Digital Financial Reporting

Reporting entities have flexibility to present disclosures differently as long as all the required disclosures are met.

Accountants creating financial reports use both quantitative measures and qualitative measures to provide such disclosures.

Quantitative measures means that you use an actual numbers disclose an amount or to show a change.  For example, "net income for the year was $1,000,000" is a quantitative measure.

Qualitative measures means not showing an actual number, but rather providing information in other ways such as using relative terms or such.  For example, disclosing an entities objective for holding or issuing derivative instruments, context necessary for understanding those instruments, strategies used to meet those objectives, and information helpful in understanding derivative activity is a qualitative measure.

Some disclosures tend to be rather objective in nature requiring little professional judgment. Other disclosures can be quite subjective, calling on an accountant to use their experience and judgment to provide the appropriate useful information.

Objective means that judgment is based on the facts of the situation and are not based on or influenced by personal feelings, preferences, tastes, or opinions. For example, balance sheets are included in financial reports and assets are part of a balance sheet is objective and there is no room for judgment.

Subjective means that judgment can be based on or influenced by personal feelings, preferences, tastes, or opinions. For example, whether a certain subsequent event is material and how to best disclose that event can be subjective, requiring significant professional judgment.

The overarching guidance to disclosing information is whether the information is useful in making useful decisions.  To be useful, the information possesses the following characteristics: relevance, reliability, comparability, and consistency.

Relevance means that the financial information makes a difference when making a decision. The information matters.

Reliability means that the financial information is free from bias and errors.

Comparability means that a standard set of financial reporting principles are used. But given options, reporting entities are free to chose between alternatives.  For example, one company might use FIFO for valuing inventories and another uses LIFO.

Consistency means that a reporting entity uses the same standard accounting principle and reporting approach/method from period to period. For example, a reporting entity cannot flip-flop between FIFO and LIFO.

A few specific aspects relating to comparability and consistency are worth pointing out because they are often confused.  Users of financial information often expect that every aspect of every reporting entities financial report to be comparable to every other reporting entity financial reports.  This is simply not the case.  Financial reports are not, and should not, be a 'form' which is filled in by an accountant.  One strength of US GAAP is its ability to let reporting entities report useful information specific to that entity.

Financial information reported by entities in the same industry sector tends to be more comparable than financial information reported by industries in different industry sectors. 

A reporting entity's disclosures from period to period tends to be very comparable.  While what disclosure information is considered useful by a given reporting entity for a given event or transaction;  once the disclosure approach is selected then the company specific disclosure of that information from period to period tends to be very consistent and comparable for any given reporting entity.

Accountants creating a financial report use disclosure rules/requirements, guiding principles, and their judgment when weaving together an appropriate financial report.

Some financial report disclosures tend to take the shape of very specific and objective quantitative measures. For example, the disclosure of earnings per share is an example of such a specific quantitative measure. These sorts of disclosures are like an "on/off" switch; either the disclosure is required or it is not and if it is required, what must be presented or disclosed is crystal clear.  There may be judgment involved in computing or measuring the amount disclosed, but the need for the disclosure itself tends to be objective.

Other disclosures take the shape of being more subjective in nature and use more qualitative measures. For example, the derivative instruments example used above, the meaning of a business acquisition or divestiture to the overall financial position of a reporting entity or which information about the acquisition or divestiture is the important information depends on many different criteria and it is the role of accountants to exercise their judgment and determine the appropriate disclosures, all things considered, using known guiding principles.

Understanding which disclosures tend to take which shape and otherwise understanding these moving pieces is critical for financial report taxonomy creation, financial report creation, and analysis of financial information expressed by these taxonomies and financial reports.

There are times when certain specific financial disclosures within two different financial reports will be very different, each reporting different facts.  Both financial disclosures being appropriate for the circumstances and both satisfy prescribed disclosure rules/requirements, both be useful, etc. 

Other times facts disclosed should be identical for reporting entities.

Understanding the difference is part of the art and science of financial reporting.

The American Institute of Certified Public Accountants (AICPA) released an exposure draft, Audit Data Standards, which I believe is very much on target. The topic of audit data standards is not really that "sexy", even less sexy than XBRL in general which tends to make most people's eyes glaze over. But these audit data standards are important at a number of levels.

Two people, Eric Cohen and Gianluca Garbellotto both of whom are named contributors to this exposure draft, have been focused on this topic for years. Their work with XBRL Global Ledger will help users of this standard realize even more benefit from digital financial reporting.

Here is why I see these audit data standards important:  They are part of the "glue" which ties a financial report to the business systems which contain the information.  Let me explain.

Years ago I was an auditor with the international firm which was called Price Waterhouse at the time (now PricewaterhouseCoopers).  As an auditor, I got to see how companies closed their books and generated their financial statements.  All the companies were doing the same thing, but each did things in slightly different ways.

When I left public accounting for my new role as the person responsible for generating financial reports, I remembered all that I had seen and synthesized what I learned into my approach to getting the financial statements out.

I did not have a name for it back then, but I eventually ended up calling what I had created a "closing book".  My closing book was a set of Microsoft Excel spreadsheets, Microsoft Access databases, and queries which tied the spreadsheets, databases, accounting system, and other odds and ends together.

The system only worked within the local network I was connected to (there was no Internet at that time), it was brittle and would not scale because I am not a very good programmer, and there were other shortfalls within the system I had created; but it was better than doing all this manually and it worked.

