.

I have perfected what I call a Disclosure Best Practices resource. I am envisioning this as a resource for intermediate accounting students to begin with. Then it could be expanded and made useful to professional accountants that external create financial reports.

This early prototype which took the form of a document helps you understand where this Disclosure Best Practice resource is going. The conceptual model of a financial report also helps you understand this resource.

There are two important things that might not be apparent by just looking at this resource.  First, the information you see is 100% machine readable as well as being readable by humans.  Second, the organization of the information into the form you see is 100% automated. That organization is achieved using metadata from the conceptual model.

If you don't understand this Disclosures Best Practices resource or why it is important, you should consider reading the document Closing the Skills Gap.

The foundation for this Disclosures Best Practices is the Reporting Checklist and Disclosure Mechanics machine-readable business rules.  Validation is run to make sure you only get examples that are of high quality.  So for example, here is the validation of the Microsoft 10-K.  Imagine that same information for (a) all public companies that report to the SEC and (b) reading the information using automated machine-based processes rather than reading the information in a web browser.

Stay tuned for more information.  If you want to keep on top of all this, participate in the campaign to improve disclosure quality.  By March 31, 2018 the 65 disclosures of the campaign and the disclosures in the Disclosure Best Practices resource should be in sync.  Then; I will start adding more disclosures to expand the set available.

Be sure you fiddle with the prototype.  There is more there than you might recognize by just glancing at it.  This is the primary entry point into the resource.  There are about 50 different disclosures.  You can see information about each of the some 50 disclosure. You can get information about economic entities that create the reports.  Information is linked to US GAAP XBRL Taxonomy information and the Accounting Standards Codification.  Topics which will enhance filtering and searching for disclosures will be incorporated. More metadata exists to filter reporting entities by reporting style, by accounting activity, by sector, by size of the company, and other such information is incorporated, but is not apparent yet.

One next step for me is to create this same resource for the XASB reporting scheme. The reason for that is to show that the conceptual model is not specific to US GAAP or even IFRS; it is general. Ultimately, it will work for all of these profiles.

#######################################################

• Here is an older version of this same prototype.
• Here is an older and slightly different version of this that uses topics.
• Here is another older version that leverages the concept arrangement patterns.
• Here is another older version that leverages topics of the ASC.
• Here is another older version that lets you compare the Level 3 Text Block and the Level 4 Disclosure Detail.