Digital Financial Reporting

In the document the Essence of Accounting I point out the mathematical model of double entry accounting and the accounting equation.  I pointed out that there are a number of financial reporting schemes and that every financial reporting scheme, by definition, follows the double entry accounting model and some form of the accounting equation.

I also pointed out that each financial reporting scheme defines some set of high-level financial report elements that make up the foundation or "key stones" of the financial reporting scheme.  This is usually done in the conceptual framework of the financial reporting scheme.  For example, the FASB defines the ten terms that form the foundation of US GAAP in SFAC 6 Elements of Financial Statements: Assets, Liabilities, Equity, Comprehensive Income, Investments by Owners, Distributions to Owners, Revenues, Expenses, Gains, Losses.

This is the same for: 

• US GAAP
• IFRS
• UK GAAP (FRS 102)
• Australian IFRS

So, you can expand the patterns down a little lower which I have summarised in my baseline PROOF representation and compared my PROOF representation to US GAAP and IFRS (working on FRS 102 now which will be forth coming).

And so while such financial reporting schemes are described inconsistently for a number of reasons, the essence of what all of these financial reporting schemes is saying, or should be saying, is very consistent.  While the definitions of the terms are not necessarily consistent, the mechanical aspects are very, very consistent.  What I mean is that US GAAP, IFRS, and UK GAAP define "Assets", "Liabilities" and "Equity" differently; each of the financial reporting schemes does define Assets, Liabilities, and Equity and it is always the case that the rule "Assets = Liabilities + Equity" in some form.

This is the true for a high-level set of about 100 or so terms, associations between terms, structures, and rules which are consistently defined or implied by US GAAP, IFRS, UK GAAP (FRS 102).  I will compile that list. Here is a prototype of that list: 

• Fundamental Accounting Concepts Associations: Human readable | Machine readable
• Fundamental Accounting Concepts Roll Up Relations: Human readable | Machine readable
• Fundamental Accounting Concepts: Machine readable

I suggest that this is also true for IPSAS, FRF for SMEs, Canadian GAAP, Australian GAAP, and many other (any other) financial reporting schemes. Why would you go through all this trouble.  Have a look at this:

Machine readable:

• Mapping between/across financial reporting schemes

Don't understand? The ZIP archive below contains an Excel spreadsheet for extracting information from the XBRL file above.  Run that and have a look at the code.  The code is pretty simplistic, but the point is you can grab information from the XBRL file. 

• Excel tool for extracting information from XBRL file

Still don't understand? You are going to want to read Computational Professional Services.  Have a look at this financial reporting metadata. This is not going to work by magic.  Start at the top, work your way down; each financial reporting scheme pretty much works the same way.

This is a huge opportunity for software engineers that understand how to communicate with accountants and accountants that understand how to communicate with software engineers.

Piece by piece, the metadata that will drive the next era of accounting, reporting, auditing, and analysis will be put together.  Here is some US GAAP and IFRS financial reports to fiddle with. HMRC does not make information available, ESMA is not up and running yet.  This is not about reporting to the SEC, this is about understanding the sorts of things you can do with reports because the information is structured, complements of 10 years of XBRL-based reports submitted to the SEC.

Will XBRL be the format used?  Maybe, maybe not.

The United Kingdom has an interesting financial reporting scheme called FRS 105, The Financial Reporting Standard applicable to the Micro-Entities Regime.  That financial reporting scheme uses the FRS 102 XBRL Taxonomy (here is an easy to read version).  Here is the actual XBRL terms: core | bus.

The UK is not modularizing the XBRL taxonomies in a manner similar to the IFRS or US GAAP XBRL Taxonomy.  They combined things together than you cannot undo, which is unfortunate.

This video from Grakn Labs, Building a Digital Tax Accountant with a Knowledge Graph Mind, walks you through how to build a tax expert system (in this case Taxfix). 

These same ideas can be used to create an accounting and financial reporting oriented expert system for US GAAP and/or IFRS or any other financial reporting scheme.

You might find it useful to check out these three articles: 

• Knowledge Graphs at a Glance
• Neo4j vs Grakn Part 1: Basics
• Neo4j vs Grakn Part 2: Semantics

Clearly a tax reporting system, an accounting related expert system, a financial reporting related expert system, an audit related expert system, or an analysis related expert system would work differently and have different work flows; but fundamentally, these are all the same sorts of things.

Here is how a software engineer and I built a working prototype expert system for creating XBRL-based financial reports and verifying that the reports were properly functioning.  This video play list shows what we created.  You can download it and try it out (email me and I will send you a license good for a year).

As The Year of the Graph puts it: Data and AI are the future; graph is the future of the future.  Structured data, artificial intelligence, graph databases all have implications for accounting, reporting, auditing, and analysis.  If you don't understand the details of what a knowledge graph is, this Knowledge Graphs at a Glance is helpful.

It is time for professional accouontants to expand their knowledge.

Start here: Computational Professional Services.