Decomposition of XBRL-based Digital Report

To effectively understand an XBRL-based digital report, you need to understand how to decompose and compose such a digital report.  In this article I am going to focus on decomposition. Decomposition is simply about taking the one big piece that is the report model and report and breaking that larger "whole" into physical, logical, and functional "parts".

I am going to use a financial statement as an example, but these same ideas relate to reports at all levels of granularity including transaction level granularity, what I call the "mezzanine level" or working paper level granularity, financial statement level granularity, or financial analysis model level of granularity.  So, the explanation of the decomposition and composition of a financial statement also works for the other levels of granularity.

As I have pointed out, a financial statement is a knowledge graph. What I mean by this is that whether the financial statement is on a clay tablet, on a piece of paper, within the technical artifact of a PDF document or electronic spreadsheet (a.k.a. "e-paper"), or an XBRL-based digital report; all of those mediums can be decomposed similarly and that decomposition can be explained using the notion of what people call a knowledge graph. If you look at its essence, we are looking at a graph.

For my initial analysis, I looked at 6,751 public company financial statements represented using XBRL submitted to the U.S. Securities and Exchange Commission (SEC). This was first done way back in 2015 but I have repeated and expanded this experimentation over many years.  This is what led to the creation of the Standard Business Report Model (SBRM) and my Seattle Method.

And I am not saying that the way I decomposed these XBRL-based digital reports is the only way to decompose these digital reports. My decomposition might not be good for all use cases, but it works for my use cases.

Looking at a mixture of the physical technical format and the logic of what is being carried within that physical format, the information that is being reported; you can decompose a financial statement into the following:

  • Report model and report
    • Networks (XBRL technical artifact; a.k.a. Group used by FASB)
    • Hypercube (XBRL technical artifact; a.k.a. Table used by FASB)
    • Component (functional part which is a combination of two XBRL technical artifacts, a Network and a Hypercube; used to differentiate hypercubes that have the same name, such as "us-gaap:StatementTable", and Networks that contain more than one Hypercube)
    • Block of Information (logical artifact which exists within a Hypercube which carries information in the form of a logical pattern of information; "set", "roll up", "roll forward", or other fundamental logical pattern of information)
    • Fragment (arbitrary logical artifact)
    • Disclosure (functional logical artifact, why the block of information was provided; to satisfy some functional requirement to provide information to a regulator)
    • Fact, Report element, Rule (seem to be both physical, logical, and functional artifacts)
This is what the above artifacts look like in one software application which implements this decomposition of the physical, logical, and functional artifacts: (You can experience this yourself using a basic viewing application here)


So, why do you need all these different views of a business report?  Well, this lets you look at the report structures from the technical perspective (Network, Hypercube, Component), or the logical perspective (Block or Information Block, Fragment), or the functional perspective (Disclosure).

The Block or Information Block is what I was really after.  Why?  Well, the Information Block is the  core pattern within all reports.  Understanding that core pattern helps you get to the next important artifact which is what is that Information Block representing?  I used to use the term "Disclosure" because I was focused on financial statements and that is what financial statements provide; disclosures.

But I have recognized that the broader notion of an "Artifact" is better. Why?  Because the artifact could be a journal or ledger which is at the transactional level; an accounting working paper or audit schedule which is between the very granular transaction but not as summarized as a financial statement; the disclosures of the financial statement (this definition of disclosure includes each of the primary statements which are disclosures, policies, and all the individual disclosures within the notes of a financial statement), and financial analysis models which could contain information from one financial statement or a set of financial statements).

And so a "Disclosure" which is a type of "Artifact" is the core "Organism" of a financial statement.  A "working paper" is a core organism also, so is a journal or ledger or a financial analysis model.  Subsidiary ledgers fit into this conceptualization also.

Facts, report elements, associations and rules that are used to describe the organisms are the "molecules" as defined by Atomic Design Methodology.  

Additional Information:

Comments

Popular posts from this blog

Professional Study Group for Digital Financial Reporting

Big Idea for 2025: Semantic Accounting and Audit Working Papers

PLATINUM Business Use Cases, Test Cases, Conformance Suite