A fundamental piece of this system was the Excel spreadsheets.  These spreadsheets organized all the information which ended up in the financial statements as I learned how to do it as an auditor.  Basically, these schedules were high-quality, well organized audit schedules which I learned to create as an auditor with Price Waterhouse.

Here is an example of the schedule which I had to support all the information related to long-term debt (click on it to see a larger screen shot):

Long-term debt related supporting information (i.e. audit schedules)

If you look at this you will notice that it has a list of debt instruments, information about each debt instrument, information about the maturities of each debt instrument, the interest accrual, and so forth.  (Remember, this is a prototype, there is more information.)

What if:

• Rather than organizing this information using a proprietary tool such as Microsoft Excel, the information was organized within a global standards such as XBRL?
• Rather then linking the information together using the cells of a spreadsheet, the information was linked semantically using the concepts of XBRL Global Ledger, the US GAAP Taxonomy (or the IFRS taxonomy), etc.
• Rather than expressing the business rules as Excel formulas, the business rules where expressed using XBRL Formula so that the business rules could be shared along with the information
• When the information was changed, that change would flow through the entire "web" of linked information, updating all the information for that change
• The links were both to internal information which exists within your organization, but also to external (with the appropriate security) information which supports the information contained in or which supports the financial report.

Said another way, what if the entire process of creating a financial report was not woven together using inconsistent information formats and droves of overworked, expensive, highly-skilled accountants re-keying and re-footing and otherwise "ticking" and "tying" everything together; but rather more standard information formats where used, what supported the information for the accountant creating the information, the internal auditor reviewing what was created, the external auditor providing independent third-party review of that information, and other views where just different views of the same information?

When people talk about paving the "last mile" of finance, it is things like the audit data standards envisioned by the AICPA which will contribute to making this vision a reality.  These audit data standards are directly related to digital financial reporting, the end product or external financial report.  The audit data standards will let you do things like navigate from the financial report all the way back to the source information in whatever system that information came from.  You can see every business rule used in the journey that information took.

More specifically, I see these audit data schedules as a set of semantic spreadsheets which are all hooked together using additional semantics such as business rules which enables information to reliably flow throughout that entire system, whether a piece of that system is internal or external (i.e. say a regulator) to your organization.  Basically an "information supply chain".

This is not some "Star Trek" type fantasy, this is the power of standards.  Think of what life would be like without the universal product code (UPC).

Raynier van Egmond and I have finalized our document Financial Report Semantics and Dynamics Theory.

While this document may not be perfect in any means, it summarizes very useful information which we began piecing together over 12 years ago when we met at the first XBRL International conference. We both have been very lucky to have been able to participate throughout the entire life of XBRL in many different and interesting ways.

Raynier and I are both generalists. Generalists are very good at tying pieces together. We hope that this and other information we have pulled together over the years which helped us to learn about XBRL will help others, particularly specialists which will likely further hone any rough edges XBRL has which could use additional polishing.

By no means did Raynier and I pull all of this information together by ourselves.  We were very fortunate to meet many, many very talented technical and business oriented people from around the world over our years of working with XBRL.

Raynier is more technical that I am, vastly more technical.  He is also probably one of the most patient people on the face of the planet earth.  I remember one evening at a Red Robin in California when Raynier explained how linkbases work to me.

We hope you find this information helpful in making digital financial reporting a reality.

It is easy to make something complicated.  It is much more challenging to make something simple.  Not simplistic; simple.  There is a difference.

Two great quotes. 

First, Albert Einstein:

Quotes by famous people (http://www.aaanything.net/)

Second, Charles Mingus:

"Anyone can make the simple complicated. Creativity is making the complicated simple."

Digital financial reporting is getting closer to reality!

I was doing some fiddling around with Google Fusion Tables.  If you have not tried them, they are worth checking out. Basically, the way it works is this.  You can create "tables" within Google Documents. A table is different than a spreadsheet.  Tables can then be "merged" or related like you relate tables in a relational database.  As fusion tables sits right now, it does not look you can configure join properties, like you can in a SQL query.

A good place to start is watching this video tutorial.  

The video walks you through the basics. Although, I could not get the merge tables part to work.  So, I just put together my own data, tried that, which worked. (This is the tutorial page for Google Fusion Tables.)

This is my end product. 

To create that, I merged (or "fused") this table of (A) US State populations for 2009 with (B) this table of the geographic area polygons of each state for plotting on a map. 

This is a visualization of the table I created (i.e. you can embed the result of what you create within a web page or share a link to your tables or merged tables): 

This shows the power of creating data, making it available as "data" and not an HTML page, PDF, or a sloppily created Excel file.  Anyone can create data files using Excel, basically all you do is "save as CSV".  What is a little harder is understanding how to create a CSV file which is easy for a computer to work with.

This also shows the power of linked data.  The semantic web.  Properly modeled XBRL-based financial information contributes to the semantic web of linkable data.

In my example, I simply took an Excel spreadsheet which had population data, linked that population data to another data set which provides the polygon which can be used to visualize the state on a map such as Google Maps, linked the two data sets together via the postal code or ID which both data sets had; and then my population data can be visualized on a map rather than just as a list.

Of course, visualizing information on a map is not something which is new.  What is new is the ease with which you can do it.  And all this will get easier, and easier, and easier.  Plus as more and more data sets are available the utility of all this will become more clear.

This is linked XBRL-based information. It is rather rudimentary, but if you look close and use your imagination a bit, you can see what is going on.  You can see taxonomy information organized by "company" but you can also view the information by "component" (i.e. the balance sheet of each company).  I hope to have some better examples of this where you don't need to use your imagination as much.